OpenCloudOS-Kernel/mm/shmem.c

3453 lines
89 KiB
C
Raw Normal View History

/*
* Resizable virtual memory filesystem for Linux.
*
* Copyright (C) 2000 Linus Torvalds.
* 2000 Transmeta Corp.
* 2000-2001 Christoph Rohland
* 2000-2001 SAP AG
* 2002 Red Hat Inc.
* Copyright (C) 2002-2011 Hugh Dickins.
* Copyright (C) 2011 Google Inc.
* Copyright (C) 2002-2005 VERITAS Software Corporation.
* Copyright (C) 2004 Andi Kleen, SuSE Labs
*
* Extended attribute support for tmpfs:
* Copyright (c) 2004, Luke Kenneth Casson Leighton <lkcl@lkcl.net>
* Copyright (c) 2004 Red Hat, Inc., James Morris <jmorris@redhat.com>
*
* tiny-shmem:
* Copyright (c) 2004, 2008 Matt Mackall <mpm@selenic.com>
*
* This file is released under the GPL.
*/
#include <linux/fs.h>
#include <linux/init.h>
#include <linux/vfs.h>
#include <linux/mount.h>
#include <linux/ramfs.h>
#include <linux/pagemap.h>
#include <linux/file.h>
#include <linux/mm.h>
#include <linux/export.h>
#include <linux/swap.h>
#include <linux/uio.h>
static struct vfsmount *shm_mnt;
#ifdef CONFIG_SHMEM
/*
* This virtual memory filesystem is heavily based on the ramfs. It
* extends ramfs by the ability to use swap and honor resource limits
* which makes it a completely usable filesystem.
*/
#include <linux/xattr.h>
#include <linux/exportfs.h>
#include <linux/posix_acl.h>
#include <linux/posix_acl_xattr.h>
#include <linux/mman.h>
#include <linux/string.h>
#include <linux/slab.h>
#include <linux/backing-dev.h>
#include <linux/shmem_fs.h>
#include <linux/writeback.h>
#include <linux/blkdev.h>
#include <linux/pagevec.h>
#include <linux/percpu_counter.h>
#include <linux/falloc.h>
#include <linux/splice.h>
#include <linux/security.h>
#include <linux/swapops.h>
#include <linux/mempolicy.h>
#include <linux/namei.h>
#include <linux/ctype.h>
#include <linux/migrate.h>
#include <linux/highmem.h>
#include <linux/seq_file.h>
#include <linux/magic.h>
#include <linux/syscalls.h>
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
#include <linux/fcntl.h>
#include <uapi/linux/memfd.h>
#include <asm/uaccess.h>
#include <asm/pgtable.h>
#define BLOCKS_PER_PAGE (PAGE_CACHE_SIZE/512)
#define VM_ACCT(size) (PAGE_CACHE_ALIGN(size) >> PAGE_SHIFT)
/* Pretend that each entry is of this size in directory's i_size */
#define BOGO_DIRENT_SIZE 20
/* Symlink up to this size is kmalloc'ed instead of using a swappable page */
#define SHORT_SYMLINK_LEN 128
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
/*
* shmem_fallocate communicates with shmem_fault or shmem_writepage via
* inode->i_private (with i_mutex making sure that it has only one user at
* a time): we would prefer not to enlarge the shmem inode just for that.
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
*/
struct shmem_falloc {
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
wait_queue_head_t *waitq; /* faults into hole wait for punch to end */
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
pgoff_t start; /* start of range currently being fallocated */
pgoff_t next; /* the next page offset to be fallocated */
pgoff_t nr_falloced; /* how many new pages have been fallocated */
pgoff_t nr_unswapped; /* how often writepage refused to swap out */
};
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
/* Flag allocation requirements to shmem_getpage */
enum sgp_type {
SGP_READ, /* don't exceed i_size, don't allocate page */
SGP_CACHE, /* don't exceed i_size, may allocate page */
SGP_DIRTY, /* like SGP_CACHE, but set new page dirty */
SGP_WRITE, /* may exceed i_size, may allocate !Uptodate page */
SGP_FALLOC, /* like SGP_WRITE, but make existing page Uptodate */
};
#ifdef CONFIG_TMPFS
static unsigned long shmem_default_max_blocks(void)
{
return totalram_pages / 2;
}
static unsigned long shmem_default_max_inodes(void)
{
return min(totalram_pages - totalhigh_pages, totalram_pages / 2);
}
#endif
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
static bool shmem_should_replace_page(struct page *page, gfp_t gfp);
static int shmem_replace_page(struct page **pagep, gfp_t gfp,
struct shmem_inode_info *info, pgoff_t index);
static int shmem_getpage_gfp(struct inode *inode, pgoff_t index,
struct page **pagep, enum sgp_type sgp, gfp_t gfp, int *fault_type);
static inline int shmem_getpage(struct inode *inode, pgoff_t index,
struct page **pagep, enum sgp_type sgp, int *fault_type)
{
return shmem_getpage_gfp(inode, index, pagep, sgp,
mapping_gfp_mask(inode->i_mapping), fault_type);
}
static inline struct shmem_sb_info *SHMEM_SB(struct super_block *sb)
{
return sb->s_fs_info;
}
/*
* shmem_file_setup pre-accounts the whole fixed size of a VM object,
* for shared memory and for shared anonymous (/dev/zero) mappings
* (unless MAP_NORESERVE and sysctl_overcommit_memory <= 1),
* consistent with the pre-accounting of private mappings ...
*/
static inline int shmem_acct_size(unsigned long flags, loff_t size)
{
return (flags & VM_NORESERVE) ?
0 : security_vm_enough_memory_mm(current->mm, VM_ACCT(size));
}
static inline void shmem_unacct_size(unsigned long flags, loff_t size)
{
if (!(flags & VM_NORESERVE))
vm_unacct_memory(VM_ACCT(size));
}
static inline int shmem_reacct_size(unsigned long flags,
loff_t oldsize, loff_t newsize)
{
if (!(flags & VM_NORESERVE)) {
if (VM_ACCT(newsize) > VM_ACCT(oldsize))
return security_vm_enough_memory_mm(current->mm,
VM_ACCT(newsize) - VM_ACCT(oldsize));
else if (VM_ACCT(newsize) < VM_ACCT(oldsize))
vm_unacct_memory(VM_ACCT(oldsize) - VM_ACCT(newsize));
}
return 0;
}
/*
* ... whereas tmpfs objects are accounted incrementally as
* pages are allocated, in order to allow huge sparse files.
* shmem_getpage reports shmem_acct_block failure as -ENOSPC not -ENOMEM,
* so that a failure on a sparse tmpfs mapping will give SIGBUS not OOM.
*/
static inline int shmem_acct_block(unsigned long flags)
{
return (flags & VM_NORESERVE) ?
security_vm_enough_memory_mm(current->mm, VM_ACCT(PAGE_CACHE_SIZE)) : 0;
}
static inline void shmem_unacct_blocks(unsigned long flags, long pages)
{
if (flags & VM_NORESERVE)
vm_unacct_memory(pages * VM_ACCT(PAGE_CACHE_SIZE));
}
static const struct super_operations shmem_ops;
static const struct address_space_operations shmem_aops;
static const struct file_operations shmem_file_operations;
static const struct inode_operations shmem_inode_operations;
static const struct inode_operations shmem_dir_inode_operations;
static const struct inode_operations shmem_special_inode_operations;
static const struct vm_operations_struct shmem_vm_ops;
static LIST_HEAD(shmem_swaplist);
static DEFINE_MUTEX(shmem_swaplist_mutex);
static int shmem_reserve_inode(struct super_block *sb)
{
struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
if (sbinfo->max_inodes) {
spin_lock(&sbinfo->stat_lock);
if (!sbinfo->free_inodes) {
spin_unlock(&sbinfo->stat_lock);
return -ENOSPC;
}
sbinfo->free_inodes--;
spin_unlock(&sbinfo->stat_lock);
}
return 0;
}
static void shmem_free_inode(struct super_block *sb)
{
struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
if (sbinfo->max_inodes) {
spin_lock(&sbinfo->stat_lock);
sbinfo->free_inodes++;
spin_unlock(&sbinfo->stat_lock);
}
}
/**
* shmem_recalc_inode - recalculate the block usage of an inode
* @inode: inode to recalc
*
* We have to calculate the free blocks since the mm can drop
* undirtied hole pages behind our back.
*
* But normally info->alloced == inode->i_mapping->nrpages + info->swapped
* So mm freed is info->alloced - (inode->i_mapping->nrpages + info->swapped)
*
* It has to be called with the spinlock held.
*/
static void shmem_recalc_inode(struct inode *inode)
{
struct shmem_inode_info *info = SHMEM_I(inode);
long freed;
freed = info->alloced - info->swapped - inode->i_mapping->nrpages;
if (freed > 0) {
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
if (sbinfo->max_blocks)
percpu_counter_add(&sbinfo->used_blocks, -freed);
info->alloced -= freed;
inode->i_blocks -= freed * BLOCKS_PER_PAGE;
shmem_unacct_blocks(info->flags, freed);
}
}
/*
* Replace item expected in radix tree by a new item, while holding tree lock.
*/
static int shmem_radix_tree_replace(struct address_space *mapping,
pgoff_t index, void *expected, void *replacement)
{
void **pslot;
mm: shmem: save one radix tree lookup when truncating swapped pages Page cache radix tree slots are usually stabilized by the page lock, but shmem's swap cookies have no such thing. Because the overall truncation loop is lockless, the swap entry is currently confirmed by a tree lookup and then deleted by another tree lookup under the same tree lock region. Use radix_tree_delete_item() instead, which does the verification and deletion with only one lookup. This also allows removing the delete-only special case from shmem_radix_tree_replace(). Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reviewed-by: Minchan Kim <minchan@kernel.org> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Bob Liu <bob.liu@oracle.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Jan Kara <jack@suse.cz> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Luigi Semenzato <semenzato@google.com> Cc: Metin Doslu <metin@citusdata.com> Cc: Michel Lespinasse <walken@google.com> Cc: Ozgun Erdogan <ozgun@citusdata.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Roman Gushchin <klamm@yandex-team.ru> Cc: Ryan Mallon <rmallon@gmail.com> Cc: Tejun Heo <tj@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-04 05:47:41 +08:00
void *item;
VM_BUG_ON(!expected);
mm: shmem: save one radix tree lookup when truncating swapped pages Page cache radix tree slots are usually stabilized by the page lock, but shmem's swap cookies have no such thing. Because the overall truncation loop is lockless, the swap entry is currently confirmed by a tree lookup and then deleted by another tree lookup under the same tree lock region. Use radix_tree_delete_item() instead, which does the verification and deletion with only one lookup. This also allows removing the delete-only special case from shmem_radix_tree_replace(). Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reviewed-by: Minchan Kim <minchan@kernel.org> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Bob Liu <bob.liu@oracle.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Jan Kara <jack@suse.cz> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Luigi Semenzato <semenzato@google.com> Cc: Metin Doslu <metin@citusdata.com> Cc: Michel Lespinasse <walken@google.com> Cc: Ozgun Erdogan <ozgun@citusdata.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Roman Gushchin <klamm@yandex-team.ru> Cc: Ryan Mallon <rmallon@gmail.com> Cc: Tejun Heo <tj@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-04 05:47:41 +08:00
VM_BUG_ON(!replacement);
pslot = radix_tree_lookup_slot(&mapping->page_tree, index);
mm: shmem: save one radix tree lookup when truncating swapped pages Page cache radix tree slots are usually stabilized by the page lock, but shmem's swap cookies have no such thing. Because the overall truncation loop is lockless, the swap entry is currently confirmed by a tree lookup and then deleted by another tree lookup under the same tree lock region. Use radix_tree_delete_item() instead, which does the verification and deletion with only one lookup. This also allows removing the delete-only special case from shmem_radix_tree_replace(). Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reviewed-by: Minchan Kim <minchan@kernel.org> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Bob Liu <bob.liu@oracle.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Jan Kara <jack@suse.cz> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Luigi Semenzato <semenzato@google.com> Cc: Metin Doslu <metin@citusdata.com> Cc: Michel Lespinasse <walken@google.com> Cc: Ozgun Erdogan <ozgun@citusdata.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Roman Gushchin <klamm@yandex-team.ru> Cc: Ryan Mallon <rmallon@gmail.com> Cc: Tejun Heo <tj@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-04 05:47:41 +08:00
if (!pslot)
return -ENOENT;
item = radix_tree_deref_slot_protected(pslot, &mapping->tree_lock);
if (item != expected)
return -ENOENT;
mm: shmem: save one radix tree lookup when truncating swapped pages Page cache radix tree slots are usually stabilized by the page lock, but shmem's swap cookies have no such thing. Because the overall truncation loop is lockless, the swap entry is currently confirmed by a tree lookup and then deleted by another tree lookup under the same tree lock region. Use radix_tree_delete_item() instead, which does the verification and deletion with only one lookup. This also allows removing the delete-only special case from shmem_radix_tree_replace(). Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reviewed-by: Minchan Kim <minchan@kernel.org> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Bob Liu <bob.liu@oracle.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Jan Kara <jack@suse.cz> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Luigi Semenzato <semenzato@google.com> Cc: Metin Doslu <metin@citusdata.com> Cc: Michel Lespinasse <walken@google.com> Cc: Ozgun Erdogan <ozgun@citusdata.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Roman Gushchin <klamm@yandex-team.ru> Cc: Ryan Mallon <rmallon@gmail.com> Cc: Tejun Heo <tj@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-04 05:47:41 +08:00
radix_tree_replace_slot(pslot, replacement);
return 0;
}
shmem: fix negative rss in memcg memory.stat When adding the page_private checks before calling shmem_replace_page(), I did realize that there is a further race, but thought it too unlikely to need a hurried fix. But independently I've been chasing why a mem cgroup's memory.stat sometimes shows negative rss after all tasks have gone: I expected it to be a stats gathering bug, but actually it's shmem swapping's fault. It's an old surprise, that when you lock_page(lookup_swap_cache(swap)), the page may have been removed from swapcache before getting the lock; or it may have been freed and reused and be back in swapcache; and it can even be using the same swap location as before (page_private same). The swapoff case is already secure against this (swap cannot be reused until the whole area has been swapped off, and a new swapped on); and shmem_getpage_gfp() is protected by shmem_add_to_page_cache()'s check for the expected radix_tree entry - but a little too late. By that time, we might have already decided to shmem_replace_page(): I don't know of a problem from that, but I'd feel more at ease not to do so spuriously. And we have already done mem_cgroup_cache_charge(), on perhaps the wrong mem cgroup: and this charge is not then undone on the error path, because PageSwapCache ends up preventing that. It's this last case which causes the occasional negative rss in memory.stat: the page is charged here as cache, but (sometimes) found to be anon when eventually it's uncharged - and in between, it's an undeserved charge on the wrong memcg. Fix this by adding an earlier check on the radix_tree entry: it's inelegant to descend the tree twice, but swapping is not the fast path, and a better solution would need a pair (try+commit) of memcg calls, and a rework of shmem_replace_page() to keep out of the swapcache. We can use the added shmem_confirm_swap() function to replace the find_get_page+page_cache_release we were already doing on the error path. And add a comment on that -EEXIST: it seems a peculiar errno to be using, but originates from its use in radix_tree_insert(). [It can be surprising to see positive rss left in a memcg's memory.stat after all tasks have gone, since it is supposed to count anonymous but not shmem. Aside from sharing anon pages via fork with a task in some other memcg, it often happens after swapping: because a swap page can't be freed while under writeback, nor while locked. So it's not an error, and these residual pages are easily freed once pressure demands.] Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Michal Hocko <mhocko@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-07-12 05:02:47 +08:00
/*
* Sometimes, before we decide whether to proceed or to fail, we must check
* that an entry was not already brought back from swap by a racing thread.
*
* Checking page is not enough: by the time a SwapCache page is locked, it
* might be reused, and again be SwapCache, using the same swap as before.
*/
static bool shmem_confirm_swap(struct address_space *mapping,
pgoff_t index, swp_entry_t swap)
{
void *item;
rcu_read_lock();
item = radix_tree_lookup(&mapping->page_tree, index);
rcu_read_unlock();
return item == swp_to_radix_entry(swap);
}
/*
* Like add_to_page_cache_locked, but error if expected item has gone.
*/
static int shmem_add_to_page_cache(struct page *page,
struct address_space *mapping,
pgoff_t index, void *expected)
{
int error;
VM_BUG_ON_PAGE(!PageLocked(page), page);
VM_BUG_ON_PAGE(!PageSwapBacked(page), page);
page_cache_get(page);
page->mapping = mapping;
page->index = index;
spin_lock_irq(&mapping->tree_lock);
if (!expected)
error = radix_tree_insert(&mapping->page_tree, index, page);
else
error = shmem_radix_tree_replace(mapping, index, expected,
page);
if (!error) {
mapping->nrpages++;
__inc_zone_page_state(page, NR_FILE_PAGES);
__inc_zone_page_state(page, NR_SHMEM);
spin_unlock_irq(&mapping->tree_lock);
} else {
page->mapping = NULL;
spin_unlock_irq(&mapping->tree_lock);
page_cache_release(page);
}
return error;
}
/*
* Like delete_from_page_cache, but substitutes swap for page.
*/
static void shmem_delete_from_page_cache(struct page *page, void *radswap)
{
struct address_space *mapping = page->mapping;
int error;
spin_lock_irq(&mapping->tree_lock);
error = shmem_radix_tree_replace(mapping, page->index, page, radswap);
page->mapping = NULL;
mapping->nrpages--;
__dec_zone_page_state(page, NR_FILE_PAGES);
__dec_zone_page_state(page, NR_SHMEM);
spin_unlock_irq(&mapping->tree_lock);
page_cache_release(page);
BUG_ON(error);
}
/*
* Remove swap entry from radix tree, free the swap and its page cache.
*/
static int shmem_free_swap(struct address_space *mapping,
pgoff_t index, void *radswap)
{
mm: shmem: save one radix tree lookup when truncating swapped pages Page cache radix tree slots are usually stabilized by the page lock, but shmem's swap cookies have no such thing. Because the overall truncation loop is lockless, the swap entry is currently confirmed by a tree lookup and then deleted by another tree lookup under the same tree lock region. Use radix_tree_delete_item() instead, which does the verification and deletion with only one lookup. This also allows removing the delete-only special case from shmem_radix_tree_replace(). Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reviewed-by: Minchan Kim <minchan@kernel.org> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Bob Liu <bob.liu@oracle.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Jan Kara <jack@suse.cz> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Luigi Semenzato <semenzato@google.com> Cc: Metin Doslu <metin@citusdata.com> Cc: Michel Lespinasse <walken@google.com> Cc: Ozgun Erdogan <ozgun@citusdata.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Roman Gushchin <klamm@yandex-team.ru> Cc: Ryan Mallon <rmallon@gmail.com> Cc: Tejun Heo <tj@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-04 05:47:41 +08:00
void *old;
spin_lock_irq(&mapping->tree_lock);
mm: shmem: save one radix tree lookup when truncating swapped pages Page cache radix tree slots are usually stabilized by the page lock, but shmem's swap cookies have no such thing. Because the overall truncation loop is lockless, the swap entry is currently confirmed by a tree lookup and then deleted by another tree lookup under the same tree lock region. Use radix_tree_delete_item() instead, which does the verification and deletion with only one lookup. This also allows removing the delete-only special case from shmem_radix_tree_replace(). Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reviewed-by: Minchan Kim <minchan@kernel.org> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Bob Liu <bob.liu@oracle.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Jan Kara <jack@suse.cz> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Luigi Semenzato <semenzato@google.com> Cc: Metin Doslu <metin@citusdata.com> Cc: Michel Lespinasse <walken@google.com> Cc: Ozgun Erdogan <ozgun@citusdata.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Roman Gushchin <klamm@yandex-team.ru> Cc: Ryan Mallon <rmallon@gmail.com> Cc: Tejun Heo <tj@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-04 05:47:41 +08:00
old = radix_tree_delete_item(&mapping->page_tree, index, radswap);
spin_unlock_irq(&mapping->tree_lock);
mm: shmem: save one radix tree lookup when truncating swapped pages Page cache radix tree slots are usually stabilized by the page lock, but shmem's swap cookies have no such thing. Because the overall truncation loop is lockless, the swap entry is currently confirmed by a tree lookup and then deleted by another tree lookup under the same tree lock region. Use radix_tree_delete_item() instead, which does the verification and deletion with only one lookup. This also allows removing the delete-only special case from shmem_radix_tree_replace(). Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Reviewed-by: Minchan Kim <minchan@kernel.org> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Mel Gorman <mgorman@suse.de> Cc: Andrea Arcangeli <aarcange@redhat.com> Cc: Bob Liu <bob.liu@oracle.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Dave Chinner <david@fromorbit.com> Cc: Greg Thelen <gthelen@google.com> Cc: Hugh Dickins <hughd@google.com> Cc: Jan Kara <jack@suse.cz> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Luigi Semenzato <semenzato@google.com> Cc: Metin Doslu <metin@citusdata.com> Cc: Michel Lespinasse <walken@google.com> Cc: Ozgun Erdogan <ozgun@citusdata.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Roman Gushchin <klamm@yandex-team.ru> Cc: Ryan Mallon <rmallon@gmail.com> Cc: Tejun Heo <tj@kernel.org> Cc: Vlastimil Babka <vbabka@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-04-04 05:47:41 +08:00
if (old != radswap)
return -ENOENT;
free_swap_and_cache(radix_to_swp_entry(radswap));
return 0;
}
SHM_UNLOCK: fix Unevictable pages stranded after swap Commit cc39c6a9bbde ("mm: account skipped entries to avoid looping in find_get_pages") correctly fixed an infinite loop; but left a problem that find_get_pages() on shmem would return 0 (appearing to callers to mean end of tree) when it meets a run of nr_pages swap entries. The only uses of find_get_pages() on shmem are via pagevec_lookup(), called from invalidate_mapping_pages(), and from shmctl SHM_UNLOCK's scan_mapping_unevictable_pages(). The first is already commented, and not worth worrying about; but the second can leave pages on the Unevictable list after an unusual sequence of swapping and locking. Fix that by using shmem_find_get_pages_and_swap() (then ignoring the swap) instead of pagevec_lookup(). But I don't want to contaminate vmscan.c with shmem internals, nor shmem.c with LRU locking. So move scan_mapping_unevictable_pages() into shmem.c, renaming it shmem_unlock_mapping(); and rename check_move_unevictable_page() to check_move_unevictable_pages(), looping down an array of pages, oftentimes under the same lock. Leave out the "rotate unevictable list" block: that's a leftover from when this was used for /proc/sys/vm/scan_unevictable_pages, whose flawed handling involved looking at pages at tail of LRU. Was there significance to the sequence first ClearPageUnevictable, then test page_evictable, then SetPageUnevictable here? I think not, we're under LRU lock, and have no barriers between those. Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Rik van Riel <riel@redhat.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michel Lespinasse <walken@google.com> Cc: <stable@vger.kernel.org> [back to 3.1 but will need respins] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-01-21 06:34:21 +08:00
/*
* SysV IPC SHM_UNLOCK restore Unevictable pages to their evictable lists.
*/
void shmem_unlock_mapping(struct address_space *mapping)
{
struct pagevec pvec;
pgoff_t indices[PAGEVEC_SIZE];
pgoff_t index = 0;
pagevec_init(&pvec, 0);
/*
* Minor point, but we might as well stop if someone else SHM_LOCKs it.
*/
while (!mapping_unevictable(mapping)) {
/*
* Avoid pagevec_lookup(): find_get_pages() returns 0 as if it
* has finished, if it hits a row of PAGEVEC_SIZE swap entries.
*/
2014-04-04 05:47:46 +08:00
pvec.nr = find_get_entries(mapping, index,
PAGEVEC_SIZE, pvec.pages, indices);
SHM_UNLOCK: fix Unevictable pages stranded after swap Commit cc39c6a9bbde ("mm: account skipped entries to avoid looping in find_get_pages") correctly fixed an infinite loop; but left a problem that find_get_pages() on shmem would return 0 (appearing to callers to mean end of tree) when it meets a run of nr_pages swap entries. The only uses of find_get_pages() on shmem are via pagevec_lookup(), called from invalidate_mapping_pages(), and from shmctl SHM_UNLOCK's scan_mapping_unevictable_pages(). The first is already commented, and not worth worrying about; but the second can leave pages on the Unevictable list after an unusual sequence of swapping and locking. Fix that by using shmem_find_get_pages_and_swap() (then ignoring the swap) instead of pagevec_lookup(). But I don't want to contaminate vmscan.c with shmem internals, nor shmem.c with LRU locking. So move scan_mapping_unevictable_pages() into shmem.c, renaming it shmem_unlock_mapping(); and rename check_move_unevictable_page() to check_move_unevictable_pages(), looping down an array of pages, oftentimes under the same lock. Leave out the "rotate unevictable list" block: that's a leftover from when this was used for /proc/sys/vm/scan_unevictable_pages, whose flawed handling involved looking at pages at tail of LRU. Was there significance to the sequence first ClearPageUnevictable, then test page_evictable, then SetPageUnevictable here? I think not, we're under LRU lock, and have no barriers between those. Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Rik van Riel <riel@redhat.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michel Lespinasse <walken@google.com> Cc: <stable@vger.kernel.org> [back to 3.1 but will need respins] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-01-21 06:34:21 +08:00
if (!pvec.nr)
break;
index = indices[pvec.nr - 1] + 1;
2014-04-04 05:47:46 +08:00
pagevec_remove_exceptionals(&pvec);
SHM_UNLOCK: fix Unevictable pages stranded after swap Commit cc39c6a9bbde ("mm: account skipped entries to avoid looping in find_get_pages") correctly fixed an infinite loop; but left a problem that find_get_pages() on shmem would return 0 (appearing to callers to mean end of tree) when it meets a run of nr_pages swap entries. The only uses of find_get_pages() on shmem are via pagevec_lookup(), called from invalidate_mapping_pages(), and from shmctl SHM_UNLOCK's scan_mapping_unevictable_pages(). The first is already commented, and not worth worrying about; but the second can leave pages on the Unevictable list after an unusual sequence of swapping and locking. Fix that by using shmem_find_get_pages_and_swap() (then ignoring the swap) instead of pagevec_lookup(). But I don't want to contaminate vmscan.c with shmem internals, nor shmem.c with LRU locking. So move scan_mapping_unevictable_pages() into shmem.c, renaming it shmem_unlock_mapping(); and rename check_move_unevictable_page() to check_move_unevictable_pages(), looping down an array of pages, oftentimes under the same lock. Leave out the "rotate unevictable list" block: that's a leftover from when this was used for /proc/sys/vm/scan_unevictable_pages, whose flawed handling involved looking at pages at tail of LRU. Was there significance to the sequence first ClearPageUnevictable, then test page_evictable, then SetPageUnevictable here? I think not, we're under LRU lock, and have no barriers between those. Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Rik van Riel <riel@redhat.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michel Lespinasse <walken@google.com> Cc: <stable@vger.kernel.org> [back to 3.1 but will need respins] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-01-21 06:34:21 +08:00
check_move_unevictable_pages(pvec.pages, pvec.nr);
pagevec_release(&pvec);
cond_resched();
}
}
/*
* Remove range of pages and swap entries from radix tree, and free them.
* If !unfalloc, truncate or punch hole; if unfalloc, undo failed fallocate.
