2019-05-19 15:08:55 +03:00
// SPDX-License-Identifier: GPL-2.0-only
2005-04-17 02:20:36 +04:00
/*
* ( C ) 1997 Linus Torvalds
2011-05-27 17:28:01 +04:00
* ( C ) 1999 Andrea Arcangeli < andrea @ suse . de > ( dynamic inode allocation )
2005-04-17 02:20:36 +04:00
*/
2011-12-07 22:17:19 +04:00
# include <linux/export.h>
2005-04-17 02:20:36 +04:00
# include <linux/fs.h>
2022-11-20 17:15:34 +03:00
# include <linux/filelock.h>
2005-04-17 02:20:36 +04:00
# include <linux/mm.h>
# include <linux/backing-dev.h>
# include <linux/hash.h>
# include <linux/swap.h>
# include <linux/security.h>
# include <linux/cdev.h>
2018-10-31 01:09:49 +03:00
# include <linux/memblock.h>
2009-05-22 01:01:26 +04:00
# include <linux/fsnotify.h>
2006-01-10 07:52:17 +03:00
# include <linux/mount.h>
2009-06-09 03:50:45 +04:00
# include <linux/posix_acl.h>
2011-05-27 17:28:01 +04:00
# include <linux/buffer_head.h> /* for inode_has_buffers */
2011-11-21 15:11:32 +04:00
# include <linux/ratelimit.h>
2013-08-28 04:17:58 +04:00
# include <linux/list_lru.h>
2018-01-29 14:41:30 +03:00
# include <linux/iversion.h>
2015-02-02 08:37:00 +03:00
# include <trace/events/writeback.h>
2011-03-22 14:23:41 +03:00
# include "internal.h"
2005-04-17 02:20:36 +04:00
2011-03-22 14:23:36 +03:00
/*
2011-05-27 17:28:01 +04:00
* Inode locking rules :
2011-03-22 14:23:36 +03:00
*
* inode - > i_lock protects :
2022-05-24 18:05:40 +03:00
* inode - > i_state , inode - > i_hash , __iget ( ) , inode - > i_io_list
2013-08-28 04:17:58 +04:00
* Inode LRU list locks protect :
2011-07-08 08:14:39 +04:00
* inode - > i_sb - > s_inode_lru , inode - > i_lru
2015-03-04 20:37:22 +03:00
* inode - > i_sb - > s_inode_list_lock protects :
* inode - > i_sb - > s_inodes , inode - > i_sb_list
2011-04-22 04:19:44 +04:00
* bdi - > wb . list_lock protects :
2015-03-04 22:07:22 +03:00
* bdi - > wb . b_ { dirty , io , more_io , dirty_time } , inode - > i_io_list
2011-03-22 14:23:42 +03:00
* inode_hash_lock protects :
* inode_hashtable , inode - > i_hash
2011-03-22 14:23:36 +03:00
*
* Lock ordering :
2011-03-22 14:23:40 +03:00
*
2015-03-04 20:37:22 +03:00
* inode - > i_sb - > s_inode_list_lock
2011-03-22 14:23:40 +03:00
* inode - > i_lock
2013-08-28 04:17:58 +04:00
* Inode LRU list locks
2011-03-22 14:23:41 +03:00
*
2011-04-22 04:19:44 +04:00
* bdi - > wb . list_lock
2011-03-22 14:23:41 +03:00
* inode - > i_lock
2011-03-22 14:23:42 +03:00
*
* inode_hash_lock
2015-03-04 20:37:22 +03:00
* inode - > i_sb - > s_inode_list_lock
2011-03-22 14:23:42 +03:00
* inode - > i_lock
*
* iunique_lock
* inode_hash_lock
2011-03-22 14:23:36 +03:00
*/
2006-03-26 13:37:24 +04:00
static unsigned int i_hash_mask __read_mostly ;
static unsigned int i_hash_shift __read_mostly ;
2011-03-22 14:23:42 +03:00
static struct hlist_head * inode_hashtable __read_mostly ;
static __cacheline_aligned_in_smp DEFINE_SPINLOCK ( inode_hash_lock ) ;
2005-04-17 02:20:36 +04:00
2011-04-06 01:51:48 +04:00
/*
* Empty aops . Can be used for the cases where the user does not
* define any of the address_space operations .
*/
const struct address_space_operations empty_aops = {
} ;
EXPORT_SYMBOL ( empty_aops ) ;
fs: bump inode and dentry counters to long
This series reworks our current object cache shrinking infrastructure in
two main ways:
* Noticing that a lot of users copy and paste their own version of LRU
lists for objects, we put some effort in providing a generic version.
It is modeled after the filesystem users: dentries, inodes, and xfs
(for various tasks), but we expect that other users could benefit in
the near future with little or no modification. Let us know if you
have any issues.
* The underlying list_lru being proposed automatically and
transparently keeps the elements in per-node lists, and is able to
manipulate the node lists individually. Given this infrastructure, we
are able to modify the up-to-now hammer called shrink_slab to proceed
with node-reclaim instead of always searching memory from all over like
it has been doing.
Per-node lru lists are also expected to lead to less contention in the lru
locks on multi-node scans, since we are now no longer fighting for a
global lock. The locks usually disappear from the profilers with this
change.
Although we have no official benchmarks for this version - be our guest to
independently evaluate this - earlier versions of this series were
performance tested (details at
http://permalink.gmane.org/gmane.linux.kernel.mm/100537) yielding no
visible performance regressions while yielding a better qualitative
behavior in NUMA machines.
