OpenCloudOS-Kernel/fs/9p/vfs_dir.c

254 lines
5.4 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-only
/*
* This file contains vfs directory ops for the 9P2000 protocol.
*
* Copyright (C) 2004 by Eric Van Hensbergen <ericvh@gmail.com>
* Copyright (C) 2002 by Ron Minnich <rminnich@lanl.gov>
*/
#include <linux/module.h>
#include <linux/errno.h>
#include <linux/fs.h>
#include <linux/file.h>
#include <linux/stat.h>
#include <linux/string.h>
#include <linux/sched.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
#include <linux/uio.h>
9p: Use fscache indexing rewrite and reenable caching Change the 9p filesystem to take account of the changes to fscache's indexing rewrite and reenable caching in 9p. The following changes have been made: (1) The fscache_netfs struct is no more, and there's no need to register the filesystem as a whole. (2) The session cookie is now an fscache_volume cookie, allocated with fscache_acquire_volume(). That takes three parameters: a string representing the "volume" in the index, a string naming the cache to use (or NULL) and a u64 that conveys coherency metadata for the volume. For 9p, I've made it render the volume name string as: "9p,<devname>,<cachetag>" where the cachetag is replaced by the aname if it wasn't supplied. This probably needs rethinking a bit as the aname can have slashes in it. It might be better to hash the cachetag and use the hash or I could substitute commas for the slashes or something. (3) The fscache_cookie_def is no more and needed information is passed directly to fscache_acquire_cookie(). The cache no longer calls back into the filesystem, but rather metadata changes are indicated at other times. fscache_acquire_cookie() is passed the same keying and coherency information as before. (4) The functions to set/reset/flush cookies are removed and fscache_use_cookie() and fscache_unuse_cookie() are used instead. fscache_use_cookie() is passed a flag to indicate if the cookie is opened for writing. fscache_unuse_cookie() is passed updates for the metadata if we changed it (ie. if the file was opened for writing). These are called when the file is opened or closed. (5) wait_on_page_bit[_killable]() is replaced with the specific wait functions for the bits waited upon. (6) I've got rid of some of the 9p-specific cache helper functions and called things like fscache_relinquish_cookie() directly as they'll optimise away if v9fs_inode_cookie() returns an unconditional NULL (which will be the case if CONFIG_9P_FSCACHE=n). (7) v9fs_vfs_setattr() is made to call fscache_resize() to change the size of the cache object. Notes: (A) We should call fscache_invalidate() if we detect that the server's copy of a file got changed by a third party, but I don't know where to do that. We don't need to do that when allocating the cookie as we get a check-and-invalidate when we initially bind to the cache object. (B) The copy-to-cache-on-writeback side of things will be handled in separate patch. Changes ======= ver #3: - Canonicalise the cookie key and coherency data to make them endianness-independent. ver #2: - Use gfpflags_allow_blocking() rather than using flag directly. - fscache_acquire_volume() now returns errors. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Jeff Layton <jlayton@kernel.org> Tested-by: Dominique Martinet <asmadeus@codewreck.org> cc: Eric Van Hensbergen <ericvh@gmail.com> cc: Latchesar Ionkov <lucho@ionkov.net> cc: v9fs-developer@lists.sourceforge.net cc: linux-cachefs@redhat.com Link: https://lore.kernel.org/r/163819664645.215744.1555314582005286846.stgit@warthog.procyon.org.uk/ # v1 Link: https://lore.kernel.org/r/163906975017.143852.3459573173204394039.stgit@warthog.procyon.org.uk/ # v2 Link: https://lore.kernel.org/r/163967178512.1823006.17377493641569138183.stgit@warthog.procyon.org.uk/ # v3 Link: https://lore.kernel.org/r/164021573143.640689.3977487095697717967.stgit@warthog.procyon.org.uk/ # v4
2020-11-18 17:06:42 +08:00
#include <linux/fscache.h>
#include <net/9p/9p.h>
#include <net/9p/client.h>
#include "v9fs.h"
#include "v9fs_vfs.h"
#include "fid.h"
/**
* struct p9_rdir - readdir accounting
* @head: start offset of current dirread buffer
* @tail: end offset of current dirread buffer
* @buf: dirread buffer
*
* private structure for keeping track of readdir
* allocated on demand
*/
struct p9_rdir {
int head;