*/
static void shmem_undo_range(struct inode *inode, loff_t lstart, loff_t lend,
bool unfalloc)
{
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
struct address_space *mapping = inode->i_mapping;
struct shmem_inode_info *info = SHMEM_I(inode);
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
pgoff_t start = (lstart + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
pgoff_t end = (lend + 1) >> PAGE_CACHE_SHIFT;
unsigned int partial_start = lstart & (PAGE_CACHE_SIZE - 1);
unsigned int partial_end = (lend + 1) & (PAGE_CACHE_SIZE - 1);
struct pagevec pvec;
pgoff_t indices[PAGEVEC_SIZE];
long nr_swaps_freed = 0;
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
pgoff_t index;
int i;
if (lend == -1)
end = -1; /* unsigned, so actually very big */
pagevec_init(&pvec, 0);
index = start;
while (index < end) {
2014-04-04 05:47:46 +08:00
pvec.nr = find_get_entries(mapping, index,
min(end - index, (pgoff_t)PAGEVEC_SIZE),
pvec.pages, indices);
if (!pvec.nr)
break;
for (i = 0; i < pagevec_count(&pvec); i++) {
struct page *page = pvec.pages[i];
index = indices[i];
if (index >= end)
break;
if (radix_tree_exceptional_entry(page)) {
if (unfalloc)
continue;
nr_swaps_freed += !shmem_free_swap(mapping,
index, page);
continue;
}
if (!trylock_page(page))
continue;
if (!unfalloc || !PageUptodate(page)) {
if (page->mapping == mapping) {
VM_BUG_ON_PAGE(PageWriteback(page), page);
truncate_inode_page(mapping, page);
}
}
unlock_page(page);
}
2014-04-04 05:47:46 +08:00
pagevec_remove_exceptionals(&pvec);
SHM_UNLOCK: fix Unevictable pages stranded after swap Commit cc39c6a9bbde ("mm: account skipped entries to avoid looping in find_get_pages") correctly fixed an infinite loop; but left a problem that find_get_pages() on shmem would return 0 (appearing to callers to mean end of tree) when it meets a run of nr_pages swap entries. The only uses of find_get_pages() on shmem are via pagevec_lookup(), called from invalidate_mapping_pages(), and from shmctl SHM_UNLOCK's scan_mapping_unevictable_pages(). The first is already commented, and not worth worrying about; but the second can leave pages on the Unevictable list after an unusual sequence of swapping and locking. Fix that by using shmem_find_get_pages_and_swap() (then ignoring the swap) instead of pagevec_lookup(). But I don't want to contaminate vmscan.c with shmem internals, nor shmem.c with LRU locking. So move scan_mapping_unevictable_pages() into shmem.c, renaming it shmem_unlock_mapping(); and rename check_move_unevictable_page() to check_move_unevictable_pages(), looping down an array of pages, oftentimes under the same lock. Leave out the "rotate unevictable list" block: that's a leftover from when this was used for /proc/sys/vm/scan_unevictable_pages, whose flawed handling involved looking at pages at tail of LRU. Was there significance to the sequence first ClearPageUnevictable, then test page_evictable, then SetPageUnevictable here? I think not, we're under LRU lock, and have no barriers between those. Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Rik van Riel <riel@redhat.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michel Lespinasse <walken@google.com> Cc: <stable@vger.kernel.org> [back to 3.1 but will need respins] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-01-21 06:34:21 +08:00
pagevec_release(&pvec);
cond_resched();
index++;
}
if (partial_start) {
struct page *page = NULL;
shmem_getpage(inode, start - 1, &page, SGP_READ, NULL);
if (page) {
unsigned int top = PAGE_CACHE_SIZE;
if (start > end) {
top = partial_end;
partial_end = 0;
}
zero_user_segment(page, partial_start, top);
set_page_dirty(page);
unlock_page(page);
page_cache_release(page);
}
}
if (partial_end) {
struct page *page = NULL;
shmem_getpage(inode, end, &page, SGP_READ, NULL);
if (page) {
zero_user_segment(page, 0, partial_end);
set_page_dirty(page);
unlock_page(page);
page_cache_release(page);
}
}
if (start >= end)
return;
index = start;
shmem: fix splicing from a hole while it's punched shmem_fault() is the actual culprit in trinity's hole-punch starvation, and the most significant cause of such problems: since a page faulted is one that then appears page_mapped(), needing unmap_mapping_range() and i_mmap_mutex to be unmapped again. But it is not the only way in which a page can be brought into a hole in the radix_tree while that hole is being punched; and Vlastimil's testing implies that if enough other processors are busy filling in the hole, then shmem_undo_range() can be kept from completing indefinitely. shmem_file_splice_read() is the main other user of SGP_CACHE, which can instantiate shmem pagecache pages in the read-only case (without holding i_mutex, so perhaps concurrently with a hole-punch). Probably it's silly not to use SGP_READ already (using the ZERO_PAGE for holes): which ought to be safe, but might bring surprises - not a change to be rushed. shmem_read_mapping_page_gfp() is an internal interface used by drivers/gpu/drm GEM (and next by uprobes): it should be okay. And shmem_file_read_iter() uses the SGP_DIRTY variant of SGP_CACHE, when called internally by the kernel (perhaps for a stacking filesystem, which might rely on holes to be reserved): it's unclear whether it could be provoked to keep hole-punch busy or not. We could apply the same umbrella as now used in shmem_fault() to shmem_file_splice_read() and the others; but it looks ugly, and use over a range raises questions - should it actually be per page? can these get starved themselves? The origin of this part of the problem is my v3.1 commit d0823576bf4b ("mm: pincer in truncate_inode_pages_range"), once it was duplicated into shmem.c. It seemed like a nice idea at the time, to ensure (barring RCU lookup fuzziness) that there's an instant when the entire hole is empty; but the indefinitely repeated scans to ensure that make it vulnerable. Revert that "enhancement" to hole-punch from shmem_undo_range(), but retain the unproblematic rescanning when it's truncating; add a couple of comments there. Remove the "indices[0] >= end" test: that is now handled satisfactorily by the inner loop, and mem_cgroup_uncharge_start()/end() are too light to be worth avoiding here. But if we do not always loop indefinitely, we do need to handle the case of swap swizzled back to page before shmem_free_swap() gets it: add a retry for that case, as suggested by Konstantin Khlebnikov; and for the case of page swizzled back to swap, as suggested by Johannes Weiner. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Suggested-by: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:13 +08:00
while (index < end) {
cond_resched();
2014-04-04 05:47:46 +08:00
pvec.nr = find_get_entries(mapping, index,
min(end - index, (pgoff_t)PAGEVEC_SIZE),
2014-04-04 05:47:46 +08:00
pvec.pages, indices);
if (!pvec.nr) {
shmem: fix splicing from a hole while it's punched shmem_fault() is the actual culprit in trinity's hole-punch starvation, and the most significant cause of such problems: since a page faulted is one that then appears page_mapped(), needing unmap_mapping_range() and i_mmap_mutex to be unmapped again. But it is not the only way in which a page can be brought into a hole in the radix_tree while that hole is being punched; and Vlastimil's testing implies that if enough other processors are busy filling in the hole, then shmem_undo_range() can be kept from completing indefinitely. shmem_file_splice_read() is the main other user of SGP_CACHE, which can instantiate shmem pagecache pages in the read-only case (without holding i_mutex, so perhaps concurrently with a hole-punch). Probably it's silly not to use SGP_READ already (using the ZERO_PAGE for holes): which ought to be safe, but might bring surprises - not a change to be rushed. shmem_read_mapping_page_gfp() is an internal interface used by drivers/gpu/drm GEM (and next by uprobes): it should be okay. And shmem_file_read_iter() uses the SGP_DIRTY variant of SGP_CACHE, when called internally by the kernel (perhaps for a stacking filesystem, which might rely on holes to be reserved): it's unclear whether it could be provoked to keep hole-punch busy or not. We could apply the same umbrella as now used in shmem_fault() to shmem_file_splice_read() and the others; but it looks ugly, and use over a range raises questions - should it actually be per page? can these get starved themselves? The origin of this part of the problem is my v3.1 commit d0823576bf4b ("mm: pincer in truncate_inode_pages_range"), once it was duplicated into shmem.c. It seemed like a nice idea at the time, to ensure (barring RCU lookup fuzziness) that there's an instant when the entire hole is empty; but the indefinitely repeated scans to ensure that make it vulnerable. Revert that "enhancement" to hole-punch from shmem_undo_range(), but retain the unproblematic rescanning when it's truncating; add a couple of comments there. Remove the "indices[0] >= end" test: that is now handled satisfactorily by the inner loop, and mem_cgroup_uncharge_start()/end() are too light to be worth avoiding here. But if we do not always loop indefinitely, we do need to handle the case of swap swizzled back to page before shmem_free_swap() gets it: add a retry for that case, as suggested by Konstantin Khlebnikov; and for the case of page swizzled back to swap, as suggested by Johannes Weiner. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Suggested-by: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:13 +08:00
/* If all gone or hole-punch or unfalloc, we're done */
if (index == start || end != -1)
break;
shmem: fix splicing from a hole while it's punched shmem_fault() is the actual culprit in trinity's hole-punch starvation, and the most significant cause of such problems: since a page faulted is one that then appears page_mapped(), needing unmap_mapping_range() and i_mmap_mutex to be unmapped again. But it is not the only way in which a page can be brought into a hole in the radix_tree while that hole is being punched; and Vlastimil's testing implies that if enough other processors are busy filling in the hole, then shmem_undo_range() can be kept from completing indefinitely. shmem_file_splice_read() is the main other user of SGP_CACHE, which can instantiate shmem pagecache pages in the read-only case (without holding i_mutex, so perhaps concurrently with a hole-punch). Probably it's silly not to use SGP_READ already (using the ZERO_PAGE for holes): which ought to be safe, but might bring surprises - not a change to be rushed. shmem_read_mapping_page_gfp() is an internal interface used by drivers/gpu/drm GEM (and next by uprobes): it should be okay. And shmem_file_read_iter() uses the SGP_DIRTY variant of SGP_CACHE, when called internally by the kernel (perhaps for a stacking filesystem, which might rely on holes to be reserved): it's unclear whether it could be provoked to keep hole-punch busy or not. We could apply the same umbrella as now used in shmem_fault() to shmem_file_splice_read() and the others; but it looks ugly, and use over a range raises questions - should it actually be per page? can these get starved themselves? The origin of this part of the problem is my v3.1 commit d0823576bf4b ("mm: pincer in truncate_inode_pages_range"), once it was duplicated into shmem.c. It seemed like a nice idea at the time, to ensure (barring RCU lookup fuzziness) that there's an instant when the entire hole is empty; but the indefinitely repeated scans to ensure that make it vulnerable. Revert that "enhancement" to hole-punch from shmem_undo_range(), but retain the unproblematic rescanning when it's truncating; add a couple of comments there. Remove the "indices[0] >= end" test: that is now handled satisfactorily by the inner loop, and mem_cgroup_uncharge_start()/end() are too light to be worth avoiding here. But if we do not always loop indefinitely, we do need to handle the case of swap swizzled back to page before shmem_free_swap() gets it: add a retry for that case, as suggested by Konstantin Khlebnikov; and for the case of page swizzled back to swap, as suggested by Johannes Weiner. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Suggested-by: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:13 +08:00
/* But if truncating, restart to make sure all gone */
index = start;
continue;
}
for (i = 0; i < pagevec_count(&pvec); i++) {
struct page *page = pvec.pages[i];
index = indices[i];
if (index >= end)
break;
if (radix_tree_exceptional_entry(page)) {
if (unfalloc)
continue;
shmem: fix splicing from a hole while it's punched shmem_fault() is the actual culprit in trinity's hole-punch starvation, and the most significant cause of such problems: since a page faulted is one that then appears page_mapped(), needing unmap_mapping_range() and i_mmap_mutex to be unmapped again. But it is not the only way in which a page can be brought into a hole in the radix_tree while that hole is being punched; and Vlastimil's testing implies that if enough other processors are busy filling in the hole, then shmem_undo_range() can be kept from completing indefinitely. shmem_file_splice_read() is the main other user of SGP_CACHE, which can instantiate shmem pagecache pages in the read-only case (without holding i_mutex, so perhaps concurrently with a hole-punch). Probably it's silly not to use SGP_READ already (using the ZERO_PAGE for holes): which ought to be safe, but might bring surprises - not a change to be rushed. shmem_read_mapping_page_gfp() is an internal interface used by drivers/gpu/drm GEM (and next by uprobes): it should be okay. And shmem_file_read_iter() uses the SGP_DIRTY variant of SGP_CACHE, when called internally by the kernel (perhaps for a stacking filesystem, which might rely on holes to be reserved): it's unclear whether it could be provoked to keep hole-punch busy or not. We could apply the same umbrella as now used in shmem_fault() to shmem_file_splice_read() and the others; but it looks ugly, and use over a range raises questions - should it actually be per page? can these get starved themselves? The origin of this part of the problem is my v3.1 commit d0823576bf4b ("mm: pincer in truncate_inode_pages_range"), once it was duplicated into shmem.c. It seemed like a nice idea at the time, to ensure (barring RCU lookup fuzziness) that there's an instant when the entire hole is empty; but the indefinitely repeated scans to ensure that make it vulnerable. Revert that "enhancement" to hole-punch from shmem_undo_range(), but retain the unproblematic rescanning when it's truncating; add a couple of comments there. Remove the "indices[0] >= end" test: that is now handled satisfactorily by the inner loop, and mem_cgroup_uncharge_start()/end() are too light to be worth avoiding here. But if we do not always loop indefinitely, we do need to handle the case of swap swizzled back to page before shmem_free_swap() gets it: add a retry for that case, as suggested by Konstantin Khlebnikov; and for the case of page swizzled back to swap, as suggested by Johannes Weiner. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Suggested-by: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:13 +08:00
if (shmem_free_swap(mapping, index, page)) {
/* Swap was replaced by page: retry */
index--;
break;
}
nr_swaps_freed++;
continue;
}
lock_page(page);
if (!unfalloc || !PageUptodate(page)) {
if (page->mapping == mapping) {
VM_BUG_ON_PAGE(PageWriteback(page), page);
truncate_inode_page(mapping, page);
shmem: fix splicing from a hole while it's punched shmem_fault() is the actual culprit in trinity's hole-punch starvation, and the most significant cause of such problems: since a page faulted is one that then appears page_mapped(), needing unmap_mapping_range() and i_mmap_mutex to be unmapped again. But it is not the only way in which a page can be brought into a hole in the radix_tree while that hole is being punched; and Vlastimil's testing implies that if enough other processors are busy filling in the hole, then shmem_undo_range() can be kept from completing indefinitely. shmem_file_splice_read() is the main other user of SGP_CACHE, which can instantiate shmem pagecache pages in the read-only case (without holding i_mutex, so perhaps concurrently with a hole-punch). Probably it's silly not to use SGP_READ already (using the ZERO_PAGE for holes): which ought to be safe, but might bring surprises - not a change to be rushed. shmem_read_mapping_page_gfp() is an internal interface used by drivers/gpu/drm GEM (and next by uprobes): it should be okay. And shmem_file_read_iter() uses the SGP_DIRTY variant of SGP_CACHE, when called internally by the kernel (perhaps for a stacking filesystem, which might rely on holes to be reserved): it's unclear whether it could be provoked to keep hole-punch busy or not. We could apply the same umbrella as now used in shmem_fault() to shmem_file_splice_read() and the others; but it looks ugly, and use over a range raises questions - should it actually be per page? can these get starved themselves? The origin of this part of the problem is my v3.1 commit d0823576bf4b ("mm: pincer in truncate_inode_pages_range"), once it was duplicated into shmem.c. It seemed like a nice idea at the time, to ensure (barring RCU lookup fuzziness) that there's an instant when the entire hole is empty; but the indefinitely repeated scans to ensure that make it vulnerable. Revert that "enhancement" to hole-punch from shmem_undo_range(), but retain the unproblematic rescanning when it's truncating; add a couple of comments there. Remove the "indices[0] >= end" test: that is now handled satisfactorily by the inner loop, and mem_cgroup_uncharge_start()/end() are too light to be worth avoiding here. But if we do not always loop indefinitely, we do need to handle the case of swap swizzled back to page before shmem_free_swap() gets it: add a retry for that case, as suggested by Konstantin Khlebnikov; and for the case of page swizzled back to swap, as suggested by Johannes Weiner. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Suggested-by: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:13 +08:00
} else {
/* Page was replaced by swap: retry */
unlock_page(page);
index--;
break;
}
}
unlock_page(page);
}
2014-04-04 05:47:46 +08:00
pagevec_remove_exceptionals(&pvec);
SHM_UNLOCK: fix Unevictable pages stranded after swap Commit cc39c6a9bbde ("mm: account skipped entries to avoid looping in find_get_pages") correctly fixed an infinite loop; but left a problem that find_get_pages() on shmem would return 0 (appearing to callers to mean end of tree) when it meets a run of nr_pages swap entries. The only uses of find_get_pages() on shmem are via pagevec_lookup(), called from invalidate_mapping_pages(), and from shmctl SHM_UNLOCK's scan_mapping_unevictable_pages(). The first is already commented, and not worth worrying about; but the second can leave pages on the Unevictable list after an unusual sequence of swapping and locking. Fix that by using shmem_find_get_pages_and_swap() (then ignoring the swap) instead of pagevec_lookup(). But I don't want to contaminate vmscan.c with shmem internals, nor shmem.c with LRU locking. So move scan_mapping_unevictable_pages() into shmem.c, renaming it shmem_unlock_mapping(); and rename check_move_unevictable_page() to check_move_unevictable_pages(), looping down an array of pages, oftentimes under the same lock. Leave out the "rotate unevictable list" block: that's a leftover from when this was used for /proc/sys/vm/scan_unevictable_pages, whose flawed handling involved looking at pages at tail of LRU. Was there significance to the sequence first ClearPageUnevictable, then test page_evictable, then SetPageUnevictable here? I think not, we're under LRU lock, and have no barriers between those. Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Rik van Riel <riel@redhat.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michel Lespinasse <walken@google.com> Cc: <stable@vger.kernel.org> [back to 3.1 but will need respins] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-01-21 06:34:21 +08:00
pagevec_release(&pvec);
index++;
}
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
spin_lock(&info->lock);
info->swapped -= nr_swaps_freed;
shmem_recalc_inode(inode);
spin_unlock(&info->lock);
}
void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend)
{
shmem_undo_range(inode, lstart, lend, false);
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
inode->i_ctime = inode->i_mtime = CURRENT_TIME;
}
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
EXPORT_SYMBOL_GPL(shmem_truncate_range);
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
static int shmem_setattr(struct dentry *dentry, struct iattr *attr)
{
struct inode *inode = d_inode(dentry);
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
struct shmem_inode_info *info = SHMEM_I(inode);
int error;
error = inode_change_ok(inode, attr);
if (error)
return error;
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
if (S_ISREG(inode->i_mode) && (attr->ia_valid & ATTR_SIZE)) {
loff_t oldsize = inode->i_size;
loff_t newsize = attr->ia_size;
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
/* protected by i_mutex */
if ((newsize < oldsize && (info->seals & F_SEAL_SHRINK)) ||
(newsize > oldsize && (info->seals & F_SEAL_GROW)))
return -EPERM;
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
if (newsize != oldsize) {
error = shmem_reacct_size(SHMEM_I(inode)->flags,
oldsize, newsize);
if (error)
return error;
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
i_size_write(inode, newsize);
inode->i_ctime = inode->i_mtime = CURRENT_TIME;
}
if (newsize <= oldsize) {
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
loff_t holebegin = round_up(newsize, PAGE_SIZE);
unmap_mapping_range(inode->i_mapping, holebegin, 0, 1);
shmem_truncate_range(inode, newsize, (loff_t)-1);
/* unmap again to remove racily COWed private pages */
unmap_mapping_range(inode->i_mapping, holebegin, 0, 1);
}
}
setattr_copy(inode, attr);
if (attr->ia_valid & ATTR_MODE)
error = posix_acl_chmod(inode, inode->i_mode);
return error;
}
static void shmem_evict_inode(struct inode *inode)
{
struct shmem_inode_info *info = SHMEM_I(inode);
if (inode->i_mapping->a_ops == &shmem_aops) {
shmem_unacct_size(info->flags, inode->i_size);
inode->i_size = 0;
shmem_truncate_range(inode, 0, (loff_t)-1);
if (!list_empty(&info->swaplist)) {
mutex_lock(&shmem_swaplist_mutex);
list_del_init(&info->swaplist);
mutex_unlock(&shmem_swaplist_mutex);
}
} else
kfree(info->symlink);
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
simple_xattrs_free(&info->xattrs);
WARN_ON(inode->i_blocks);
shmem_free_inode(inode->i_sb);
clear_inode(inode);
}
/*
* If swap found in inode, free it and move page from swapcache to filecache.
*/
static int shmem_unuse_inode(struct shmem_inode_info *info,
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
swp_entry_t swap, struct page **pagep)
{
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
struct address_space *mapping = info->vfs_inode.i_mapping;
void *radswap;
pgoff_t index;
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
gfp_t gfp;
int error = 0;
radswap = swp_to_radix_entry(swap);
tmpfs radix_tree: locate_item to speed up swapoff We have already acknowledged that swapoff of a tmpfs file is slower than it was before conversion to the generic radix_tree: a little slower there will be acceptable, if the hotter paths are faster. But it was a shock to find swapoff of a 500MB file 20 times slower on my laptop, taking 10 minutes; and at that rate it significantly slows down my testing. Now, most of that turned out to be overhead from PROVE_LOCKING and PROVE_RCU: without those it was only 4 times slower than before; and more realistic tests on other machines don't fare as badly. I've tried a number of things to improve it, including tagging the swap entries, then doing lookup by tag: I'd expected that to halve the time, but in practice it's erratic, and often counter-productive. The only change I've so far found to make a consistent improvement, is to short-circuit the way we go back and forth, gang lookup packing entries into the array supplied, then shmem scanning that array for the target entry. Scanning in place doubles the speed, so it's now only twice as slow as before (or three times slower when the PROVEs are on). So, add radix_tree_locate_item() as an expedient, once-off, single-caller hack to do the lookup directly in place. #ifdef it on CONFIG_SHMEM and CONFIG_SWAP, as much to document its limited applicability as save space in other configurations. And, sadly, #include sched.h for cond_resched(). Signed-off-by: Hugh Dickins <hughd@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:27 +08:00
index = radix_tree_locate_item(&mapping->page_tree, radswap);
if (index == -1)
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
return -EAGAIN; /* tell shmem_unuse we found nothing */
tmpfs: fix shmem_swaplist races Intensive swapoff testing shows shmem_unuse spinning on an entry in shmem_swaplist pointing to itself: how does that come about? Days pass... First guess is this: shmem_delete_inode tests list_empty without taking the global mutex (so the swapping case doesn't slow down the common case); but there's an instant in shmem_unuse_inode's list_move_tail when the list entry may appear empty (a rare case, because it's actually moving the head not the the list member). So there's a danger of leaving the inode on the swaplist when it's freed, then reinitialized to point to itself when reused. Fix that by skipping the list_move_tail when it's a no-op, which happens to plug this. But this same spinning then surfaces on another machine. Ah, I'd never suspected it, but shmem_writepage's swaplist manipulation is unsafe: though we still hold page lock, which would hold off inode deletion if the page were in pagecache, it doesn't hold off once it's in swapcache (free_swap_and_cache doesn't wait on locked pages). Hmm: we could put the the inode on swaplist earlier, but then shmem_unuse_inode could never prune unswapped inodes. Fix this with an igrab before dropping info->lock, as in shmem_unuse_inode; though I am a little uneasy about the iput which has to follow - it works, and I see nothing wrong with it, but it is surprising that shmem inode deletion may now occur below shmem_writepage. Revisit this fix later? And while we're looking at these races: the way shmem_unuse tests swapped without holding info->lock looks unsafe, if we've more than one swap area: a racing shmem_writepage on another page of the same inode could be putting it in swapcache, just as we're deciding to remove the inode from swaplist - there's a danger of going on swap without being listed, so a later swapoff would hang, being unable to locate the entry. Move that test and removal down into shmem_unuse_inode, once info->lock is held. Signed-off-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:55 +08:00
/*
* Move _head_ to start search for next from here.
* But be careful: shmem_evict_inode checks list_empty without taking
tmpfs: fix shmem_swaplist races Intensive swapoff testing shows shmem_unuse spinning on an entry in shmem_swaplist pointing to itself: how does that come about? Days pass... First guess is this: shmem_delete_inode tests list_empty without taking the global mutex (so the swapping case doesn't slow down the common case); but there's an instant in shmem_unuse_inode's list_move_tail when the list entry may appear empty (a rare case, because it's actually moving the head not the the list member). So there's a danger of leaving the inode on the swaplist when it's freed, then reinitialized to point to itself when reused. Fix that by skipping the list_move_tail when it's a no-op, which happens to plug this. But this same spinning then surfaces on another machine. Ah, I'd never suspected it, but shmem_writepage's swaplist manipulation is unsafe: though we still hold page lock, which would hold off inode deletion if the page were in pagecache, it doesn't hold off once it's in swapcache (free_swap_and_cache doesn't wait on locked pages). Hmm: we could put the the inode on swaplist earlier, but then shmem_unuse_inode could never prune unswapped inodes. Fix this with an igrab before dropping info->lock, as in shmem_unuse_inode; though I am a little uneasy about the iput which has to follow - it works, and I see nothing wrong with it, but it is surprising that shmem inode deletion may now occur below shmem_writepage. Revisit this fix later? And while we're looking at these races: the way shmem_unuse tests swapped without holding info->lock looks unsafe, if we've more than one swap area: a racing shmem_writepage on another page of the same inode could be putting it in swapcache, just as we're deciding to remove the inode from swaplist - there's a danger of going on swap without being listed, so a later swapoff would hang, being unable to locate the entry. Move that test and removal down into shmem_unuse_inode, once info->lock is held. Signed-off-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:55 +08:00
* mutex, and there's an instant in list_move_tail when info->swaplist
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
* would appear empty, if it were the only one on shmem_swaplist.
tmpfs: fix shmem_swaplist races Intensive swapoff testing shows shmem_unuse spinning on an entry in shmem_swaplist pointing to itself: how does that come about? Days pass... First guess is this: shmem_delete_inode tests list_empty without taking the global mutex (so the swapping case doesn't slow down the common case); but there's an instant in shmem_unuse_inode's list_move_tail when the list entry may appear empty (a rare case, because it's actually moving the head not the the list member). So there's a danger of leaving the inode on the swaplist when it's freed, then reinitialized to point to itself when reused. Fix that by skipping the list_move_tail when it's a no-op, which happens to plug this. But this same spinning then surfaces on another machine. Ah, I'd never suspected it, but shmem_writepage's swaplist manipulation is unsafe: though we still hold page lock, which would hold off inode deletion if the page were in pagecache, it doesn't hold off once it's in swapcache (free_swap_and_cache doesn't wait on locked pages). Hmm: we could put the the inode on swaplist earlier, but then shmem_unuse_inode could never prune unswapped inodes. Fix this with an igrab before dropping info->lock, as in shmem_unuse_inode; though I am a little uneasy about the iput which has to follow - it works, and I see nothing wrong with it, but it is surprising that shmem inode deletion may now occur below shmem_writepage. Revisit this fix later? And while we're looking at these races: the way shmem_unuse tests swapped without holding info->lock looks unsafe, if we've more than one swap area: a racing shmem_writepage on another page of the same inode could be putting it in swapcache, just as we're deciding to remove the inode from swaplist - there's a danger of going on swap without being listed, so a later swapoff would hang, being unable to locate the entry. Move that test and removal down into shmem_unuse_inode, once info->lock is held. Signed-off-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:55 +08:00
*/
if (shmem_swaplist.next != &info->swaplist)
list_move_tail(&shmem_swaplist, &info->swaplist);
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
gfp = mapping_gfp_mask(mapping);
if (shmem_should_replace_page(*pagep, gfp)) {
mutex_unlock(&shmem_swaplist_mutex);
error = shmem_replace_page(pagep, gfp, info, index);
mutex_lock(&shmem_swaplist_mutex);
/*
* We needed to drop mutex to make that restrictive page
* allocation, but the inode might have been freed while we
* dropped it: although a racing shmem_evict_inode() cannot
* complete without emptying the radix_tree, our page lock
* on this swapcache page is not enough to prevent that -
* free_swap_and_cache() of our swap entry will only
* trylock_page(), removing swap from radix_tree whatever.
*
* We must not proceed to shmem_add_to_page_cache() if the
* inode has been freed, but of course we cannot rely on
* inode or mapping or info to check that. However, we can
* safely check if our swap entry is still in use (and here
* it can't have got reused for another page): if it's still
* in use, then the inode cannot have been freed yet, and we
* can safely proceed (if it's no longer in use, that tells
* nothing about the inode, but we don't need to unuse swap).
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
*/
if (!page_swapcount(*pagep))
error = -ENOENT;
}
memcg: handle swap caches SwapCache support for memory resource controller (memcg) Before mem+swap controller, memcg itself should handle SwapCache in proper way. This is cut-out from it. In current memcg, SwapCache is just leaked and the user can create tons of SwapCache. This is a leak of account and should be handled. SwapCache accounting is done as following. charge (anon) - charged when it's mapped. (because of readahead, charge at add_to_swap_cache() is not sane) uncharge (anon) - uncharged when it's dropped from swapcache and fully unmapped. means it's not uncharged at unmap. Note: delete from swap cache at swap-in is done after rmap information is established. charge (shmem) - charged at swap-in. this prevents charge at add_to_page_cache(). uncharge (shmem) - uncharged when it's dropped from swapcache and not on shmem's radix-tree. at migration, check against 'old page' is modified to handle shmem. Comparing to the old version discussed (and caused troubles), we have advantages of - PCG_USED bit. - simple migrating handling. So, situation is much easier than several months ago, maybe. [hugh@veritas.com: memcg: handle swap caches build fix] Reviewed-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Tested-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Hugh Dickins <hugh@veritas.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Pavel Emelyanov <xemul@openvz.org> Signed-off-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-01-08 10:07:56 +08:00
/*
* We rely on shmem_swaplist_mutex, not only to protect the swaplist,
* but also to hold up shmem_evict_inode(): so inode cannot be freed
* beneath us (pagelock doesn't help until the page is in pagecache).
memcg: handle swap caches SwapCache support for memory resource controller (memcg) Before mem+swap controller, memcg itself should handle SwapCache in proper way. This is cut-out from it. In current memcg, SwapCache is just leaked and the user can create tons of SwapCache. This is a leak of account and should be handled. SwapCache accounting is done as following. charge (anon) - charged when it's mapped. (because of readahead, charge at add_to_swap_cache() is not sane) uncharge (anon) - uncharged when it's dropped from swapcache and fully unmapped. means it's not uncharged at unmap. Note: delete from swap cache at swap-in is done after rmap information is established. charge (shmem) - charged at swap-in. this prevents charge at add_to_page_cache(). uncharge (shmem) - uncharged when it's dropped from swapcache and not on shmem's radix-tree. at migration, check against 'old page' is modified to handle shmem. Comparing to the old version discussed (and caused troubles), we have advantages of - PCG_USED bit. - simple migrating handling. So, situation is much easier than several months ago, maybe. [hugh@veritas.com: memcg: handle swap caches build fix] Reviewed-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Tested-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Hugh Dickins <hugh@veritas.com> Cc: Li Zefan <lizf@cn.fujitsu.com> Cc: Balbir Singh <balbir@in.ibm.com> Cc: Pavel Emelyanov <xemul@openvz.org> Signed-off-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-01-08 10:07:56 +08:00
*/
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
if (!error)
error = shmem_add_to_page_cache(*pagep, mapping, index,
radswap);
if (error != -ENOMEM) {
/*
* Truncation and eviction use free_swap_and_cache(), which
* only does trylock page: if we raced, best clean up here.
*/
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
delete_from_swap_cache(*pagep);
set_page_dirty(*pagep);
if (!error) {
spin_lock(&info->lock);
info->swapped--;
spin_unlock(&info->lock);
swap_free(swap);
}
}
return error;
}
/*
* Search through swapped inodes to find and replace swap by page.
*/
int shmem_unuse(swp_entry_t swap, struct page *page)
{
struct list_head *this, *next;
struct shmem_inode_info *info;
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
struct mem_cgroup *memcg;
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
int error = 0;
/*
* There's a faint possibility that swap page was replaced before
* caller locked it: caller will come back later with the right page.
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
*/
if (unlikely(!PageSwapCache(page) || page_private(page) != swap.val))
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
goto out;
/*
* Charge page using GFP_KERNEL while we can wait, before taking
* the shmem_swaplist_mutex which might hold up shmem_writepage().
* Charged back to the user (not to caller) when swap account is used.
*/
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
error = mem_cgroup_try_charge(page, current->mm, GFP_KERNEL, &memcg);
if (error)
goto out;
/* No radix_tree_preload: swap entry keeps a place for page in tree */
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
error = -EAGAIN;
mutex_lock(&shmem_swaplist_mutex);
list_for_each_safe(this, next, &shmem_swaplist) {
info = list_entry(this, struct shmem_inode_info, swaplist);
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
if (info->swapped)
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
error = shmem_unuse_inode(info, swap, &page);
else
list_del_init(&info->swaplist);
cond_resched();
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
if (error != -EAGAIN)
break;
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
/* found nothing in this: move on to search the next */
}
mutex_unlock(&shmem_swaplist_mutex);
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
if (error) {
if (error != -ENOMEM)
error = 0;
mem_cgroup_cancel_charge(page, memcg);
} else
mem_cgroup_commit_charge(page, memcg, true);
out:
unlock_page(page);
page_cache_release(page);
return error;
}
/*
* Move the page from the page cache to the swap cache.
*/
static int shmem_writepage(struct page *page, struct writeback_control *wbc)
{
struct shmem_inode_info *info;
struct address_space *mapping;
struct inode *inode;
swp_entry_t swap;
pgoff_t index;
BUG_ON(!PageLocked(page));
mapping = page->mapping;
index = page->index;
inode = mapping->host;
info = SHMEM_I(inode);
if (info->flags & VM_LOCKED)
goto redirty;
if (!total_swap_pages)
goto redirty;
/*
* Our capabilities prevent regular writeback or sync from ever calling
* shmem_writepage; but a stacking filesystem might use ->writepage of
* its underlying filesystem, in which case tmpfs should write out to
* swap only in response to memory pressure, and not for the writeback
* threads or sync.
*/
if (!wbc->for_reclaim) {
WARN_ON_ONCE(1); /* Still happens? Tell us about it! */
goto redirty;
}
/*
* This is somewhat ridiculous, but without plumbing a SWAP_MAP_FALLOC
* value into swapfile.c, the only way we can correctly account for a
* fallocated page arriving here is now to initialize it and write it.
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
*
* That's okay for a page already fallocated earlier, but if we have
* not yet completed the fallocation, then (a) we want to keep track
* of this page in case we have to undo it, and (b) it may not be a
* good idea to continue anyway, once we're pushing into swap. So
* reactivate the page, and let shmem_fallocate() quit when too many.