With this infrastructure in place, we can use the list_lru entry point to
provide memcg isolation and per-memcg targeted reclaim. Historically,
those two pieces of work have been posted together. This version presents
only the infrastructure work, deferring the memcg work for a later time,
so we can focus on getting this part tested. You can see more about the
history of such work at http://lwn.net/Articles/552769/
Dave Chinner (18):
dcache: convert dentry_stat.nr_unused to per-cpu counters
dentry: move to per-sb LRU locks
dcache: remove dentries from LRU before putting on dispose list
mm: new shrinker API
shrinker: convert superblock shrinkers to new API
list: add a new LRU list type
inode: convert inode lru list to generic lru list code.
dcache: convert to use new lru list infrastructure
list_lru: per-node list infrastructure
shrinker: add node awareness
fs: convert inode and dentry shrinking to be node aware
xfs: convert buftarg LRU to generic code
xfs: rework buffer dispose list tracking
xfs: convert dquot cache lru to list_lru
fs: convert fs shrinkers to new scan/count API
drivers: convert shrinkers to new count/scan API
shrinker: convert remaining shrinkers to count/scan API
shrinker: Kill old ->shrink API.
Glauber Costa (7):
fs: bump inode and dentry counters to long
super: fix calculation of shrinkable objects for small numbers
list_lru: per-node API
vmscan: per-node deferred work
i915: bail out earlier when shrinker cannot acquire mutex
hugepage: convert huge zero page shrinker to new shrinker API
list_lru: dynamically adjust node arrays
This patch:
There are situations in very large machines in which we can have a large
quantity of dirty inodes, unused dentries, etc. This is particularly true
when umounting a filesystem, where eventually since every live object will
eventually be discarded.
Dave Chinner reported a problem with this while experimenting with the
shrinker revamp patchset. So we believe it is time for a change. This
patch just moves int to longs. Machines where it matters should have a
big long anyway.
Signed-off-by: Glauber Costa <glommer@openvz.org>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Arve Hjønnevåg <arve@android.com>
Cc: Carlos Maiolino <cmaiolino@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: J. Bruce Fields <bfields@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-08-28 04:17:53 +04:00
static DEFINE_PER_CPU ( unsigned long , nr_inodes ) ;
static DEFINE_PER_CPU ( unsigned long , nr_unused ) ;
2010-10-23 13:03:02 +04:00
2009-03-31 18:05:54 +04:00
static struct kmem_cache * inode_cachep __read_mostly ;
2005-04-17 02:20:36 +04:00
fs: bump inode and dentry counters to long
This series reworks our current object cache shrinking infrastructure in
two main ways:
* Noticing that a lot of users copy and paste their own version of LRU
lists for objects, we put some effort in providing a generic version.
It is modeled after the filesystem users: dentries, inodes, and xfs
(for various tasks), but we expect that other users could benefit in
the near future with little or no modification. Let us know if you
have any issues.
* The underlying list_lru being proposed automatically and
transparently keeps the elements in per-node lists, and is able to
manipulate the node lists individually. Given this infrastructure, we
are able to modify the up-to-now hammer called shrink_slab to proceed
with node-reclaim instead of always searching memory from all over like
it has been doing.
Per-node lru lists are also expected to lead to less contention in the lru
locks on multi-node scans, since we are now no longer fighting for a
global lock. The locks usually disappear from the profilers with this
change.
Although we have no official benchmarks for this version - be our guest to
independently evaluate this - earlier versions of this series were
performance tested (details at
http://permalink.gmane.org/gmane.linux.kernel.mm/100537) yielding no
visible performance regressions while yielding a better qualitative
behavior in NUMA machines.
With this infrastructure in place, we can use the list_lru entry point to
provide memcg isolation and per-memcg targeted reclaim. Historically,
those two pieces of work have been posted together. This version presents
only the infrastructure work, deferring the memcg work for a later time,
so we can focus on getting this part tested. You can see more about the
history of such work at http://lwn.net/Articles/552769/
Dave Chinner (18):
dcache: convert dentry_stat.nr_unused to per-cpu counters
dentry: move to per-sb LRU locks
dcache: remove dentries from LRU before putting on dispose list
mm: new shrinker API
shrinker: convert superblock shrinkers to new API
list: add a new LRU list type
inode: convert inode lru list to generic lru list code.
dcache: convert to use new lru list infrastructure
list_lru: per-node list infrastructure
shrinker: add node awareness
fs: convert inode and dentry shrinking to be node aware
xfs: convert buftarg LRU to generic code
xfs: rework buffer dispose list tracking
xfs: convert dquot cache lru to list_lru
fs: convert fs shrinkers to new scan/count API
drivers: convert shrinkers to new count/scan API
shrinker: convert remaining shrinkers to count/scan API
shrinker: Kill old ->shrink API.
Glauber Costa (7):
fs: bump inode and dentry counters to long
super: fix calculation of shrinkable objects for small numbers
list_lru: per-node API
vmscan: per-node deferred work
i915: bail out earlier when shrinker cannot acquire mutex
hugepage: convert huge zero page shrinker to new shrinker API
list_lru: dynamically adjust node arrays
This patch:
There are situations in very large machines in which we can have a large
quantity of dirty inodes, unused dentries, etc. This is particularly true
when umounting a filesystem, where eventually since every live object will
eventually be discarded.
Dave Chinner reported a problem with this while experimenting with the
shrinker revamp patchset. So we believe it is time for a change. This
patch just moves int to longs. Machines where it matters should have a
big long anyway.
Signed-off-by: Glauber Costa <glommer@openvz.org>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Arve Hjønnevåg <arve@android.com>
Cc: Carlos Maiolino <cmaiolino@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: J. Bruce Fields <bfields@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-08-28 04:17:53 +04:00
static long get_nr_inodes ( void )
2010-10-23 13:03:02 +04:00
{
fs: use fast counters for vfs caches
percpu_counter library generates quite nasty code, so unless you need
to dynamically allocate counters or take fast approximate value, a
simple per cpu set of counters is much better.
The percpu_counter can never be made to work as well, because it has an
indirection from pointer to percpu memory, and it can't use direct
this_cpu_inc interfaces because it doesn't use static PER_CPU data, so
code will always be worse.
