Flatten out nested code structure in huge_pte_offset()
and huge_pte_alloc().
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adds support for 16GB hugepage size. To use this page size
use kernel parameters as:
default_hugepagesz=16G hugepagesz=16G hugepages=10
Testing:
Tested with the stream benchmark which allocates 48G of
arrays backed by 16G hugepages and does RW operation on
them in parallel.
Orabug: 25362942
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
get_user_pages() is used to do direct IO. It already
handles the case where the address range is backed
by PMD huge pages. This patch now adds the case where
the range could be backed by PUD huge pages.
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pull sparc fixes from David Miller:
- Fix symbol version generation for assembler on sparc, from
Nagarathnam Muthusamy.
- Fix compound page handling in gup_huge_pmd(), from Nitin Gupta.
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc:
sparc64: Fix gup_huge_pmd
Adding the type of exported symbols
sed regex in Makefile.build requires line break between exported symbols
Adding asm-prototypes.h for genksyms to generate crc
A poisoned or migrated hugepage is stored as a swap entry in the page
tables. On architectures that support hugepages consisting of
contiguous page table entries (such as on arm64) this leads to ambiguity
in determining the page table entry to return in huge_pte_offset() when
a poisoned entry is encountered.
Let's remove the ambiguity by adding a size parameter to convey
additional information about the requested address. Also fixup the
definition/usage of huge_pte_offset() throughout the tree.
Link: http://lkml.kernel.org/r/20170522133604.11392-4-punit.agrawal@arm.com
Signed-off-by: Punit Agrawal <punit.agrawal@arm.com>
Acked-by: Steve Capper <steve.capper@arm.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: James Hogan <james.hogan@imgtec.com> (odd fixer:METAG ARCHITECTURE)
Cc: Ralf Baechle <ralf@linux-mips.org> (supporter:MIPS)
Cc: "James E.J. Bottomley" <jejb@parisc-linux.org>
Cc: Helge Deller <deller@gmx.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Rich Felker <dalias@libc.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Chris Metcalf <cmetcalf@mellanox.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: "Aneesh Kumar K.V" <aneesh.kumar@linux.vnet.ibm.com>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The function assumes that each PMD points to head of a
huge page. This is not correct as a PMD can point to
start of any 8M region with a, say 256M, hugepage. The
fix ensures that it points to the correct head of any PMD
huge page.
Cc: Julian Calaby <julian.calaby@gmail.com>
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Stack guard page is a useful feature to reduce a risk of stack smashing
into a different mapping. We have been using a single page gap which
is sufficient to prevent having stack adjacent to a different mapping.
But this seems to be insufficient in the light of the stack usage in
userspace. E.g. glibc uses as large as 64kB alloca() in many commonly
used functions. Others use constructs liks gid_t buffer[NGROUPS_MAX]
which is 256kB or stack strings with MAX_ARG_STRLEN.
This will become especially dangerous for suid binaries and the default
no limit for the stack size limit because those applications can be
tricked to consume a large portion of the stack and a single glibc call
could jump over the guard page. These attacks are not theoretical,
unfortunatelly.
Make those attacks less probable by increasing the stack guard gap
to 1MB (on systems with 4k pages; but make it depend on the page size
because systems with larger base pages might cap stack allocations in
the PAGE_SIZE units) which should cover larger alloca() and VLA stack
allocations. It is obviously not a full fix because the problem is
somehow inherent, but it should reduce attack space a lot.
One could argue that the gap size should be configurable from userspace,
but that can be done later when somebody finds that the new 1MB is wrong
for some special case applications. For now, add a kernel command line
option (stack_guard_gap) to specify the stack gap size (in page units).
Implementation wise, first delete all the old code for stack guard page:
because although we could get away with accounting one extra page in a
stack vma, accounting a larger gap can break userspace - case in point,
a program run with "ulimit -S -v 20000" failed when the 1MB gap was
counted for RLIMIT_AS; similar problems could come with RLIMIT_MLOCK
and strict non-overcommit mode.