*/
if (!PageUptodate(page)) {
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
if (inode->i_private) {
struct shmem_falloc *shmem_falloc;
spin_lock(&inode->i_lock);
shmem_falloc = inode->i_private;
if (shmem_falloc &&
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
!shmem_falloc->waitq &&
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
index >= shmem_falloc->start &&
index < shmem_falloc->next)
shmem_falloc->nr_unswapped++;
else
shmem_falloc = NULL;
spin_unlock(&inode->i_lock);
if (shmem_falloc)
goto redirty;
}
clear_highpage(page);
flush_dcache_page(page);
SetPageUptodate(page);
}
swap = get_swap_page();
if (!swap.val)
goto redirty;
tmpfs: fix race between umount and writepage Konstanin Khlebnikov reports that a dangerous race between umount and shmem_writepage can be reproduced by this script: for i in {1..300} ; do mkdir $i while true ; do mount -t tmpfs none $i dd if=/dev/zero of=$i/test bs=1M count=$(($RANDOM % 100)) umount $i done & done on a 6xCPU node with 8Gb RAM: kernel very unstable after this accident. =) Kernel log: VFS: Busy inodes after unmount of tmpfs. Self-destruct in 5 seconds. Have a nice day... WARNING: at lib/list_debug.c:53 __list_del_entry+0x8d/0x98() list_del corruption. prev->next should be ffff880222fdaac8, but was (null) Pid: 11222, comm: mount.tmpfs Not tainted 2.6.39-rc2+ #4 Call Trace: warn_slowpath_common+0x80/0x98 warn_slowpath_fmt+0x41/0x43 __list_del_entry+0x8d/0x98 evict+0x50/0x113 iput+0x138/0x141 ... BUG: unable to handle kernel paging request at ffffffffffffffff IP: shmem_free_blocks+0x18/0x4c Pid: 10422, comm: dd Tainted: G W 2.6.39-rc2+ #4 Call Trace: shmem_recalc_inode+0x61/0x66 shmem_writepage+0xba/0x1dc pageout+0x13c/0x24c shrink_page_list+0x28e/0x4be shrink_inactive_list+0x21f/0x382 ... shmem_writepage() calls igrab() on the inode for the page which came from page reclaim, to add it later into shmem_swaplist for swapoff operation. This igrab() can race with super-block deactivating process: shrink_inactive_list() deactivate_super() pageout() tmpfs_fs_type->kill_sb() shmem_writepage() kill_litter_super() generic_shutdown_super() evict_inodes() igrab() atomic_read(&inode->i_count) skip-inode iput() if (!list_empty(&sb->s_inodes)) printk("VFS: Busy inodes after... This igrap-iput pair was added in commit 1b1b32f2c6f6 "tmpfs: fix shmem_swaplist races" based on incorrect assumptions: igrab() protects the inode from concurrent eviction by deletion, but it does nothing to protect it from concurrent unmounting, which goes ahead despite the raised i_count. So this use of igrab() was wrong all along, but the race made much worse in 2.6.37 when commit 63997e98a3be "split invalidate_inodes()" replaced two attempts at invalidate_inodes() by a single evict_inodes(). Konstantin posted a plausible patch, raising sb->s_active too: I'm unsure whether it was correct or not; but burnt once by igrab(), I am sure that we don't want to rely more deeply upon externals here. Fix it by adding the inode to shmem_swaplist earlier, while the page lock on page in page cache still secures the inode against eviction, without artifically raising i_count. It was originally added later because shmem_unuse_inode() is liable to remove an inode from the list while it's unswapped; but we can guard against that by taking spinlock before dropping mutex. Reported-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Signed-off-by: Hugh Dickins <hughd@google.com> Tested-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-12 06:13:36 +08:00
/*
* Add inode to shmem_unuse()'s list of swapped-out inodes,
* if it's not already there. Do it now before the page is
* moved to swap cache, when its pagelock no longer protects
tmpfs: fix race between umount and writepage Konstanin Khlebnikov reports that a dangerous race between umount and shmem_writepage can be reproduced by this script: for i in {1..300} ; do mkdir $i while true ; do mount -t tmpfs none $i dd if=/dev/zero of=$i/test bs=1M count=$(($RANDOM % 100)) umount $i done & done on a 6xCPU node with 8Gb RAM: kernel very unstable after this accident. =) Kernel log: VFS: Busy inodes after unmount of tmpfs. Self-destruct in 5 seconds. Have a nice day... WARNING: at lib/list_debug.c:53 __list_del_entry+0x8d/0x98() list_del corruption. prev->next should be ffff880222fdaac8, but was (null) Pid: 11222, comm: mount.tmpfs Not tainted 2.6.39-rc2+ #4 Call Trace: warn_slowpath_common+0x80/0x98 warn_slowpath_fmt+0x41/0x43 __list_del_entry+0x8d/0x98 evict+0x50/0x113 iput+0x138/0x141 ... BUG: unable to handle kernel paging request at ffffffffffffffff IP: shmem_free_blocks+0x18/0x4c Pid: 10422, comm: dd Tainted: G W 2.6.39-rc2+ #4 Call Trace: shmem_recalc_inode+0x61/0x66 shmem_writepage+0xba/0x1dc pageout+0x13c/0x24c shrink_page_list+0x28e/0x4be shrink_inactive_list+0x21f/0x382 ... shmem_writepage() calls igrab() on the inode for the page which came from page reclaim, to add it later into shmem_swaplist for swapoff operation. This igrab() can race with super-block deactivating process: shrink_inactive_list() deactivate_super() pageout() tmpfs_fs_type->kill_sb() shmem_writepage() kill_litter_super() generic_shutdown_super() evict_inodes() igrab() atomic_read(&inode->i_count) skip-inode iput() if (!list_empty(&sb->s_inodes)) printk("VFS: Busy inodes after... This igrap-iput pair was added in commit 1b1b32f2c6f6 "tmpfs: fix shmem_swaplist races" based on incorrect assumptions: igrab() protects the inode from concurrent eviction by deletion, but it does nothing to protect it from concurrent unmounting, which goes ahead despite the raised i_count. So this use of igrab() was wrong all along, but the race made much worse in 2.6.37 when commit 63997e98a3be "split invalidate_inodes()" replaced two attempts at invalidate_inodes() by a single evict_inodes(). Konstantin posted a plausible patch, raising sb->s_active too: I'm unsure whether it was correct or not; but burnt once by igrab(), I am sure that we don't want to rely more deeply upon externals here. Fix it by adding the inode to shmem_swaplist earlier, while the page lock on page in page cache still secures the inode against eviction, without artifically raising i_count. It was originally added later because shmem_unuse_inode() is liable to remove an inode from the list while it's unswapped; but we can guard against that by taking spinlock before dropping mutex. Reported-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Signed-off-by: Hugh Dickins <hughd@google.com> Tested-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-12 06:13:36 +08:00
* the inode from eviction. But don't unlock the mutex until
* we've incremented swapped, because shmem_unuse_inode() will
* prune a !swapped inode from the swaplist under this mutex.
tmpfs: fix race between umount and writepage Konstanin Khlebnikov reports that a dangerous race between umount and shmem_writepage can be reproduced by this script: for i in {1..300} ; do mkdir $i while true ; do mount -t tmpfs none $i dd if=/dev/zero of=$i/test bs=1M count=$(($RANDOM % 100)) umount $i done & done on a 6xCPU node with 8Gb RAM: kernel very unstable after this accident. =) Kernel log: VFS: Busy inodes after unmount of tmpfs. Self-destruct in 5 seconds. Have a nice day... WARNING: at lib/list_debug.c:53 __list_del_entry+0x8d/0x98() list_del corruption. prev->next should be ffff880222fdaac8, but was (null) Pid: 11222, comm: mount.tmpfs Not tainted 2.6.39-rc2+ #4 Call Trace: warn_slowpath_common+0x80/0x98 warn_slowpath_fmt+0x41/0x43 __list_del_entry+0x8d/0x98 evict+0x50/0x113 iput+0x138/0x141 ... BUG: unable to handle kernel paging request at ffffffffffffffff IP: shmem_free_blocks+0x18/0x4c Pid: 10422, comm: dd Tainted: G W 2.6.39-rc2+ #4 Call Trace: shmem_recalc_inode+0x61/0x66 shmem_writepage+0xba/0x1dc pageout+0x13c/0x24c shrink_page_list+0x28e/0x4be shrink_inactive_list+0x21f/0x382 ... shmem_writepage() calls igrab() on the inode for the page which came from page reclaim, to add it later into shmem_swaplist for swapoff operation. This igrab() can race with super-block deactivating process: shrink_inactive_list() deactivate_super() pageout() tmpfs_fs_type->kill_sb() shmem_writepage() kill_litter_super() generic_shutdown_super() evict_inodes() igrab() atomic_read(&inode->i_count) skip-inode iput() if (!list_empty(&sb->s_inodes)) printk("VFS: Busy inodes after... This igrap-iput pair was added in commit 1b1b32f2c6f6 "tmpfs: fix shmem_swaplist races" based on incorrect assumptions: igrab() protects the inode from concurrent eviction by deletion, but it does nothing to protect it from concurrent unmounting, which goes ahead despite the raised i_count. So this use of igrab() was wrong all along, but the race made much worse in 2.6.37 when commit 63997e98a3be "split invalidate_inodes()" replaced two attempts at invalidate_inodes() by a single evict_inodes(). Konstantin posted a plausible patch, raising sb->s_active too: I'm unsure whether it was correct or not; but burnt once by igrab(), I am sure that we don't want to rely more deeply upon externals here. Fix it by adding the inode to shmem_swaplist earlier, while the page lock on page in page cache still secures the inode against eviction, without artifically raising i_count. It was originally added later because shmem_unuse_inode() is liable to remove an inode from the list while it's unswapped; but we can guard against that by taking spinlock before dropping mutex. Reported-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Signed-off-by: Hugh Dickins <hughd@google.com> Tested-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-12 06:13:36 +08:00
*/
mutex_lock(&shmem_swaplist_mutex);
if (list_empty(&info->swaplist))
list_add_tail(&info->swaplist, &shmem_swaplist);
tmpfs: fix race between umount and writepage Konstanin Khlebnikov reports that a dangerous race between umount and shmem_writepage can be reproduced by this script: for i in {1..300} ; do mkdir $i while true ; do mount -t tmpfs none $i dd if=/dev/zero of=$i/test bs=1M count=$(($RANDOM % 100)) umount $i done & done on a 6xCPU node with 8Gb RAM: kernel very unstable after this accident. =) Kernel log: VFS: Busy inodes after unmount of tmpfs. Self-destruct in 5 seconds. Have a nice day... WARNING: at lib/list_debug.c:53 __list_del_entry+0x8d/0x98() list_del corruption. prev->next should be ffff880222fdaac8, but was (null) Pid: 11222, comm: mount.tmpfs Not tainted 2.6.39-rc2+ #4 Call Trace: warn_slowpath_common+0x80/0x98 warn_slowpath_fmt+0x41/0x43 __list_del_entry+0x8d/0x98 evict+0x50/0x113 iput+0x138/0x141 ... BUG: unable to handle kernel paging request at ffffffffffffffff IP: shmem_free_blocks+0x18/0x4c Pid: 10422, comm: dd Tainted: G W 2.6.39-rc2+ #4 Call Trace: shmem_recalc_inode+0x61/0x66 shmem_writepage+0xba/0x1dc pageout+0x13c/0x24c shrink_page_list+0x28e/0x4be shrink_inactive_list+0x21f/0x382 ... shmem_writepage() calls igrab() on the inode for the page which came from page reclaim, to add it later into shmem_swaplist for swapoff operation. This igrab() can race with super-block deactivating process: shrink_inactive_list() deactivate_super() pageout() tmpfs_fs_type->kill_sb() shmem_writepage() kill_litter_super() generic_shutdown_super() evict_inodes() igrab() atomic_read(&inode->i_count) skip-inode iput() if (!list_empty(&sb->s_inodes)) printk("VFS: Busy inodes after... This igrap-iput pair was added in commit 1b1b32f2c6f6 "tmpfs: fix shmem_swaplist races" based on incorrect assumptions: igrab() protects the inode from concurrent eviction by deletion, but it does nothing to protect it from concurrent unmounting, which goes ahead despite the raised i_count. So this use of igrab() was wrong all along, but the race made much worse in 2.6.37 when commit 63997e98a3be "split invalidate_inodes()" replaced two attempts at invalidate_inodes() by a single evict_inodes(). Konstantin posted a plausible patch, raising sb->s_active too: I'm unsure whether it was correct or not; but burnt once by igrab(), I am sure that we don't want to rely more deeply upon externals here. Fix it by adding the inode to shmem_swaplist earlier, while the page lock on page in page cache still secures the inode against eviction, without artifically raising i_count. It was originally added later because shmem_unuse_inode() is liable to remove an inode from the list while it's unswapped; but we can guard against that by taking spinlock before dropping mutex. Reported-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Signed-off-by: Hugh Dickins <hughd@google.com> Tested-by: Konstantin Khlebnikov <khlebnikov@openvz.org> Cc: <stable@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-12 06:13:36 +08:00
if (add_to_swap_cache(page, swap, GFP_ATOMIC) == 0) {
swap_shmem_alloc(swap);
shmem_delete_from_page_cache(page, swp_to_radix_entry(swap));
spin_lock(&info->lock);
info->swapped++;
shmem_recalc_inode(inode);
spin_unlock(&info->lock);
mutex_unlock(&shmem_swaplist_mutex);
BUG_ON(page_mapped(page));
shmem: writepage directly to swap Synopsis: if shmem_writepage calls swap_writepage directly, most shmem swap loads benefit, and a catastrophic interaction between SLUB and some flash storage is avoided. shmem_writepage() has always been peculiar in making no attempt to write: it has just transferred a shmem page from file cache to swap cache, then let that page make its way around the LRU again before being written and freed. The idea was that people use tmpfs because they want those pages to stay in RAM; so although we give it an overflow to swap, we should resist writing too soon, giving those pages a second chance before they can be reclaimed. That was always questionable, and I've toyed with this patch for years; but never had a clear justification to depart from the original design. It became more questionable in 2.6.28, when the split LRU patches classed shmem and tmpfs pages as SwapBacked rather than as file_cache: that in itself gives them more resistance to reclaim than normal file pages. I prepared this patch for 2.6.29, but the merge window arrived before I'd completed gathering statistics to justify sending it in. Then while comparing SLQB against SLUB, running SLUB on a laptop I'd habitually used with SLAB, I found SLUB to run my tmpfs kbuild swapping tests five times slower than SLAB or SLQB - other machines slower too, but nowhere near so bad. Simpler "cp -a" swapping tests showed the same. slub_max_order=0 brings sanity to all, but heavy swapping is too far from normal to justify such a tuning. The crucial factor on that laptop turns out to be that I'm using an SD card for swap. What happens is this: By default, SLUB uses order-2 pages for shmem_inode_cache (and many other fs inodes), so creating tmpfs files under memory pressure brings lumpy reclaim into play. One subpage of the order is chosen from the bottom of the LRU as usual, then the other three picked out from their random positions on the LRUs. In a tmpfs load, many of these pages will be ones which already passed through shmem_writepage, so already have swap allocated. And though their offsets on swap were probably allocated sequentially, now that the pages are picked off at random, their swap offsets are scattered. But the flash storage on the SD card is very sensitive to having its writes merged: once swap is written at scattered offsets, performance falls apart. Rotating disk seeks increase too, but less disastrously. So: stop giving shmem/tmpfs pages a second pass around the LRU, write them out to swap as soon as their swap has been allocated. It's surely possible to devise an artificial load which runs faster the old way, one whose sizing is such that the tmpfs pages on their second pass are the ones that are wanted again, and other pages not. But I've not yet found such a load: on all machines, under the loads I've tried, immediate swap_writepage speeds up shmem swapping: especially when using the SLUB allocator (and more effectively than slub_max_order=0), but also with the others; and it also reduces the variance between runs. How much faster varies widely: a factor of five is rare, 5% is common. One load which might have suffered: imagine a swapping shmem load in a limited mem_cgroup on a machine with plenty of memory. Before 2.6.29 the swapcache was not charged, and such a load would have run quickest with the shmem swapcache never written to swap. But now swapcache is charged, so even this load benefits from shmem_writepage directly to swap. Apologies for the #ifndef CONFIG_SWAP swap_writepage() stub in swap.h: it's silly because that will never get called; but refactoring shmem.c sensibly according to CONFIG_SWAP will be a separate task. Signed-off-by: Hugh Dickins <hugh@veritas.com> Acked-by: Pekka Enberg <penberg@cs.helsinki.fi> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-01 06:23:33 +08:00
swap_writepage(page, wbc);
return 0;
}
mutex_unlock(&shmem_swaplist_mutex);
mm: memcontrol: rewrite uncharge API The memcg uncharging code that is involved towards the end of a page's lifetime - truncation, reclaim, swapout, migration - is impressively complicated and fragile. Because anonymous and file pages were always charged before they had their page->mapping established, uncharges had to happen when the page type could still be known from the context; as in unmap for anonymous, page cache removal for file and shmem pages, and swap cache truncation for swap pages. However, these operations happen well before the page is actually freed, and so a lot of synchronization is necessary: - Charging, uncharging, page migration, and charge migration all need to take a per-page bit spinlock as they could race with uncharging. - Swap cache truncation happens during both swap-in and swap-out, and possibly repeatedly before the page is actually freed. This means that the memcg swapout code is called from many contexts that make no sense and it has to figure out the direction from page state to make sure memory and memory+swap are always correctly charged. - On page migration, the old page might be unmapped but then reused, so memcg code has to prevent untimely uncharging in that case. Because this code - which should be a simple charge transfer - is so special-cased, it is not reusable for replace_page_cache(). But now that charged pages always have a page->mapping, introduce mem_cgroup_uncharge(), which is called after the final put_page(), when we know for sure that nobody is looking at the page anymore. For page migration, introduce mem_cgroup_migrate(), which is called after the migration is successful and the new page is fully rmapped. Because the old page is no longer uncharged after migration, prevent double charges by decoupling the page's memcg association (PCG_USED and pc->mem_cgroup) from the page holding an actual charge. The new bits PCG_MEM and PCG_MEMSW represent the respective charges and are transferred to the new page during migration. mem_cgroup_migrate() is suitable for replace_page_cache() as well, which gets rid of mem_cgroup_replace_page_cache(). However, care needs to be taken because both the source and the target page can already be charged and on the LRU when fuse is splicing: grab the page lock on the charge moving side to prevent changing pc->mem_cgroup of a page under migration. Also, the lruvecs of both pages change as we uncharge the old and charge the new during migration, and putback may race with us, so grab the lru lock and isolate the pages iff on LRU to prevent races and ensure the pages are on the right lruvec afterward. Swap accounting is massively simplified: because the page is no longer uncharged as early as swap cache deletion, a new mem_cgroup_swapout() can transfer the page's memory+swap charge (PCG_MEMSW) to the swap entry before the final put_page() in page reclaim. Finally, page_cgroup changes are now protected by whatever protection the page itself offers: anonymous pages are charged under the page table lock, whereas page cache insertions, swapin, and migration hold the page lock. Uncharging happens under full exclusion with no outstanding references. Charging and uncharging also ensure that the page is off-LRU, which serializes against charge migration. Remove the very costly page_cgroup lock and set pc->flags non-atomically. [mhocko@suse.cz: mem_cgroup_charge_statistics needs preempt_disable] [vdavydov@parallels.com: fix flags definition] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Cc: Hugh Dickins <hughd@google.com> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Tested-by: Jet Chen <jet.chen@intel.com> Acked-by: Michal Hocko <mhocko@suse.cz> Tested-by: Felipe Balbi <balbi@ti.com> Signed-off-by: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:22 +08:00
swapcache_free(swap);
redirty:
set_page_dirty(page);
if (wbc->for_reclaim)
return AOP_WRITEPAGE_ACTIVATE; /* Return with page locked */
unlock_page(page);
return 0;
}
#ifdef CONFIG_NUMA
#ifdef CONFIG_TMPFS
mempolicy: use struct mempolicy pointer in shmem_sb_info This patch replaces the mempolicy mode, mode_flags, and nodemask in the shmem_sb_info struct with a struct mempolicy pointer, initialized to NULL. This removes dependency on the details of mempolicy from shmem.c and hugetlbfs inode.c and simplifies the interfaces. mpol_parse_str() in mempolicy.c is changed to return, via a pointer to a pointer arg, a struct mempolicy pointer on success. For MPOL_DEFAULT, the returned pointer is NULL. Further, mpol_parse_str() now takes a 'no_context' argument that causes the input nodemask to be stored in the w.user_nodemask of the created mempolicy for use when the mempolicy is installed in a tmpfs inode shared policy tree. At that time, any cpuset contextualization is applied to the original input nodemask. This preserves the previous behavior where the input nodemask was stored in the superblock. We can think of the returned mempolicy as "context free". Because mpol_parse_str() is now calling mpol_new(), we can remove from mpol_to_str() the semantic checks that mpol_new() already performs. Add 'no_context' parameter to mpol_to_str() to specify that it should format the nodemask in w.user_nodemask for 'bind' and 'interleave' policies. Change mpol_shared_policy_init() to take a pointer to a "context free" struct mempolicy and to create a new, "contextualized" mempolicy using the mode, mode_flags and user_nodemask from the input mempolicy. Note: we know that the mempolicy passed to mpol_to_str() or mpol_shared_policy_init() from a tmpfs superblock is "context free". This is currently the only instance thereof. However, if we found more uses for this concept, and introduced any ambiguity as to whether a mempolicy was context free or not, we could add another internal mode flag to identify context free mempolicies. Then, we could remove the 'no_context' argument from mpol_to_str(). Added shmem_get_sbmpol() to return a reference counted superblock mempolicy, if one exists, to pass to mpol_shared_policy_init(). We must add the reference under the sb stat_lock to prevent races with replacement of the mpol by remount. This reference is removed in mpol_shared_policy_init(). [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: yet another build fix] Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:26 +08:00
static void shmem_show_mpol(struct seq_file *seq, struct mempolicy *mpol)
{
mempolicy: rework shmem mpol parsing and display mm/shmem.c currently contains functions to parse and display memory policy strings for the tmpfs 'mpol' mount option. Move this to mm/mempolicy.c with the rest of the mempolicy support. With subsequent patches, we'll be able to remove knowledge of the details [mode, flags, policy, ...] completely from shmem.c 1) replace shmem_parse_mpol() in mm/shmem.c with mpol_parse_str() in mm/mempolicy.c. Rework to use the policy_types[] array [used by mpol_to_str()] to look up mode by name. 2) use mpol_to_str() to format policy for shmem_show_mpol(). mpol_to_str() expects a pointer to a struct mempolicy, so temporarily construct one. This will be replaced with a reference to a struct mempolicy in the tmpfs superblock in a subsequent patch. NOTE 1: I changed mpol_to_str() to use a colon ':' rather than an equal sign '=' as the nodemask delimiter to match mpol_parse_str() and the tmpfs/shmem mpol mount option formatting that now uses mpol_to_str(). This is a user visible change to numa_maps, but then the addition of the mode flags already changed the display. It makes sense to me to have the mounts and numa_maps display the policy in the same format. However, if anyone objects strongly, I can pass the desired nodemask delimeter as an arg to mpol_to_str(). Note 2: Like show_numa_map(), I don't check the return code from mpol_to_str(). I do use a longer buffer than the one provided by show_numa_map(), which seems to have sufficed so far. Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:23 +08:00
char buffer[64];
mempolicy: use struct mempolicy pointer in shmem_sb_info This patch replaces the mempolicy mode, mode_flags, and nodemask in the shmem_sb_info struct with a struct mempolicy pointer, initialized to NULL. This removes dependency on the details of mempolicy from shmem.c and hugetlbfs inode.c and simplifies the interfaces. mpol_parse_str() in mempolicy.c is changed to return, via a pointer to a pointer arg, a struct mempolicy pointer on success. For MPOL_DEFAULT, the returned pointer is NULL. Further, mpol_parse_str() now takes a 'no_context' argument that causes the input nodemask to be stored in the w.user_nodemask of the created mempolicy for use when the mempolicy is installed in a tmpfs inode shared policy tree. At that time, any cpuset contextualization is applied to the original input nodemask. This preserves the previous behavior where the input nodemask was stored in the superblock. We can think of the returned mempolicy as "context free". Because mpol_parse_str() is now calling mpol_new(), we can remove from mpol_to_str() the semantic checks that mpol_new() already performs. Add 'no_context' parameter to mpol_to_str() to specify that it should format the nodemask in w.user_nodemask for 'bind' and 'interleave' policies. Change mpol_shared_policy_init() to take a pointer to a "context free" struct mempolicy and to create a new, "contextualized" mempolicy using the mode, mode_flags and user_nodemask from the input mempolicy. Note: we know that the mempolicy passed to mpol_to_str() or mpol_shared_policy_init() from a tmpfs superblock is "context free". This is currently the only instance thereof. However, if we found more uses for this concept, and introduced any ambiguity as to whether a mempolicy was context free or not, we could add another internal mode flag to identify context free mempolicies. Then, we could remove the 'no_context' argument from mpol_to_str(). Added shmem_get_sbmpol() to return a reference counted superblock mempolicy, if one exists, to pass to mpol_shared_policy_init(). We must add the reference under the sb stat_lock to prevent races with replacement of the mpol by remount. This reference is removed in mpol_shared_policy_init(). [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: yet another build fix] Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:26 +08:00
if (!mpol || mpol->mode == MPOL_DEFAULT)
mempolicy: rework shmem mpol parsing and display mm/shmem.c currently contains functions to parse and display memory policy strings for the tmpfs 'mpol' mount option. Move this to mm/mempolicy.c with the rest of the mempolicy support. With subsequent patches, we'll be able to remove knowledge of the details [mode, flags, policy, ...] completely from shmem.c 1) replace shmem_parse_mpol() in mm/shmem.c with mpol_parse_str() in mm/mempolicy.c. Rework to use the policy_types[] array [used by mpol_to_str()] to look up mode by name. 2) use mpol_to_str() to format policy for shmem_show_mpol(). mpol_to_str() expects a pointer to a struct mempolicy, so temporarily construct one. This will be replaced with a reference to a struct mempolicy in the tmpfs superblock in a subsequent patch. NOTE 1: I changed mpol_to_str() to use a colon ':' rather than an equal sign '=' as the nodemask delimiter to match mpol_parse_str() and the tmpfs/shmem mpol mount option formatting that now uses mpol_to_str(). This is a user visible change to numa_maps, but then the addition of the mode flags already changed the display. It makes sense to me to have the mounts and numa_maps display the policy in the same format. However, if anyone objects strongly, I can pass the desired nodemask delimeter as an arg to mpol_to_str(). Note 2: Like show_numa_map(), I don't check the return code from mpol_to_str(). I do use a longer buffer than the one provided by show_numa_map(), which seems to have sufficed so far. Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:23 +08:00
return; /* show nothing */
mpol_to_str(buffer, sizeof(buffer), mpol);
mempolicy: rework shmem mpol parsing and display mm/shmem.c currently contains functions to parse and display memory policy strings for the tmpfs 'mpol' mount option. Move this to mm/mempolicy.c with the rest of the mempolicy support. With subsequent patches, we'll be able to remove knowledge of the details [mode, flags, policy, ...] completely from shmem.c 1) replace shmem_parse_mpol() in mm/shmem.c with mpol_parse_str() in mm/mempolicy.c. Rework to use the policy_types[] array [used by mpol_to_str()] to look up mode by name. 2) use mpol_to_str() to format policy for shmem_show_mpol(). mpol_to_str() expects a pointer to a struct mempolicy, so temporarily construct one. This will be replaced with a reference to a struct mempolicy in the tmpfs superblock in a subsequent patch. NOTE 1: I changed mpol_to_str() to use a colon ':' rather than an equal sign '=' as the nodemask delimiter to match mpol_parse_str() and the tmpfs/shmem mpol mount option formatting that now uses mpol_to_str(). This is a user visible change to numa_maps, but then the addition of the mode flags already changed the display. It makes sense to me to have the mounts and numa_maps display the policy in the same format. However, if anyone objects strongly, I can pass the desired nodemask delimeter as an arg to mpol_to_str(). Note 2: Like show_numa_map(), I don't check the return code from mpol_to_str(). I do use a longer buffer than the one provided by show_numa_map(), which seems to have sufficed so far. Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:23 +08:00
seq_printf(seq, ",mpol=%s", buffer);
}
mempolicy: use struct mempolicy pointer in shmem_sb_info This patch replaces the mempolicy mode, mode_flags, and nodemask in the shmem_sb_info struct with a struct mempolicy pointer, initialized to NULL. This removes dependency on the details of mempolicy from shmem.c and hugetlbfs inode.c and simplifies the interfaces. mpol_parse_str() in mempolicy.c is changed to return, via a pointer to a pointer arg, a struct mempolicy pointer on success. For MPOL_DEFAULT, the returned pointer is NULL. Further, mpol_parse_str() now takes a 'no_context' argument that causes the input nodemask to be stored in the w.user_nodemask of the created mempolicy for use when the mempolicy is installed in a tmpfs inode shared policy tree. At that time, any cpuset contextualization is applied to the original input nodemask. This preserves the previous behavior where the input nodemask was stored in the superblock. We can think of the returned mempolicy as "context free". Because mpol_parse_str() is now calling mpol_new(), we can remove from mpol_to_str() the semantic checks that mpol_new() already performs. Add 'no_context' parameter to mpol_to_str() to specify that it should format the nodemask in w.user_nodemask for 'bind' and 'interleave' policies. Change mpol_shared_policy_init() to take a pointer to a "context free" struct mempolicy and to create a new, "contextualized" mempolicy using the mode, mode_flags and user_nodemask from the input mempolicy. Note: we know that the mempolicy passed to mpol_to_str() or mpol_shared_policy_init() from a tmpfs superblock is "context free". This is currently the only instance thereof. However, if we found more uses for this concept, and introduced any ambiguity as to whether a mempolicy was context free or not, we could add another internal mode flag to identify context free mempolicies. Then, we could remove the 'no_context' argument from mpol_to_str(). Added shmem_get_sbmpol() to return a reference counted superblock mempolicy, if one exists, to pass to mpol_shared_policy_init(). We must add the reference under the sb stat_lock to prevent races with replacement of the mpol by remount. This reference is removed in mpol_shared_policy_init(). [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: yet another build fix] Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:26 +08:00
static struct mempolicy *shmem_get_sbmpol(struct shmem_sb_info *sbinfo)
{
struct mempolicy *mpol = NULL;
if (sbinfo->mpol) {
spin_lock(&sbinfo->stat_lock); /* prevent replace/use races */
mpol = sbinfo->mpol;
mpol_get(mpol);
spin_unlock(&sbinfo->stat_lock);
}
return mpol;
}
#endif /* CONFIG_TMPFS */
static struct page *shmem_swapin(swp_entry_t swap, gfp_t gfp,
struct shmem_inode_info *info, pgoff_t index)
{
struct vm_area_struct pvma;
struct page *page;
mempolicy: rework mempolicy Reference Counting [yet again] After further discussion with Christoph Lameter, it has become clear that my earlier attempts to clean up the mempolicy reference counting were a bit of overkill in some areas, resulting in superflous ref/unref in what are usually fast paths. In other areas, further inspection reveals that I botched the unref for interleave policies. A separate patch, suitable for upstream/stable trees, fixes up the known errors in the previous attempt to fix reference counting. This patch reworks the memory policy referencing counting and, one hopes, simplifies the code. Maybe I'll get it right this time. See the update to the numa_memory_policy.txt document for a discussion of memory policy reference counting that motivates this patch. Summary: Lookup of mempolicy, based on (vma, address) need only add a reference for shared policy, and we need only unref the policy when finished for shared policies. So, this patch backs out all of the unneeded extra reference counting added by my previous attempt. It then unrefs only shared policies when we're finished with them, using the mpol_cond_put() [conditional put] helper function introduced by this patch. Note that shmem_swapin() calls read_swap_cache_async() with a dummy vma containing just the policy. read_swap_cache_async() can call alloc_page_vma() multiple times, so we can't let alloc_page_vma() unref the shared policy in this case. To avoid this, we make a copy of any non-null shared policy and remove the MPOL_F_SHARED flag from the copy. This copy occurs before reading a page [or multiple pages] from swap, so the overhead should not be an issue here. I introduced a new static inline function "mpol_cond_copy()" to copy the shared policy to an on-stack policy and remove the flags that would require a conditional free. The current implementation of mpol_cond_copy() assumes that the struct mempolicy contains no pointers to dynamically allocated structures that must be duplicated or reference counted during copy. Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:16 +08:00
/* Create a pseudo vma that just contains the policy */
swapin_readahead: excise NUMA bogosity For three years swapin_readahead has been cluttered with fanciful CONFIG_NUMA code, advancing addr, and stepping on to the next vma at the boundary, to line up the mempolicy for each page allocation. It _might_ be a good idea to allocate swap more according to vma layout; but the fact is, that's not how we do it at all, 2.6 even less than 2.4: swap is allocated as needed for pages as they sink to the bottom of the inactive LRUs. Sometimes that may match vma layout, but not so often that it's worth going to these misleading vma->vm_next lengths: rip all that out. Originally I intended to retain the incrementation of addr, but correct its initial value: valid_swaphandles generally supplies an offset below the target addr (this is readaround rather than readahead), but addr has not been adjusted accordingly, so in the interleave case it has usually been allocating the target page from the "wrong" node (though that may not matter very much). But look at the equivalent shmem_swapin code: either by oversight or by design, though it has all the apparatus for choosing a new mempolicy per page, it uses the same idx throughout, choosing the same mempolicy and interleave node for each page of the cluster. Which is actually a much better strategy: each node has its own LRUs and its own kswapd, so if you're betting on any particular relationship between swap and node, the best bet is that nearby swap entries belong to pages from the same node - even when the mempolicy of the target page is to interleave. And examining a map of nodes corresponding to swap entries on a numa=fake system bears this out. (We could later tweak swap allocation to make it even more likely, but this patch is merely about removing cruft.) So, neither adjust nor increment addr in swapin_readahead, and then shmem_swapin can use it too; the pseudo-vma to pass policy need only be set up once per cluster, and so few fields of pvma are used, let's skip the memset - from shmem_alloc_page also. Signed-off-by: Hugh Dickins <hugh@veritas.com> Acked-by: Rik van Riel <riel@redhat.com> Cc: Andi Kleen <ak@suse.de> Cc: Christoph Lameter <clameter@sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:40 +08:00