In the fastpath, it is the difference between this:
incl %gs:nr_dentry # nr_dentry
and this:
movl percpu_counter_batch(%rip), %edx # percpu_counter_batch,
movl $1, %esi #,
movq $nr_dentry, %rdi #,
call __percpu_counter_add # (plus I clobber registers)
__percpu_counter_add:
pushq %rbp #
movq %rsp, %rbp #,
subq $32, %rsp #,
movq %rbx, -24(%rbp) #,
movq %r12, -16(%rbp) #,
movq %r13, -8(%rbp) #,
movq %rdi, %rbx # fbc, fbc
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
incl -8124(%rax) # <variable>.preempt_count
movq 32(%rdi), %r12 # <variable>.counters, tcp_ptr__
#APP
# 78 "lib/percpu_counter.c" 1
add %gs:this_cpu_off, %r12 # this_cpu_off, tcp_ptr__
# 0 "" 2
#NO_APP
movslq (%r12),%r13 #* tcp_ptr__, tmp73
movslq %edx,%rax # batch, batch
addq %rsi, %r13 # amount, count
cmpq %rax, %r13 # batch, count
jge .L27 #,
negl %edx # tmp76
movslq %edx,%rdx # tmp76, tmp77
cmpq %rdx, %r13 # tmp77, count
jg .L28 #,
.L27:
movq %rbx, %rdi # fbc,
call _raw_spin_lock #
addq %r13, 8(%rbx) # count, <variable>.count
movq %rbx, %rdi # fbc,
movl $0, (%r12) #,* tcp_ptr__
call _raw_spin_unlock #
.L29:
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
decl -8124(%rax) # <variable>.preempt_count
movq -8136(%rax), %rax #, D.14625
testb $8, %al #, D.14625
jne .L32 #,
.L31:
movq -24(%rbp), %rbx #,
movq -16(%rbp), %r12 #,
movq -8(%rbp), %r13 #,
leave
ret
.p2align 4,,10
.p2align 3
.L28:
movl %r13d, (%r12) # count,*
jmp .L29 #
.L32:
call preempt_schedule #
.p2align 4,,6
jmp .L31 #
.size __percpu_counter_add, .-__percpu_counter_add
.p2align 4,,15
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 09:49:19 +03:00
int i ;
fs: bump inode and dentry counters to long
This series reworks our current object cache shrinking infrastructure in
two main ways:
* Noticing that a lot of users copy and paste their own version of LRU
lists for objects, we put some effort in providing a generic version.
It is modeled after the filesystem users: dentries, inodes, and xfs
(for various tasks), but we expect that other users could benefit in
the near future with little or no modification. Let us know if you
have any issues.
* The underlying list_lru being proposed automatically and
transparently keeps the elements in per-node lists, and is able to
manipulate the node lists individually. Given this infrastructure, we
are able to modify the up-to-now hammer called shrink_slab to proceed
with node-reclaim instead of always searching memory from all over like
it has been doing.
Per-node lru lists are also expected to lead to less contention in the lru
locks on multi-node scans, since we are now no longer fighting for a
global lock. The locks usually disappear from the profilers with this
change.
Although we have no official benchmarks for this version - be our guest to
independently evaluate this - earlier versions of this series were
performance tested (details at
http://permalink.gmane.org/gmane.linux.kernel.mm/100537) yielding no
visible performance regressions while yielding a better qualitative
behavior in NUMA machines.
With this infrastructure in place, we can use the list_lru entry point to
provide memcg isolation and per-memcg targeted reclaim. Historically,
those two pieces of work have been posted together. This version presents
only the infrastructure work, deferring the memcg work for a later time,
so we can focus on getting this part tested. You can see more about the
history of such work at http://lwn.net/Articles/552769/
Dave Chinner (18):
dcache: convert dentry_stat.nr_unused to per-cpu counters
dentry: move to per-sb LRU locks
dcache: remove dentries from LRU before putting on dispose list
mm: new shrinker API
shrinker: convert superblock shrinkers to new API
list: add a new LRU list type
inode: convert inode lru list to generic lru list code.
dcache: convert to use new lru list infrastructure
list_lru: per-node list infrastructure
shrinker: add node awareness
fs: convert inode and dentry shrinking to be node aware
xfs: convert buftarg LRU to generic code
xfs: rework buffer dispose list tracking
xfs: convert dquot cache lru to list_lru
fs: convert fs shrinkers to new scan/count API
drivers: convert shrinkers to new count/scan API
shrinker: convert remaining shrinkers to count/scan API
shrinker: Kill old ->shrink API.
Glauber Costa (7):
fs: bump inode and dentry counters to long
super: fix calculation of shrinkable objects for small numbers
list_lru: per-node API
vmscan: per-node deferred work
i915: bail out earlier when shrinker cannot acquire mutex
hugepage: convert huge zero page shrinker to new shrinker API
list_lru: dynamically adjust node arrays
This patch:
There are situations in very large machines in which we can have a large
quantity of dirty inodes, unused dentries, etc. This is particularly true
when umounting a filesystem, where eventually since every live object will
eventually be discarded.
Dave Chinner reported a problem with this while experimenting with the
shrinker revamp patchset. So we believe it is time for a change. This
patch just moves int to longs. Machines where it matters should have a
big long anyway.
Signed-off-by: Glauber Costa <glommer@openvz.org>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Arve Hjønnevåg <arve@android.com>
Cc: Carlos Maiolino <cmaiolino@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: J. Bruce Fields <bfields@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-08-28 04:17:53 +04:00
long sum = 0 ;
fs: use fast counters for vfs caches
percpu_counter library generates quite nasty code, so unless you need
to dynamically allocate counters or take fast approximate value, a
simple per cpu set of counters is much better.
The percpu_counter can never be made to work as well, because it has an
indirection from pointer to percpu memory, and it can't use direct
this_cpu_inc interfaces because it doesn't use static PER_CPU data, so
code will always be worse.