Instead of keeping gap inside the stack vma, maintain the stack guard
gap as a gap between vmas: using vm_start_gap() in place of vm_start
(or vm_end_gap() in place of vm_end if VM_GROWSUP) in just those few
places which need to respect the gap - mainly arch_get_unmapped_area(),
and and the vma tree's subtree_gap support for that.
Original-patch-by: Oleg Nesterov <oleg@redhat.com>
Original-patch-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Hugh Dickins <hughd@google.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Tested-by: Helge Deller <deller@gmx.de> # parisc
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The old method that is using xcall and softint to get new context id is
deleted, as it is replaced by a method of using per_cpu_secondary_mm
without xcall to perform the context wrap.
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Reviewed-by: Bob Picco <bob.picco@oracle.com>
Reviewed-by: Steven Sistare <steven.sistare@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The current wrap implementation has a race issue: it is called outside of
the ctx_alloc_lock, and also does not wait for all CPUs to complete the
wrap. This means that a thread can get a new context with a new version
and another thread might still be running with the same context. The
problem is especially severe on CPUs with shared TLBs, like sun4v. I used
the following test to very quickly reproduce the problem:
- start over 8K processes (must be more than context IDs)
- write and read values at a memory location in every process.
Very quickly memory corruptions start happening, and what we read back
does not equal what we wrote.
Several approaches were explored before settling on this one:
Approach 1:
Move smp_new_mmu_context_version() inside ctx_alloc_lock, and wait for
every process to complete the wrap. (Note: every CPU must WAIT before
leaving smp_new_mmu_context_version_client() until every one arrives).
This approach ends up with deadlocks, as some threads own locks which other
threads are waiting for, and they never receive softint until these threads
exit smp_new_mmu_context_version_client(). Since we do not allow the exit,
deadlock happens.
Approach 2:
Handle wrap right during mondo interrupt. Use etrap/rtrap to enter into
into C code, and issue new versions to every CPU.
This approach adds some overhead to runtime: in switch_mm() we must add
some checks to make sure that versions have not changed due to wrap while
we were loading the new secondary context. (could be protected by PSTATE_IE
but that degrades performance as on M7 and older CPUs as it takes 50 cycles
for each access). Also, we still need a global per-cpu array of MMs to know
where we need to load new contexts, otherwise we can change context to a
thread that is going way (if we received mondo between switch_mm() and
switch_to() time). Finally, there are some issues with window registers in
rtrap() when context IDs are changed during CPU mondo time.
The approach in this patch is the simplest and has almost no impact on
runtime. We use the array with mm's where last secondary contexts were
loaded onto CPUs and bump their versions to the new generation without
changing context IDs. If a new process comes in to get a context ID, it
will go through get_new_mmu_context() because of version mismatch. But the
running processes do not need to be interrupted. And wrap is quicker as we
do not need to xcall and wait for everyone to receive and complete wrap.
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Reviewed-by: Bob Picco <bob.picco@oracle.com>
Reviewed-by: Steven Sistare <steven.sistare@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The new wrap is going to use information from this array to figure out
mm's that currently have valid secondary contexts setup.
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Reviewed-by: Bob Picco <bob.picco@oracle.com>
Reviewed-by: Steven Sistare <steven.sistare@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
CTX_FIRST_VERSION defines the first context version, but also it defines
first context. This patch redefines it to only include the first context
version.
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Reviewed-by: Bob Picco <bob.picco@oracle.com>
Reviewed-by: Steven Sistare <steven.sistare@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
After a wrap (getting a new context version) a process must get a new
context id, which means that we would need to flush the context id from
the TLB before running for the first time with this ID on every CPU. But,
we use mm_cpumask to determine if this process has been running on this CPU
before, and this mask is not reset after a wrap. So, there are two possible
fixes for this issue:
1. Clear mm cpumask whenever mm gets a new context id
2. Unconditionally flush context every time process is running on a CPU
This patch implements the first solution
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Reviewed-by: Bob Picco <bob.picco@oracle.com>
Reviewed-by: Steven Sistare <steven.sistare@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
hugetlb_bad_size needs to be called on invalid values. Also change the
pr_warn to a pr_err to better align with other platforms.