pvma.vm_start = 0;
/* Bias interleave by inode number to distribute better across nodes */
pvma.vm_pgoff = index + info->vfs_inode.i_ino;
swapin_readahead: excise NUMA bogosity For three years swapin_readahead has been cluttered with fanciful CONFIG_NUMA code, advancing addr, and stepping on to the next vma at the boundary, to line up the mempolicy for each page allocation. It _might_ be a good idea to allocate swap more according to vma layout; but the fact is, that's not how we do it at all, 2.6 even less than 2.4: swap is allocated as needed for pages as they sink to the bottom of the inactive LRUs. Sometimes that may match vma layout, but not so often that it's worth going to these misleading vma->vm_next lengths: rip all that out. Originally I intended to retain the incrementation of addr, but correct its initial value: valid_swaphandles generally supplies an offset below the target addr (this is readaround rather than readahead), but addr has not been adjusted accordingly, so in the interleave case it has usually been allocating the target page from the "wrong" node (though that may not matter very much). But look at the equivalent shmem_swapin code: either by oversight or by design, though it has all the apparatus for choosing a new mempolicy per page, it uses the same idx throughout, choosing the same mempolicy and interleave node for each page of the cluster. Which is actually a much better strategy: each node has its own LRUs and its own kswapd, so if you're betting on any particular relationship between swap and node, the best bet is that nearby swap entries belong to pages from the same node - even when the mempolicy of the target page is to interleave. And examining a map of nodes corresponding to swap entries on a numa=fake system bears this out. (We could later tweak swap allocation to make it even more likely, but this patch is merely about removing cruft.) So, neither adjust nor increment addr in swapin_readahead, and then shmem_swapin can use it too; the pseudo-vma to pass policy need only be set up once per cluster, and so few fields of pvma are used, let's skip the memset - from shmem_alloc_page also. Signed-off-by: Hugh Dickins <hugh@veritas.com> Acked-by: Rik van Riel <riel@redhat.com> Cc: Andi Kleen <ak@suse.de> Cc: Christoph Lameter <clameter@sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:40 +08:00
pvma.vm_ops = NULL;
pvma.vm_policy = mpol_shared_policy_lookup(&info->policy, index);
page = swapin_readahead(swap, gfp, &pvma, 0);
/* Drop reference taken by mpol_shared_policy_lookup() */
mpol_cond_put(pvma.vm_policy);
return page;
}
static struct page *shmem_alloc_page(gfp_t gfp,
struct shmem_inode_info *info, pgoff_t index)
{
struct vm_area_struct pvma;
struct page *page;
swapin_readahead: excise NUMA bogosity For three years swapin_readahead has been cluttered with fanciful CONFIG_NUMA code, advancing addr, and stepping on to the next vma at the boundary, to line up the mempolicy for each page allocation. It _might_ be a good idea to allocate swap more according to vma layout; but the fact is, that's not how we do it at all, 2.6 even less than 2.4: swap is allocated as needed for pages as they sink to the bottom of the inactive LRUs. Sometimes that may match vma layout, but not so often that it's worth going to these misleading vma->vm_next lengths: rip all that out. Originally I intended to retain the incrementation of addr, but correct its initial value: valid_swaphandles generally supplies an offset below the target addr (this is readaround rather than readahead), but addr has not been adjusted accordingly, so in the interleave case it has usually been allocating the target page from the "wrong" node (though that may not matter very much). But look at the equivalent shmem_swapin code: either by oversight or by design, though it has all the apparatus for choosing a new mempolicy per page, it uses the same idx throughout, choosing the same mempolicy and interleave node for each page of the cluster. Which is actually a much better strategy: each node has its own LRUs and its own kswapd, so if you're betting on any particular relationship between swap and node, the best bet is that nearby swap entries belong to pages from the same node - even when the mempolicy of the target page is to interleave. And examining a map of nodes corresponding to swap entries on a numa=fake system bears this out. (We could later tweak swap allocation to make it even more likely, but this patch is merely about removing cruft.) So, neither adjust nor increment addr in swapin_readahead, and then shmem_swapin can use it too; the pseudo-vma to pass policy need only be set up once per cluster, and so few fields of pvma are used, let's skip the memset - from shmem_alloc_page also. Signed-off-by: Hugh Dickins <hugh@veritas.com> Acked-by: Rik van Riel <riel@redhat.com> Cc: Andi Kleen <ak@suse.de> Cc: Christoph Lameter <clameter@sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:40 +08:00
/* Create a pseudo vma that just contains the policy */
pvma.vm_start = 0;
/* Bias interleave by inode number to distribute better across nodes */
pvma.vm_pgoff = index + info->vfs_inode.i_ino;
swapin_readahead: excise NUMA bogosity For three years swapin_readahead has been cluttered with fanciful CONFIG_NUMA code, advancing addr, and stepping on to the next vma at the boundary, to line up the mempolicy for each page allocation. It _might_ be a good idea to allocate swap more according to vma layout; but the fact is, that's not how we do it at all, 2.6 even less than 2.4: swap is allocated as needed for pages as they sink to the bottom of the inactive LRUs. Sometimes that may match vma layout, but not so often that it's worth going to these misleading vma->vm_next lengths: rip all that out. Originally I intended to retain the incrementation of addr, but correct its initial value: valid_swaphandles generally supplies an offset below the target addr (this is readaround rather than readahead), but addr has not been adjusted accordingly, so in the interleave case it has usually been allocating the target page from the "wrong" node (though that may not matter very much). But look at the equivalent shmem_swapin code: either by oversight or by design, though it has all the apparatus for choosing a new mempolicy per page, it uses the same idx throughout, choosing the same mempolicy and interleave node for each page of the cluster. Which is actually a much better strategy: each node has its own LRUs and its own kswapd, so if you're betting on any particular relationship between swap and node, the best bet is that nearby swap entries belong to pages from the same node - even when the mempolicy of the target page is to interleave. And examining a map of nodes corresponding to swap entries on a numa=fake system bears this out. (We could later tweak swap allocation to make it even more likely, but this patch is merely about removing cruft.) So, neither adjust nor increment addr in swapin_readahead, and then shmem_swapin can use it too; the pseudo-vma to pass policy need only be set up once per cluster, and so few fields of pvma are used, let's skip the memset - from shmem_alloc_page also. Signed-off-by: Hugh Dickins <hugh@veritas.com> Acked-by: Rik van Riel <riel@redhat.com> Cc: Andi Kleen <ak@suse.de> Cc: Christoph Lameter <clameter@sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-02-05 14:28:40 +08:00
pvma.vm_ops = NULL;
pvma.vm_policy = mpol_shared_policy_lookup(&info->policy, index);
mempolicy: rework mempolicy Reference Counting [yet again] After further discussion with Christoph Lameter, it has become clear that my earlier attempts to clean up the mempolicy reference counting were a bit of overkill in some areas, resulting in superflous ref/unref in what are usually fast paths. In other areas, further inspection reveals that I botched the unref for interleave policies. A separate patch, suitable for upstream/stable trees, fixes up the known errors in the previous attempt to fix reference counting. This patch reworks the memory policy referencing counting and, one hopes, simplifies the code. Maybe I'll get it right this time. See the update to the numa_memory_policy.txt document for a discussion of memory policy reference counting that motivates this patch. Summary: Lookup of mempolicy, based on (vma, address) need only add a reference for shared policy, and we need only unref the policy when finished for shared policies. So, this patch backs out all of the unneeded extra reference counting added by my previous attempt. It then unrefs only shared policies when we're finished with them, using the mpol_cond_put() [conditional put] helper function introduced by this patch. Note that shmem_swapin() calls read_swap_cache_async() with a dummy vma containing just the policy. read_swap_cache_async() can call alloc_page_vma() multiple times, so we can't let alloc_page_vma() unref the shared policy in this case. To avoid this, we make a copy of any non-null shared policy and remove the MPOL_F_SHARED flag from the copy. This copy occurs before reading a page [or multiple pages] from swap, so the overhead should not be an issue here. I introduced a new static inline function "mpol_cond_copy()" to copy the shared policy to an on-stack policy and remove the flags that would require a conditional free. The current implementation of mpol_cond_copy() assumes that the struct mempolicy contains no pointers to dynamically allocated structures that must be duplicated or reference counted during copy. Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:16 +08:00
page = alloc_page_vma(gfp, &pvma, 0);
/* Drop reference taken by mpol_shared_policy_lookup() */
mpol_cond_put(pvma.vm_policy);
return page;
}
#else /* !CONFIG_NUMA */
#ifdef CONFIG_TMPFS
static inline void shmem_show_mpol(struct seq_file *seq, struct mempolicy *mpol)
{
}
#endif /* CONFIG_TMPFS */
static inline struct page *shmem_swapin(swp_entry_t swap, gfp_t gfp,
struct shmem_inode_info *info, pgoff_t index)
{
return swapin_readahead(swap, gfp, NULL, 0);
}
static inline struct page *shmem_alloc_page(gfp_t gfp,
struct shmem_inode_info *info, pgoff_t index)
{
return alloc_page(gfp);
}
#endif /* CONFIG_NUMA */
mempolicy: use struct mempolicy pointer in shmem_sb_info This patch replaces the mempolicy mode, mode_flags, and nodemask in the shmem_sb_info struct with a struct mempolicy pointer, initialized to NULL. This removes dependency on the details of mempolicy from shmem.c and hugetlbfs inode.c and simplifies the interfaces. mpol_parse_str() in mempolicy.c is changed to return, via a pointer to a pointer arg, a struct mempolicy pointer on success. For MPOL_DEFAULT, the returned pointer is NULL. Further, mpol_parse_str() now takes a 'no_context' argument that causes the input nodemask to be stored in the w.user_nodemask of the created mempolicy for use when the mempolicy is installed in a tmpfs inode shared policy tree. At that time, any cpuset contextualization is applied to the original input nodemask. This preserves the previous behavior where the input nodemask was stored in the superblock. We can think of the returned mempolicy as "context free". Because mpol_parse_str() is now calling mpol_new(), we can remove from mpol_to_str() the semantic checks that mpol_new() already performs. Add 'no_context' parameter to mpol_to_str() to specify that it should format the nodemask in w.user_nodemask for 'bind' and 'interleave' policies. Change mpol_shared_policy_init() to take a pointer to a "context free" struct mempolicy and to create a new, "contextualized" mempolicy using the mode, mode_flags and user_nodemask from the input mempolicy. Note: we know that the mempolicy passed to mpol_to_str() or mpol_shared_policy_init() from a tmpfs superblock is "context free". This is currently the only instance thereof. However, if we found more uses for this concept, and introduced any ambiguity as to whether a mempolicy was context free or not, we could add another internal mode flag to identify context free mempolicies. Then, we could remove the 'no_context' argument from mpol_to_str(). Added shmem_get_sbmpol() to return a reference counted superblock mempolicy, if one exists, to pass to mpol_shared_policy_init(). We must add the reference under the sb stat_lock to prevent races with replacement of the mpol by remount. This reference is removed in mpol_shared_policy_init(). [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: yet another build fix] Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:26 +08:00
#if !defined(CONFIG_NUMA) || !defined(CONFIG_TMPFS)
static inline struct mempolicy *shmem_get_sbmpol(struct shmem_sb_info *sbinfo)
{
return NULL;
}
#endif
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
/*
* When a page is moved from swapcache to shmem filecache (either by the
* usual swapin of shmem_getpage_gfp(), or by the less common swapoff of
* shmem_unuse_inode()), it may have been read in earlier from swap, in
* ignorance of the mapping it belongs to. If that mapping has special
* constraints (like the gma500 GEM driver, which requires RAM below 4GB),
* we may need to copy to a suitable page before moving to filecache.
*
* In a future release, this may well be extended to respect cpuset and
* NUMA mempolicy, and applied also to anonymous pages in do_swap_page();
* but for now it is a simple matter of zone.
*/
static bool shmem_should_replace_page(struct page *page, gfp_t gfp)
{
return page_zonenum(page) > gfp_zone(gfp);
}
static int shmem_replace_page(struct page **pagep, gfp_t gfp,
struct shmem_inode_info *info, pgoff_t index)
{
struct page *oldpage, *newpage;
struct address_space *swap_mapping;
pgoff_t swap_index;
int error;
oldpage = *pagep;
swap_index = page_private(oldpage);
swap_mapping = page_mapping(oldpage);
/*
* We have arrived here because our zones are constrained, so don't
* limit chance of success by further cpuset and node constraints.
*/
gfp &= ~GFP_CONSTRAINT_MASK;
newpage = shmem_alloc_page(gfp, info, index);
if (!newpage)
return -ENOMEM;
page_cache_get(newpage);
copy_highpage(newpage, oldpage);
flush_dcache_page(newpage);
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
__set_page_locked(newpage);
SetPageUptodate(newpage);
SetPageSwapBacked(newpage);
set_page_private(newpage, swap_index);
SetPageSwapCache(newpage);
/*
* Our caller will very soon move newpage out of swapcache, but it's
* a nice clean interface for us to replace oldpage by newpage there.
*/
spin_lock_irq(&swap_mapping->tree_lock);
error = shmem_radix_tree_replace(swap_mapping, swap_index, oldpage,
newpage);
if (!error) {
__inc_zone_page_state(newpage, NR_FILE_PAGES);
__dec_zone_page_state(oldpage, NR_FILE_PAGES);
}
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
spin_unlock_irq(&swap_mapping->tree_lock);
if (unlikely(error)) {
/*
* Is this possible? I think not, now that our callers check
* both PageSwapCache and page_private after getting page lock;
* but be defensive. Reverse old to newpage for clear and free.
*/
oldpage = newpage;
} else {
mem_cgroup_migrate(oldpage, newpage, true);
lru_cache_add_anon(newpage);
*pagep = newpage;
}
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
ClearPageSwapCache(oldpage);
set_page_private(oldpage, 0);
unlock_page(oldpage);
page_cache_release(oldpage);
page_cache_release(oldpage);
return error;
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
}
/*
* shmem_getpage_gfp - find page in cache, or get from swap, or allocate
*
* If we allocate a new one we do not mark it dirty. That's up to the
* vm. If we swap it in we mark it dirty since we also free the swap
* entry since a page cannot live in both the swap and page cache
*/
static int shmem_getpage_gfp(struct inode *inode, pgoff_t index,
struct page **pagep, enum sgp_type sgp, gfp_t gfp, int *fault_type)
{
struct address_space *mapping = inode->i_mapping;
struct shmem_inode_info *info;
struct shmem_sb_info *sbinfo;
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
struct mem_cgroup *memcg;
struct page *page;
swp_entry_t swap;
int error;
int once = 0;
int alloced = 0;
if (index > (MAX_LFS_FILESIZE >> PAGE_CACHE_SHIFT))
return -EFBIG;
repeat:
swap.val = 0;
2014-04-04 05:47:46 +08:00
page = find_lock_entry(mapping, index);
if (radix_tree_exceptional_entry(page)) {
swap = radix_to_swp_entry(page);
page = NULL;
}
if (sgp != SGP_WRITE && sgp != SGP_FALLOC &&
((loff_t)index << PAGE_CACHE_SHIFT) >= i_size_read(inode)) {
error = -EINVAL;
goto failed;
}
if (page && sgp == SGP_WRITE)
mark_page_accessed(page);
/* fallocated page? */
if (page && !PageUptodate(page)) {
if (sgp != SGP_READ)
goto clear;
unlock_page(page);
page_cache_release(page);
page = NULL;
}
if (page || (sgp == SGP_READ && !swap.val)) {
*pagep = page;
return 0;
}
/*
* Fast cache lookup did not find it:
* bring it back from swap or allocate.
*/
info = SHMEM_I(inode);
sbinfo = SHMEM_SB(inode->i_sb);
if (swap.val) {
/* Look it up and read it in.. */
page = lookup_swap_cache(swap);
if (!page) {
/* here we actually do the io */
if (fault_type)
*fault_type |= VM_FAULT_MAJOR;
page = shmem_swapin(swap, gfp, info, index);
if (!page) {
error = -ENOMEM;
goto failed;
}
}
/* We have to do this with page locked to prevent races */
lock_page(page);
if (!PageSwapCache(page) || page_private(page) != swap.val ||
shmem: fix negative rss in memcg memory.stat When adding the page_private checks before calling shmem_replace_page(), I did realize that there is a further race, but thought it too unlikely to need a hurried fix. But independently I've been chasing why a mem cgroup's memory.stat sometimes shows negative rss after all tasks have gone: I expected it to be a stats gathering bug, but actually it's shmem swapping's fault. It's an old surprise, that when you lock_page(lookup_swap_cache(swap)), the page may have been removed from swapcache before getting the lock; or it may have been freed and reused and be back in swapcache; and it can even be using the same swap location as before (page_private same). The swapoff case is already secure against this (swap cannot be reused until the whole area has been swapped off, and a new swapped on); and shmem_getpage_gfp() is protected by shmem_add_to_page_cache()'s check for the expected radix_tree entry - but a little too late. By that time, we might have already decided to shmem_replace_page(): I don't know of a problem from that, but I'd feel more at ease not to do so spuriously. And we have already done mem_cgroup_cache_charge(), on perhaps the wrong mem cgroup: and this charge is not then undone on the error path, because PageSwapCache ends up preventing that. It's this last case which causes the occasional negative rss in memory.stat: the page is charged here as cache, but (sometimes) found to be anon when eventually it's uncharged - and in between, it's an undeserved charge on the wrong memcg. Fix this by adding an earlier check on the radix_tree entry: it's inelegant to descend the tree twice, but swapping is not the fast path, and a better solution would need a pair (try+commit) of memcg calls, and a rework of shmem_replace_page() to keep out of the swapcache. We can use the added shmem_confirm_swap() function to replace the find_get_page+page_cache_release we were already doing on the error path. And add a comment on that -EEXIST: it seems a peculiar errno to be using, but originates from its use in radix_tree_insert(). [It can be surprising to see positive rss left in a memcg's memory.stat after all tasks have gone, since it is supposed to count anonymous but not shmem. Aside from sharing anon pages via fork with a task in some other memcg, it often happens after swapping: because a swap page can't be freed while under writeback, nor while locked. So it's not an error, and these residual pages are easily freed once pressure demands.] Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Michal Hocko <mhocko@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-07-12 05:02:47 +08:00
!shmem_confirm_swap(mapping, index, swap)) {
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
error = -EEXIST; /* try again */
shmem: fix negative rss in memcg memory.stat When adding the page_private checks before calling shmem_replace_page(), I did realize that there is a further race, but thought it too unlikely to need a hurried fix. But independently I've been chasing why a mem cgroup's memory.stat sometimes shows negative rss after all tasks have gone: I expected it to be a stats gathering bug, but actually it's shmem swapping's fault. It's an old surprise, that when you lock_page(lookup_swap_cache(swap)), the page may have been removed from swapcache before getting the lock; or it may have been freed and reused and be back in swapcache; and it can even be using the same swap location as before (page_private same). The swapoff case is already secure against this (swap cannot be reused until the whole area has been swapped off, and a new swapped on); and shmem_getpage_gfp() is protected by shmem_add_to_page_cache()'s check for the expected radix_tree entry - but a little too late. By that time, we might have already decided to shmem_replace_page(): I don't know of a problem from that, but I'd feel more at ease not to do so spuriously. And we have already done mem_cgroup_cache_charge(), on perhaps the wrong mem cgroup: and this charge is not then undone on the error path, because PageSwapCache ends up preventing that. It's this last case which causes the occasional negative rss in memory.stat: the page is charged here as cache, but (sometimes) found to be anon when eventually it's uncharged - and in between, it's an undeserved charge on the wrong memcg. Fix this by adding an earlier check on the radix_tree entry: it's inelegant to descend the tree twice, but swapping is not the fast path, and a better solution would need a pair (try+commit) of memcg calls, and a rework of shmem_replace_page() to keep out of the swapcache. We can use the added shmem_confirm_swap() function to replace the find_get_page+page_cache_release we were already doing on the error path. And add a comment on that -EEXIST: it seems a peculiar errno to be using, but originates from its use in radix_tree_insert(). [It can be surprising to see positive rss left in a memcg's memory.stat after all tasks have gone, since it is supposed to count anonymous but not shmem. Aside from sharing anon pages via fork with a task in some other memcg, it often happens after swapping: because a swap page can't be freed while under writeback, nor while locked. So it's not an error, and these residual pages are easily freed once pressure demands.] Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Michal Hocko <mhocko@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-07-12 05:02:47 +08:00
goto unlock;
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
}
if (!PageUptodate(page)) {
error = -EIO;
goto failed;
}
wait_on_page_writeback(page);
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
if (shmem_should_replace_page(page, gfp)) {
error = shmem_replace_page(&page, gfp, info, index);
if (error)
goto failed;
}
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
error = mem_cgroup_try_charge(page, current->mm, gfp, &memcg);
shmem: fix negative rss in memcg memory.stat When adding the page_private checks before calling shmem_replace_page(), I did realize that there is a further race, but thought it too unlikely to need a hurried fix. But independently I've been chasing why a mem cgroup's memory.stat sometimes shows negative rss after all tasks have gone: I expected it to be a stats gathering bug, but actually it's shmem swapping's fault. It's an old surprise, that when you lock_page(lookup_swap_cache(swap)), the page may have been removed from swapcache before getting the lock; or it may have been freed and reused and be back in swapcache; and it can even be using the same swap location as before (page_private same). The swapoff case is already secure against this (swap cannot be reused until the whole area has been swapped off, and a new swapped on); and shmem_getpage_gfp() is protected by shmem_add_to_page_cache()'s check for the expected radix_tree entry - but a little too late. By that time, we might have already decided to shmem_replace_page(): I don't know of a problem from that, but I'd feel more at ease not to do so spuriously. And we have already done mem_cgroup_cache_charge(), on perhaps the wrong mem cgroup: and this charge is not then undone on the error path, because PageSwapCache ends up preventing that. It's this last case which causes the occasional negative rss in memory.stat: the page is charged here as cache, but (sometimes) found to be anon when eventually it's uncharged - and in between, it's an undeserved charge on the wrong memcg. Fix this by adding an earlier check on the radix_tree entry: it's inelegant to descend the tree twice, but swapping is not the fast path, and a better solution would need a pair (try+commit) of memcg calls, and a rework of shmem_replace_page() to keep out of the swapcache. We can use the added shmem_confirm_swap() function to replace the find_get_page+page_cache_release we were already doing on the error path. And add a comment on that -EEXIST: it seems a peculiar errno to be using, but originates from its use in radix_tree_insert(). [It can be surprising to see positive rss left in a memcg's memory.stat after all tasks have gone, since it is supposed to count anonymous but not shmem. Aside from sharing anon pages via fork with a task in some other memcg, it often happens after swapping: because a swap page can't be freed while under writeback, nor while locked. So it's not an error, and these residual pages are easily freed once pressure demands.] Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Michal Hocko <mhocko@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-07-12 05:02:47 +08:00
if (!error) {
error = shmem_add_to_page_cache(page, mapping, index,
swp_to_radix_entry(swap));
/*
* We already confirmed swap under page lock, and make
* no memory allocation here, so usually no possibility
* of error; but free_swap_and_cache() only trylocks a
* page, so it is just possible that the entry has been
* truncated or holepunched since swap was confirmed.
* shmem_undo_range() will have done some of the
* unaccounting, now delete_from_swap_cache() will do
* the rest.
* Reset swap.val? No, leave it so "failed" goes back to
* "repeat": reading a hole and writing should succeed.
*/
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
if (error) {
mem_cgroup_cancel_charge(page, memcg);
delete_from_swap_cache(page);
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
}
shmem: fix negative rss in memcg memory.stat When adding the page_private checks before calling shmem_replace_page(), I did realize that there is a further race, but thought it too unlikely to need a hurried fix. But independently I've been chasing why a mem cgroup's memory.stat sometimes shows negative rss after all tasks have gone: I expected it to be a stats gathering bug, but actually it's shmem swapping's fault. It's an old surprise, that when you lock_page(lookup_swap_cache(swap)), the page may have been removed from swapcache before getting the lock; or it may have been freed and reused and be back in swapcache; and it can even be using the same swap location as before (page_private same). The swapoff case is already secure against this (swap cannot be reused until the whole area has been swapped off, and a new swapped on); and shmem_getpage_gfp() is protected by shmem_add_to_page_cache()'s check for the expected radix_tree entry - but a little too late. By that time, we might have already decided to shmem_replace_page(): I don't know of a problem from that, but I'd feel more at ease not to do so spuriously. And we have already done mem_cgroup_cache_charge(), on perhaps the wrong mem cgroup: and this charge is not then undone on the error path, because PageSwapCache ends up preventing that. It's this last case which causes the occasional negative rss in memory.stat: the page is charged here as cache, but (sometimes) found to be anon when eventually it's uncharged - and in between, it's an undeserved charge on the wrong memcg. Fix this by adding an earlier check on the radix_tree entry: it's inelegant to descend the tree twice, but swapping is not the fast path, and a better solution would need a pair (try+commit) of memcg calls, and a rework of shmem_replace_page() to keep out of the swapcache. We can use the added shmem_confirm_swap() function to replace the find_get_page+page_cache_release we were already doing on the error path. And add a comment on that -EEXIST: it seems a peculiar errno to be using, but originates from its use in radix_tree_insert(). [It can be surprising to see positive rss left in a memcg's memory.stat after all tasks have gone, since it is supposed to count anonymous but not shmem. Aside from sharing anon pages via fork with a task in some other memcg, it often happens after swapping: because a swap page can't be freed while under writeback, nor while locked. So it's not an error, and these residual pages are easily freed once pressure demands.] Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Michal Hocko <mhocko@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-07-12 05:02:47 +08:00
}
if (error)
goto failed;
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
mem_cgroup_commit_charge(page, memcg, true);
spin_lock(&info->lock);
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
info->swapped--;
shmem_recalc_inode(inode);
spin_unlock(&info->lock);
if (sgp == SGP_WRITE)
mark_page_accessed(page);
delete_from_swap_cache(page);
set_page_dirty(page);
swap_free(swap);
} else {
if (shmem_acct_block(info->flags)) {
error = -ENOSPC;
goto failed;
}
if (sbinfo->max_blocks) {
if (percpu_counter_compare(&sbinfo->used_blocks,
sbinfo->max_blocks) >= 0) {
error = -ENOSPC;
goto unacct;
}
percpu_counter_inc(&sbinfo->used_blocks);
}
page = shmem_alloc_page(gfp, info, index);
if (!page) {
error = -ENOMEM;
goto decused;
}
__SetPageSwapBacked(page);
__set_page_locked(page);
if (sgp == SGP_WRITE)
__SetPageReferenced(page);
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
error = mem_cgroup_try_charge(page, current->mm, gfp, &memcg);
if (error)
goto decused;
lib/radix-tree.c: make radix_tree_node_alloc() work correctly within interrupt With users of radix_tree_preload() run from interrupt (block/blk-ioc.c is one such possible user), the following race can happen: radix_tree_preload() ... radix_tree_insert() radix_tree_node_alloc() if (rtp->nr) { ret = rtp->nodes[rtp->nr - 1]; <interrupt> ... radix_tree_preload() ... radix_tree_insert() radix_tree_node_alloc() if (rtp->nr) { ret = rtp->nodes[rtp->nr - 1]; And we give out one radix tree node twice. That clearly results in radix tree corruption with different results (usually OOPS) depending on which two users of radix tree race. We fix the problem by making radix_tree_node_alloc() always allocate fresh radix tree nodes when in interrupt. Using preloading when in interrupt doesn't make sense since all the allocations have to be atomic anyway and we cannot steal nodes from process-context users because some users rely on radix_tree_insert() succeeding after radix_tree_preload(). in_interrupt() check is somewhat ugly but we cannot simply key off passed gfp_mask as that is acquired from root_gfp_mask() and thus the same for all preload users. Another part of the fix is to avoid node preallocation in radix_tree_preload() when passed gfp_mask doesn't allow waiting. Again, preallocation in such case doesn't make sense and when preallocation would happen in interrupt we could possibly leak some allocated nodes. However, some users of radix_tree_preload() require following radix_tree_insert() to succeed. To avoid unexpected effects for these users, radix_tree_preload() only warns if passed gfp mask doesn't allow waiting and we provide a new function radix_tree_maybe_preload() for those users which get different gfp mask from different call sites and which are prepared to handle radix_tree_insert() failure. Signed-off-by: Jan Kara <jack@suse.cz> Cc: Jens Axboe <jaxboe@fusionio.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-09-12 05:26:05 +08:00
error = radix_tree_maybe_preload(gfp & GFP_RECLAIM_MASK);
if (!error) {
error = shmem_add_to_page_cache(page, mapping, index,
NULL);
radix_tree_preload_end();
}
if (error) {
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
mem_cgroup_cancel_charge(page, memcg);
goto decused;
}
mm: memcontrol: rewrite charge API These patches rework memcg charge lifetime to integrate more naturally with the lifetime of user pages. This drastically simplifies the code and reduces charging and uncharging overhead. The most expensive part of charging and uncharging is the page_cgroup bit spinlock, which is removed entirely after this series. Here are the top-10 profile entries of a stress test that reads a 128G sparse file on a freshly booted box, without even a dedicated cgroup (i.e. executing in the root memcg). Before: 15.36% cat [kernel.kallsyms] [k] copy_user_generic_string 13.31% cat [kernel.kallsyms] [k] memset 11.48% cat [kernel.kallsyms] [k] do_mpage_readpage 4.23% cat [kernel.kallsyms] [k] get_page_from_freelist 2.38% cat [kernel.kallsyms] [k] put_page 2.32% cat [kernel.kallsyms] [k] __mem_cgroup_commit_charge 2.18% kswapd0 [kernel.kallsyms] [k] __mem_cgroup_uncharge_common 1.92% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.86% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.62% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn After: 15.67% cat [kernel.kallsyms] [k] copy_user_generic_string 13.48% cat [kernel.kallsyms] [k] memset 11.42% cat [kernel.kallsyms] [k] do_mpage_readpage 3.98% cat [kernel.kallsyms] [k] get_page_from_freelist 2.46% cat [kernel.kallsyms] [k] put_page 2.13% kswapd0 [kernel.kallsyms] [k] shrink_page_list 1.88% cat [kernel.kallsyms] [k] __radix_tree_lookup 1.67% cat [kernel.kallsyms] [k] __pagevec_lru_add_fn 1.39% kswapd0 [kernel.kallsyms] [k] free_pcppages_bulk 1.30% cat [kernel.kallsyms] [k] kfree As you can see, the memcg footprint has shrunk quite a bit. text data bss dec hex filename 37970 9892 400 48262 bc86 mm/memcontrol.o.old 35239 9892 400 45531 b1db mm/memcontrol.o This patch (of 4): The memcg charge API charges pages before they are rmapped - i.e. have an actual "type" - and so every callsite needs its own set of charge and uncharge functions to know what type is being operated on. Worse, uncharge has to happen from a context that is still type-specific, rather than at the end of the page's lifetime with exclusive access, and so requires a lot of synchronization. Rewrite the charge API to provide a generic set of try_charge(), commit_charge() and cancel_charge() transaction operations, much like what's currently done for swap-in: mem_cgroup_try_charge() attempts to reserve a charge, reclaiming pages from the memcg if necessary. mem_cgroup_commit_charge() commits the page to the charge once it has a valid page->mapping and PageAnon() reliably tells the type. mem_cgroup_cancel_charge() aborts the transaction. This reduces the charge API and enables subsequent patches to drastically simplify uncharging. As pages need to be committed after rmap is established but before they are added to the LRU, page_add_new_anon_rmap() must stop doing LRU additions again. Revive lru_cache_add_active_or_unevictable(). [hughd@google.com: fix shmem_unuse] [hughd@google.com: Add comments on the private use of -EAGAIN] Signed-off-by: Johannes Weiner <hannes@cmpxchg.org> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> Cc: Vladimir Davydov <vdavydov@parallels.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:19:20 +08:00
mem_cgroup_commit_charge(page, memcg, false);
lru_cache_add_anon(page);
spin_lock(&info->lock);
info->alloced++;
inode->i_blocks += BLOCKS_PER_PAGE;
shmem_recalc_inode(inode);
spin_unlock(&info->lock);
alloced = true;
/*
* Let SGP_FALLOC use the SGP_WRITE optimization on a new page.