In the fastpath, it is the difference between this:
incl %gs:nr_dentry # nr_dentry
and this:
movl percpu_counter_batch(%rip), %edx # percpu_counter_batch,
movl $1, %esi #,
movq $nr_dentry, %rdi #,
call __percpu_counter_add # (plus I clobber registers)
__percpu_counter_add:
pushq %rbp #
movq %rsp, %rbp #,
subq $32, %rsp #,
movq %rbx, -24(%rbp) #,
movq %r12, -16(%rbp) #,
movq %r13, -8(%rbp) #,
movq %rdi, %rbx # fbc, fbc
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
incl -8124(%rax) # <variable>.preempt_count
movq 32(%rdi), %r12 # <variable>.counters, tcp_ptr__
#APP
# 78 "lib/percpu_counter.c" 1
add %gs:this_cpu_off, %r12 # this_cpu_off, tcp_ptr__
# 0 "" 2
#NO_APP
movslq (%r12),%r13 #* tcp_ptr__, tmp73
movslq %edx,%rax # batch, batch
addq %rsi, %r13 # amount, count
cmpq %rax, %r13 # batch, count
jge .L27 #,
negl %edx # tmp76
movslq %edx,%rdx # tmp76, tmp77
cmpq %rdx, %r13 # tmp77, count
jg .L28 #,
.L27:
movq %rbx, %rdi # fbc,
call _raw_spin_lock #
addq %r13, 8(%rbx) # count, <variable>.count
movq %rbx, %rdi # fbc,
movl $0, (%r12) #,* tcp_ptr__
call _raw_spin_unlock #
.L29:
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
decl -8124(%rax) # <variable>.preempt_count
movq -8136(%rax), %rax #, D.14625
testb $8, %al #, D.14625
jne .L32 #,
.L31:
movq -24(%rbp), %rbx #,
movq -16(%rbp), %r12 #,
movq -8(%rbp), %r13 #,
leave
ret
.p2align 4,,10
.p2align 3
.L28:
movl %r13d, (%r12) # count,*
jmp .L29 #
.L32:
call preempt_schedule #
.p2align 4,,6
jmp .L31 #
.size __percpu_counter_add, .-__percpu_counter_add
.p2align 4,,15
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 09:49:19 +03:00
for_each_possible_cpu ( i )
sum + = per_cpu ( nr_inodes , i ) ;
return sum < 0 ? 0 : sum ;
2010-10-23 13:03:02 +04:00
}
fs: bump inode and dentry counters to long
This series reworks our current object cache shrinking infrastructure in
two main ways:
* Noticing that a lot of users copy and paste their own version of LRU
lists for objects, we put some effort in providing a generic version.
It is modeled after the filesystem users: dentries, inodes, and xfs
(for various tasks), but we expect that other users could benefit in
the near future with little or no modification. Let us know if you
have any issues.
* The underlying list_lru being proposed automatically and
transparently keeps the elements in per-node lists, and is able to
manipulate the node lists individually. Given this infrastructure, we
are able to modify the up-to-now hammer called shrink_slab to proceed
with node-reclaim instead of always searching memory from all over like
it has been doing.
Per-node lru lists are also expected to lead to less contention in the lru
locks on multi-node scans, since we are now no longer fighting for a
global lock. The locks usually disappear from the profilers with this
change.
Although we have no official benchmarks for this version - be our guest to
independently evaluate this - earlier versions of this series were
performance tested (details at
http://permalink.gmane.org/gmane.linux.kernel.mm/100537) yielding no
visible performance regressions while yielding a better qualitative
behavior in NUMA machines.
With this infrastructure in place, we can use the list_lru entry point to
provide memcg isolation and per-memcg targeted reclaim. Historically,
those two pieces of work have been posted together. This version presents
only the infrastructure work, deferring the memcg work for a later time,
so we can focus on getting this part tested. You can see more about the
history of such work at http://lwn.net/Articles/552769/
Dave Chinner (18):
dcache: convert dentry_stat.nr_unused to per-cpu counters
dentry: move to per-sb LRU locks
dcache: remove dentries from LRU before putting on dispose list
mm: new shrinker API
shrinker: convert superblock shrinkers to new API
list: add a new LRU list type
inode: convert inode lru list to generic lru list code.
dcache: convert to use new lru list infrastructure
list_lru: per-node list infrastructure
shrinker: add node awareness
fs: convert inode and dentry shrinking to be node aware
xfs: convert buftarg LRU to generic code
xfs: rework buffer dispose list tracking
xfs: convert dquot cache lru to list_lru
fs: convert fs shrinkers to new scan/count API
drivers: convert shrinkers to new count/scan API
shrinker: convert remaining shrinkers to count/scan API
shrinker: Kill old ->shrink API.
Glauber Costa (7):
fs: bump inode and dentry counters to long
super: fix calculation of shrinkable objects for small numbers
list_lru: per-node API
vmscan: per-node deferred work
i915: bail out earlier when shrinker cannot acquire mutex
hugepage: convert huge zero page shrinker to new shrinker API
list_lru: dynamically adjust node arrays
This patch:
There are situations in very large machines in which we can have a large
quantity of dirty inodes, unused dentries, etc. This is particularly true
when umounting a filesystem, where eventually since every live object will
eventually be discarded.
Dave Chinner reported a problem with this while experimenting with the
shrinker revamp patchset. So we believe it is time for a change. This
patch just moves int to longs. Machines where it matters should have a
big long anyway.
Signed-off-by: Glauber Costa <glommer@openvz.org>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Arve Hjønnevåg <arve@android.com>
Cc: Carlos Maiolino <cmaiolino@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: J. Bruce Fields <bfields@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-08-28 04:17:53 +04:00
static inline long get_nr_inodes_unused ( void )
2010-10-23 13:03:02 +04:00
{
2011-07-08 08:14:38 +04:00
int i ;
fs: bump inode and dentry counters to long
This series reworks our current object cache shrinking infrastructure in
two main ways:
* Noticing that a lot of users copy and paste their own version of LRU
lists for objects, we put some effort in providing a generic version.