Signed-off-by: Liam R. Howlett <Liam.Howlett@Oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a TSB grows beyond its current capacity, a new TSB is allocated
and copy_tsb is called to copy entries from the old TSB to the new.
A hash shift based on page size is used to calculate the index of an
entry in the TSB. copy_tsb has hard coded PAGE_SHIFT in these
calculations. However, for huge page TSBs the value REAL_HPAGE_SHIFT
should be used. As a result, when copy_tsb is called for a huge page
TSB the entries are placed at the incorrect index in the newly
allocated TSB. When doing hardware table walk, the MMU does not
match these entries and we end up in the TSB miss handling code.
This code will then create and write an entry to the correct index
in the TSB. We take a performance hit for the table walk miss and
recreation of these entries.
Pass a new parameter to copy_tsb that is the page size shift to be
used when copying the TSB.
Suggested-by: Anthony Yznaga <anthony.yznaga@oracle.com>
Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Greetings,
GCC 7 introduced the -Wstringop-overflow flag to detect buffer overflows
in calls to string handling functions [1][2]. Due to the way
``empty_zero_page'' is declared in arch/sparc/include/setup.h, this
causes a warning to trigger at compile time in the function mem_init(),
which is subsequently converted to an error. The ensuing patch fixes
this issue and aligns the declaration of empty_zero_page to that of
other architectures. Thank you.
Cheers,
Orlando.
[1] https://gcc.gnu.org/ml/gcc-patches/2016-10/msg02308.html
[2] https://gcc.gnu.org/gcc-7/changes.html
Signed-off-by: Orlando Arias <oarias@knights.ucf.edu>
--------------------------------------------------------------------------------
Signed-off-by: David S. Miller <davem@davemloft.net>
Make sure the start adderess is aligned to PMD_SIZE
boundary when freeing page table backing a hugepage
region. The issue was causing segfaults when a region
backed by 64K pages was unmapped since such a region
is in general not PMD_SIZE aligned.
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sparc32:allmodconfig fails to build with the following error.
ERROR: "vac_cache_size" [drivers/infiniband/sw/rxe/rdma_rxe.ko] undefined!
Fixes: cb88645596 ("infiniband: Fix alignment of mmap cookies ...")
Cc: Jason Gunthorpe <jgunthorpe@obsidianresearch.com>
Cc: Doug Ledford <dledford@redhat.com>
Signed-off-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
The memory corruption was happening due to incorrect
TLB/TSB flushing of hugepages.
Reported-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
I encountered this bug when using /proc/kcore to examine the kernel. Plus a
coworker inquired about debugging tools. We computed pa but did
not use it during the maximum physical address bits test. Instead we used
the identity mapped virtual address which will always fail this test.
I believe the defect came in here:
[bpicco@zareason linus.git]$ git describe --contains bb4e6e85da
v3.18-rc1~87^2~4
.
Signed-off-by: Bob Picco <bob.picco@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Update code that relied on sched.h including various MM types for them.
This will allow us to remove the <linux/mm_types.h> include from <linux/sched.h>.
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
We are going to split <linux/sched/debug.h> out of <linux/sched.h>, which
will have to be picked up from other headers and a couple of .c files.
Create a trivial placeholder <linux/sched/debug.h> file that just
maps to <linux/sched.h> to make this patch obviously correct and
bisectable.
Include the new header in the files that are going to need it.
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
We are going to split more MM APIs out of <linux/sched.h>, which
will have to be picked up from a couple of .c files.
The APIs that we are going to move are:
arch_pick_mmap_layout()
arch_get_unmapped_area()
arch_get_unmapped_area_topdown()
mm_update_next_owner()
Include the header in the files that are going to need it.
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Patch "sparc64: Add 64K page size support"
unconditionally used __flush_huge_tsb_one_entry()
which is available only when hugetlb support is
enabled.
Another issue was incorrect TSB flushing for 64K
pages in flush_tsb_user().
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In add_node_ranges() when memblock resize happens, the iterator keeps using
the previous freed array. This bug cause hangs on machine where there are
over 128 memory blocks during boot. For example, on machines where memory
interleaving is small.