*/
if (sgp == SGP_FALLOC)
sgp = SGP_WRITE;
clear:
/*
* Let SGP_WRITE caller clear ends if write does not fill page;
* but SGP_FALLOC on a page fallocated earlier must initialize
* it now, lest undo on failure cancel our earlier guarantee.
*/
if (sgp != SGP_WRITE) {
clear_highpage(page);
flush_dcache_page(page);
SetPageUptodate(page);
}
if (sgp == SGP_DIRTY)
set_page_dirty(page);
}
shmem: replace page if mapping excludes its zone The GMA500 GPU driver uses GEM shmem objects, but with a new twist: the backing RAM has to be below 4GB. Not a problem while the boards supported only 4GB: but now Intel's D2700MUD boards support 8GB, and their GMA3600 is managed by the GMA500 driver. shmem/tmpfs has never pretended to support hardware restrictions on the backing memory, but it might have appeared to do so before v3.1, and even now it works fine until a page is swapped out then back in. When read_cache_page_gfp() supplied a freshly allocated page for copy, that compensated for whatever choice might have been made by earlier swapin readahead; but swapoff was likely to destroy the illusion. We'd like to continue to support GMA500, so now add a new shmem_should_replace_page() check on the zone when about to move a page from swapcache to filecache (in swapin and swapoff cases), with shmem_replace_page() to allocate and substitute a suitable page (given gma500/gem.c's mapping_set_gfp_mask GFP_KERNEL | __GFP_DMA32). This does involve a minor extension to mem_cgroup_replace_page_cache() (the page may or may not have already been charged); and I've removed a comment and call to mem_cgroup_uncharge_cache_page(), which in fact is always a no-op while PageSwapCache. Also removed optimization of an unlikely path in shmem_getpage_gfp(), now that we need to check PageSwapCache more carefully (a racing caller might already have made the copy). And at one point shmem_unuse_inode() needs to use the hitherto private page_swapcount(), to guard against racing with inode eviction. It would make sense to extend shmem_should_replace_page(), to cover cpuset and NUMA mempolicy restrictions too, but set that aside for now: needs a cleanup of shmem mempolicy handling, and more testing, and ought to handle swap faults in do_swap_page() as well as shmem. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Alan Cox <alan@lxorguk.ukuu.org.uk> Cc: Stephane Marchesin <marcheu@chromium.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Dave Airlie <airlied@gmail.com> Cc: Daniel Vetter <daniel@ffwll.ch> Cc: Rob Clark <rob.clark@linaro.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:38 +08:00
/* Perhaps the file has been truncated since we checked */
if (sgp != SGP_WRITE && sgp != SGP_FALLOC &&
((loff_t)index << PAGE_CACHE_SHIFT) >= i_size_read(inode)) {
error = -EINVAL;
if (alloced)
goto trunc;
else
goto failed;
}
*pagep = page;
return 0;
/*
* Error recovery.
*/
trunc:
info = SHMEM_I(inode);
ClearPageDirty(page);
delete_from_page_cache(page);
spin_lock(&info->lock);
info->alloced--;
inode->i_blocks -= BLOCKS_PER_PAGE;
spin_unlock(&info->lock);
decused:
sbinfo = SHMEM_SB(inode->i_sb);
if (sbinfo->max_blocks)
percpu_counter_add(&sbinfo->used_blocks, -1);
unacct:
shmem_unacct_blocks(info->flags, 1);
failed:
shmem: fix negative rss in memcg memory.stat When adding the page_private checks before calling shmem_replace_page(), I did realize that there is a further race, but thought it too unlikely to need a hurried fix. But independently I've been chasing why a mem cgroup's memory.stat sometimes shows negative rss after all tasks have gone: I expected it to be a stats gathering bug, but actually it's shmem swapping's fault. It's an old surprise, that when you lock_page(lookup_swap_cache(swap)), the page may have been removed from swapcache before getting the lock; or it may have been freed and reused and be back in swapcache; and it can even be using the same swap location as before (page_private same). The swapoff case is already secure against this (swap cannot be reused until the whole area has been swapped off, and a new swapped on); and shmem_getpage_gfp() is protected by shmem_add_to_page_cache()'s check for the expected radix_tree entry - but a little too late. By that time, we might have already decided to shmem_replace_page(): I don't know of a problem from that, but I'd feel more at ease not to do so spuriously. And we have already done mem_cgroup_cache_charge(), on perhaps the wrong mem cgroup: and this charge is not then undone on the error path, because PageSwapCache ends up preventing that. It's this last case which causes the occasional negative rss in memory.stat: the page is charged here as cache, but (sometimes) found to be anon when eventually it's uncharged - and in between, it's an undeserved charge on the wrong memcg. Fix this by adding an earlier check on the radix_tree entry: it's inelegant to descend the tree twice, but swapping is not the fast path, and a better solution would need a pair (try+commit) of memcg calls, and a rework of shmem_replace_page() to keep out of the swapcache. We can use the added shmem_confirm_swap() function to replace the find_get_page+page_cache_release we were already doing on the error path. And add a comment on that -EEXIST: it seems a peculiar errno to be using, but originates from its use in radix_tree_insert(). [It can be surprising to see positive rss left in a memcg's memory.stat after all tasks have gone, since it is supposed to count anonymous but not shmem. Aside from sharing anon pages via fork with a task in some other memcg, it often happens after swapping: because a swap page can't be freed while under writeback, nor while locked. So it's not an error, and these residual pages are easily freed once pressure demands.] Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Michal Hocko <mhocko@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-07-12 05:02:47 +08:00
if (swap.val && error != -EINVAL &&
!shmem_confirm_swap(mapping, index, swap))
error = -EEXIST;
unlock:
if (page) {
unlock_page(page);
page_cache_release(page);
}
if (error == -ENOSPC && !once++) {
info = SHMEM_I(inode);
spin_lock(&info->lock);
shmem_recalc_inode(inode);
spin_unlock(&info->lock);
goto repeat;
}
shmem: fix negative rss in memcg memory.stat When adding the page_private checks before calling shmem_replace_page(), I did realize that there is a further race, but thought it too unlikely to need a hurried fix. But independently I've been chasing why a mem cgroup's memory.stat sometimes shows negative rss after all tasks have gone: I expected it to be a stats gathering bug, but actually it's shmem swapping's fault. It's an old surprise, that when you lock_page(lookup_swap_cache(swap)), the page may have been removed from swapcache before getting the lock; or it may have been freed and reused and be back in swapcache; and it can even be using the same swap location as before (page_private same). The swapoff case is already secure against this (swap cannot be reused until the whole area has been swapped off, and a new swapped on); and shmem_getpage_gfp() is protected by shmem_add_to_page_cache()'s check for the expected radix_tree entry - but a little too late. By that time, we might have already decided to shmem_replace_page(): I don't know of a problem from that, but I'd feel more at ease not to do so spuriously. And we have already done mem_cgroup_cache_charge(), on perhaps the wrong mem cgroup: and this charge is not then undone on the error path, because PageSwapCache ends up preventing that. It's this last case which causes the occasional negative rss in memory.stat: the page is charged here as cache, but (sometimes) found to be anon when eventually it's uncharged - and in between, it's an undeserved charge on the wrong memcg. Fix this by adding an earlier check on the radix_tree entry: it's inelegant to descend the tree twice, but swapping is not the fast path, and a better solution would need a pair (try+commit) of memcg calls, and a rework of shmem_replace_page() to keep out of the swapcache. We can use the added shmem_confirm_swap() function to replace the find_get_page+page_cache_release we were already doing on the error path. And add a comment on that -EEXIST: it seems a peculiar errno to be using, but originates from its use in radix_tree_insert(). [It can be surprising to see positive rss left in a memcg's memory.stat after all tasks have gone, since it is supposed to count anonymous but not shmem. Aside from sharing anon pages via fork with a task in some other memcg, it often happens after swapping: because a swap page can't be freed while under writeback, nor while locked. So it's not an error, and these residual pages are easily freed once pressure demands.] Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Johannes Weiner <hannes@cmpxchg.org> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: Michal Hocko <mhocko@suse.cz> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-07-12 05:02:47 +08:00
if (error == -EEXIST) /* from above or from radix_tree_insert */
goto repeat;
return error;
}
static int shmem_fault(struct vm_area_struct *vma, struct vm_fault *vmf)
{
struct inode *inode = file_inode(vma->vm_file);
int error;
int ret = VM_FAULT_LOCKED;
/*
* Trinity finds that probing a hole which tmpfs is punching can
* prevent the hole-punch from ever completing: which in turn
* locks writers out with its hold on i_mutex. So refrain from
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
* faulting pages into the hole while it's being punched. Although
* shmem_undo_range() does remove the additions, it may be unable to
* keep up, as each new page needs its own unmap_mapping_range() call,
* and the i_mmap tree grows ever slower to scan if new vmas are added.
*
* It does not matter if we sometimes reach this check just before the
* hole-punch begins, so that one fault then races with the punch:
* we just need to make racing faults a rare case.
*
* The implementation below would be much simpler if we just used a
* standard mutex or completion: but we cannot take i_mutex in fault,
* and bloating every shmem inode for this unlikely case would be sad.
*/
if (unlikely(inode->i_private)) {
struct shmem_falloc *shmem_falloc;
spin_lock(&inode->i_lock);
shmem_falloc = inode->i_private;
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
if (shmem_falloc &&
shmem_falloc->waitq &&
vmf->pgoff >= shmem_falloc->start &&
vmf->pgoff < shmem_falloc->next) {
wait_queue_head_t *shmem_falloc_waitq;
DEFINE_WAIT(shmem_fault_wait);
ret = VM_FAULT_NOPAGE;
if ((vmf->flags & FAULT_FLAG_ALLOW_RETRY) &&
!(vmf->flags & FAULT_FLAG_RETRY_NOWAIT)) {
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
/* It's polite to up mmap_sem if we can */
up_read(&vma->vm_mm->mmap_sem);
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
ret = VM_FAULT_RETRY;
}
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
shmem_falloc_waitq = shmem_falloc->waitq;
prepare_to_wait(shmem_falloc_waitq, &shmem_fault_wait,
TASK_UNINTERRUPTIBLE);
spin_unlock(&inode->i_lock);
schedule();
/*
* shmem_falloc_waitq points into the shmem_fallocate()
* stack of the hole-punching task: shmem_falloc_waitq
* is usually invalid by the time we reach here, but
* finish_wait() does not dereference it in that case;
* though i_lock needed lest racing with wake_up_all().
*/
spin_lock(&inode->i_lock);
finish_wait(shmem_falloc_waitq, &shmem_fault_wait);
spin_unlock(&inode->i_lock);
return ret;
}
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
spin_unlock(&inode->i_lock);
}
error = shmem_getpage(inode, vmf->pgoff, &vmf->page, SGP_CACHE, &ret);
if (error)
return ((error == -ENOMEM) ? VM_FAULT_OOM : VM_FAULT_SIGBUS);
memcg: add the pagefault count into memcg stats Two new stats in per-memcg memory.stat which tracks the number of page faults and number of major page faults. "pgfault" "pgmajfault" They are different from "pgpgin"/"pgpgout" stat which count number of pages charged/discharged to the cgroup and have no meaning of reading/ writing page to disk. It is valuable to track the two stats for both measuring application's performance as well as the efficiency of the kernel page reclaim path. Counting pagefaults per process is useful, but we also need the aggregated value since processes are monitored and controlled in cgroup basis in memcg. Functional test: check the total number of pgfault/pgmajfault of all memcgs and compare with global vmstat value: $ cat /proc/vmstat | grep fault pgfault 1070751 pgmajfault 553 $ cat /dev/cgroup/memory.stat | grep fault pgfault 1071138 pgmajfault 553 total_pgfault 1071142 total_pgmajfault 553 $ cat /dev/cgroup/A/memory.stat | grep fault pgfault 199 pgmajfault 0 total_pgfault 199 total_pgmajfault 0 Performance test: run page fault test(pft) wit 16 thread on faulting in 15G anon pages in 16G container. There is no regression noticed on the "flt/cpu/s" Sample output from pft: TAG pft:anon-sys-default: Gb Thr CLine User System Wall flt/cpu/s fault/wsec 15 16 1 0.67s 233.41s 14.76s 16798.546 266356.260 +-------------------------------------------------------------------------+ N Min Max Median Avg Stddev x 10 16682.962 17344.027 16913.524 16928.812 166.5362 + 10 16695.568 16923.896 16820.604 16824.652 84.816568 No difference proven at 95.0% confidence [akpm@linux-foundation.org: fix build] [hughd@google.com: shmem fix] Signed-off-by: Ying Han <yinghan@google.com> Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Reviewed-by: Minchan Kim <minchan.kim@gmail.com> Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Acked-by: Balbir Singh <balbir@linux.vnet.ibm.com> Signed-off-by: Hugh Dickins <hughd@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-27 07:25:38 +08:00
if (ret & VM_FAULT_MAJOR) {
count_vm_event(PGMAJFAULT);
mem_cgroup_count_vm_event(vma->vm_mm, PGMAJFAULT);
}
return ret;
}
#ifdef CONFIG_NUMA
static int shmem_set_policy(struct vm_area_struct *vma, struct mempolicy *mpol)
{
struct inode *inode = file_inode(vma->vm_file);
return mpol_set_shared_policy(&SHMEM_I(inode)->policy, vma, mpol);
}
static struct mempolicy *shmem_get_policy(struct vm_area_struct *vma,
unsigned long addr)
{
struct inode *inode = file_inode(vma->vm_file);
pgoff_t index;
index = ((addr - vma->vm_start) >> PAGE_SHIFT) + vma->vm_pgoff;
return mpol_shared_policy_lookup(&SHMEM_I(inode)->policy, index);
}
#endif
int shmem_lock(struct file *file, int lock, struct user_struct *user)
{
struct inode *inode = file_inode(file);
struct shmem_inode_info *info = SHMEM_I(inode);
int retval = -ENOMEM;
spin_lock(&info->lock);
if (lock && !(info->flags & VM_LOCKED)) {
if (!user_shm_lock(inode->i_size, user))
goto out_nomem;
info->flags |= VM_LOCKED;
mapping_set_unevictable(file->f_mapping);
}
if (!lock && (info->flags & VM_LOCKED) && user) {
user_shm_unlock(inode->i_size, user);
info->flags &= ~VM_LOCKED;
mapping_clear_unevictable(file->f_mapping);
}
retval = 0;
out_nomem:
spin_unlock(&info->lock);
return retval;
}
static int shmem_mmap(struct file *file, struct vm_area_struct *vma)
{
file_accessed(file);
vma->vm_ops = &shmem_vm_ops;
return 0;
}
static struct inode *shmem_get_inode(struct super_block *sb, const struct inode *dir,
umode_t mode, dev_t dev, unsigned long flags)
{
struct inode *inode;
struct shmem_inode_info *info;
struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
if (shmem_reserve_inode(sb))
return NULL;
inode = new_inode(sb);
if (inode) {
inode->i_ino = get_next_ino();
inode_init_owner(inode, dir, mode);
inode->i_blocks = 0;
inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME;
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
inode->i_generation = get_seconds();
info = SHMEM_I(inode);
memset(info, 0, (char *)inode - (char *)info);
spin_lock_init(&info->lock);
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
info->seals = F_SEAL_SEAL;
info->flags = flags & VM_NORESERVE;
INIT_LIST_HEAD(&info->swaplist);
simple_xattrs_init(&info->xattrs);
cache_no_acl(inode);
switch (mode & S_IFMT) {
default:
inode->i_op = &shmem_special_inode_operations;
init_special_inode(inode, mode, dev);
break;
case S_IFREG:
inode->i_mapping->a_ops = &shmem_aops;
inode->i_op = &shmem_inode_operations;
inode->i_fop = &shmem_file_operations;
mempolicy: use struct mempolicy pointer in shmem_sb_info This patch replaces the mempolicy mode, mode_flags, and nodemask in the shmem_sb_info struct with a struct mempolicy pointer, initialized to NULL. This removes dependency on the details of mempolicy from shmem.c and hugetlbfs inode.c and simplifies the interfaces. mpol_parse_str() in mempolicy.c is changed to return, via a pointer to a pointer arg, a struct mempolicy pointer on success. For MPOL_DEFAULT, the returned pointer is NULL. Further, mpol_parse_str() now takes a 'no_context' argument that causes the input nodemask to be stored in the w.user_nodemask of the created mempolicy for use when the mempolicy is installed in a tmpfs inode shared policy tree. At that time, any cpuset contextualization is applied to the original input nodemask. This preserves the previous behavior where the input nodemask was stored in the superblock. We can think of the returned mempolicy as "context free". Because mpol_parse_str() is now calling mpol_new(), we can remove from mpol_to_str() the semantic checks that mpol_new() already performs. Add 'no_context' parameter to mpol_to_str() to specify that it should format the nodemask in w.user_nodemask for 'bind' and 'interleave' policies. Change mpol_shared_policy_init() to take a pointer to a "context free" struct mempolicy and to create a new, "contextualized" mempolicy using the mode, mode_flags and user_nodemask from the input mempolicy. Note: we know that the mempolicy passed to mpol_to_str() or mpol_shared_policy_init() from a tmpfs superblock is "context free". This is currently the only instance thereof. However, if we found more uses for this concept, and introduced any ambiguity as to whether a mempolicy was context free or not, we could add another internal mode flag to identify context free mempolicies. Then, we could remove the 'no_context' argument from mpol_to_str(). Added shmem_get_sbmpol() to return a reference counted superblock mempolicy, if one exists, to pass to mpol_shared_policy_init(). We must add the reference under the sb stat_lock to prevent races with replacement of the mpol by remount. This reference is removed in mpol_shared_policy_init(). [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: yet another build fix] Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:26 +08:00
mpol_shared_policy_init(&info->policy,
shmem_get_sbmpol(sbinfo));
break;
case S_IFDIR:
inc_nlink(inode);
/* Some things misbehave if size == 0 on a directory */
inode->i_size = 2 * BOGO_DIRENT_SIZE;
inode->i_op = &shmem_dir_inode_operations;
inode->i_fop = &simple_dir_operations;
break;
case S_IFLNK:
/*
* Must not load anything in the rbtree,
* mpol_free_shared_policy will not be called.
*/
mempolicy: use struct mempolicy pointer in shmem_sb_info This patch replaces the mempolicy mode, mode_flags, and nodemask in the shmem_sb_info struct with a struct mempolicy pointer, initialized to NULL. This removes dependency on the details of mempolicy from shmem.c and hugetlbfs inode.c and simplifies the interfaces. mpol_parse_str() in mempolicy.c is changed to return, via a pointer to a pointer arg, a struct mempolicy pointer on success. For MPOL_DEFAULT, the returned pointer is NULL. Further, mpol_parse_str() now takes a 'no_context' argument that causes the input nodemask to be stored in the w.user_nodemask of the created mempolicy for use when the mempolicy is installed in a tmpfs inode shared policy tree. At that time, any cpuset contextualization is applied to the original input nodemask. This preserves the previous behavior where the input nodemask was stored in the superblock. We can think of the returned mempolicy as "context free". Because mpol_parse_str() is now calling mpol_new(), we can remove from mpol_to_str() the semantic checks that mpol_new() already performs. Add 'no_context' parameter to mpol_to_str() to specify that it should format the nodemask in w.user_nodemask for 'bind' and 'interleave' policies. Change mpol_shared_policy_init() to take a pointer to a "context free" struct mempolicy and to create a new, "contextualized" mempolicy using the mode, mode_flags and user_nodemask from the input mempolicy. Note: we know that the mempolicy passed to mpol_to_str() or mpol_shared_policy_init() from a tmpfs superblock is "context free". This is currently the only instance thereof. However, if we found more uses for this concept, and introduced any ambiguity as to whether a mempolicy was context free or not, we could add another internal mode flag to identify context free mempolicies. Then, we could remove the 'no_context' argument from mpol_to_str(). Added shmem_get_sbmpol() to return a reference counted superblock mempolicy, if one exists, to pass to mpol_shared_policy_init(). We must add the reference under the sb stat_lock to prevent races with replacement of the mpol by remount. This reference is removed in mpol_shared_policy_init(). [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: yet another build fix] Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:26 +08:00
mpol_shared_policy_init(&info->policy, NULL);
break;
}
} else
shmem_free_inode(sb);
return inode;
}
2014-04-04 05:47:46 +08:00
bool shmem_mapping(struct address_space *mapping)
{
mm: shmem: check for mapping owner before dereferencing mapping->host can be NULL and shouldn't be dereferenced before being checked. [ 1295.741844] GPF could be caused by NULL-ptr deref or user memory accessgeneral protection fault: 0000 [#1] SMP KASAN [ 1295.746387] Dumping ftrace buffer: [ 1295.748217] (ftrace buffer empty) [ 1295.749527] Modules linked in: [ 1295.750268] CPU: 62 PID: 23410 Comm: trinity-c70 Not tainted 3.19.0-next-20150219-sasha-00045-g9130270f #1939 [ 1295.750268] task: ffff8803a49db000 ti: ffff8803a4dc8000 task.ti: ffff8803a4dc8000 [ 1295.750268] RIP: shmem_mapping (mm/shmem.c:1458) [ 1295.750268] RSP: 0000:ffff8803a4dcfbf8 EFLAGS: 00010206 [ 1295.750268] RAX: dffffc0000000000 RBX: 0000000000000000 RCX: 00000000000f2804 [ 1295.750268] RDX: 0000000000000005 RSI: 0400000000000794 RDI: 0000000000000028 [ 1295.750268] RBP: ffff8803a4dcfc08 R08: 0000000000000000 R09: 00000000031de000 [ 1295.750268] R10: dffffc0000000000 R11: 00000000031c1000 R12: 0400000000000794 [ 1295.750268] R13: 00000000031c2000 R14: 00000000031de000 R15: ffff880e3bdc1000 [ 1295.750268] FS: 00007f8703c7e700(0000) GS:ffff881164800000(0000) knlGS:0000000000000000 [ 1295.750268] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 1295.750268] CR2: 0000000004e58000 CR3: 00000003a9f3c000 CR4: 00000000000007a0 [ 1295.750268] DR0: ffffffff81000000 DR1: 0000009494949494 DR2: 0000000000000000 [ 1295.750268] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 00000000000d0602 [ 1295.750268] Stack: [ 1295.750268] ffff8803a4dcfec8 ffffffffbb1dc770 ffff8803a4dcfc38 ffffffffad6f230b [ 1295.750268] ffffffffad6f2b0d 0000014100000000 ffff88001e17c08b ffff880d9453fe08 [ 1295.750268] ffff8803a4dcfd18 ffffffffad6f2ce2 ffff8803a49dbcd8 ffff8803a49dbce0 [ 1295.750268] Call Trace: [ 1295.750268] mincore_page (mm/mincore.c:61) [ 1295.750268] ? mincore_pte_range (include/linux/spinlock.h:312 mm/mincore.c:131) [ 1295.750268] mincore_pte_range (mm/mincore.c:151) [ 1295.750268] ? mincore_unmapped_range (mm/mincore.c:113) [ 1295.750268] __walk_page_range (mm/pagewalk.c:51 mm/pagewalk.c:90 mm/pagewalk.c:116 mm/pagewalk.c:204) [ 1295.750268] walk_page_range (mm/pagewalk.c:275) [ 1295.750268] SyS_mincore (mm/mincore.c:191 mm/mincore.c:253 mm/mincore.c:220) [ 1295.750268] ? mincore_pte_range (mm/mincore.c:220) [ 1295.750268] ? mincore_unmapped_range (mm/mincore.c:113) [ 1295.750268] ? __mincore_unmapped_range (mm/mincore.c:105) [ 1295.750268] ? ptlock_free (mm/mincore.c:24) [ 1295.750268] ? syscall_trace_enter (arch/x86/kernel/ptrace.c:1610) [ 1295.750268] ia32_do_call (arch/x86/ia32/ia32entry.S:446) [ 1295.750268] Code: e5 48 c1 ea 03 53 48 89 fb 48 83 ec 08 80 3c 02 00 75 4f 48 b8 00 00 00 00 00 fc ff df 48 8b 1b 48 8d 7b 28 48 89 fa 48 c1 ea 03 <80> 3c 02 00 75 3f 48 b8 00 00 00 00 00 fc ff df 48 8b 5b 28 48 All code ======== 0: e5 48 in $0x48,%eax 2: c1 ea 03 shr $0x3,%edx 5: 53 push %rbx 6: 48 89 fb mov %rdi,%rbx 9: 48 83 ec 08 sub $0x8,%rsp d: 80 3c 02 00 cmpb $0x0,(%rdx,%rax,1) 11: 75 4f jne 0x62 13: 48 b8 00 00 00 00 00 movabs $0xdffffc0000000000,%rax 1a: fc ff df 1d: 48 8b 1b mov (%rbx),%rbx 20: 48 8d 7b 28 lea 0x28(%rbx),%rdi 24: 48 89 fa mov %rdi,%rdx 27: 48 c1 ea 03 shr $0x3,%rdx 2b:* 80 3c 02 00 cmpb $0x0,(%rdx,%rax,1) <-- trapping instruction 2f: 75 3f jne 0x70 31: 48 b8 00 00 00 00 00 movabs $0xdffffc0000000000,%rax 38: fc ff df 3b: 48 8b 5b 28 mov 0x28(%rbx),%rbx 3f: 48 rex.W ... Code starting with the faulting instruction =========================================== 0: 80 3c 02 00 cmpb $0x0,(%rdx,%rax,1) 4: 75 3f jne 0x45 6: 48 b8 00 00 00 00 00 movabs $0xdffffc0000000000,%rax d: fc ff df 10: 48 8b 5b 28 mov 0x28(%rbx),%rbx 14: 48 rex.W ... [ 1295.750268] RIP shmem_mapping (mm/shmem.c:1458) [ 1295.750268] RSP <ffff8803a4dcfbf8> Fixes: 97b713ba3e ("fs: kill BDI_CAP_SWAP_BACKED") Signed-off-by: Sasha Levin <sasha.levin@oracle.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jens Axboe <axboe@fb.com>
2015-02-23 18:38:00 +08:00
if (!mapping->host)
return false;
return mapping->host->i_sb->s_op == &shmem_ops;
2014-04-04 05:47:46 +08:00
}
#ifdef CONFIG_TMPFS
static const struct inode_operations shmem_symlink_inode_operations;
static const struct inode_operations shmem_short_symlink_operations;
#ifdef CONFIG_TMPFS_XATTR
static int shmem_initxattrs(struct inode *, const struct xattr *, void *);
#else
#define shmem_initxattrs NULL
#endif
static int
shmem_write_begin(struct file *file, struct address_space *mapping,
loff_t pos, unsigned len, unsigned flags,
struct page **pagep, void **fsdata)
{
struct inode *inode = mapping->host;
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
struct shmem_inode_info *info = SHMEM_I(inode);
pgoff_t index = pos >> PAGE_CACHE_SHIFT;
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
/* i_mutex is held by caller */
if (unlikely(info->seals)) {
if (info->seals & F_SEAL_WRITE)
return -EPERM;
if ((info->seals & F_SEAL_GROW) && pos + len > inode->i_size)
return -EPERM;
}
return shmem_getpage(inode, index, pagep, SGP_WRITE, NULL);
}
static int
shmem_write_end(struct file *file, struct address_space *mapping,
loff_t pos, unsigned len, unsigned copied,
struct page *page, void *fsdata)
{
struct inode *inode = mapping->host;
if (pos + copied > inode->i_size)
i_size_write(inode, pos + copied);
if (!PageUptodate(page)) {
if (copied < PAGE_CACHE_SIZE) {
unsigned from = pos & (PAGE_CACHE_SIZE - 1);
zero_user_segments(page, 0, from,
from + copied, PAGE_CACHE_SIZE);
}
SetPageUptodate(page);
}
set_page_dirty(page);
unlock_page(page);
page_cache_release(page);
return copied;
}
static ssize_t shmem_file_read_iter(struct kiocb *iocb, struct iov_iter *to)
{
struct file *file = iocb->ki_filp;
struct inode *inode = file_inode(file);
struct address_space *mapping = inode->i_mapping;
pgoff_t index;
unsigned long offset;
enum sgp_type sgp = SGP_READ;
int error = 0;
ssize_t retval = 0;
loff_t *ppos = &iocb->ki_pos;
/*
* Might this read be for a stacking filesystem? Then when reading
* holes of a sparse file, we actually need to allocate those pages,
* and even mark them dirty, so it cannot exceed the max_blocks limit.
*/
if (!iter_is_iovec(to))
sgp = SGP_DIRTY;
index = *ppos >> PAGE_CACHE_SHIFT;
offset = *ppos & ~PAGE_CACHE_MASK;
for (;;) {
struct page *page = NULL;
pgoff_t end_index;
unsigned long nr, ret;
loff_t i_size = i_size_read(inode);
end_index = i_size >> PAGE_CACHE_SHIFT;
if (index > end_index)
break;
if (index == end_index) {
nr = i_size & ~PAGE_CACHE_MASK;
if (nr <= offset)
break;
}
error = shmem_getpage(inode, index, &page, sgp, NULL);
if (error) {
if (error == -EINVAL)
error = 0;
break;
}
if (page)
unlock_page(page);
/*
* We must evaluate after, since reads (unlike writes)
* are called without i_mutex protection against truncate
*/
nr = PAGE_CACHE_SIZE;
i_size = i_size_read(inode);
end_index = i_size >> PAGE_CACHE_SHIFT;
if (index == end_index) {
nr = i_size & ~PAGE_CACHE_MASK;
if (nr <= offset) {
if (page)
page_cache_release(page);
break;
}
}
nr -= offset;
if (page) {
/*
* If users can be writing to this page using arbitrary
* virtual addresses, take care about potential aliasing
* before reading the page on the kernel side.
*/
if (mapping_writably_mapped(mapping))
flush_dcache_page(page);
/*
* Mark the page accessed if we read the beginning.
*/
if (!offset)
mark_page_accessed(page);
2005-10-30 09:16:12 +08:00
} else {
page = ZERO_PAGE(0);
2005-10-30 09:16:12 +08:00
page_cache_get(page);
}
/*
* Ok, we have the page, and it's up-to-date, so
* now we can copy it to user space...