It is modeled after the filesystem users: dentries, inodes, and xfs
(for various tasks), but we expect that other users could benefit in
the near future with little or no modification. Let us know if you
have any issues.
* The underlying list_lru being proposed automatically and
transparently keeps the elements in per-node lists, and is able to
manipulate the node lists individually. Given this infrastructure, we
are able to modify the up-to-now hammer called shrink_slab to proceed
with node-reclaim instead of always searching memory from all over like
it has been doing.
Per-node lru lists are also expected to lead to less contention in the lru
locks on multi-node scans, since we are now no longer fighting for a
global lock. The locks usually disappear from the profilers with this
change.
Although we have no official benchmarks for this version - be our guest to
independently evaluate this - earlier versions of this series were
performance tested (details at
http://permalink.gmane.org/gmane.linux.kernel.mm/100537) yielding no
visible performance regressions while yielding a better qualitative
behavior in NUMA machines.
With this infrastructure in place, we can use the list_lru entry point to
provide memcg isolation and per-memcg targeted reclaim. Historically,
those two pieces of work have been posted together. This version presents
only the infrastructure work, deferring the memcg work for a later time,
so we can focus on getting this part tested. You can see more about the
history of such work at http://lwn.net/Articles/552769/
Dave Chinner (18):
dcache: convert dentry_stat.nr_unused to per-cpu counters
dentry: move to per-sb LRU locks
dcache: remove dentries from LRU before putting on dispose list
mm: new shrinker API
shrinker: convert superblock shrinkers to new API
list: add a new LRU list type
inode: convert inode lru list to generic lru list code.
dcache: convert to use new lru list infrastructure
list_lru: per-node list infrastructure
shrinker: add node awareness
fs: convert inode and dentry shrinking to be node aware
xfs: convert buftarg LRU to generic code
xfs: rework buffer dispose list tracking
xfs: convert dquot cache lru to list_lru
fs: convert fs shrinkers to new scan/count API
drivers: convert shrinkers to new count/scan API
shrinker: convert remaining shrinkers to count/scan API
shrinker: Kill old ->shrink API.
Glauber Costa (7):
fs: bump inode and dentry counters to long
super: fix calculation of shrinkable objects for small numbers
list_lru: per-node API
vmscan: per-node deferred work
i915: bail out earlier when shrinker cannot acquire mutex
hugepage: convert huge zero page shrinker to new shrinker API
list_lru: dynamically adjust node arrays
This patch:
There are situations in very large machines in which we can have a large
quantity of dirty inodes, unused dentries, etc. This is particularly true
when umounting a filesystem, where eventually since every live object will
eventually be discarded.
Dave Chinner reported a problem with this while experimenting with the
shrinker revamp patchset. So we believe it is time for a change. This
patch just moves int to longs. Machines where it matters should have a
big long anyway.
Signed-off-by: Glauber Costa <glommer@openvz.org>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Arve Hjønnevåg <arve@android.com>
Cc: Carlos Maiolino <cmaiolino@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: J. Bruce Fields <bfields@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-08-28 04:17:53 +04:00
long sum = 0 ;
2011-07-08 08:14:38 +04:00
for_each_possible_cpu ( i )
sum + = per_cpu ( nr_unused , i ) ;
return sum < 0 ? 0 : sum ;
2010-10-23 13:03:02 +04:00
}
fs: bump inode and dentry counters to long
This series reworks our current object cache shrinking infrastructure in
two main ways:
* Noticing that a lot of users copy and paste their own version of LRU
lists for objects, we put some effort in providing a generic version.
It is modeled after the filesystem users: dentries, inodes, and xfs
(for various tasks), but we expect that other users could benefit in
the near future with little or no modification. Let us know if you
have any issues.
* The underlying list_lru being proposed automatically and
transparently keeps the elements in per-node lists, and is able to
manipulate the node lists individually. Given this infrastructure, we
are able to modify the up-to-now hammer called shrink_slab to proceed
with node-reclaim instead of always searching memory from all over like
it has been doing.
Per-node lru lists are also expected to lead to less contention in the lru
locks on multi-node scans, since we are now no longer fighting for a
global lock. The locks usually disappear from the profilers with this
change.
Although we have no official benchmarks for this version - be our guest to
independently evaluate this - earlier versions of this series were
performance tested (details at
http://permalink.gmane.org/gmane.linux.kernel.mm/100537) yielding no
visible performance regressions while yielding a better qualitative
behavior in NUMA machines.
With this infrastructure in place, we can use the list_lru entry point to
provide memcg isolation and per-memcg targeted reclaim. Historically,
those two pieces of work have been posted together. This version presents
only the infrastructure work, deferring the memcg work for a later time,
so we can focus on getting this part tested. You can see more about the
history of such work at http://lwn.net/Articles/552769/
Dave Chinner (18):
dcache: convert dentry_stat.nr_unused to per-cpu counters
dentry: move to per-sb LRU locks
dcache: remove dentries from LRU before putting on dispose list
mm: new shrinker API
shrinker: convert superblock shrinkers to new API
list: add a new LRU list type
inode: convert inode lru list to generic lru list code.
dcache: convert to use new lru list infrastructure
list_lru: per-node list infrastructure
shrinker: add node awareness
fs: convert inode and dentry shrinking to be node aware
xfs: convert buftarg LRU to generic code
xfs: rework buffer dispose list tracking
xfs: convert dquot cache lru to list_lru
fs: convert fs shrinkers to new scan/count API
drivers: convert shrinkers to new count/scan API
shrinker: convert remaining shrinkers to count/scan API
shrinker: Kill old ->shrink API.