The problem is seen on T4-4 because it cant have 2T of memory, and memory
is interleaved at 8G. So we have 2T/8G = 256 regions to set node IDs. The
starting size of regions array is 128. Thus, we have to double at least one
time (actually we have to double twice because some memory is already
reserved and thus we need more than 256 regions). We start using an
incorrect pointer to the array after the first doubling.
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Signed-off-by: Babu Moger <babu.moger@oracle.com>
Reviewed-by: Babu Moger <babu.moger@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
add_node_ranges() takes 2.6s - 3.6s per 1T of boot time. On machine with 6T
memory it takes 15.4s, on 32T it would take 82s-115s of boot time.
This function sets NUMA ids for memory blocks, and scans the whole memory a
page at a time to do so. But, we could use values in latency groups mask
and match to determine the boundaries without checking every single page.
With the fix the add_node_ranges() time is reduced from 15.4s down to 0.2s
on machine with 6T memory.
Signed-off-by: Pavel Tatashin <pasha.tatashin@oracle.com>
Reviewed-by: Babu Moger <babu.moger@oracle.com>
Reviewed-by: Bob Picco <bob.picco@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch depends on:
[v6] sparc64: Multi-page size support
- Testing
Tested on Sonoma by running stream benchmark instance which allocated
48G worth of 64K pages.
boot params: default_hugepagesz=64K hugepagesz=64K hugepages=1310720
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for using multiple hugepage sizes simultaneously
on mainline. Currently, support for 256M has been added which
can be used along with 8M pages.
Page tables are set like this (e.g. for 256M page):
VA + (8M * x) -> PA + (8M * x) (sz bit = 256M) where x in [0, 31]
and TSB is set similarly:
VA + (4M * x) -> PA + (4M * x) (sz bit = 256M) where x in [0, 63]
- Testing
Tested on Sonoma (which supports 256M pages) by running stream
benchmark instances in parallel: one instance uses 8M pages and
another uses 256M pages, consuming 48G each.
Boot params used:
default_hugepagesz=256M hugepagesz=256M hugepages=300 hugepagesz=8M
hugepages=10000
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The objects viking_ops, viking_sun4d_smp_ops and smp_cachetlb_ops of
type sparc32_cachetlb_ops are not modified anywhere after getting modified
in the init functions. Inside init their reference is also stored in a
pointer of type const struct sparc32_cachetlb_ops *. So these structures
are never modified after init, therefore add __ro_after to the declaration
of these structures.
Signed-off-by: Bhumika Goyal <bhumirks@gmail.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
We have a generic implementation for quite some time already. If there
is any arch specific information to be printed then we should add a
callback called from the generic code rather than duplicate the whole
show_mem.
The current code has resulted in the code duplication and the output
divergence which is both confusing and adds maintainance costs.
Let's just get rid of this mess.
Link: http://lkml.kernel.org/r/20170117091543.25850-4-mhocko@kernel.org
Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Guan Xuetao <gxt@mprc.pku.edu.cn> [UniCore32]
Acked-by: Helge Deller <deller@gmx.de> [for parisc]
Acked-by: Chris Metcalf <cmetcalf@mellanox.com> [for tile]
Acked-by: Mel Gorman <mgorman@suse.de>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: "James E.J. Bottomley" <jejb@parisc-linux.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Hillf Danton <hillf.zj@alibaba-inc.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: David Rientjes <rientjes@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This file was using module.h for search_exception_table. We've
now separated that content out into its own file "extable.h" so
now move over to that. Unlike most other instances, we can't
delete the module.h include here since the file needs that for
the within_module_init definition.
Reported-by: kbuild test robot <lkp@intel.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: sparclinux@vger.kernel.org
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
This was entirely automated, using the script by Al:
PATT='^[[:blank:]]*#[[:blank:]]*include[[:blank:]]*<asm/uaccess.h>'
sed -i -e "s!$PATT!#include <linux/uaccess.h>!" \
$(git grep -l "$PATT"|grep -v ^include/linux/uaccess.h)
to do the replacement at the end of the merge window.