int tail;
uint8_t buf[];
};
/**
* dt_type - return file type
* @mistat: mistat structure
*
*/
static inline int dt_type(struct p9_wstat *mistat)
{
unsigned long perm = mistat->mode;
int rettype = DT_REG;
if (perm & P9_DMDIR)
rettype = DT_DIR;
if (perm & P9_DMSYMLINK)
rettype = DT_LNK;
return rettype;
}
/**
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
* v9fs_alloc_rdir_buf - Allocate buffer used for read and readdir
* @filp: opened file structure
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
* @buflen: Length in bytes of buffer to allocate
*
*/
static struct p9_rdir *v9fs_alloc_rdir_buf(struct file *filp, int buflen)
{
struct p9_fid *fid = filp->private_data;
if (!fid->rdir)
fid->rdir = kzalloc(sizeof(struct p9_rdir) + buflen, GFP_KERNEL);
return fid->rdir;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
}
/**
* v9fs_dir_readdir - iterate through a directory
* @file: opened file structure
* @ctx: actor we feed the entries to
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
*
*/
static int v9fs_dir_readdir(struct file *file, struct dir_context *ctx)
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
{
bool over;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
struct p9_wstat st;
int err = 0;
struct p9_fid *fid;
int buflen;
struct p9_rdir *rdir;
struct kvec kvec;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
p9_debug(P9_DEBUG_VFS, "name %pD\n", file);
fid = file->private_data;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
buflen = fid->clnt->msize - P9_IOHDRSZ;
rdir = v9fs_alloc_rdir_buf(file, buflen);
if (!rdir)
return -ENOMEM;
kvec.iov_base = rdir->buf;
kvec.iov_len = buflen;
while (1) {
if (rdir->tail == rdir->head) {
struct iov_iter to;
int n;
iov_iter_kvec(&to, ITER_DEST, &kvec, 1, buflen);
n = p9_client_read(file->private_data, ctx->pos, &to,
&err);
if (err)
return err;
if (n == 0)
return 0;
rdir->head = 0;
rdir->tail = n;
}
while (rdir->head < rdir->tail) {
err = p9stat_read(fid->clnt, rdir->buf + rdir->head,
rdir->tail - rdir->head, &st);
if (err <= 0) {
p9_debug(P9_DEBUG_VFS, "returned %d\n", err);
return -EIO;
}
over = !dir_emit(ctx, st.name, strlen(st.name),
v9fs_qid2ino(&st.qid), dt_type(&st));
p9stat_free(&st);
if (over)
return 0;
rdir->head += err;
ctx->pos += err;
}
}
}
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
/**
* v9fs_dir_readdir_dotl - iterate through a directory
* @file: opened file structure
* @ctx: actor we feed the entries to
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
*
*/
static int v9fs_dir_readdir_dotl(struct file *file, struct dir_context *ctx)
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
{
int err = 0;
struct p9_fid *fid;
int buflen;
struct p9_rdir *rdir;
struct p9_dirent curdirent;
p9_debug(P9_DEBUG_VFS, "name %pD\n", file);
fid = file->private_data;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
buflen = fid->clnt->msize - P9_READDIRHDRSZ;
rdir = v9fs_alloc_rdir_buf(file, buflen);
if (!rdir)
return -ENOMEM;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
while (1) {
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
if (rdir->tail == rdir->head) {
err = p9_client_readdir(fid, rdir->buf, buflen,
ctx->pos);
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
if (err <= 0)
return err;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
rdir->head = 0;
rdir->tail = err;
}
while (rdir->head < rdir->tail) {
err = p9dirent_read(fid->clnt, rdir->buf + rdir->head,
rdir->tail - rdir->head,
&curdirent);
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
if (err < 0) {
p9_debug(P9_DEBUG_VFS, "returned %d\n", err);
return -EIO;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
}
if (!dir_emit(ctx, curdirent.d_name,
strlen(curdirent.d_name),
v9fs_qid2ino(&curdirent.qid),
curdirent.d_type))
return 0;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
ctx->pos = curdirent.d_off;