Glauber Costa (7):
fs: bump inode and dentry counters to long
super: fix calculation of shrinkable objects for small numbers
list_lru: per-node API
vmscan: per-node deferred work
i915: bail out earlier when shrinker cannot acquire mutex
hugepage: convert huge zero page shrinker to new shrinker API
list_lru: dynamically adjust node arrays
This patch:
There are situations in very large machines in which we can have a large
quantity of dirty inodes, unused dentries, etc. This is particularly true
when umounting a filesystem, where eventually since every live object will
eventually be discarded.
Dave Chinner reported a problem with this while experimenting with the
shrinker revamp patchset. So we believe it is time for a change. This
patch just moves int to longs. Machines where it matters should have a
big long anyway.
Signed-off-by: Glauber Costa <glommer@openvz.org>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Arve Hjønnevåg <arve@android.com>
Cc: Carlos Maiolino <cmaiolino@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: J. Bruce Fields <bfields@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-08-28 04:17:53 +04:00
long get_nr_dirty_inodes ( void )
2010-10-23 13:03:02 +04:00
{
fs: use fast counters for vfs caches
percpu_counter library generates quite nasty code, so unless you need
to dynamically allocate counters or take fast approximate value, a
simple per cpu set of counters is much better.
The percpu_counter can never be made to work as well, because it has an
indirection from pointer to percpu memory, and it can't use direct
this_cpu_inc interfaces because it doesn't use static PER_CPU data, so
code will always be worse.
In the fastpath, it is the difference between this:
incl %gs:nr_dentry # nr_dentry
and this:
movl percpu_counter_batch(%rip), %edx # percpu_counter_batch,
movl $1, %esi #,
movq $nr_dentry, %rdi #,
call __percpu_counter_add # (plus I clobber registers)
__percpu_counter_add:
pushq %rbp #
movq %rsp, %rbp #,
subq $32, %rsp #,
movq %rbx, -24(%rbp) #,
movq %r12, -16(%rbp) #,
movq %r13, -8(%rbp) #,
movq %rdi, %rbx # fbc, fbc
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
incl -8124(%rax) # <variable>.preempt_count
movq 32(%rdi), %r12 # <variable>.counters, tcp_ptr__
#APP
# 78 "lib/percpu_counter.c" 1
add %gs:this_cpu_off, %r12 # this_cpu_off, tcp_ptr__
# 0 "" 2
#NO_APP
movslq (%r12),%r13 #* tcp_ptr__, tmp73
movslq %edx,%rax # batch, batch
addq %rsi, %r13 # amount, count
cmpq %rax, %r13 # batch, count
jge .L27 #,
negl %edx # tmp76
movslq %edx,%rdx # tmp76, tmp77
cmpq %rdx, %r13 # tmp77, count
jg .L28 #,
.L27:
movq %rbx, %rdi # fbc,
call _raw_spin_lock #
addq %r13, 8(%rbx) # count, <variable>.count
movq %rbx, %rdi # fbc,
movl $0, (%r12) #,* tcp_ptr__
call _raw_spin_unlock #
.L29:
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
decl -8124(%rax) # <variable>.preempt_count
movq -8136(%rax), %rax #, D.14625
testb $8, %al #, D.14625
jne .L32 #,
.L31:
movq -24(%rbp), %rbx #,
movq -16(%rbp), %r12 #,
movq -8(%rbp), %r13 #,
leave
ret
.p2align 4,,10
.p2align 3
.L28:
movl %r13d, (%r12) # count,*
jmp .L29 #
.L32:
call preempt_schedule #
.p2align 4,,6
jmp .L31 #
.size __percpu_counter_add, .-__percpu_counter_add
.p2align 4,,15
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 09:49:19 +03:00
/* not actually dirty inodes, but a wild approximation */
fs: bump inode and dentry counters to long
This series reworks our current object cache shrinking infrastructure in
two main ways:
* Noticing that a lot of users copy and paste their own version of LRU
lists for objects, we put some effort in providing a generic version.
It is modeled after the filesystem users: dentries, inodes, and xfs
(for various tasks), but we expect that other users could benefit in
the near future with little or no modification. Let us know if you
have any issues.
* The underlying list_lru being proposed automatically and
transparently keeps the elements in per-node lists, and is able to
manipulate the node lists individually. Given this infrastructure, we
are able to modify the up-to-now hammer called shrink_slab to proceed
with node-reclaim instead of always searching memory from all over like
it has been doing.
Per-node lru lists are also expected to lead to less contention in the lru
locks on multi-node scans, since we are now no longer fighting for a
global lock. The locks usually disappear from the profilers with this
change.
Although we have no official benchmarks for this version - be our guest to
independently evaluate this - earlier versions of this series were
performance tested (details at
http://permalink.gmane.org/gmane.linux.kernel.mm/100537) yielding no
visible performance regressions while yielding a better qualitative
behavior in NUMA machines.
With this infrastructure in place, we can use the list_lru entry point to
provide memcg isolation and per-memcg targeted reclaim. Historically,
those two pieces of work have been posted together. This version presents
only the infrastructure work, deferring the memcg work for a later time,
so we can focus on getting this part tested. You can see more about the
history of such work at http://lwn.net/Articles/552769/
Dave Chinner (18):
dcache: convert dentry_stat.nr_unused to per-cpu counters
dentry: move to per-sb LRU locks
dcache: remove dentries from LRU before putting on dispose list
mm: new shrinker API
shrinker: convert superblock shrinkers to new API
list: add a new LRU list type
inode: convert inode lru list to generic lru list code.
dcache: convert to use new lru list infrastructure
list_lru: per-node list infrastructure
shrinker: add node awareness
fs: convert inode and dentry shrinking to be node aware
xfs: convert buftarg LRU to generic code
xfs: rework buffer dispose list tracking
xfs: convert dquot cache lru to list_lru
fs: convert fs shrinkers to new scan/count API
drivers: convert shrinkers to new count/scan API
shrinker: convert remaining shrinkers to count/scan API
shrinker: Kill old ->shrink API.