*/
ret = copy_page_to_iter(page, offset, nr, to);
retval += ret;
offset += ret;
index += offset >> PAGE_CACHE_SHIFT;
offset &= ~PAGE_CACHE_MASK;
page_cache_release(page);
if (!iov_iter_count(to))
break;
if (ret < nr) {
error = -EFAULT;
break;
}
cond_resched();
}
*ppos = ((loff_t) index << PAGE_CACHE_SHIFT) + offset;
file_accessed(file);
return retval ? retval : error;
}
static ssize_t shmem_file_splice_read(struct file *in, loff_t *ppos,
struct pipe_inode_info *pipe, size_t len,
unsigned int flags)
{
struct address_space *mapping = in->f_mapping;
struct inode *inode = mapping->host;
unsigned int loff, nr_pages, req_pages;
struct page *pages[PIPE_DEF_BUFFERS];
struct partial_page partial[PIPE_DEF_BUFFERS];
struct page *page;
pgoff_t index, end_index;
loff_t isize, left;
int error, page_nr;
struct splice_pipe_desc spd = {
.pages = pages,
.partial = partial,
.nr_pages_max = PIPE_DEF_BUFFERS,
.flags = flags,
.ops = &page_cache_pipe_buf_ops,
.spd_release = spd_release_page,
};
isize = i_size_read(inode);
if (unlikely(*ppos >= isize))
return 0;
left = isize - *ppos;
if (unlikely(left < len))
len = left;
if (splice_grow_spd(pipe, &spd))
return -ENOMEM;
index = *ppos >> PAGE_CACHE_SHIFT;
loff = *ppos & ~PAGE_CACHE_MASK;
req_pages = (len + loff + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
nr_pages = min(req_pages, spd.nr_pages_max);
spd.nr_pages = find_get_pages_contig(mapping, index,
nr_pages, spd.pages);
index += spd.nr_pages;
error = 0;
while (spd.nr_pages < nr_pages) {
error = shmem_getpage(inode, index, &page, SGP_CACHE, NULL);
if (error)
break;
unlock_page(page);
spd.pages[spd.nr_pages++] = page;
index++;
}
index = *ppos >> PAGE_CACHE_SHIFT;
nr_pages = spd.nr_pages;
spd.nr_pages = 0;
for (page_nr = 0; page_nr < nr_pages; page_nr++) {
unsigned int this_len;
if (!len)
break;
this_len = min_t(unsigned long, len, PAGE_CACHE_SIZE - loff);
page = spd.pages[page_nr];
if (!PageUptodate(page) || page->mapping != mapping) {
error = shmem_getpage(inode, index, &page,
SGP_CACHE, NULL);
if (error)
break;
unlock_page(page);
page_cache_release(spd.pages[page_nr]);
spd.pages[page_nr] = page;
}
isize = i_size_read(inode);
end_index = (isize - 1) >> PAGE_CACHE_SHIFT;
if (unlikely(!isize || index > end_index))
break;
if (end_index == index) {
unsigned int plen;
plen = ((isize - 1) & ~PAGE_CACHE_MASK) + 1;
if (plen <= loff)
break;
this_len = min(this_len, plen - loff);
len = this_len;
}
spd.partial[page_nr].offset = loff;
spd.partial[page_nr].len = this_len;
len -= this_len;
loff = 0;
spd.nr_pages++;
index++;
}
while (page_nr < nr_pages)
page_cache_release(spd.pages[page_nr++]);
if (spd.nr_pages)
error = splice_to_pipe(pipe, &spd);
splice_shrink_spd(&spd);
if (error > 0) {
*ppos += error;
file_accessed(in);
}
return error;
}
/*
* llseek SEEK_DATA or SEEK_HOLE through the radix_tree.
*/
static pgoff_t shmem_seek_hole_data(struct address_space *mapping,
pgoff_t index, pgoff_t end, int whence)
{
struct page *page;
struct pagevec pvec;
pgoff_t indices[PAGEVEC_SIZE];
bool done = false;
int i;
pagevec_init(&pvec, 0);
pvec.nr = 1; /* start small: we may be there already */
while (!done) {
2014-04-04 05:47:46 +08:00
pvec.nr = find_get_entries(mapping, index,
pvec.nr, pvec.pages, indices);
if (!pvec.nr) {
if (whence == SEEK_DATA)
index = end;
break;
}
for (i = 0; i < pvec.nr; i++, index++) {
if (index < indices[i]) {
if (whence == SEEK_HOLE) {
done = true;
break;
}
index = indices[i];
}
page = pvec.pages[i];
if (page && !radix_tree_exceptional_entry(page)) {
if (!PageUptodate(page))
page = NULL;
}
if (index >= end ||
(page && whence == SEEK_DATA) ||
(!page && whence == SEEK_HOLE)) {
done = true;
break;
}
}
2014-04-04 05:47:46 +08:00
pagevec_remove_exceptionals(&pvec);
pagevec_release(&pvec);
pvec.nr = PAGEVEC_SIZE;
cond_resched();
}
return index;
}
static loff_t shmem_file_llseek(struct file *file, loff_t offset, int whence)
{
struct address_space *mapping = file->f_mapping;
struct inode *inode = mapping->host;
pgoff_t start, end;
loff_t new_offset;
if (whence != SEEK_DATA && whence != SEEK_HOLE)
return generic_file_llseek_size(file, offset, whence,
MAX_LFS_FILESIZE, i_size_read(inode));
mutex_lock(&inode->i_mutex);
/* We're holding i_mutex so we can access i_size directly */
if (offset < 0)
offset = -EINVAL;
else if (offset >= inode->i_size)
offset = -ENXIO;
else {
start = offset >> PAGE_CACHE_SHIFT;
end = (inode->i_size + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
new_offset = shmem_seek_hole_data(mapping, start, end, whence);
new_offset <<= PAGE_CACHE_SHIFT;
if (new_offset > offset) {
if (new_offset < inode->i_size)
offset = new_offset;
else if (whence == SEEK_DATA)
offset = -ENXIO;
else
offset = inode->i_size;
}
}
if (offset >= 0)
offset = vfs_setpos(file, offset, MAX_LFS_FILESIZE);
mutex_unlock(&inode->i_mutex);
return offset;
}
/*
* We need a tag: a new tag would expand every radix_tree_node by 8 bytes,
* so reuse a tag which we firmly believe is never set or cleared on shmem.
*/
#define SHMEM_TAG_PINNED PAGECACHE_TAG_TOWRITE
#define LAST_SCAN 4 /* about 150ms max */
static void shmem_tag_pins(struct address_space *mapping)
{
struct radix_tree_iter iter;
void **slot;
pgoff_t start;
struct page *page;
lru_add_drain();
start = 0;
rcu_read_lock();
restart:
radix_tree_for_each_slot(slot, &mapping->page_tree, &iter, start) {
page = radix_tree_deref_slot(slot);
if (!page || radix_tree_exception(page)) {
if (radix_tree_deref_retry(page))
goto restart;
} else if (page_count(page) - page_mapcount(page) > 1) {
spin_lock_irq(&mapping->tree_lock);
radix_tree_tag_set(&mapping->page_tree, iter.index,
SHMEM_TAG_PINNED);
spin_unlock_irq(&mapping->tree_lock);
}
if (need_resched()) {
cond_resched_rcu();
start = iter.index + 1;
goto restart;
}
}
rcu_read_unlock();
}
/*
* Setting SEAL_WRITE requires us to verify there's no pending writer. However,
* via get_user_pages(), drivers might have some pending I/O without any active
* user-space mappings (eg., direct-IO, AIO). Therefore, we look at all pages
* and see whether it has an elevated ref-count. If so, we tag them and wait for
* them to be dropped.
* The caller must guarantee that no new user will acquire writable references
* to those pages to avoid races.
*/
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
static int shmem_wait_for_pins(struct address_space *mapping)
{
struct radix_tree_iter iter;
void **slot;
pgoff_t start;
struct page *page;
int error, scan;
shmem_tag_pins(mapping);
error = 0;
for (scan = 0; scan <= LAST_SCAN; scan++) {
if (!radix_tree_tagged(&mapping->page_tree, SHMEM_TAG_PINNED))
break;
if (!scan)
lru_add_drain_all();
else if (schedule_timeout_killable((HZ << scan) / 200))
scan = LAST_SCAN;
start = 0;
rcu_read_lock();
restart:
radix_tree_for_each_tagged(slot, &mapping->page_tree, &iter,
start, SHMEM_TAG_PINNED) {
page = radix_tree_deref_slot(slot);
if (radix_tree_exception(page)) {
if (radix_tree_deref_retry(page))
goto restart;
page = NULL;
}
if (page &&
page_count(page) - page_mapcount(page) != 1) {
if (scan < LAST_SCAN)
goto continue_resched;
/*
* On the last scan, we clean up all those tags
* we inserted; but make a note that we still
* found pages pinned.
*/
error = -EBUSY;
}
spin_lock_irq(&mapping->tree_lock);
radix_tree_tag_clear(&mapping->page_tree,
iter.index, SHMEM_TAG_PINNED);
spin_unlock_irq(&mapping->tree_lock);
continue_resched:
if (need_resched()) {
cond_resched_rcu();
start = iter.index + 1;
goto restart;
}
}
rcu_read_unlock();
}
return error;
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
}
#define F_ALL_SEALS (F_SEAL_SEAL | \
F_SEAL_SHRINK | \
F_SEAL_GROW | \
F_SEAL_WRITE)
int shmem_add_seals(struct file *file, unsigned int seals)
{
struct inode *inode = file_inode(file);
struct shmem_inode_info *info = SHMEM_I(inode);
int error;
/*
* SEALING
* Sealing allows multiple parties to share a shmem-file but restrict
* access to a specific subset of file operations. Seals can only be
* added, but never removed. This way, mutually untrusted parties can
* share common memory regions with a well-defined policy. A malicious
* peer can thus never perform unwanted operations on a shared object.
*
* Seals are only supported on special shmem-files and always affect
* the whole underlying inode. Once a seal is set, it may prevent some
* kinds of access to the file. Currently, the following seals are
* defined:
* SEAL_SEAL: Prevent further seals from being set on this file
* SEAL_SHRINK: Prevent the file from shrinking
* SEAL_GROW: Prevent the file from growing
* SEAL_WRITE: Prevent write access to the file
*
* As we don't require any trust relationship between two parties, we
* must prevent seals from being removed. Therefore, sealing a file
* only adds a given set of seals to the file, it never touches
* existing seals. Furthermore, the "setting seals"-operation can be
* sealed itself, which basically prevents any further seal from being
* added.
*
* Semantics of sealing are only defined on volatile files. Only
* anonymous shmem files support sealing. More importantly, seals are
* never written to disk. Therefore, there's no plan to support it on
* other file types.
*/
if (file->f_op != &shmem_file_operations)
return -EINVAL;
if (!(file->f_mode & FMODE_WRITE))
return -EPERM;
if (seals & ~(unsigned int)F_ALL_SEALS)
return -EINVAL;
mutex_lock(&inode->i_mutex);
if (info->seals & F_SEAL_SEAL) {
error = -EPERM;
goto unlock;
}
if ((seals & F_SEAL_WRITE) && !(info->seals & F_SEAL_WRITE)) {
error = mapping_deny_writable(file->f_mapping);
if (error)
goto unlock;
error = shmem_wait_for_pins(file->f_mapping);
if (error) {
mapping_allow_writable(file->f_mapping);
goto unlock;
}
}
info->seals |= seals;
error = 0;
unlock:
mutex_unlock(&inode->i_mutex);
return error;
}
EXPORT_SYMBOL_GPL(shmem_add_seals);
int shmem_get_seals(struct file *file)
{
if (file->f_op != &shmem_file_operations)
return -EINVAL;
return SHMEM_I(file_inode(file))->seals;
}
EXPORT_SYMBOL_GPL(shmem_get_seals);
long shmem_fcntl(struct file *file, unsigned int cmd, unsigned long arg)
{
long error;
switch (cmd) {
case F_ADD_SEALS:
/* disallow upper 32bit */
if (arg > UINT_MAX)
return -EINVAL;
error = shmem_add_seals(file, arg);
break;
case F_GET_SEALS:
error = shmem_get_seals(file);
break;
default:
error = -EINVAL;
break;
}
return error;
}
static long shmem_fallocate(struct file *file, int mode, loff_t offset,
loff_t len)
{
struct inode *inode = file_inode(file);
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb);
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
struct shmem_inode_info *info = SHMEM_I(inode);
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
struct shmem_falloc shmem_falloc;
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
pgoff_t start, index, end;
int error;
if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE))
return -EOPNOTSUPP;
mutex_lock(&inode->i_mutex);
if (mode & FALLOC_FL_PUNCH_HOLE) {
struct address_space *mapping = file->f_mapping;
loff_t unmap_start = round_up(offset, PAGE_SIZE);
loff_t unmap_end = round_down(offset + len, PAGE_SIZE) - 1;
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
DECLARE_WAIT_QUEUE_HEAD_ONSTACK(shmem_falloc_waitq);
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
/* protected by i_mutex */
if (info->seals & F_SEAL_WRITE) {
error = -EPERM;
goto out;
}
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
shmem_falloc.waitq = &shmem_falloc_waitq;
shmem_falloc.start = unmap_start >> PAGE_SHIFT;
shmem_falloc.next = (unmap_end + 1) >> PAGE_SHIFT;
spin_lock(&inode->i_lock);
inode->i_private = &shmem_falloc;
spin_unlock(&inode->i_lock);
if ((u64)unmap_end > (u64)unmap_start)
unmap_mapping_range(mapping, unmap_start,
1 + unmap_end - unmap_start, 0);
shmem_truncate_range(inode, offset, offset + len - 1);
/* No need to unmap again: hole-punching leaves COWed pages */
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
spin_lock(&inode->i_lock);
inode->i_private = NULL;
wake_up_all(&shmem_falloc_waitq);
spin_unlock(&inode->i_lock);
error = 0;
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
goto out;
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
}
/* We need to check rlimit even when FALLOC_FL_KEEP_SIZE */
error = inode_newsize_ok(inode, offset + len);
if (error)
goto out;
shm: add sealing API If two processes share a common memory region, they usually want some guarantees to allow safe access. This often includes: - one side cannot overwrite data while the other reads it - one side cannot shrink the buffer while the other accesses it - one side cannot grow the buffer beyond previously set boundaries If there is a trust-relationship between both parties, there is no need for policy enforcement. However, if there's no trust relationship (eg., for general-purpose IPC) sharing memory-regions is highly fragile and often not possible without local copies. Look at the following two use-cases: 1) A graphics client wants to share its rendering-buffer with a graphics-server. The memory-region is allocated by the client for read/write access and a second FD is passed to the server. While scanning out from the memory region, the server has no guarantee that the client doesn't shrink the buffer at any time, requiring rather cumbersome SIGBUS handling. 2) A process wants to perform an RPC on another process. To avoid huge bandwidth consumption, zero-copy is preferred. After a message is assembled in-memory and a FD is passed to the remote side, both sides want to be sure that neither modifies this shared copy, anymore. The source may have put sensible data into the message without a separate copy and the target may want to parse the message inline, to avoid a local copy. While SIGBUS handling, POSIX mandatory locking and MAP_DENYWRITE provide ways to achieve most of this, the first one is unproportionally ugly to use in libraries and the latter two are broken/racy or even disabled due to denial of service attacks. This patch introduces the concept of SEALING. If you seal a file, a specific set of operations is blocked on that file forever. Unlike locks, seals can only be set, never removed. Hence, once you verified a specific set of seals is set, you're guaranteed that no-one can perform the blocked operations on this file, anymore. An initial set of SEALS is introduced by this patch: - SHRINK: If SEAL_SHRINK is set, the file in question cannot be reduced in size. This affects ftruncate() and open(O_TRUNC). - GROW: If SEAL_GROW is set, the file in question cannot be increased in size. This affects ftruncate(), fallocate() and write(). - WRITE: If SEAL_WRITE is set, no write operations (besides resizing) are possible. This affects fallocate(PUNCH_HOLE), mmap() and write(). - SEAL: If SEAL_SEAL is set, no further seals can be added to a file. This basically prevents the F_ADD_SEAL operation on a file and can be set to prevent others from adding further seals that you don't want. The described use-cases can easily use these seals to provide safe use without any trust-relationship: 1) The graphics server can verify that a passed file-descriptor has SEAL_SHRINK set. This allows safe scanout, while the client is allowed to increase buffer size for window-resizing on-the-fly. Concurrent writes are explicitly allowed. 2) For general-purpose IPC, both processes can verify that SEAL_SHRINK, SEAL_GROW and SEAL_WRITE are set. This guarantees that neither process can modify the data while the other side parses it. Furthermore, it guarantees that even with writable FDs passed to the peer, it cannot increase the size to hit memory-limits of the source process (in case the file-storage is accounted to the source). The new API is an extension to fcntl(), adding two new commands: F_GET_SEALS: Return a bitset describing the seals on the file. This can be called on any FD if the underlying file supports sealing. F_ADD_SEALS: Change the seals of a given file. This requires WRITE access to the file and F_SEAL_SEAL may not already be set. Furthermore, the underlying file must support sealing and there may not be any existing shared mapping of that file. Otherwise, EBADF/EPERM is returned. The given seals are _added_ to the existing set of seals on the file. You cannot remove seals again. The fcntl() handler is currently specific to shmem and disabled on all files. A file needs to explicitly support sealing for this interface to work. A separate syscall is added in a follow-up, which creates files that support sealing. There is no intention to support this on other file-systems. Semantics are unclear for non-volatile files and we lack any use-case right now. Therefore, the implementation is specific to shmem. Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: Michael Kerrisk <mtk.manpages@gmail.com> Cc: Ryan Lortie <desrt@desrt.ca> Cc: Lennart Poettering <lennart@poettering.net> Cc: Daniel Mack <zonque@gmail.com> Cc: Andy Lutomirski <luto@amacapital.net> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-09 05:25:27 +08:00
if ((info->seals & F_SEAL_GROW) && offset + len > inode->i_size) {
error = -EPERM;
goto out;
}
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
start = offset >> PAGE_CACHE_SHIFT;
end = (offset + len + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
/* Try to avoid a swapstorm if len is impossible to satisfy */
if (sbinfo->max_blocks && end - start > sbinfo->max_blocks) {
error = -ENOSPC;
goto out;
}
shmem: fix faulting into a hole, not taking i_mutex Commit f00cdc6df7d7 ("shmem: fix faulting into a hole while it's punched") was buggy: Sasha sent a lockdep report to remind us that grabbing i_mutex in the fault path is a no-no (write syscall may already hold i_mutex while faulting user buffer). We tried a completely different approach (see following patch) but that proved inadequate: good enough for a rational workload, but not good enough against trinity - which forks off so many mappings of the object that contention on i_mmap_mutex while hole-puncher holds i_mutex builds into serious starvation when concurrent faults force the puncher to fall back to single-page unmap_mapping_range() searches of the i_mmap tree. So return to the original umbrella approach, but keep away from i_mutex this time. We really don't want to bloat every shmem inode with a new mutex or completion, just to protect this unlikely case from trinity. So extend the original with wait_queue_head on stack at the hole-punch end, and wait_queue item on the stack at the fault end. This involves further use of i_lock to guard against the races: lockdep has been happy so far, and I see fs/inode.c:unlock_new_inode() holds i_lock around wake_up_bit(), which is comparable to what we do here. i_lock is more convenient, but we could switch to shmem's info->lock. This issue has been tagged with CVE-2014-4171, which will require commit f00cdc6df7d7 and this and the following patch to be backported: we suggest to 3.1+, though in fact the trinity forkbomb effect might go back as far as 2.6.16, when madvise(,,MADV_REMOVE) came in - or might not, since much has changed, with i_mmap_mutex a spinlock before 3.0. Anyone running trinity on 3.0 and earlier? I don't think we need care. Signed-off-by: Hugh Dickins <hughd@google.com> Reported-by: Sasha Levin <sasha.levin@oracle.com> Tested-by: Sasha Levin <sasha.levin@oracle.com> Cc: Vlastimil Babka <vbabka@suse.cz> Cc: Konstantin Khlebnikov <koct9i@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Lukas Czerner <lczerner@redhat.com> Cc: Dave Jones <davej@redhat.com> Cc: <stable@vger.kernel.org> [3.1+] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-07-24 05:00:10 +08:00
shmem_falloc.waitq = NULL;
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
shmem_falloc.start = start;
shmem_falloc.next = start;
shmem_falloc.nr_falloced = 0;
shmem_falloc.nr_unswapped = 0;
spin_lock(&inode->i_lock);
inode->i_private = &shmem_falloc;
spin_unlock(&inode->i_lock);
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
for (index = start; index < end; index++) {
struct page *page;
/*
* Good, the fallocate(2) manpage permits EINTR: we may have
* been interrupted because we are using up too much memory.
*/
if (signal_pending(current))
error = -EINTR;
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
else if (shmem_falloc.nr_unswapped > shmem_falloc.nr_falloced)
error = -ENOMEM;
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
else
error = shmem_getpage(inode, index, &page, SGP_FALLOC,
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
NULL);
if (error) {
/* Remove the !PageUptodate pages we added */
shmem_undo_range(inode,
(loff_t)start << PAGE_CACHE_SHIFT,
(loff_t)index << PAGE_CACHE_SHIFT, true);
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
goto undone;
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
}
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
/*
* Inform shmem_writepage() how far we have reached.
* No need for lock or barrier: we have the page lock.
*/
shmem_falloc.next++;
if (!PageUptodate(page))
shmem_falloc.nr_falloced++;
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
/*
* If !PageUptodate, leave it that way so that freeable pages
* can be recognized if we need to rollback on error later.
* But set_page_dirty so that memory pressure will swap rather
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
* than free the pages we are allocating (and SGP_CACHE pages
* might still be clean: we now need to mark those dirty too).
*/
set_page_dirty(page);
unlock_page(page);
page_cache_release(page);
cond_resched();
}
if (!(mode & FALLOC_FL_KEEP_SIZE) && offset + len > inode->i_size)
i_size_write(inode, offset + len);
inode->i_ctime = CURRENT_TIME;
tmpfs: quit when fallocate fills memory As it stands, a large fallocate() on tmpfs is liable to fill memory with pages, freed on failure except when they run into swap, at which point they become fixed into the file despite the failure. That feels quite wrong, to be consuming resources precisely when they're in short supply. Go the other way instead: shmem_fallocate() indicate the range it has fallocated to shmem_writepage(), keeping count of pages it's allocating; shmem_writepage() reactivate instead of swapping out pages fallocated by this syscall (but happily swap out those from earlier occasions), keeping count; shmem_fallocate() compare counts and give up once the reactivated pages have started to coming back to writepage (approximately: some zones would in fact recycle faster than others). This is a little unusual, but works well: although we could consider the failure to swap as a bug, and fix it later with SWAP_MAP_FALLOC handling added in swapfile.c and memcontrol.c, I doubt that we shall ever want to. (If there's no swap, an over-large fallocate() on tmpfs is limited in the same way as writing: stopped by rlimit, or by tmpfs mount size if that was set sensibly, or by __vm_enough_memory() heuristics if OVERCOMMIT_GUESS or OVERCOMMIT_NEVER. If OVERCOMMIT_ALWAYS, then it is liable to OOM-kill others as writing would, but stops and frees if interrupted.) Now that everything is freed on failure, we can then skip updating ctime. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:42 +08:00
undone:
spin_lock(&inode->i_lock);
inode->i_private = NULL;
spin_unlock(&inode->i_lock);
tmpfs: support fallocate preallocation The systemd plumbers expressed a wish that tmpfs support preallocation. Cong Wang wrote a patch, but several kernel guys expressed scepticism: https://lkml.org/lkml/2011/11/18/137 Christoph Hellwig: What for exactly? Please explain why preallocating on tmpfs would make any sense. Kay Sievers: To be able to safely use mmap(), regarding SIGBUS, on files on the /dev/shm filesystem. The glibc fallback loop for -ENOSYS [or -EOPNOTSUPP] on fallocate is just ugly. Hugh Dickins: If tmpfs is going to support fallocate(FALLOC_FL_PUNCH_HOLE), it would seem perverse to permit the deallocation but fail the allocation. Christoph Hellwig: Agreed. Now that we do have shmem_fallocate() for hole-punching, plumb in basic support for preallocation mode too. It's fairly straightforward (though quite a few details needed attention), except for when it fails part way through. What a pity that fallocate(2) was not specified to return the length allocated, permitting short fallocations! As it is, when it fails part way through, we ought to free what has just been allocated by this system call; but must be very sure not to free any allocated earlier, or any allocated by racing accesses (not all excluded by i_mutex). But we cannot distinguish them: so in this patch simply leak allocations on partial failure (they will be freed later if the file is removed). An attractive alternative approach would have been for fallocate() not to allocate pages at all, but note reservations by entries in the radix-tree. But that would give less assurance, and, critically, would be hard to fit with mem cgroups (who owns the reservations?): allocating pages lets fallocate() behave in just the same way as write(). Based-on-patch-by: Cong Wang <amwang@redhat.com> Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Cc: Cong Wang <amwang@redhat.com> Cc: Kay Sievers <kay@vrfy.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-05-30 06:06:41 +08:00
out:
mutex_unlock(&inode->i_mutex);
return error;
}
static int shmem_statfs(struct dentry *dentry, struct kstatfs *buf)
{
struct shmem_sb_info *sbinfo = SHMEM_SB(dentry->d_sb);
buf->f_type = TMPFS_MAGIC;
buf->f_bsize = PAGE_CACHE_SIZE;
buf->f_namelen = NAME_MAX;
if (sbinfo->max_blocks) {
buf->f_blocks = sbinfo->max_blocks;
buf->f_bavail =
buf->f_bfree = sbinfo->max_blocks -
percpu_counter_sum(&sbinfo->used_blocks);
}
if (sbinfo->max_inodes) {
buf->f_files = sbinfo->max_inodes;
buf->f_ffree = sbinfo->free_inodes;
}
/* else leave those fields 0 like simple_statfs */
return 0;
}
/*
* File creation. Allocate an inode, and we're done..
*/
static int
shmem_mknod(struct inode *dir, struct dentry *dentry, umode_t mode, dev_t dev)
{
struct inode *inode;
int error = -ENOSPC;
inode = shmem_get_inode(dir->i_sb, dir, mode, dev, VM_NORESERVE);
if (inode) {
error = simple_acl_create(dir, inode);
if (error)
goto out_iput;
error = security_inode_init_security(inode, dir,
&dentry->d_name,
shmem_initxattrs, NULL);
if (error && error != -EOPNOTSUPP)
goto out_iput;
error = 0;
dir->i_size += BOGO_DIRENT_SIZE;
dir->i_ctime = dir->i_mtime = CURRENT_TIME;
d_instantiate(dentry, inode);
dget(dentry); /* Extra count - pin the dentry in core */
}
return error;
out_iput:
iput(inode);
return error;
}
static int
shmem_tmpfile(struct inode *dir, struct dentry *dentry, umode_t mode)
{
struct inode *inode;
int error = -ENOSPC;
inode = shmem_get_inode(dir->i_sb, dir, mode, 0, VM_NORESERVE);
if (inode) {
error = security_inode_init_security(inode, dir,
NULL,
shmem_initxattrs, NULL);
if (error && error != -EOPNOTSUPP)
goto out_iput;
error = simple_acl_create(dir, inode);
if (error)
goto out_iput;
d_tmpfile(dentry, inode);
}
return error;
out_iput:
iput(inode);
return error;
}
static int shmem_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode)
{
int error;
if ((error = shmem_mknod(dir, dentry, mode | S_IFDIR, 0)))
return error;
inc_nlink(dir);
return 0;
}
static int shmem_create(struct inode *dir, struct dentry *dentry, umode_t mode,
bool excl)
{
return shmem_mknod(dir, dentry, mode | S_IFREG, 0);
}
/*
* Link a file..
*/
static int shmem_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry)
{
struct inode *inode = d_inode(old_dentry);
int ret;
/*
* No ordinary (disk based) filesystem counts links as inodes;
* but each new link needs a new dentry, pinning lowmem, and
* tmpfs dentries cannot be pruned until they are unlinked.
*/
ret = shmem_reserve_inode(inode->i_sb);
if (ret)
goto out;
dir->i_size += BOGO_DIRENT_SIZE;
inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME;
inc_nlink(inode);
ihold(inode); /* New dentry reference */
dget(dentry); /* Extra pinning count for the created dentry */
d_instantiate(dentry, inode);
out:
return ret;
}
static int shmem_unlink(struct inode *dir, struct dentry *dentry)
{
struct inode *inode = d_inode(dentry);
if (inode->i_nlink > 1 && !S_ISDIR(inode->i_mode))
shmem_free_inode(inode->i_sb);
dir->i_size -= BOGO_DIRENT_SIZE;
inode->i_ctime = dir->i_ctime = dir->i_mtime = CURRENT_TIME;
drop_nlink(inode);
dput(dentry); /* Undo the count from "create" - this does all the work */
return 0;
}
static int shmem_rmdir(struct inode *dir, struct dentry *dentry)
{
if (!simple_empty(dentry))
return -ENOTEMPTY;
drop_nlink(d_inode(dentry));
drop_nlink(dir);
return shmem_unlink(dir, dentry);
}
static int shmem_exchange(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry)
{
VFS: (Scripted) Convert S_ISLNK/DIR/REG(dentry->d_inode) to d_is_*(dentry) Convert the following where appropriate: (1) S_ISLNK(dentry->d_inode) to d_is_symlink(dentry). (2) S_ISREG(dentry->d_inode) to d_is_reg(dentry). (3) S_ISDIR(dentry->d_inode) to d_is_dir(dentry). This is actually more complicated than it appears as some calls should be converted to d_can_lookup() instead. The difference is whether the directory in question is a real dir with a ->lookup op or whether it's a fake dir with a ->d_automount op. In some circumstances, we can subsume checks for dentry->d_inode not being NULL into this, provided we the code isn't in a filesystem that expects d_inode to be NULL if the dirent really *is* negative (ie. if we're going to use d_inode() rather than d_backing_inode() to get the inode pointer). Note that the dentry type field may be set to something other than DCACHE_MISS_TYPE when d_inode is NULL in the case of unionmount, where the VFS manages the fall-through from a negative dentry to a lower layer. In such a case, the dentry type of the negative union dentry is set to the same as the type of the lower dentry. However, if you know d_inode is not NULL at the call site, then you can use the d_is_xxx() functions even in a filesystem. There is one further complication: a 0,0 chardev dentry may be labelled DCACHE_WHITEOUT_TYPE rather than DCACHE_SPECIAL_TYPE. Strictly, this was intended for special directory entry types that don't have attached inodes. The following perl+coccinelle script was used: use strict; my @callers; open($fd, 'git grep -l \'S_IS[A-Z].*->d_inode\' |') || die "Can't grep for S_ISDIR and co. callers"; @callers = <$fd>; close($fd); unless (@callers) { print "No matches\n"; exit(0); } my @cocci = ( '@@', 'expression E;', '@@', '', '- S_ISLNK(E->d_inode->i_mode)', '+ d_is_symlink(E)', '', '@@', 'expression E;', '@@', '', '- S_ISDIR(E->d_inode->i_mode)', '+ d_is_dir(E)', '', '@@', 'expression E;', '@@', '', '- S_ISREG(E->d_inode->i_mode)', '+ d_is_reg(E)' ); my $coccifile = "tmp.sp.cocci"; open($fd, ">$coccifile") || die $coccifile; print($fd "$_\n") || die $coccifile foreach (@cocci); close($fd); foreach my $file (@callers) { chomp $file; print "Processing ", $file, "\n"; system("spatch", "--sp-file", $coccifile, $file, "--in-place", "--no-show-diff") == 0 || die "spatch failed"; } [AV: overlayfs parts skipped] Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-01-29 20:02:35 +08:00
bool old_is_dir = d_is_dir(old_dentry);
bool new_is_dir = d_is_dir(new_dentry);
if (old_dir != new_dir && old_is_dir != new_is_dir) {
if (old_is_dir) {
drop_nlink(old_dir);
inc_nlink(new_dir);
} else {
drop_nlink(new_dir);
inc_nlink(old_dir);
}
}
old_dir->i_ctime = old_dir->i_mtime =
new_dir->i_ctime = new_dir->i_mtime =
d_inode(old_dentry)->i_ctime =
d_inode(new_dentry)->i_ctime = CURRENT_TIME;
return 0;
}
static int shmem_whiteout(struct inode *old_dir, struct dentry *old_dentry)
{
struct dentry *whiteout;
int error;
whiteout = d_alloc(old_dentry->d_parent, &old_dentry->d_name);
if (!whiteout)
return -ENOMEM;
error = shmem_mknod(old_dir, whiteout,
S_IFCHR | WHITEOUT_MODE, WHITEOUT_DEV);
dput(whiteout);
if (error)
return error;
/*
* Cheat and hash the whiteout while the old dentry is still in
* place, instead of playing games with FS_RENAME_DOES_D_MOVE.
*
* d_lookup() will consistently find one of them at this point,
* not sure which one, but that isn't even important.
*/
d_rehash(whiteout);
return 0;
}
/*
* The VFS layer already does all the dentry stuff for rename,
* we just have to decrement the usage count for the target if
* it exists so that the VFS layer correctly free's it when it
* gets overwritten.