9p: readdir implementation for 9p2000.L This patch implements the kernel part of readdir() implementation for 9p2000.L Change from V3: Instead of inode, server now sends qids for each dirent SYNOPSIS size[4] Treaddir tag[2] fid[4] offset[8] count[4] size[4] Rreaddir tag[2] count[4] data[count] DESCRIPTION The readdir request asks the server to read the directory specified by 'fid' at an offset specified by 'offset' and return as many dirent structures as possible that fit into count bytes. Each dirent structure is laid out as follows. qid.type[1] the type of the file (directory, etc.), represented as a bit vector corresponding to the high 8 bits of the file's mode word. qid.vers[4] version number for given path qid.path[8] the file server's unique identification for the file offset[8] offset into the next dirent. type[1] type of this directory entry. name[256] name of this directory entry. This patch adds v9fs_dir_readdir_dotl() as the readdir() call for 9p2000.L. This function sends P9_TREADDIR command to the server. In response the server sends a buffer filled with dirent structures. This is different from the existing v9fs_dir_readdir() call which receives stat structures from the server. This results in significant speedup of readdir() on large directories. For example, doing 'ls >/dev/null' on a directory with 10000 files on my laptop takes 1.088 seconds with the existing code, but only takes 0.339 seconds with the new readdir. Signed-off-by: Sripathi Kodi <sripathik@in.ibm.com> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> Signed-off-by: Eric Van Hensbergen <ericvh@gmail.com>
2010-06-04 21:41:26 +08:00
rdir->head += err;
}
}
}
/**
* v9fs_dir_release - close a directory or a file
* @inode: inode of the directory or file
* @filp: file pointer to a directory or file
*
*/
int v9fs_dir_release(struct inode *inode, struct file *filp)
{
9p: Use fscache indexing rewrite and reenable caching Change the 9p filesystem to take account of the changes to fscache's indexing rewrite and reenable caching in 9p. The following changes have been made: (1) The fscache_netfs struct is no more, and there's no need to register the filesystem as a whole. (2) The session cookie is now an fscache_volume cookie, allocated with fscache_acquire_volume(). That takes three parameters: a string representing the "volume" in the index, a string naming the cache to use (or NULL) and a u64 that conveys coherency metadata for the volume. For 9p, I've made it render the volume name string as: "9p,<devname>,<cachetag>" where the cachetag is replaced by the aname if it wasn't supplied. This probably needs rethinking a bit as the aname can have slashes in it. It might be better to hash the cachetag and use the hash or I could substitute commas for the slashes or something. (3) The fscache_cookie_def is no more and needed information is passed directly to fscache_acquire_cookie(). The cache no longer calls back into the filesystem, but rather metadata changes are indicated at other times. fscache_acquire_cookie() is passed the same keying and coherency information as before. (4) The functions to set/reset/flush cookies are removed and fscache_use_cookie() and fscache_unuse_cookie() are used instead. fscache_use_cookie() is passed a flag to indicate if the cookie is opened for writing. fscache_unuse_cookie() is passed updates for the metadata if we changed it (ie. if the file was opened for writing). These are called when the file is opened or closed. (5) wait_on_page_bit[_killable]() is replaced with the specific wait functions for the bits waited upon. (6) I've got rid of some of the 9p-specific cache helper functions and called things like fscache_relinquish_cookie() directly as they'll optimise away if v9fs_inode_cookie() returns an unconditional NULL (which will be the case if CONFIG_9P_FSCACHE=n). (7) v9fs_vfs_setattr() is made to call fscache_resize() to change the size of the cache object. Notes: (A) We should call fscache_invalidate() if we detect that the server's copy of a file got changed by a third party, but I don't know where to do that. We don't need to do that when allocating the cookie as we get a check-and-invalidate when we initially bind to the cache object. (B) The copy-to-cache-on-writeback side of things will be handled in separate patch. Changes ======= ver #3: - Canonicalise the cookie key and coherency data to make them endianness-independent. ver #2: - Use gfpflags_allow_blocking() rather than using flag directly. - fscache_acquire_volume() now returns errors. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Jeff Layton <jlayton@kernel.org> Tested-by: Dominique Martinet <asmadeus@codewreck.org> cc: Eric Van Hensbergen <ericvh@gmail.com> cc: Latchesar Ionkov <lucho@ionkov.net> cc: v9fs-developer@lists.sourceforge.net cc: linux-cachefs@redhat.com Link: https://lore.kernel.org/r/163819664645.215744.1555314582005286846.stgit@warthog.procyon.org.uk/ # v1 Link: https://lore.kernel.org/r/163906975017.143852.3459573173204394039.stgit@warthog.procyon.org.uk/ # v2 Link: https://lore.kernel.org/r/163967178512.1823006.17377493641569138183.stgit@warthog.procyon.org.uk/ # v3 Link: https://lore.kernel.org/r/164021573143.640689.3977487095697717967.stgit@warthog.procyon.org.uk/ # v4