Glauber Costa (7):
fs: bump inode and dentry counters to long
super: fix calculation of shrinkable objects for small numbers
list_lru: per-node API
vmscan: per-node deferred work
i915: bail out earlier when shrinker cannot acquire mutex
hugepage: convert huge zero page shrinker to new shrinker API
list_lru: dynamically adjust node arrays
This patch:
There are situations in very large machines in which we can have a large
quantity of dirty inodes, unused dentries, etc. This is particularly true
when umounting a filesystem, where eventually since every live object will
eventually be discarded.
Dave Chinner reported a problem with this while experimenting with the
shrinker revamp patchset. So we believe it is time for a change. This
patch just moves int to longs. Machines where it matters should have a
big long anyway.
Signed-off-by: Glauber Costa <glommer@openvz.org>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Arve Hjønnevåg <arve@android.com>
Cc: Carlos Maiolino <cmaiolino@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: J. Bruce Fields <bfields@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-08-28 04:17:53 +04:00
long nr_dirty = get_nr_inodes ( ) - get_nr_inodes_unused ( ) ;
2010-10-23 13:03:02 +04:00
return nr_dirty > 0 ? nr_dirty : 0 ;
}
/*
* Handle nr_inode sysctl
*/
# ifdef CONFIG_SYSCTL
2022-01-22 09:12:52 +03:00
/*
* Statistics gathering . .
*/
static struct inodes_stat_t inodes_stat ;
static int proc_nr_inodes ( struct ctl_table * table , int write , void * buffer ,
size_t * lenp , loff_t * ppos )
2010-10-23 13:03:02 +04:00
{
inodes_stat . nr_inodes = get_nr_inodes ( ) ;
2011-07-08 08:14:38 +04:00
inodes_stat . nr_unused = get_nr_inodes_unused ( ) ;
fs: bump inode and dentry counters to long
This series reworks our current object cache shrinking infrastructure in
two main ways:
* Noticing that a lot of users copy and paste their own version of LRU
lists for objects, we put some effort in providing a generic version.
It is modeled after the filesystem users: dentries, inodes, and xfs
(for various tasks), but we expect that other users could benefit in
the near future with little or no modification. Let us know if you
have any issues.
* The underlying list_lru being proposed automatically and
transparently keeps the elements in per-node lists, and is able to
manipulate the node lists individually. Given this infrastructure, we
are able to modify the up-to-now hammer called shrink_slab to proceed
with node-reclaim instead of always searching memory from all over like
it has been doing.
Per-node lru lists are also expected to lead to less contention in the lru
locks on multi-node scans, since we are now no longer fighting for a
global lock. The locks usually disappear from the profilers with this
change.
Although we have no official benchmarks for this version - be our guest to
independently evaluate this - earlier versions of this series were
performance tested (details at
http://permalink.gmane.org/gmane.linux.kernel.mm/100537) yielding no
visible performance regressions while yielding a better qualitative
behavior in NUMA machines.
With this infrastructure in place, we can use the list_lru entry point to
provide memcg isolation and per-memcg targeted reclaim. Historically,
those two pieces of work have been posted together. This version presents
only the infrastructure work, deferring the memcg work for a later time,
so we can focus on getting this part tested. You can see more about the
history of such work at http://lwn.net/Articles/552769/
Dave Chinner (18):
dcache: convert dentry_stat.nr_unused to per-cpu counters
dentry: move to per-sb LRU locks
dcache: remove dentries from LRU before putting on dispose list
mm: new shrinker API
shrinker: convert superblock shrinkers to new API
list: add a new LRU list type
inode: convert inode lru list to generic lru list code.
dcache: convert to use new lru list infrastructure
list_lru: per-node list infrastructure
shrinker: add node awareness
fs: convert inode and dentry shrinking to be node aware
xfs: convert buftarg LRU to generic code
xfs: rework buffer dispose list tracking
xfs: convert dquot cache lru to list_lru
fs: convert fs shrinkers to new scan/count API
drivers: convert shrinkers to new count/scan API
shrinker: convert remaining shrinkers to count/scan API
shrinker: Kill old ->shrink API.
Glauber Costa (7):
fs: bump inode and dentry counters to long
super: fix calculation of shrinkable objects for small numbers
list_lru: per-node API
vmscan: per-node deferred work
i915: bail out earlier when shrinker cannot acquire mutex
hugepage: convert huge zero page shrinker to new shrinker API
list_lru: dynamically adjust node arrays
This patch:
There are situations in very large machines in which we can have a large
quantity of dirty inodes, unused dentries, etc. This is particularly true
when umounting a filesystem, where eventually since every live object will
eventually be discarded.
Dave Chinner reported a problem with this while experimenting with the
shrinker revamp patchset. So we believe it is time for a change. This
patch just moves int to longs. Machines where it matters should have a
big long anyway.