*/
static int shmem_rename2(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry, unsigned int flags)
{
struct inode *inode = d_inode(old_dentry);
int they_are_dirs = S_ISDIR(inode->i_mode);
if (flags & ~(RENAME_NOREPLACE | RENAME_EXCHANGE | RENAME_WHITEOUT))
return -EINVAL;
if (flags & RENAME_EXCHANGE)
return shmem_exchange(old_dir, old_dentry, new_dir, new_dentry);
if (!simple_empty(new_dentry))
return -ENOTEMPTY;
if (flags & RENAME_WHITEOUT) {
int error;
error = shmem_whiteout(old_dir, old_dentry);
if (error)
return error;
}
if (d_really_is_positive(new_dentry)) {
(void) shmem_unlink(new_dir, new_dentry);
if (they_are_dirs) {
drop_nlink(d_inode(new_dentry));
drop_nlink(old_dir);
}
} else if (they_are_dirs) {
drop_nlink(old_dir);
inc_nlink(new_dir);
}
old_dir->i_size -= BOGO_DIRENT_SIZE;
new_dir->i_size += BOGO_DIRENT_SIZE;
old_dir->i_ctime = old_dir->i_mtime =
new_dir->i_ctime = new_dir->i_mtime =
inode->i_ctime = CURRENT_TIME;
return 0;
}
static int shmem_symlink(struct inode *dir, struct dentry *dentry, const char *symname)
{
int error;
int len;
struct inode *inode;
struct page *page;
char *kaddr;
struct shmem_inode_info *info;
len = strlen(symname) + 1;
if (len > PAGE_CACHE_SIZE)
return -ENAMETOOLONG;
inode = shmem_get_inode(dir->i_sb, dir, S_IFLNK|S_IRWXUGO, 0, VM_NORESERVE);
if (!inode)
return -ENOSPC;
error = security_inode_init_security(inode, dir, &dentry->d_name,
shmem_initxattrs, NULL);
if (error) {
if (error != -EOPNOTSUPP) {
iput(inode);
return error;
}
error = 0;
}
info = SHMEM_I(inode);
inode->i_size = len-1;
if (len <= SHORT_SYMLINK_LEN) {
info->symlink = kmemdup(symname, len, GFP_KERNEL);
if (!info->symlink) {
iput(inode);
return -ENOMEM;
}
inode->i_op = &shmem_short_symlink_operations;
inode->i_link = info->symlink;
} else {
error = shmem_getpage(inode, 0, &page, SGP_WRITE, NULL);
if (error) {
iput(inode);
return error;
}
inode->i_mapping->a_ops = &shmem_aops;
inode->i_op = &shmem_symlink_inode_operations;
kaddr = kmap_atomic(page);
memcpy(kaddr, symname, len);
kunmap_atomic(kaddr);
SetPageUptodate(page);
set_page_dirty(page);
unlock_page(page);
page_cache_release(page);
}
dir->i_size += BOGO_DIRENT_SIZE;
dir->i_ctime = dir->i_mtime = CURRENT_TIME;
d_instantiate(dentry, inode);
dget(dentry);
return 0;
}
static const char *shmem_follow_link(struct dentry *dentry, void **cookie)
{
struct page *page = NULL;
int error = shmem_getpage(d_inode(dentry), 0, &page, SGP_READ, NULL);
if (error)
return ERR_PTR(error);
unlock_page(page);
*cookie = page;
return kmap(page);
}
static void shmem_put_link(struct inode *unused, void *cookie)
{
struct page *page = cookie;
kunmap(page);
mark_page_accessed(page);
page_cache_release(page);
}
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#ifdef CONFIG_TMPFS_XATTR
/*
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
* Superblocks without xattr inode operations may get some security.* xattr
* support from the LSM "for free". As soon as we have any other xattrs
* like ACLs, we also need to implement the security.* handlers at
* filesystem level, though.
*/
/*
* Callback for security_inode_init_security() for acquiring xattrs.
*/
static int shmem_initxattrs(struct inode *inode,
const struct xattr *xattr_array,
void *fs_info)
{
struct shmem_inode_info *info = SHMEM_I(inode);
const struct xattr *xattr;
struct simple_xattr *new_xattr;
size_t len;
for (xattr = xattr_array; xattr->name != NULL; xattr++) {
new_xattr = simple_xattr_alloc(xattr->value, xattr->value_len);
if (!new_xattr)
return -ENOMEM;
len = strlen(xattr->name) + 1;
new_xattr->name = kmalloc(XATTR_SECURITY_PREFIX_LEN + len,
GFP_KERNEL);
if (!new_xattr->name) {
kfree(new_xattr);
return -ENOMEM;
}
memcpy(new_xattr->name, XATTR_SECURITY_PREFIX,
XATTR_SECURITY_PREFIX_LEN);
memcpy(new_xattr->name + XATTR_SECURITY_PREFIX_LEN,
xattr->name, len);
simple_xattr_list_add(&info->xattrs, new_xattr);
}
return 0;
}
static const struct xattr_handler *shmem_xattr_handlers[] = {
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#ifdef CONFIG_TMPFS_POSIX_ACL
&posix_acl_access_xattr_handler,
&posix_acl_default_xattr_handler,
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#endif
NULL
};
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
static int shmem_xattr_validate(const char *name)
{
struct { const char *prefix; size_t len; } arr[] = {
{ XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN },
{ XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN }
};
int i;
for (i = 0; i < ARRAY_SIZE(arr); i++) {
size_t preflen = arr[i].len;
if (strncmp(name, arr[i].prefix, preflen) == 0) {
if (!name[preflen])
return -EINVAL;
return 0;
}
}
return -EOPNOTSUPP;
}
static ssize_t shmem_getxattr(struct dentry *dentry, const char *name,
void *buffer, size_t size)
{
struct shmem_inode_info *info = SHMEM_I(d_inode(dentry));
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
int err;
/*
* If this is a request for a synthetic attribute in the system.*
* namespace use the generic infrastructure to resolve a handler
* for it via sb->s_xattr.
*/
if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
return generic_getxattr(dentry, name, buffer, size);
err = shmem_xattr_validate(name);
if (err)
return err;
return simple_xattr_get(&info->xattrs, name, buffer, size);
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
}
static int shmem_setxattr(struct dentry *dentry, const char *name,
const void *value, size_t size, int flags)
{
struct shmem_inode_info *info = SHMEM_I(d_inode(dentry));
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
int err;
/*
* If this is a request for a synthetic attribute in the system.*
* namespace use the generic infrastructure to resolve a handler
* for it via sb->s_xattr.
*/
if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
return generic_setxattr(dentry, name, value, size, flags);
err = shmem_xattr_validate(name);
if (err)
return err;
return simple_xattr_set(&info->xattrs, name, value, size, flags);
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
}
static int shmem_removexattr(struct dentry *dentry, const char *name)
{
struct shmem_inode_info *info = SHMEM_I(d_inode(dentry));
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
int err;
/*
* If this is a request for a synthetic attribute in the system.*
* namespace use the generic infrastructure to resolve a handler
* for it via sb->s_xattr.
*/
if (!strncmp(name, XATTR_SYSTEM_PREFIX, XATTR_SYSTEM_PREFIX_LEN))
return generic_removexattr(dentry, name);
err = shmem_xattr_validate(name);
if (err)
return err;
return simple_xattr_remove(&info->xattrs, name);
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
}
static ssize_t shmem_listxattr(struct dentry *dentry, char *buffer, size_t size)
{
struct shmem_inode_info *info = SHMEM_I(d_inode(dentry));
return simple_xattr_list(&info->xattrs, buffer, size);
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
}
#endif /* CONFIG_TMPFS_XATTR */
static const struct inode_operations shmem_short_symlink_operations = {
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
.readlink = generic_readlink,
.follow_link = simple_follow_link,
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#ifdef CONFIG_TMPFS_XATTR
.setxattr = shmem_setxattr,
.getxattr = shmem_getxattr,
.listxattr = shmem_listxattr,
.removexattr = shmem_removexattr,
#endif
};
static const struct inode_operations shmem_symlink_inode_operations = {
.readlink = generic_readlink,
.follow_link = shmem_follow_link,
.put_link = shmem_put_link,
#ifdef CONFIG_TMPFS_XATTR
.setxattr = shmem_setxattr,
.getxattr = shmem_getxattr,
.listxattr = shmem_listxattr,
.removexattr = shmem_removexattr,
#endif
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
};
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
static struct dentry *shmem_get_parent(struct dentry *child)
{
return ERR_PTR(-ESTALE);
}
static int shmem_match(struct inode *ino, void *vfh)
{
__u32 *fh = vfh;
__u64 inum = fh[2];
inum = (inum << 32) | fh[1];
return ino->i_ino == inum && fh[0] == ino->i_generation;
}
static struct dentry *shmem_fh_to_dentry(struct super_block *sb,
struct fid *fid, int fh_len, int fh_type)
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
{
struct inode *inode;
struct dentry *dentry = NULL;
u64 inum;
if (fh_len < 3)
return NULL;
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
inum = fid->raw[2];
inum = (inum << 32) | fid->raw[1];
inode = ilookup5(sb, (unsigned long)(inum + fid->raw[0]),
shmem_match, fid->raw);
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
if (inode) {
dentry = d_find_alias(inode);
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
iput(inode);
}
return dentry;
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
}
static int shmem_encode_fh(struct inode *inode, __u32 *fh, int *len,
struct inode *parent)
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
{
if (*len < 3) {
*len = 3;
return FILEID_INVALID;
}
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
if (inode_unhashed(inode)) {
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
/* Unfortunately insert_inode_hash is not idempotent,
* so as we hash inodes here rather than at creation
* time, we need a lock to ensure we only try
* to do it once
*/
static DEFINE_SPINLOCK(lock);
spin_lock(&lock);
if (inode_unhashed(inode))
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
__insert_inode_hash(inode,
inode->i_ino + inode->i_generation);
spin_unlock(&lock);
}
fh[0] = inode->i_generation;
fh[1] = inode->i_ino;
fh[2] = ((__u64)inode->i_ino) >> 32;
*len = 3;
return 1;
}
static const struct export_operations shmem_export_ops = {
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
.get_parent = shmem_get_parent,
.encode_fh = shmem_encode_fh,
.fh_to_dentry = shmem_fh_to_dentry,
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
};
static int shmem_parse_options(char *options, struct shmem_sb_info *sbinfo,
bool remount)
{
char *this_char, *value, *rest;
struct mempolicy *mpol = NULL;
uid_t uid;
gid_t gid;
while (options != NULL) {
this_char = options;
for (;;) {
/*
* NUL-terminate this option: unfortunately,
* mount options form a comma-separated list,
* but mpol's nodelist may also contain commas.
*/
options = strchr(options, ',');
if (options == NULL)
break;
options++;
if (!isdigit(*options)) {
options[-1] = '\0';
break;
}
}
if (!*this_char)
continue;
if ((value = strchr(this_char,'=')) != NULL) {
*value++ = 0;
} else {
printk(KERN_ERR
"tmpfs: No value for mount option '%s'\n",
this_char);
goto error;
}
if (!strcmp(this_char,"size")) {
unsigned long long size;
size = memparse(value,&rest);
if (*rest == '%') {
size <<= PAGE_SHIFT;
size *= totalram_pages;
do_div(size, 100);
rest++;
}
if (*rest)
goto bad_val;
sbinfo->max_blocks =
DIV_ROUND_UP(size, PAGE_CACHE_SIZE);
} else if (!strcmp(this_char,"nr_blocks")) {
sbinfo->max_blocks = memparse(value, &rest);
if (*rest)
goto bad_val;
} else if (!strcmp(this_char,"nr_inodes")) {
sbinfo->max_inodes = memparse(value, &rest);
if (*rest)
goto bad_val;
} else if (!strcmp(this_char,"mode")) {
if (remount)
continue;
sbinfo->mode = simple_strtoul(value, &rest, 8) & 07777;
if (*rest)
goto bad_val;
} else if (!strcmp(this_char,"uid")) {
if (remount)
continue;
uid = simple_strtoul(value, &rest, 0);
if (*rest)
goto bad_val;
sbinfo->uid = make_kuid(current_user_ns(), uid);
if (!uid_valid(sbinfo->uid))
goto bad_val;
} else if (!strcmp(this_char,"gid")) {
if (remount)
continue;
gid = simple_strtoul(value, &rest, 0);
if (*rest)
goto bad_val;
sbinfo->gid = make_kgid(current_user_ns(), gid);
if (!gid_valid(sbinfo->gid))
goto bad_val;
} else if (!strcmp(this_char,"mpol")) {
mpol_put(mpol);
mpol = NULL;
if (mpol_parse_str(value, &mpol))
goto bad_val;
} else {
printk(KERN_ERR "tmpfs: Bad mount option %s\n",
this_char);
goto error;
}
}
sbinfo->mpol = mpol;
return 0;
bad_val:
printk(KERN_ERR "tmpfs: Bad value '%s' for mount option '%s'\n",
value, this_char);
error:
mpol_put(mpol);
return 1;
}
static int shmem_remount_fs(struct super_block *sb, int *flags, char *data)
{
struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
struct shmem_sb_info config = *sbinfo;
unsigned long inodes;
int error = -EINVAL;
tmpfs: fix use-after-free of mempolicy object The tmpfs remount logic preserves filesystem mempolicy if the mpol=M option is not specified in the remount request. A new policy can be specified if mpol=M is given. Before this patch remounting an mpol bound tmpfs without specifying mpol= mount option in the remount request would set the filesystem's mempolicy object to a freed mempolicy object. To reproduce the problem boot a DEBUG_PAGEALLOC kernel and run: # mkdir /tmp/x # mount -t tmpfs -o size=100M,mpol=interleave nodev /tmp/x # grep /tmp/x /proc/mounts nodev /tmp/x tmpfs rw,relatime,size=102400k,mpol=interleave:0-3 0 0 # mount -o remount,size=200M nodev /tmp/x # grep /tmp/x /proc/mounts nodev /tmp/x tmpfs rw,relatime,size=204800k,mpol=??? 0 0 # note ? garbage in mpol=... output above # dd if=/dev/zero of=/tmp/x/f count=1 # panic here Panic: BUG: unable to handle kernel NULL pointer dereference at (null) IP: [< (null)>] (null) [...] Oops: 0010 [#1] SMP DEBUG_PAGEALLOC Call Trace: mpol_shared_policy_init+0xa5/0x160 shmem_get_inode+0x209/0x270 shmem_mknod+0x3e/0xf0 shmem_create+0x18/0x20 vfs_create+0xb5/0x130 do_last+0x9a1/0xea0 path_openat+0xb3/0x4d0 do_filp_open+0x42/0xa0 do_sys_open+0xfe/0x1e0 compat_sys_open+0x1b/0x20 cstar_dispatch+0x7/0x1f Non-debug kernels will not crash immediately because referencing the dangling mpol will not cause a fault. Instead the filesystem will reference a freed mempolicy object, which will cause unpredictable behavior. The problem boils down to a dropped mpol reference below if shmem_parse_options() does not allocate a new mpol: config = *sbinfo shmem_parse_options(data, &config, true) mpol_put(sbinfo->mpol) sbinfo->mpol = config.mpol /* BUG: saves unreferenced mpol */ This patch avoids the crash by not releasing the mempolicy if shmem_parse_options() doesn't create a new mpol. How far back does this issue go? I see it in both 2.6.36 and 3.3. I did not look back further. Signed-off-by: Greg Thelen <gthelen@google.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-23 08:36:01 +08:00
config.mpol = NULL;
if (shmem_parse_options(data, &config, true))
return error;
spin_lock(&sbinfo->stat_lock);
inodes = sbinfo->max_inodes - sbinfo->free_inodes;
if (percpu_counter_compare(&sbinfo->used_blocks, config.max_blocks) > 0)
goto out;
if (config.max_inodes < inodes)
goto out;
/*
* Those tests disallow limited->unlimited while any are in use;
* but we must separately disallow unlimited->limited, because
* in that case we have no record of how much is already in use.
*/
if (config.max_blocks && !sbinfo->max_blocks)
goto out;
if (config.max_inodes && !sbinfo->max_inodes)
goto out;
error = 0;
sbinfo->max_blocks = config.max_blocks;
sbinfo->max_inodes = config.max_inodes;
sbinfo->free_inodes = config.max_inodes - inodes;
mempolicy: use struct mempolicy pointer in shmem_sb_info This patch replaces the mempolicy mode, mode_flags, and nodemask in the shmem_sb_info struct with a struct mempolicy pointer, initialized to NULL. This removes dependency on the details of mempolicy from shmem.c and hugetlbfs inode.c and simplifies the interfaces. mpol_parse_str() in mempolicy.c is changed to return, via a pointer to a pointer arg, a struct mempolicy pointer on success. For MPOL_DEFAULT, the returned pointer is NULL. Further, mpol_parse_str() now takes a 'no_context' argument that causes the input nodemask to be stored in the w.user_nodemask of the created mempolicy for use when the mempolicy is installed in a tmpfs inode shared policy tree. At that time, any cpuset contextualization is applied to the original input nodemask. This preserves the previous behavior where the input nodemask was stored in the superblock. We can think of the returned mempolicy as "context free". Because mpol_parse_str() is now calling mpol_new(), we can remove from mpol_to_str() the semantic checks that mpol_new() already performs. Add 'no_context' parameter to mpol_to_str() to specify that it should format the nodemask in w.user_nodemask for 'bind' and 'interleave' policies. Change mpol_shared_policy_init() to take a pointer to a "context free" struct mempolicy and to create a new, "contextualized" mempolicy using the mode, mode_flags and user_nodemask from the input mempolicy. Note: we know that the mempolicy passed to mpol_to_str() or mpol_shared_policy_init() from a tmpfs superblock is "context free". This is currently the only instance thereof. However, if we found more uses for this concept, and introduced any ambiguity as to whether a mempolicy was context free or not, we could add another internal mode flag to identify context free mempolicies. Then, we could remove the 'no_context' argument from mpol_to_str(). Added shmem_get_sbmpol() to return a reference counted superblock mempolicy, if one exists, to pass to mpol_shared_policy_init(). We must add the reference under the sb stat_lock to prevent races with replacement of the mpol by remount. This reference is removed in mpol_shared_policy_init(). [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: yet another build fix] Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:26 +08:00
tmpfs: fix use-after-free of mempolicy object The tmpfs remount logic preserves filesystem mempolicy if the mpol=M option is not specified in the remount request. A new policy can be specified if mpol=M is given. Before this patch remounting an mpol bound tmpfs without specifying mpol= mount option in the remount request would set the filesystem's mempolicy object to a freed mempolicy object. To reproduce the problem boot a DEBUG_PAGEALLOC kernel and run: # mkdir /tmp/x # mount -t tmpfs -o size=100M,mpol=interleave nodev /tmp/x # grep /tmp/x /proc/mounts nodev /tmp/x tmpfs rw,relatime,size=102400k,mpol=interleave:0-3 0 0 # mount -o remount,size=200M nodev /tmp/x # grep /tmp/x /proc/mounts nodev /tmp/x tmpfs rw,relatime,size=204800k,mpol=??? 0 0 # note ? garbage in mpol=... output above # dd if=/dev/zero of=/tmp/x/f count=1 # panic here Panic: BUG: unable to handle kernel NULL pointer dereference at (null) IP: [< (null)>] (null) [...] Oops: 0010 [#1] SMP DEBUG_PAGEALLOC Call Trace: mpol_shared_policy_init+0xa5/0x160 shmem_get_inode+0x209/0x270 shmem_mknod+0x3e/0xf0 shmem_create+0x18/0x20 vfs_create+0xb5/0x130 do_last+0x9a1/0xea0 path_openat+0xb3/0x4d0 do_filp_open+0x42/0xa0 do_sys_open+0xfe/0x1e0 compat_sys_open+0x1b/0x20 cstar_dispatch+0x7/0x1f Non-debug kernels will not crash immediately because referencing the dangling mpol will not cause a fault. Instead the filesystem will reference a freed mempolicy object, which will cause unpredictable behavior. The problem boils down to a dropped mpol reference below if shmem_parse_options() does not allocate a new mpol: config = *sbinfo shmem_parse_options(data, &config, true) mpol_put(sbinfo->mpol) sbinfo->mpol = config.mpol /* BUG: saves unreferenced mpol */ This patch avoids the crash by not releasing the mempolicy if shmem_parse_options() doesn't create a new mpol. How far back does this issue go? I see it in both 2.6.36 and 3.3. I did not look back further. Signed-off-by: Greg Thelen <gthelen@google.com> Acked-by: Hugh Dickins <hughd@google.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-23 08:36:01 +08:00
/*
* Preserve previous mempolicy unless mpol remount option was specified.
*/
if (config.mpol) {
mpol_put(sbinfo->mpol);
sbinfo->mpol = config.mpol; /* transfers initial ref */
}
out:
spin_unlock(&sbinfo->stat_lock);
return error;
}
static int shmem_show_options(struct seq_file *seq, struct dentry *root)
{
struct shmem_sb_info *sbinfo = SHMEM_SB(root->d_sb);
if (sbinfo->max_blocks != shmem_default_max_blocks())
seq_printf(seq, ",size=%luk",
sbinfo->max_blocks << (PAGE_CACHE_SHIFT - 10));
if (sbinfo->max_inodes != shmem_default_max_inodes())
seq_printf(seq, ",nr_inodes=%lu", sbinfo->max_inodes);
if (sbinfo->mode != (S_IRWXUGO | S_ISVTX))
seq_printf(seq, ",mode=%03ho", sbinfo->mode);
if (!uid_eq(sbinfo->uid, GLOBAL_ROOT_UID))
seq_printf(seq, ",uid=%u",
from_kuid_munged(&init_user_ns, sbinfo->uid));
if (!gid_eq(sbinfo->gid, GLOBAL_ROOT_GID))
seq_printf(seq, ",gid=%u",
from_kgid_munged(&init_user_ns, sbinfo->gid));
mempolicy: use struct mempolicy pointer in shmem_sb_info This patch replaces the mempolicy mode, mode_flags, and nodemask in the shmem_sb_info struct with a struct mempolicy pointer, initialized to NULL. This removes dependency on the details of mempolicy from shmem.c and hugetlbfs inode.c and simplifies the interfaces. mpol_parse_str() in mempolicy.c is changed to return, via a pointer to a pointer arg, a struct mempolicy pointer on success. For MPOL_DEFAULT, the returned pointer is NULL. Further, mpol_parse_str() now takes a 'no_context' argument that causes the input nodemask to be stored in the w.user_nodemask of the created mempolicy for use when the mempolicy is installed in a tmpfs inode shared policy tree. At that time, any cpuset contextualization is applied to the original input nodemask. This preserves the previous behavior where the input nodemask was stored in the superblock. We can think of the returned mempolicy as "context free". Because mpol_parse_str() is now calling mpol_new(), we can remove from mpol_to_str() the semantic checks that mpol_new() already performs. Add 'no_context' parameter to mpol_to_str() to specify that it should format the nodemask in w.user_nodemask for 'bind' and 'interleave' policies. Change mpol_shared_policy_init() to take a pointer to a "context free" struct mempolicy and to create a new, "contextualized" mempolicy using the mode, mode_flags and user_nodemask from the input mempolicy. Note: we know that the mempolicy passed to mpol_to_str() or mpol_shared_policy_init() from a tmpfs superblock is "context free". This is currently the only instance thereof. However, if we found more uses for this concept, and introduced any ambiguity as to whether a mempolicy was context free or not, we could add another internal mode flag to identify context free mempolicies. Then, we could remove the 'no_context' argument from mpol_to_str(). Added shmem_get_sbmpol() to return a reference counted superblock mempolicy, if one exists, to pass to mpol_shared_policy_init(). We must add the reference under the sb stat_lock to prevent races with replacement of the mpol by remount. This reference is removed in mpol_shared_policy_init(). [akpm@linux-foundation.org: build fix] [akpm@linux-foundation.org: another build fix] [akpm@linux-foundation.org: yet another build fix] Signed-off-by: Lee Schermerhorn <lee.schermerhorn@hp.com> Cc: Christoph Lameter <clameter@sgi.com> Cc: David Rientjes <rientjes@google.com> Cc: Mel Gorman <mel@csn.ul.ie> Cc: Andi Kleen <ak@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2008-04-28 17:13:26 +08:00
shmem_show_mpol(seq, sbinfo->mpol);
return 0;
}
#define MFD_NAME_PREFIX "memfd:"
#define MFD_NAME_PREFIX_LEN (sizeof(MFD_NAME_PREFIX) - 1)
#define MFD_NAME_MAX_LEN (NAME_MAX - MFD_NAME_PREFIX_LEN)
#define MFD_ALL_FLAGS (MFD_CLOEXEC | MFD_ALLOW_SEALING)
SYSCALL_DEFINE2(memfd_create,
const char __user *, uname,
unsigned int, flags)
{
struct shmem_inode_info *info;
struct file *file;
int fd, error;
char *name;
long len;
if (flags & ~(unsigned int)MFD_ALL_FLAGS)
return -EINVAL;
/* length includes terminating zero */
len = strnlen_user(uname, MFD_NAME_MAX_LEN + 1);
if (len <= 0)
return -EFAULT;
if (len > MFD_NAME_MAX_LEN + 1)
return -EINVAL;
name = kmalloc(len + MFD_NAME_PREFIX_LEN, GFP_TEMPORARY);
if (!name)
return -ENOMEM;
strcpy(name, MFD_NAME_PREFIX);
if (copy_from_user(&name[MFD_NAME_PREFIX_LEN], uname, len)) {
error = -EFAULT;
goto err_name;
}
/* terminating-zero may have changed after strnlen_user() returned */
if (name[len + MFD_NAME_PREFIX_LEN - 1]) {
error = -EFAULT;
goto err_name;
}
fd = get_unused_fd_flags((flags & MFD_CLOEXEC) ? O_CLOEXEC : 0);
if (fd < 0) {
error = fd;
goto err_name;
}
file = shmem_file_setup(name, 0, VM_NORESERVE);
if (IS_ERR(file)) {
error = PTR_ERR(file);
goto err_fd;
}
info = SHMEM_I(file_inode(file));
file->f_mode |= FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE;
file->f_flags |= O_RDWR | O_LARGEFILE;
if (flags & MFD_ALLOW_SEALING)
info->seals &= ~F_SEAL_SEAL;
fd_install(fd, file);
kfree(name);
return fd;
err_fd:
put_unused_fd(fd);
err_name:
kfree(name);
return error;
}
#endif /* CONFIG_TMPFS */
static void shmem_put_super(struct super_block *sb)
{
struct shmem_sb_info *sbinfo = SHMEM_SB(sb);
percpu_counter_destroy(&sbinfo->used_blocks);
mpol_put(sbinfo->mpol);
kfree(sbinfo);
sb->s_fs_info = NULL;
}
Driver Core: devtmpfs - kernel-maintained tmpfs-based /dev Devtmpfs lets the kernel create a tmpfs instance called devtmpfs very early at kernel initialization, before any driver-core device is registered. Every device with a major/minor will provide a device node in devtmpfs. Devtmpfs can be changed and altered by userspace at any time, and in any way needed - just like today's udev-mounted tmpfs. Unmodified udev versions will run just fine on top of it, and will recognize an already existing kernel-created device node and use it. The default node permissions are root:root 0600. Proper permissions and user/group ownership, meaningful symlinks, all other policy still needs to be applied by userspace. If a node is created by devtmps, devtmpfs will remove the device node when the device goes away. If the device node was created by userspace, or the devtmpfs created node was replaced by userspace, it will no longer be removed by devtmpfs. If it is requested to auto-mount it, it makes init=/bin/sh work without any further userspace support. /dev will be fully populated and dynamic, and always reflect the current device state of the kernel. With the commonly used dynamic device numbers, it solves the problem where static devices nodes may point to the wrong devices. It is intended to make the initial bootup logic simpler and more robust, by de-coupling the creation of the inital environment, to reliably run userspace processes, from a complex userspace bootstrap logic to provide a working /dev. Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: Jan Blunck <jblunck@suse.de> Tested-By: Harald Hoyer <harald@redhat.com> Tested-By: Scott James Remnant <scott@ubuntu.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2009-04-30 21:23:42 +08:00
int shmem_fill_super(struct super_block *sb, void *data, int silent)
{
struct inode *inode;
struct shmem_sb_info *sbinfo;
int err = -ENOMEM;
/* Round up to L1_CACHE_BYTES to resist false sharing */
shmem: initialize struct shmem_sb_info to zero Fixes the following kmemcheck false positive (the compiler is using a 32-bit mov to load the 16-bit sbinfo->mode in shmem_fill_super): [ 0.337000] Total of 1 processors activated (3088.38 BogoMIPS). [ 0.352000] CPU0 attaching NULL sched-domain. [ 0.360000] WARNING: kmemcheck: Caught 32-bit read from uninitialized memory (9f8020fc) [ 0.361000] a44240820000000041f6998100000000000000000000000000000000ff030000 [ 0.368000] i i i i i i i i i i i i i i i i u u u u i i i i i i i i i i u u [ 0.375000] ^ [ 0.376000] [ 0.377000] Pid: 9, comm: khelper Not tainted (2.6.31-tip #206) P4DC6 [ 0.378000] EIP: 0060:[<810a3a95>] EFLAGS: 00010246 CPU: 0 [ 0.379000] EIP is at shmem_fill_super+0xb5/0x120 [ 0.380000] EAX: 00000000 EBX: 9f845400 ECX: 824042a4 EDX: 8199f641 [ 0.381000] ESI: 9f8020c0 EDI: 9f845400 EBP: 9f81af68 ESP: 81cd6eec [ 0.382000] DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068 [ 0.383000] CR0: 8005003b CR2: 9f806200 CR3: 01ccd000 CR4: 000006d0 [ 0.384000] DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000 [ 0.385000] DR6: ffff4ff0 DR7: 00000400 [ 0.386000] [<810c25fc>] get_sb_nodev+0x3c/0x80 [ 0.388000] [<810a3514>] shmem_get_sb+0x14/0x20 [ 0.390000] [<810c207f>] vfs_kern_mount+0x4f/0x120 [ 0.392000] [<81b2849e>] init_tmpfs+0x7e/0xb0 [ 0.394000] [<81b11597>] do_basic_setup+0x17/0x30 [ 0.396000] [<81b11907>] kernel_init+0x57/0xa0 [ 0.398000] [<810039b7>] kernel_thread_helper+0x7/0x10 [ 0.400000] [<ffffffff>] 0xffffffff [ 0.402000] khelper used greatest stack depth: 2820 bytes left [ 0.407000] calling init_mmap_min_addr+0x0/0x10 @ 1 [ 0.408000] initcall init_mmap_min_addr+0x0/0x10 returned 0 after 0 usecs Reported-by: Ingo Molnar <mingo@elte.hu> Analysed-by: Vegard Nossum <vegard.nossum@gmail.com> Signed-off-by: Pekka Enberg <penberg@cs.helsinki.fi> Acked-by: Hugh Dickins <hugh.dickins@tiscali.co.uk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-09-22 08:03:50 +08:00
sbinfo = kzalloc(max((int)sizeof(struct shmem_sb_info),
L1_CACHE_BYTES), GFP_KERNEL);
if (!sbinfo)
return -ENOMEM;
sbinfo->mode = S_IRWXUGO | S_ISVTX;
sbinfo->uid = current_fsuid();
sbinfo->gid = current_fsgid();
sb->s_fs_info = sbinfo;
#ifdef CONFIG_TMPFS
/*
* Per default we only allow half of the physical ram per
* tmpfs instance, limiting inodes to one per page of lowmem;
* but the internal instance is left unlimited.