Requested-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
A compile warning is introduced by a commit to fix the find_node().
This patch fix the compile warning by moving find_node() into __init
section. Because find_node() is only used by memblock_nid_range() which
is only used by a __init add_node_ranges(). find_node() and
memblock_nid_range() should also be inside __init section.
Signed-off-by: Thomas Tai <thomas.tai@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When booting up LDOM, find_node() warns that a physical address
doesn't match a NUMA node.
WARNING: CPU: 0 PID: 0 at arch/sparc/mm/init_64.c:835
find_node+0xf4/0x120 find_node: A physical address doesn't
match a NUMA node rule. Some physical memory will be
owned by node 0.Modules linked in:
CPU: 0 PID: 0 Comm: swapper Not tainted 4.9.0-rc3 #4
Call Trace:
[0000000000468ba0] __warn+0xc0/0xe0
[0000000000468c74] warn_slowpath_fmt+0x34/0x60
[00000000004592f4] find_node+0xf4/0x120
[0000000000dd0774] add_node_ranges+0x38/0xe4
[0000000000dd0b1c] numa_parse_mdesc+0x268/0x2e4
[0000000000dd0e9c] bootmem_init+0xb8/0x160
[0000000000dd174c] paging_init+0x808/0x8fc
[0000000000dcb0d0] setup_arch+0x2c8/0x2f0
[0000000000dc68a0] start_kernel+0x48/0x424
[0000000000dcb374] start_early_boot+0x27c/0x28c
[0000000000a32c08] tlb_fixup_done+0x4c/0x64
[0000000000027f08] 0x27f08
It is because linux use an internal structure node_masks[] to
keep the best memory latency node only. However, LDOM mdesc can
contain single latency-group with multiple memory latency nodes.
If the address doesn't match the best latency node within
node_masks[], it should check for an alternative via mdesc.
The warning message should only be printed if the address
doesn't match any node_masks[] nor within mdesc. To minimize
the impact of searching mdesc every time, the last matched
mask and index is stored in a variable.
Signed-off-by: Thomas Tai <thomas.tai@oracle.com>
Reviewed-by: Chris Hyser <chris.hyser@oracle.com>
Reviewed-by: Liam Merwick <liam.merwick@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When the vmalloc area gets fragmented, and because the firmware
mapping area sits between where modules live and the vmalloc area, we
can sometimes receive requests for enormous kernel TLB range flushes.
When this happens the cpu just spins flushing billions of pages and
this triggers the NMI watchdog and other problems.
We took care of this on the TSB side by doing a linear scan of the
table once we pass a certain threshold.
Do something similar for the TLB flush, however we are limited by
the TLB flush facilities provided by the different chip variants.
First of all we use an (mostly arbitrary) cut-off of 256K which is
about 32 pages. This can be tuned in the future.
The huge range code path for each chip works as follows:
1) On spitfire we flush all non-locked TLB entries using diagnostic
acceses.
2) On cheetah we use the "flush all" TLB flush.
3) On sun4v/hypervisor we do a TLB context flush on context 0, which
unlike previous chips does not remove "permanent" or locked
entries.
We could probably do something better on spitfire, such as limiting
the flush to kernel TLB entries or even doing range comparisons.
However that probably isn't worth it since those chips are old and
the TLB only had 64 entries.
Reported-by: James Clarke <jrtc27@jrtc27.com>
Tested-by: James Clarke <jrtc27@jrtc27.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Just like the non-cross-call TLB flush handlers, the cross-call ones need
to avoid doing PC-relative branches outside of their code blocks.
Signed-off-by: David S. Miller <davem@davemloft.net>
If the number of pages we are flushing is more than twice the number
of entries in the TSB, just scan the TSB table for matches rather
than probing each and every page in the range.
Based upon a patch and report by James Clarke.
Signed-off-by: David S. Miller <davem@davemloft.net>
When we copy code over to patch another piece of code, we can only use
PC-relative branches that target code within that piece of code.
Such PC-relative branches cannot be made to external symbols because
the patch moves the location of the code and thus modifies the
relative address of external symbols.