2020-11-18 17:06:42 +08:00
struct v9fs_inode *v9inode = V9FS_I(inode);
struct p9_fid *fid;
9p: Use fscache indexing rewrite and reenable caching Change the 9p filesystem to take account of the changes to fscache's indexing rewrite and reenable caching in 9p. The following changes have been made: (1) The fscache_netfs struct is no more, and there's no need to register the filesystem as a whole. (2) The session cookie is now an fscache_volume cookie, allocated with fscache_acquire_volume(). That takes three parameters: a string representing the "volume" in the index, a string naming the cache to use (or NULL) and a u64 that conveys coherency metadata for the volume. For 9p, I've made it render the volume name string as: "9p,<devname>,<cachetag>" where the cachetag is replaced by the aname if it wasn't supplied. This probably needs rethinking a bit as the aname can have slashes in it. It might be better to hash the cachetag and use the hash or I could substitute commas for the slashes or something. (3) The fscache_cookie_def is no more and needed information is passed directly to fscache_acquire_cookie(). The cache no longer calls back into the filesystem, but rather metadata changes are indicated at other times. fscache_acquire_cookie() is passed the same keying and coherency information as before. (4) The functions to set/reset/flush cookies are removed and fscache_use_cookie() and fscache_unuse_cookie() are used instead. fscache_use_cookie() is passed a flag to indicate if the cookie is opened for writing. fscache_unuse_cookie() is passed updates for the metadata if we changed it (ie. if the file was opened for writing). These are called when the file is opened or closed. (5) wait_on_page_bit[_killable]() is replaced with the specific wait functions for the bits waited upon. (6) I've got rid of some of the 9p-specific cache helper functions and called things like fscache_relinquish_cookie() directly as they'll optimise away if v9fs_inode_cookie() returns an unconditional NULL (which will be the case if CONFIG_9P_FSCACHE=n). (7) v9fs_vfs_setattr() is made to call fscache_resize() to change the size of the cache object. Notes: (A) We should call fscache_invalidate() if we detect that the server's copy of a file got changed by a third party, but I don't know where to do that. We don't need to do that when allocating the cookie as we get a check-and-invalidate when we initially bind to the cache object. (B) The copy-to-cache-on-writeback side of things will be handled in separate patch. Changes ======= ver #3: - Canonicalise the cookie key and coherency data to make them endianness-independent. ver #2: - Use gfpflags_allow_blocking() rather than using flag directly. - fscache_acquire_volume() now returns errors. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Jeff Layton <jlayton@kernel.org> Tested-by: Dominique Martinet <asmadeus@codewreck.org> cc: Eric Van Hensbergen <ericvh@gmail.com> cc: Latchesar Ionkov <lucho@ionkov.net> cc: v9fs-developer@lists.sourceforge.net cc: linux-cachefs@redhat.com Link: https://lore.kernel.org/r/163819664645.215744.1555314582005286846.stgit@warthog.procyon.org.uk/ # v1 Link: https://lore.kernel.org/r/163906975017.143852.3459573173204394039.stgit@warthog.procyon.org.uk/ # v2 Link: https://lore.kernel.org/r/163967178512.1823006.17377493641569138183.stgit@warthog.procyon.org.uk/ # v3 Link: https://lore.kernel.org/r/164021573143.640689.3977487095697717967.stgit@warthog.procyon.org.uk/ # v4
2020-11-18 17:06:42 +08:00
__le32 version;
loff_t i_size;
int retval = 0;
fid = filp->private_data;
p9_debug(P9_DEBUG_VFS, "inode: %p filp: %p fid: %d\n",
inode, filp, fid ? fid->fid : -1);