*/
if (!(sb->s_flags & MS_KERNMOUNT)) {
sbinfo->max_blocks = shmem_default_max_blocks();
sbinfo->max_inodes = shmem_default_max_inodes();
if (shmem_parse_options(data, sbinfo, false)) {
err = -EINVAL;
goto failed;
}
} else {
sb->s_flags |= MS_NOUSER;
}
[PATCH] knfsd: add nfs-export support to tmpfs We need to encode a decode the 'file' part of a handle. We simply use the inode number and generation number to construct the filehandle. The generation number is the time when the file was created. As inode numbers cycle through the full 32 bits before being reused, there is no real chance of the same inum being allocated to different files in the same second so this is suitably unique. Using time-of-day rather than e.g. jiffies makes it less likely that the same filehandle can be created after a reboot. In order to be able to decode a filehandle we need to be able to lookup by inum, which means that the inode needs to be added to the inode hash table (tmpfs doesn't currently hash inodes as there is never a need to lookup by inum). To avoid overhead when not exporting, we only hash an inode when it is first exported. This requires a lock to ensure it isn't hashed twice. This code is separate from the patch posted in June06 from Atal Shargorodsky which provided the same functionality, but does borrow slightly from it. Locking comment: Most filesystems that hash their inodes do so at the point where the 'struct inode' is initialised, and that has suitable locking (I_NEW). Here in shmem, we are hashing the inode later, the first time we need an NFS file handle for it. We no longer have I_NEW to ensure only one thread tries to add it to the hash table. Cc: Atal Shargorodsky <atal@codefidence.com> Cc: Gilad Ben-Yossef <gilad@codefidence.com> Signed-off-by: David M. Grimes <dgrimes@navisite.com> Signed-off-by: Neil Brown <neilb@suse.de> Acked-by: Hugh Dickins <hugh@veritas.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-10-17 15:09:45 +08:00
sb->s_export_op = &shmem_export_ops;
sb->s_flags |= MS_NOSEC;
#else
sb->s_flags |= MS_NOUSER;
#endif
spin_lock_init(&sbinfo->stat_lock);
if (percpu_counter_init(&sbinfo->used_blocks, 0, GFP_KERNEL))
goto failed;
sbinfo->free_inodes = sbinfo->max_inodes;
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
sb->s_maxbytes = MAX_LFS_FILESIZE;
sb->s_blocksize = PAGE_CACHE_SIZE;
sb->s_blocksize_bits = PAGE_CACHE_SHIFT;
sb->s_magic = TMPFS_MAGIC;
sb->s_op = &shmem_ops;
sb->s_time_gran = 1;
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#ifdef CONFIG_TMPFS_XATTR
sb->s_xattr = shmem_xattr_handlers;
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#endif
#ifdef CONFIG_TMPFS_POSIX_ACL
sb->s_flags |= MS_POSIXACL;
#endif
inode = shmem_get_inode(sb, NULL, S_IFDIR | sbinfo->mode, 0, VM_NORESERVE);
if (!inode)
goto failed;
inode->i_uid = sbinfo->uid;
inode->i_gid = sbinfo->gid;
sb->s_root = d_make_root(inode);
if (!sb->s_root)
goto failed;
return 0;
failed:
shmem_put_super(sb);
return err;
}
static struct kmem_cache *shmem_inode_cachep;
static struct inode *shmem_alloc_inode(struct super_block *sb)
{
struct shmem_inode_info *info;
info = kmem_cache_alloc(shmem_inode_cachep, GFP_KERNEL);
if (!info)
return NULL;
return &info->vfs_inode;
}
static void shmem_destroy_callback(struct rcu_head *head)
2011-01-07 14:49:49 +08:00
{
struct inode *inode = container_of(head, struct inode, i_rcu);
kmem_cache_free(shmem_inode_cachep, SHMEM_I(inode));
}
static void shmem_destroy_inode(struct inode *inode)
{
if (S_ISREG(inode->i_mode))
mpol_free_shared_policy(&SHMEM_I(inode)->policy);
call_rcu(&inode->i_rcu, shmem_destroy_callback);
}
static void shmem_init_inode(void *foo)
{
struct shmem_inode_info *info = foo;
inode_init_once(&info->vfs_inode);
}
static int shmem_init_inodecache(void)
{
shmem_inode_cachep = kmem_cache_create("shmem_inode_cache",
sizeof(struct shmem_inode_info),
0, SLAB_PANIC, shmem_init_inode);
return 0;
}
static void shmem_destroy_inodecache(void)
{
kmem_cache_destroy(shmem_inode_cachep);
}
static const struct address_space_operations shmem_aops = {
.writepage = shmem_writepage,
.set_page_dirty = __set_page_dirty_no_writeback,
#ifdef CONFIG_TMPFS
.write_begin = shmem_write_begin,
.write_end = shmem_write_end,
#endif
#ifdef CONFIG_MIGRATION
.migratepage = migrate_page,
#endif
.error_remove_page = generic_error_remove_page,
};
static const struct file_operations shmem_file_operations = {
.mmap = shmem_mmap,
#ifdef CONFIG_TMPFS
.llseek = shmem_file_llseek,
.read_iter = shmem_file_read_iter,
.write_iter = generic_file_write_iter,
.fsync = noop_fsync,
.splice_read = shmem_file_splice_read,
.splice_write = iter_file_splice_write,
.fallocate = shmem_fallocate,
#endif
};
static const struct inode_operations shmem_inode_operations = {
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
.setattr = shmem_setattr,
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#ifdef CONFIG_TMPFS_XATTR
.setxattr = shmem_setxattr,
.getxattr = shmem_getxattr,
.listxattr = shmem_listxattr,
.removexattr = shmem_removexattr,
.set_acl = simple_set_acl,
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#endif
};
static const struct inode_operations shmem_dir_inode_operations = {
#ifdef CONFIG_TMPFS
.create = shmem_create,
.lookup = simple_lookup,
.link = shmem_link,
.unlink = shmem_unlink,
.symlink = shmem_symlink,
.mkdir = shmem_mkdir,
.rmdir = shmem_rmdir,
.mknod = shmem_mknod,
.rename2 = shmem_rename2,
.tmpfile = shmem_tmpfile,
#endif
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#ifdef CONFIG_TMPFS_XATTR
.setxattr = shmem_setxattr,
.getxattr = shmem_getxattr,
.listxattr = shmem_listxattr,
.removexattr = shmem_removexattr,
#endif
#ifdef CONFIG_TMPFS_POSIX_ACL
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
.setattr = shmem_setattr,
.set_acl = simple_set_acl,
#endif
};
static const struct inode_operations shmem_special_inode_operations = {
tmpfs: implement generic xattr support Implement generic xattrs for tmpfs filesystems. The Feodra project, while trying to replace suid apps with file capabilities, realized that tmpfs, which is used on the build systems, does not support file capabilities and thus cannot be used to build packages which use file capabilities. Xattrs are also needed for overlayfs. The xattr interface is a bit odd. If a filesystem does not implement any {get,set,list}xattr functions the VFS will call into some random LSM hooks and the running LSM can then implement some method for handling xattrs. SELinux for example provides a method to support security.selinux but no other security.* xattrs. As it stands today when one enables CONFIG_TMPFS_POSIX_ACL tmpfs will have xattr handler routines specifically to handle acls. Because of this tmpfs would loose the VFS/LSM helpers to support the running LSM. To make up for that tmpfs had stub functions that did nothing but call into the LSM hooks which implement the helpers. This new patch does not use the LSM fallback functions and instead just implements a native get/set/list xattr feature for the full security.* and trusted.* namespace like a normal filesystem. This means that tmpfs can now support both security.selinux and security.capability, which was not previously possible. The basic implementation is that I attach a: struct shmem_xattr { struct list_head list; /* anchored by shmem_inode_info->xattr_list */ char *name; size_t size; char value[0]; }; Into the struct shmem_inode_info for each xattr that is set. This implementation could easily support the user.* namespace as well, except some care needs to be taken to prevent large amounts of unswappable memory being allocated for unprivileged users. [mszeredi@suse.cz: new config option, suport trusted.*, support symlinks] Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: Miklos Szeredi <mszeredi@suse.cz> Acked-by: Serge Hallyn <serge.hallyn@ubuntu.com> Tested-by: Serge Hallyn <serge.hallyn@ubuntu.com> Cc: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Hugh Dickins <hughd@google.com> Tested-by: Jordi Pujol <jordipujolp@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-05-25 08:12:39 +08:00
#ifdef CONFIG_TMPFS_XATTR
.setxattr = shmem_setxattr,
.getxattr = shmem_getxattr,
.listxattr = shmem_listxattr,
.removexattr = shmem_removexattr,
#endif
#ifdef CONFIG_TMPFS_POSIX_ACL
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
.setattr = shmem_setattr,
.set_acl = simple_set_acl,
#endif
};
static const struct super_operations shmem_ops = {
.alloc_inode = shmem_alloc_inode,
.destroy_inode = shmem_destroy_inode,
#ifdef CONFIG_TMPFS
.statfs = shmem_statfs,
.remount_fs = shmem_remount_fs,
.show_options = shmem_show_options,
#endif
.evict_inode = shmem_evict_inode,
.drop_inode = generic_delete_inode,
.put_super = shmem_put_super,
};
static const struct vm_operations_struct shmem_vm_ops = {
mm: merge populate and nopage into fault (fixes nonlinear) Nonlinear mappings are (AFAIKS) simply a virtual memory concept that encodes the virtual address -> file offset differently from linear mappings. ->populate is a layering violation because the filesystem/pagecache code should need to know anything about the virtual memory mapping. The hitch here is that the ->nopage handler didn't pass down enough information (ie. pgoff). But it is more logical to pass pgoff rather than have the ->nopage function calculate it itself anyway (because that's a similar layering violation). Having the populate handler install the pte itself is likewise a nasty thing to be doing. This patch introduces a new fault handler that replaces ->nopage and ->populate and (later) ->nopfn. Most of the old mechanism is still in place so there is a lot of duplication and nice cleanups that can be removed if everyone switches over. The rationale for doing this in the first place is that nonlinear mappings are subject to the pagefault vs invalidate/truncate race too, and it seemed stupid to duplicate the synchronisation logic rather than just consolidate the two. After this patch, MAP_NONBLOCK no longer sets up ptes for pages present in pagecache. Seems like a fringe functionality anyway. NOPAGE_REFAULT is removed. This should be implemented with ->fault, and no users have hit mainline yet. [akpm@linux-foundation.org: cleanup] [randy.dunlap@oracle.com: doc. fixes for readahead] [akpm@linux-foundation.org: build fix] Signed-off-by: Nick Piggin <npiggin@suse.de> Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: Mark Fasheh <mark.fasheh@oracle.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-19 16:46:59 +08:00
.fault = shmem_fault,
.map_pages = filemap_map_pages,
#ifdef CONFIG_NUMA
.set_policy = shmem_set_policy,
.get_policy = shmem_get_policy,
#endif
};
static struct dentry *shmem_mount(struct file_system_type *fs_type,
int flags, const char *dev_name, void *data)
{
return mount_nodev(fs_type, flags, data, shmem_fill_super);
}
static struct file_system_type shmem_fs_type = {
.owner = THIS_MODULE,
.name = "tmpfs",
.mount = shmem_mount,
.kill_sb = kill_litter_super,
.fs_flags = FS_USERNS_MOUNT,
};
int __init shmem_init(void)
{
int error;
/* If rootfs called this, don't re-init */
if (shmem_inode_cachep)
return 0;
error = shmem_init_inodecache();
if (error)
goto out3;
error = register_filesystem(&shmem_fs_type);
if (error) {
printk(KERN_ERR "Could not register tmpfs\n");
goto out2;
}
shm_mnt = kern_mount(&shmem_fs_type);
if (IS_ERR(shm_mnt)) {
error = PTR_ERR(shm_mnt);
printk(KERN_ERR "Could not kern_mount tmpfs\n");
goto out1;
}
return 0;
out1:
unregister_filesystem(&shmem_fs_type);
out2:
shmem_destroy_inodecache();
out3:
shm_mnt = ERR_PTR(error);
return error;
}
#else /* !CONFIG_SHMEM */
/*
* tiny-shmem: simple shmemfs and tmpfs using ramfs code
*
* This is intended for small system where the benefits of the full
* shmem code (swap-backed and resource-limited) are outweighed by
* their complexity. On systems without swap this code should be
* effectively equivalent, but much lighter weight.
*/
static struct file_system_type shmem_fs_type = {
.name = "tmpfs",
.mount = ramfs_mount,
.kill_sb = kill_litter_super,
.fs_flags = FS_USERNS_MOUNT,
};
int __init shmem_init(void)
{
BUG_ON(register_filesystem(&shmem_fs_type) != 0);
shm_mnt = kern_mount(&shmem_fs_type);
BUG_ON(IS_ERR(shm_mnt));
return 0;
}
int shmem_unuse(swp_entry_t swap, struct page *page)
{
return 0;
}
int shmem_lock(struct file *file, int lock, struct user_struct *user)
{
return 0;
}
SHM_UNLOCK: fix Unevictable pages stranded after swap Commit cc39c6a9bbde ("mm: account skipped entries to avoid looping in find_get_pages") correctly fixed an infinite loop; but left a problem that find_get_pages() on shmem would return 0 (appearing to callers to mean end of tree) when it meets a run of nr_pages swap entries. The only uses of find_get_pages() on shmem are via pagevec_lookup(), called from invalidate_mapping_pages(), and from shmctl SHM_UNLOCK's scan_mapping_unevictable_pages(). The first is already commented, and not worth worrying about; but the second can leave pages on the Unevictable list after an unusual sequence of swapping and locking. Fix that by using shmem_find_get_pages_and_swap() (then ignoring the swap) instead of pagevec_lookup(). But I don't want to contaminate vmscan.c with shmem internals, nor shmem.c with LRU locking. So move scan_mapping_unevictable_pages() into shmem.c, renaming it shmem_unlock_mapping(); and rename check_move_unevictable_page() to check_move_unevictable_pages(), looping down an array of pages, oftentimes under the same lock. Leave out the "rotate unevictable list" block: that's a leftover from when this was used for /proc/sys/vm/scan_unevictable_pages, whose flawed handling involved looking at pages at tail of LRU. Was there significance to the sequence first ClearPageUnevictable, then test page_evictable, then SetPageUnevictable here? I think not, we're under LRU lock, and have no barriers between those. Signed-off-by: Hugh Dickins <hughd@google.com> Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Cc: Minchan Kim <minchan.kim@gmail.com> Cc: Rik van Riel <riel@redhat.com> Cc: Shaohua Li <shaohua.li@intel.com> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michel Lespinasse <walken@google.com> Cc: <stable@vger.kernel.org> [back to 3.1 but will need respins] Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2012-01-21 06:34:21 +08:00
void shmem_unlock_mapping(struct address_space *mapping)
{
}
void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend)
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
{
truncate_inode_pages_range(inode->i_mapping, lstart, lend);
tmpfs: take control of its truncate_range 2.6.35's new truncate convention gave tmpfs the opportunity to control its file truncation, no longer enforced from outside by vmtruncate(). We shall want to build upon that, to handle pagecache and swap together. Slightly redefine the ->truncate_range interface: let it now be called between the unmap_mapping_range()s, with the filesystem responsible for doing the truncate_inode_pages_range() from it - just as the filesystem is nowadays responsible for doing that from its ->setattr. Let's rename shmem_notify_change() to shmem_setattr(). Instead of calling the generic truncate_setsize(), bring that code in so we can call shmem_truncate_range() - which will later be updated to perform its own variant of truncate_inode_pages_range(). Remove the punch_hole unmap_mapping_range() from shmem_truncate_range(): now that the COW's unmap_mapping_range() comes after ->truncate_range, there is no need to call it a third time. Export shmem_truncate_range() and add it to the list in shmem_fs.h, so that i915_gem_object_truncate() can call it explicitly in future; get this patch in first, then update drm/i915 once this is available (until then, i915 will just be doing the truncate_inode_pages() twice). Though introduced five years ago, no other filesystem is implementing ->truncate_range, and its only other user is madvise(,,MADV_REMOVE): we expect to convert it to fallocate(,FALLOC_FL_PUNCH_HOLE,,) shortly, whereupon ->truncate_range can be removed from inode_operations - shmem_truncate_range() will help i915 across that transition too. Signed-off-by: Hugh Dickins <hughd@google.com> Cc: Christoph Hellwig <hch@infradead.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-06-28 07:18:03 +08:00
}
EXPORT_SYMBOL_GPL(shmem_truncate_range);
#define shmem_vm_ops generic_file_vm_ops
#define shmem_file_operations ramfs_file_operations
#define shmem_get_inode(sb, dir, mode, dev, flags) ramfs_get_inode(sb, dir, mode, dev)
#define shmem_acct_size(flags, size) 0
#define shmem_unacct_size(flags, size) do {} while (0)
#endif /* CONFIG_SHMEM */
/* common code */
static struct dentry_operations anon_ops = {
.d_dname = simple_dname
};
security: shmem: implement kernel private shmem inodes We have a problem where the big_key key storage implementation uses a shmem backed inode to hold the key contents. Because of this detail of implementation LSM checks are being done between processes trying to read the keys and the tmpfs backed inode. The LSM checks are already being handled on the key interface level and should not be enforced at the inode level (since the inode is an implementation detail, not a part of the security model) This patch implements a new function shmem_kernel_file_setup() which returns the equivalent to shmem_file_setup() only the underlying inode has S_PRIVATE set. This means that all LSM checks for the inode in question are skipped. It should only be used for kernel internal operations where the inode is not exposed to userspace without proper LSM checking. It is possible that some other users of shmem_file_setup() should use the new interface, but this has not been explored. Reproducing this bug is a little bit difficult. The steps I used on Fedora are: (1) Turn off selinux enforcing: setenforce 0 (2) Create a huge key k=`dd if=/dev/zero bs=8192 count=1 | keyctl padd big_key test-key @s` (3) Access the key in another context: runcon system_u:system_r:httpd_t:s0-s0:c0.c1023 keyctl print $k >/dev/null (4) Examine the audit logs: ausearch -m AVC -i --subject httpd_t | audit2allow If the last command's output includes a line that looks like: allow httpd_t user_tmpfs_t:file { open read }; There was an inode check between httpd and the tmpfs filesystem. With this patch no such denial will be seen. (NOTE! you should clear your audit log if you have tested for this previously) (Please return you box to enforcing) Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: David Howells <dhowells@redhat.com> cc: Hugh Dickins <hughd@google.com> cc: linux-mm@kvack.org
2013-12-02 19:24:19 +08:00
static struct file *__shmem_file_setup(const char *name, loff_t size,
unsigned long flags, unsigned int i_flags)
{
struct file *res;
struct inode *inode;
struct path path;
struct super_block *sb;
struct qstr this;
if (IS_ERR(shm_mnt))
return ERR_CAST(shm_mnt);
tmpfs: demolish old swap vector support The maximum size of a shmem/tmpfs file has been limited by the maximum size of its triple-indirect swap vector. With 4kB page size, maximum filesize was just over 2TB on a 32-bit kernel, but sadly one eighth of that on a 64-bit kernel. (With 8kB page size, maximum filesize was just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel, MAX_LFS_FILESIZE being then more restrictive than swap vector layout.) It's a shame that tmpfs should be more restrictive than ramfs, and this limitation has now been noticed. Add another level to the swap vector? No, it became obscure and hard to maintain, once I complicated it to make use of highmem pages nine years ago: better choose another way. Surely, if 2.4 had had the radix tree pagecache introduced in 2.5, then tmpfs would never have invented its own peculiar radix tree: we would have fitted swap entries into the common radix tree instead, in much the same way as we fit swap entries into page tables. And why should each file have a separate radix tree for its pages and for its swap entries? The swap entries are required precisely where and when the pages are not. We want to put them together in a single radix tree: which can then avoid much of the locking which was needed to prevent them from being exchanged underneath us. This also avoids the waste of memory devoted to swap vectors, first in the shmem_inode itself, then at least two more pages once a file grew beyond 16 data pages (pages accounted by df and du, but not by memcg). Allocated upfront, to avoid allocation when under swapping pressure, but pure waste when CONFIG_SWAP is not set - I have never spattered around the ifdefs to prevent that, preferring this move to sharing the common radix tree instead. There are three downsides to sharing the radix tree. One, that it binds tmpfs more tightly to the rest of mm, either requiring knowledge of swap entries in radix tree there, or duplication of its code here in shmem.c. I believe that the simplications and memory savings (and probable higher performance, not yet measured) justify that. Two, that on HIGHMEM systems with SWAP enabled, it's the lowmem radix nodes that cannot be freed under memory pressure - whereas before it was the less precious highmem swap vector pages that could not be freed. I'm hoping that 64-bit has now been accessible for long enough, that the highmem argument has grown much less persuasive. Three, that swapoff is slower than it used to be on tmpfs files, since it's using a simple generic mechanism not tailored to it: I find this noticeable, and shall want to improve, but maybe nobody else will notice. So... now remove most of the old swap vector code from shmem.c. But, for the moment, keep the simple i_direct vector of 16 pages, with simple accessors shmem_put_swap() and shmem_get_swap(), as a toy implementation to help mark where swap needs to be handled in subsequent patches. Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Rik van Riel <riel@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-04 07:21:20 +08:00
if (size < 0 || size > MAX_LFS_FILESIZE)
return ERR_PTR(-EINVAL);
if (shmem_acct_size(flags, size))
return ERR_PTR(-ENOMEM);
res = ERR_PTR(-ENOMEM);
this.name = name;
this.len = strlen(name);
this.hash = 0; /* will go */
sb = shm_mnt->mnt_sb;
path.mnt = mntget(shm_mnt);
path.dentry = d_alloc_pseudo(sb, &this);
if (!path.dentry)
goto put_memory;
d_set_d_op(path.dentry, &anon_ops);
res = ERR_PTR(-ENOSPC);
inode = shmem_get_inode(sb, NULL, S_IFREG | S_IRWXUGO, 0, flags);
if (!inode)
goto put_memory;
security: shmem: implement kernel private shmem inodes We have a problem where the big_key key storage implementation uses a shmem backed inode to hold the key contents. Because of this detail of implementation LSM checks are being done between processes trying to read the keys and the tmpfs backed inode. The LSM checks are already being handled on the key interface level and should not be enforced at the inode level (since the inode is an implementation detail, not a part of the security model) This patch implements a new function shmem_kernel_file_setup() which returns the equivalent to shmem_file_setup() only the underlying inode has S_PRIVATE set. This means that all LSM checks for the inode in question are skipped. It should only be used for kernel internal operations where the inode is not exposed to userspace without proper LSM checking. It is possible that some other users of shmem_file_setup() should use the new interface, but this has not been explored. Reproducing this bug is a little bit difficult. The steps I used on Fedora are: (1) Turn off selinux enforcing: setenforce 0 (2) Create a huge key k=`dd if=/dev/zero bs=8192 count=1 | keyctl padd big_key test-key @s` (3) Access the key in another context: runcon system_u:system_r:httpd_t:s0-s0:c0.c1023 keyctl print $k >/dev/null (4) Examine the audit logs: ausearch -m AVC -i --subject httpd_t | audit2allow If the last command's output includes a line that looks like: allow httpd_t user_tmpfs_t:file { open read }; There was an inode check between httpd and the tmpfs filesystem. With this patch no such denial will be seen. (NOTE! you should clear your audit log if you have tested for this previously) (Please return you box to enforcing) Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: David Howells <dhowells@redhat.com> cc: Hugh Dickins <hughd@google.com> cc: linux-mm@kvack.org
2013-12-02 19:24:19 +08:00
inode->i_flags |= i_flags;
d_instantiate(path.dentry, inode);
inode->i_size = size;
clear_nlink(inode); /* It is unlinked */
res = ERR_PTR(ramfs_nommu_expand_for_mapping(inode, size));
if (IS_ERR(res))
goto put_path;
res = alloc_file(&path, FMODE_WRITE | FMODE_READ,
&shmem_file_operations);
if (IS_ERR(res))
goto put_path;
return res;
put_memory:
shmem_unacct_size(flags, size);
put_path:
path_put(&path);
return res;
}
security: shmem: implement kernel private shmem inodes We have a problem where the big_key key storage implementation uses a shmem backed inode to hold the key contents. Because of this detail of implementation LSM checks are being done between processes trying to read the keys and the tmpfs backed inode. The LSM checks are already being handled on the key interface level and should not be enforced at the inode level (since the inode is an implementation detail, not a part of the security model) This patch implements a new function shmem_kernel_file_setup() which returns the equivalent to shmem_file_setup() only the underlying inode has S_PRIVATE set. This means that all LSM checks for the inode in question are skipped. It should only be used for kernel internal operations where the inode is not exposed to userspace without proper LSM checking. It is possible that some other users of shmem_file_setup() should use the new interface, but this has not been explored. Reproducing this bug is a little bit difficult. The steps I used on Fedora are: (1) Turn off selinux enforcing: setenforce 0 (2) Create a huge key k=`dd if=/dev/zero bs=8192 count=1 | keyctl padd big_key test-key @s` (3) Access the key in another context: runcon system_u:system_r:httpd_t:s0-s0:c0.c1023 keyctl print $k >/dev/null (4) Examine the audit logs: ausearch -m AVC -i --subject httpd_t | audit2allow If the last command's output includes a line that looks like: allow httpd_t user_tmpfs_t:file { open read }; There was an inode check between httpd and the tmpfs filesystem. With this patch no such denial will be seen. (NOTE! you should clear your audit log if you have tested for this previously) (Please return you box to enforcing) Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: David Howells <dhowells@redhat.com> cc: Hugh Dickins <hughd@google.com> cc: linux-mm@kvack.org
2013-12-02 19:24:19 +08:00
/**
* shmem_kernel_file_setup - get an unlinked file living in tmpfs which must be
* kernel internal. There will be NO LSM permission checks against the
* underlying inode. So users of this interface must do LSM checks at a
ipc: use private shmem or hugetlbfs inodes for shm segments. The shm implementation internally uses shmem or hugetlbfs inodes for shm segments. As these inodes are never directly exposed to userspace and only accessed through the shm operations which are already hooked by security modules, mark the inodes with the S_PRIVATE flag so that inode security initialization and permission checking is skipped. This was motivated by the following lockdep warning: ====================================================== [ INFO: possible circular locking dependency detected ] 4.2.0-0.rc3.git0.1.fc24.x86_64+debug #1 Tainted: G W ------------------------------------------------------- httpd/1597 is trying to acquire lock: (&ids->rwsem){+++++.}, at: shm_close+0x34/0x130 but task is already holding lock: (&mm->mmap_sem){++++++}, at: SyS_shmdt+0x4b/0x180 which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #3 (&mm->mmap_sem){++++++}: lock_acquire+0xc7/0x270 __might_fault+0x7a/0xa0 filldir+0x9e/0x130 xfs_dir2_block_getdents.isra.12+0x198/0x1c0 [xfs] xfs_readdir+0x1b4/0x330 [xfs] xfs_file_readdir+0x2b/0x30 [xfs] iterate_dir+0x97/0x130 SyS_getdents+0x91/0x120 entry_SYSCALL_64_fastpath+0x12/0x76 -> #2 (&xfs_dir_ilock_class){++++.+}: lock_acquire+0xc7/0x270 down_read_nested+0x57/0xa0 xfs_ilock+0x167/0x350 [xfs] xfs_ilock_attr_map_shared+0x38/0x50 [xfs] xfs_attr_get+0xbd/0x190 [xfs] xfs_xattr_get+0x3d/0x70 [xfs] generic_getxattr+0x4f/0x70 inode_doinit_with_dentry+0x162/0x670 sb_finish_set_opts+0xd9/0x230 selinux_set_mnt_opts+0x35c/0x660 superblock_doinit+0x77/0xf0 delayed_superblock_init+0x10/0x20 iterate_supers+0xb3/0x110 selinux_complete_init+0x2f/0x40 security_load_policy+0x103/0x600 sel_write_load+0xc1/0x750 __vfs_write+0x37/0x100 vfs_write+0xa9/0x1a0 SyS_write+0x58/0xd0 entry_SYSCALL_64_fastpath+0x12/0x76 ... Signed-off-by: Stephen Smalley <sds@tycho.nsa.gov> Reported-by: Morten Stevens <mstevens@fedoraproject.org> Acked-by: Hugh Dickins <hughd@google.com> Acked-by: Paul Moore <paul@paul-moore.com> Cc: Manfred Spraul <manfred@colorfullife.com> Cc: Davidlohr Bueso <dave@stgolabs.net> Cc: Prarit Bhargava <prarit@redhat.com> Cc: Eric Paris <eparis@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-08-07 06:46:55 +08:00
* higher layer. The users are the big_key and shm implementations. LSM
* checks are provided at the key or shm level rather than the inode.
security: shmem: implement kernel private shmem inodes We have a problem where the big_key key storage implementation uses a shmem backed inode to hold the key contents. Because of this detail of implementation LSM checks are being done between processes trying to read the keys and the tmpfs backed inode. The LSM checks are already being handled on the key interface level and should not be enforced at the inode level (since the inode is an implementation detail, not a part of the security model) This patch implements a new function shmem_kernel_file_setup() which returns the equivalent to shmem_file_setup() only the underlying inode has S_PRIVATE set. This means that all LSM checks for the inode in question are skipped. It should only be used for kernel internal operations where the inode is not exposed to userspace without proper LSM checking. It is possible that some other users of shmem_file_setup() should use the new interface, but this has not been explored. Reproducing this bug is a little bit difficult. The steps I used on Fedora are: (1) Turn off selinux enforcing: setenforce 0 (2) Create a huge key k=`dd if=/dev/zero bs=8192 count=1 | keyctl padd big_key test-key @s` (3) Access the key in another context: runcon system_u:system_r:httpd_t:s0-s0:c0.c1023 keyctl print $k >/dev/null (4) Examine the audit logs: ausearch -m AVC -i --subject httpd_t | audit2allow If the last command's output includes a line that looks like: allow httpd_t user_tmpfs_t:file { open read }; There was an inode check between httpd and the tmpfs filesystem. With this patch no such denial will be seen. (NOTE! you should clear your audit log if you have tested for this previously) (Please return you box to enforcing) Signed-off-by: Eric Paris <eparis@redhat.com> Signed-off-by: David Howells <dhowells@redhat.com> cc: Hugh Dickins <hughd@google.com> cc: linux-mm@kvack.org
2013-12-02 19:24:19 +08:00
* @name: name for dentry (to be seen in /proc/<pid>/maps
* @size: size to be set for the file
* @flags: VM_NORESERVE suppresses pre-accounting of the entire object size
*/
struct file *shmem_kernel_file_setup(const char *name, loff_t size, unsigned long flags)
{
return __shmem_file_setup(name, size, flags, S_PRIVATE);
}
/**
* shmem_file_setup - get an unlinked file living in tmpfs
* @name: name for dentry (to be seen in /proc/<pid>/maps
* @size: size to be set for the file
* @flags: VM_NORESERVE suppresses pre-accounting of the entire object size
*/
struct file *shmem_file_setup(const char *name, loff_t size, unsigned long flags)
{
return __shmem_file_setup(name, size, flags, 0);
}
EXPORT_SYMBOL_GPL(shmem_file_setup);
/**
* shmem_zero_setup - setup a shared anonymous mapping
* @vma: the vma to be mmapped is prepared by do_mmap_pgoff
*/
int shmem_zero_setup(struct vm_area_struct *vma)
{
struct file *file;
loff_t size = vma->vm_end - vma->vm_start;
/*
* Cloning a new file under mmap_sem leads to a lock ordering conflict
* between XFS directory reading and selinux: since this file is only
* accessible to the user through its mapping, use S_PRIVATE flag to
* bypass file security, in the same way as shmem_kernel_file_setup().
*/
file = __shmem_file_setup("dev/zero", size, vma->vm_flags, S_PRIVATE);
if (IS_ERR(file))
return PTR_ERR(file);
if (vma->vm_file)
fput(vma->vm_file);
vma->vm_file = file;
vma->vm_ops = &shmem_vm_ops;
return 0;
}
/**
* shmem_read_mapping_page_gfp - read into page cache, using specified page allocation flags.
* @mapping: the page's address_space
* @index: the page index
* @gfp: the page allocator flags to use if allocating
*
* This behaves as a tmpfs "read_cache_page_gfp(mapping, index, gfp)",
* with any new page allocations done using the specified allocation flags.
* But read_cache_page_gfp() uses the ->readpage() method: which does not
* suit tmpfs, since it may have pages in swapcache, and needs to find those
* for itself; although drivers/gpu/drm i915 and ttm rely upon this support.
*
* i915_gem_object_get_pages_gtt() mixes __GFP_NORETRY | __GFP_NOWARN in
* with the mapping_gfp_mask(), to avoid OOMing the machine unnecessarily.
*/
struct page *shmem_read_mapping_page_gfp(struct address_space *mapping,
pgoff_t index, gfp_t gfp)
{
#ifdef CONFIG_SHMEM
struct inode *inode = mapping->host;
struct page *page;
int error;
BUG_ON(mapping->a_ops != &shmem_aops);
error = shmem_getpage_gfp(inode, index, &page, SGP_CACHE, gfp, NULL);
if (error)
page = ERR_PTR(error);
else
unlock_page(page);
return page;
#else
/*
* The tiny !SHMEM case uses ramfs without swap
*/
return read_cache_page_gfp(mapping, index, gfp);
#endif
}
EXPORT_SYMBOL_GPL(shmem_read_mapping_page_gfp);