Use an absolute jmpl to fix this problem.
Signed-off-by: David S. Miller <davem@davemloft.net>
This removes the 'write' and 'force' use from get_user_pages_unlocked()
and replaces them with 'gup_flags' to make the use of FOLL_FORCE
explicit in callers as use of this flag can result in surprising
behaviour (and hence bugs) within the mm subsystem.
Signed-off-by: Lorenzo Stoakes <lstoakes@gmail.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Michal Hocko <mhocko@suse.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
These files were only including module.h for exception table
related functions. We've now separated that content out into its
own file "extable.h" so now move over to that and avoid all the
extra header content in module.h that we don't really need to compile
these files.
Cc: "David S. Miller" <davem@davemloft.net>
Cc: sparclinux@vger.kernel.org
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, irq stack bootmem is allocated for all possible cpus
before nr_cpus value changes the list of possible cpus. As a result,
there is unnecessary wastage of bootmemory.
Move the irq stack bootmem allocation so that it happens after
possible cpu list is modified based on nr_cpus value.
Signed-off-by: Atish Patra <atish.patra@oracle.com>
Reviewed-by: Bob Picco <bob.picco@oracle.com>
Reviewed-by: Vijay Kumar <vijay.ac.kumar@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit af1b1a9b36 ("sparc64 mm: Fix base TSB sizing when hugetlb
pages are used") addressed the difference between hugetlb and THP
pages when computing TSB sizes. The following additional issues
were also discovered while working with the code.
In order to save memory, THP makes use of a huge zero page. This huge
zero page does not count against a task's RSS, but it does consume TSB
entries. This is similar to hugetlb pages. Therefore, count huge
zero page entries in hugetlb_pte_count.
Accounting of THP pages is done in the routine set_pmd_at().
Unfortunately, this does not catch the case where a THP page is split.
To handle this case, decrement the count in pmdp_invalidate().
pmdp_invalidate is only called when splitting a THP. However, 'sanity
checks' are added in case it is ever called for other purposes.
A more general issue exists with HPAGE_SIZE accounting.
hugetlb_pte_count tracks the number of HPAGE_SIZE (8M) pages. This
value is used to size the TSB for HPAGE_SIZE pages. However,
each HPAGE_SIZE page consists of two REAL_HPAGE_SIZE (4M) pages.
The TSB contains an entry for each REAL_HPAGE_SIZE page. Therefore,
the number of REAL_HPAGE_SIZE pages should be used to size the huge
page TSB. A new compile time constant REAL_HPAGE_PER_HPAGE is used
to multiply hugetlb_pte_count before sizing the TSB.
Changes from V1
- Fixed build issue if hugetlb or THP not configured
Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
To fix:
WARNING: vmlinux.o(.text.unlikely+0x580): Section mismatch in
reference from the function find_numa_latencies_for_group() to the
function .init.text:find_mlgroup()
The function find_numa_latencies_for_group() references the
function __init find_mlgroup(). This is often because
find_numa_latencies_for_group lacks a __init annotation or the
annotation of find_mlgroup is wrong.
It turns out find_numa_latencies_for_group is only called from:
static int __init numa_parse_mdesc(void)
and hence we can tag find_numa_latencies_for_group with __init.
In doing so we see that find_best_numa_node_for_mlgroup is only
called from within __init and hence can also be marked with __init.
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Nitin Gupta <nitin.m.gupta@oracle.com>
Cc: Chris Hyser <chris.hyser@oracle.com>
Cc: Santosh Shilimkar <santosh.shilimkar@oracle.com>
Cc: sparclinux@vger.kernel.org
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
For PMD aligned (8M) hugepages, we currently allocate
all four page table levels which is wasteful. We now
allocate till PMD level only which saves memory usage
from page tables.
Also, when freeing page table for 8M hugepage backed region,
make sure we don't try to access non-existent PTE level.