9p: add refcount to p9_fid struct Fix race issue in fid contention. Eric's and Greg's patch offer a mechanism to fix open-unlink-f*syscall bug in 9p. But there is race issue in fid parallel accesses. As Greg's patch stores all of fids from opened files into according inode, so all the lookup fid ops can retrieve fid from inode preferentially. But there is no mechanism to handle the fid contention issue. For example, there are two threads get the same fid in the same time and one of them clunk the fid before the other thread ready to discard the fid. In this scenario, it will lead to some fatal problems, even kernel core dump. I introduce a mechanism to fix this race issue. A counter field introduced into p9_fid struct to store the reference counter to the fid. When a fid is allocated from the inode or dentry, the counter will increase, and will decrease at the end of its occupation. It is guaranteed that the fid won't be clunked before the reference counter go down to 0, then we can avoid the clunked fid to be used. tests: race issue test from the old test case: for file in {01..50}; do touch f.${file}; done seq 1 1000 | xargs -n 1 -P 50 -I{} cat f.* > /dev/null open-unlink-f*syscall test: I have tested for f*syscall include: ftruncate fstat fchown fchmod faccessat. Link: http://lkml.kernel.org/r/20200923141146.90046-5-jianyong.wu@arm.com Fixes: 478ba09edc1f ("fs/9p: search open fids first") Signed-off-by: Jianyong Wu <jianyong.wu@arm.com> Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
2020-09-23 22:11:46 +08:00
if (fid) {
if ((S_ISREG(inode->i_mode)) && (filp->f_mode & FMODE_WRITE))
retval = filemap_fdatawrite(inode->i_mapping);
9p: add refcount to p9_fid struct Fix race issue in fid contention. Eric's and Greg's patch offer a mechanism to fix open-unlink-f*syscall bug in 9p. But there is race issue in fid parallel accesses. As Greg's patch stores all of fids from opened files into according inode, so all the lookup fid ops can retrieve fid from inode preferentially. But there is no mechanism to handle the fid contention issue. For example, there are two threads get the same fid in the same time and one of them clunk the fid before the other thread ready to discard the fid. In this scenario, it will lead to some fatal problems, even kernel core dump. I introduce a mechanism to fix this race issue. A counter field introduced into p9_fid struct to store the reference counter to the fid. When a fid is allocated from the inode or dentry, the counter will increase, and will decrease at the end of its occupation. It is guaranteed that the fid won't be clunked before the reference counter go down to 0, then we can avoid the clunked fid to be used. tests: race issue test from the old test case: for file in {01..50}; do touch f.${file}; done seq 1 1000 | xargs -n 1 -P 50 -I{} cat f.* > /dev/null open-unlink-f*syscall test: I have tested for f*syscall include: ftruncate fstat fchown fchmod faccessat. Link: http://lkml.kernel.org/r/20200923141146.90046-5-jianyong.wu@arm.com Fixes: 478ba09edc1f ("fs/9p: search open fids first") Signed-off-by: Jianyong Wu <jianyong.wu@arm.com> Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
2020-09-23 22:11:46 +08:00
spin_lock(&inode->i_lock);
hlist_del(&fid->ilist);
spin_unlock(&inode->i_lock);
retval = p9_fid_put(fid);
9p: add refcount to p9_fid struct Fix race issue in fid contention. Eric's and Greg's patch offer a mechanism to fix open-unlink-f*syscall bug in 9p. But there is race issue in fid parallel accesses. As Greg's patch stores all of fids from opened files into according inode, so all the lookup fid ops can retrieve fid from inode preferentially. But there is no mechanism to handle the fid contention issue. For example, there are two threads get the same fid in the same time and one of them clunk the fid before the other thread ready to discard the fid. In this scenario, it will lead to some fatal problems, even kernel core dump. I introduce a mechanism to fix this race issue. A counter field introduced into p9_fid struct to store the reference counter to the fid. When a fid is allocated from the inode or dentry, the counter will increase, and will decrease at the end of its occupation. It is guaranteed that the fid won't be clunked before the reference counter go down to 0, then we can avoid the clunked fid to be used. tests: race issue test from the old test case: for file in {01..50}; do touch f.${file}; done seq 1 1000 | xargs -n 1 -P 50 -I{} cat f.* > /dev/null open-unlink-f*syscall test: I have tested for f*syscall include: ftruncate fstat fchown fchmod faccessat. Link: http://lkml.kernel.org/r/20200923141146.90046-5-jianyong.wu@arm.com Fixes: 478ba09edc1f ("fs/9p: search open fids first") Signed-off-by: Jianyong Wu <jianyong.wu@arm.com> Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>
2020-09-23 22:11:46 +08:00
}
9p: Use fscache indexing rewrite and reenable caching Change the 9p filesystem to take account of the changes to fscache's indexing rewrite and reenable caching in 9p. The following changes have been made: (1) The fscache_netfs struct is no more, and there's no need to register the filesystem as a whole. (2) The session cookie is now an fscache_volume cookie, allocated with fscache_acquire_volume(). That takes three parameters: a string representing the "volume" in the index, a string naming the cache to use (or NULL) and a u64 that conveys coherency metadata for the volume. For 9p, I've made it render the volume name string as: "9p,<devname>,<cachetag>" where the cachetag is replaced by the aname if it wasn't supplied. This probably needs rethinking a bit as the aname can have slashes in it. It might be better to hash the cachetag and use the hash or I could substitute commas for the slashes or something. (3) The fscache_cookie_def is no more and needed information is passed directly to fscache_acquire_cookie(). The cache no longer calls back into the filesystem, but rather metadata changes are indicated at other times. fscache_acquire_cookie() is passed the same keying and coherency information as before. (4) The functions to set/reset/flush cookies are removed and fscache_use_cookie() and fscache_unuse_cookie() are used instead. fscache_use_cookie() is passed a flag to indicate if the cookie is opened for writing. fscache_unuse_cookie() is passed updates for the metadata if we changed it (ie. if the file was opened for writing). These are called when the file is opened or closed. (5) wait_on_page_bit[_killable]() is replaced with the specific wait functions for the bits waited upon. (6) I've got rid of some of the 9p-specific cache helper functions and called things like fscache_relinquish_cookie() directly as they'll optimise away if v9fs_inode_cookie() returns an unconditional NULL (which will be the case if CONFIG_9P_FSCACHE=n). (7) v9fs_vfs_setattr() is made to call fscache_resize() to change the size of the cache object. Notes: (A) We should call fscache_invalidate() if we detect that the server's copy of a file got changed by a third party, but I don't know where to do that. We don't need to do that when allocating the cookie as we get a check-and-invalidate when we initially bind to the cache object. (B) The copy-to-cache-on-writeback side of things will be handled in separate patch. Changes ======= ver #3: - Canonicalise the cookie key and coherency data to make them endianness-independent. ver #2: - Use gfpflags_allow_blocking() rather than using flag directly. - fscache_acquire_volume() now returns errors. Signed-off-by: David Howells <dhowells@redhat.com> Acked-by: Jeff Layton <jlayton@kernel.org> Tested-by: Dominique Martinet <asmadeus@codewreck.org> cc: Eric Van Hensbergen <ericvh@gmail.com> cc: Latchesar Ionkov <lucho@ionkov.net> cc: v9fs-developer@lists.sourceforge.net cc: linux-cachefs@redhat.com Link: https://lore.kernel.org/r/163819664645.215744.1555314582005286846.stgit@warthog.procyon.org.uk/ # v1 Link: https://lore.kernel.org/r/163906975017.143852.3459573173204394039.stgit@warthog.procyon.org.uk/ # v2 Link: https://lore.kernel.org/r/163967178512.1823006.17377493641569138183.stgit@warthog.procyon.org.uk/ # v3 Link: https://lore.kernel.org/r/164021573143.640689.3977487095697717967.stgit@warthog.procyon.org.uk/ # v4
2020-11-18 17:06:42 +08:00
if ((filp->f_mode & FMODE_WRITE)) {
version = cpu_to_le32(v9inode->qid.version);
i_size = i_size_read(inode);
fscache_unuse_cookie(v9fs_inode_cookie(v9inode),
&version, &i_size);
} else {
fscache_unuse_cookie(v9fs_inode_cookie(v9inode), NULL, NULL);
}
return retval;
}
const struct file_operations v9fs_dir_operations = {
.read = generic_read_dir,
.llseek = generic_file_llseek,
.iterate_shared = v9fs_dir_readdir,
.open = v9fs_file_open,
.release = v9fs_dir_release,
};
const struct file_operations v9fs_dir_operations_dotl = {
.read = generic_read_dir,
.llseek = generic_file_llseek,
.iterate_shared = v9fs_dir_readdir_dotl,
.open = v9fs_file_open,
.release = v9fs_dir_release,
.fsync = v9fs_file_fsync_dotl,
};