Orabug: 22630259
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
do_sparc64_fault() calculates both the base and huge page RSS sizes and
uses this information in calls to tsb_grow(). The calculation for base
page TSB size is not correct if the task uses hugetlb pages. hugetlb
pages are not accounted for in RSS, therefore the call to get_mm_rss(mm)
does not include hugetlb pages. However, the number of pages based on
huge_pte_count (which does include hugetlb pages) is subtracted from
this value. This will result in an artificially small and often negative
RSS calculation. The base TSB size is then often set to max_tsb_size
as the passed RSS is unsigned, so a negative value looks really big.
THP pages are also accounted for in huge_pte_count, and THP pages are
accounted for in RSS so the calculation in do_sparc64_fault() is correct
if a task only uses THP pages.
A single huge_pte_count is not sufficient for TSB sizing if both hugetlb
and THP pages can be used. Instead of a single counter, use two: one
for hugetlb and one for THP.
Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is the third version of the patchset previously sent [1]. I have
basically only rebased it on top of 4.7-rc1 tree and dropped "dm: get
rid of superfluous gfp flags" which went through dm tree. I am sending
it now because it is tree wide and chances for conflicts are reduced
considerably when we want to target rc2. I plan to send the next step
and rename the flag and move to a better semantic later during this
release cycle so we will have a new semantic ready for 4.8 merge window
hopefully.
Motivation:
While working on something unrelated I've checked the current usage of
__GFP_REPEAT in the tree. It seems that a majority of the usage is and
always has been bogus because __GFP_REPEAT has always been about costly
high order allocations while we are using it for order-0 or very small
orders very often. It seems that a big pile of them is just a
copy&paste when a code has been adopted from one arch to another.
I think it makes some sense to get rid of them because they are just
making the semantic more unclear. Please note that GFP_REPEAT is
documented as
* __GFP_REPEAT: Try hard to allocate the memory, but the allocation attempt
* _might_ fail. This depends upon the particular VM implementation.
while !costly requests have basically nofail semantic. So one could
reasonably expect that order-0 request with __GFP_REPEAT will not loop
for ever. This is not implemented right now though.
I would like to move on with __GFP_REPEAT and define a better semantic
for it.
$ git grep __GFP_REPEAT origin/master | wc -l
111
$ git grep __GFP_REPEAT | wc -l
36
So we are down to the third after this patch series. The remaining
places really seem to be relying on __GFP_REPEAT due to large allocation
requests. This still needs some double checking which I will do later
after all the simple ones are sorted out.
I am touching a lot of arch specific code here and I hope I got it right
but as a matter of fact I even didn't compile test for some archs as I
do not have cross compiler for them. Patches should be quite trivial to
review for stupid compile mistakes though. The tricky parts are usually
hidden by macro definitions and thats where I would appreciate help from
arch maintainers.
[1] http://lkml.kernel.org/r/1461849846-27209-1-git-send-email-mhocko@kernel.org
This patch (of 19):
__GFP_REPEAT has a rather weak semantic but since it has been introduced
around 2.6.12 it has been ignored for low order allocations. Yet we
have the full kernel tree with its usage for apparently order-0
allocations. This is really confusing because __GFP_REPEAT is
explicitly documented to allow allocation failures which is a weaker
semantic than the current order-0 has (basically nofail).
Let's simply drop __GFP_REPEAT from those places. This would allow to
identify place which really need allocator to retry harder and formulate
a more specific semantic for what the flag is supposed to do actually.
Link: http://lkml.kernel.org/r/1464599699-30131-2-git-send-email-mhocko@kernel.org
Signed-off-by: Michal Hocko <mhocko@suse.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: "James E.J. Bottomley" <jejb@parisc-linux.org>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Chen Liqin <liqin.linux@gmail.com>
Cc: Chris Metcalf <cmetcalf@mellanox.com> [for tile]
Cc: Guan Xuetao <gxt@mprc.pku.edu.cn>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Helge Deller <deller@gmx.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: John Crispin <blogic@openwrt.org>
Cc: Lennox Wu <lennox.wu@gmail.com>
Cc: Ley Foon Tan <lftan@altera.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Matt Fleming <matt@codeblueprint.co.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: Rich Felker <dalias@libc.org>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vineet Gupta <vgupta@synopsys.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Nitin Gupta