Signed-off-by: Glauber Costa <glommer@openvz.org>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: "Theodore Ts'o" <tytso@mit.edu>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: Arve Hjønnevåg <arve@android.com>
Cc: Carlos Maiolino <cmaiolino@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Chuck Lever <chuck.lever@oracle.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Dave Chinner <dchinner@redhat.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Gleb Natapov <gleb@redhat.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: J. Bruce Fields <bfields@redhat.com>
Cc: Jan Kara <jack@suse.cz>
Cc: Jerome Glisse <jglisse@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Kent Overstreet <koverstreet@google.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Mel Gorman <mgorman@suse.de>
Cc: Steven Whitehouse <swhiteho@redhat.com>
Cc: Thomas Hellstrom <thellstrom@vmware.com>
Cc: Trond Myklebust <Trond.Myklebust@netapp.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2013-08-28 04:17:53 +04:00
return proc_doulongvec_minmax ( table , write , buffer , lenp , ppos ) ;
2010-10-23 13:03:02 +04:00
}
2022-01-22 09:12:52 +03:00
static struct ctl_table inodes_sysctls [ ] = {
{
. procname = " inode-nr " ,
. data = & inodes_stat ,
. maxlen = 2 * sizeof ( long ) ,
. mode = 0444 ,
. proc_handler = proc_nr_inodes ,
} ,
{
. procname = " inode-state " ,
. data = & inodes_stat ,
. maxlen = 7 * sizeof ( long ) ,
. mode = 0444 ,
. proc_handler = proc_nr_inodes ,
} ,
{ }
} ;
static int __init init_fs_inode_sysctls ( void )
{
register_sysctl_init ( " fs " , inodes_sysctls ) ;
return 0 ;
}
early_initcall ( init_fs_inode_sysctls ) ;
2010-10-23 13:03:02 +04:00
# endif
2014-11-19 07:38:21 +03:00
static int no_open ( struct inode * inode , struct file * file )
{
return - ENXIO ;
}
2008-10-30 09:32:23 +03:00
/**
2017-05-09 01:57:53 +03:00
* inode_init_always - perform inode structure initialisation
2009-01-07 01:41:13 +03:00
* @ sb : superblock inode belongs to
* @ inode : inode to initialise
2008-10-30 09:32:23 +03:00
*
* These are initializations that need to be done on every inode
* allocation as the fields are not initialised by slab allocation .
*/
2009-08-07 21:38:25 +04:00
int inode_init_always ( struct super_block * sb , struct inode * inode )
2005-04-17 02:20:36 +04:00
{
2009-09-22 04:01:11 +04:00
static const struct inode_operations empty_iops ;
2014-11-19 07:38:21 +03:00
static const struct file_operations no_open_fops = { . open = no_open } ;
2009-03-31 18:05:54 +04:00
struct address_space * const mapping = & inode - > i_data ;
2008-10-30 09:32:23 +03:00
inode - > i_sb = sb ;
inode - > i_blkbits = sb - > s_blocksize_bits ;
inode - > i_flags = 0 ;
2020-03-04 13:28:31 +03:00
atomic64_set ( & inode - > i_sequence , 0 ) ;
2008-10-30 09:32:23 +03:00
atomic_set ( & inode - > i_count , 1 ) ;
inode - > i_op = & empty_iops ;
2014-11-19 07:38:21 +03:00
inode - > i_fop = & no_open_fops ;
fs/inode.c: make inode_init_always() initialize i_ino to 0
Currently inode_init_always() doesn't initialize i_ino to 0. This is
unexpected because unlike the other inode fields that aren't initialized
by inode_init_always(), i_ino isn't guaranteed to end up back at its
initial value after the inode is freed. Only one filesystem (XFS)
actually sets set i_ino back to 0 when freeing its inodes.
So, callers of new_inode() see some random previous i_ino. Normally
that's fine, since normally i_ino isn't accessed before being set.
There can be edge cases where that isn't necessarily true, though.
The one I've run into is that on ext4, when creating an encrypted file,
the new file's encryption key has to be set up prior to the jbd2
transaction, and thus prior to i_ino being set. If something goes
wrong, fs/crypto/ may log warning or error messages, which normally
include i_ino. So it needs to know whether it is valid to include i_ino
yet or not. Also, on some files i_ino needs to be hashed for use in the
crypto, so fs/crypto/ needs to know whether that can be done yet or not.
There are ways this could be worked around, either in fs/crypto/ or in
fs/ext4/. But, it seems there's no reason not to just fix
inode_init_always() to do the expected thing and initialize i_ino to 0.
So, do that, and also remove the initialization in jfs_fill_super() that
becomes redundant.
Signed-off-by: Eric Biggers <ebiggers@google.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2020-10-31 03:44:20 +03:00
inode - > i_ino = 0 ;
2011-10-28 16:13:30 +04:00
inode - > __i_nlink = 1 ;
2011-08-07 09:45:50 +04:00
inode - > i_opflags = 0 ;
2016-09-29 18:48:39 +03:00
if ( sb - > s_xattr )
inode - > i_opflags | = IOP_XATTR ;
2012-02-08 19:07:50 +04:00
i_uid_write ( inode , 0 ) ;
i_gid_write ( inode , 0 ) ;
2008-10-30 09:32:23 +03:00
atomic_set ( & inode - > i_writecount , 0 ) ;
inode - > i_size = 0 ;
fs: add fcntl() interface for setting/getting write life time hints
Define a set of write life time hints:
RWH_WRITE_LIFE_NOT_SET No hint information set
RWH_WRITE_LIFE_NONE No hints about write life time
RWH_WRITE_LIFE_SHORT Data written has a short life time
RWH_WRITE_LIFE_MEDIUM Data written has a medium life time
RWH_WRITE_LIFE_LONG Data written has a long life time
RWH_WRITE_LIFE_EXTREME Data written has an extremely long life time
The intent is for these values to be relative to each other, no
absolute meaning should be attached to these flag names.
Add an fcntl interface for querying these flags, and also for
setting them as well:
F_GET_RW_HINT Returns the read/write hint set on the
underlying inode.
F_SET_RW_HINT Set one of the above write hints on the
underlying inode.
F_GET_FILE_RW_HINT Returns the read/write hint set on the
file descriptor.
F_SET_FILE_RW_HINT Set one of the above write hints on the
file descriptor.
The user passes in a 64-bit pointer to get/set these values, and
the interface returns 0/-1 on success/error.
Sample program testing/implementing basic setting/getting of write
hints is below.
Add support for storing the write life time hint in the inode flags
and in struct file as well, and pass them to the kiocb flags. If
both a file and its corresponding inode has a write hint, then we
use the one in the file, if available. The file hint can be used
for sync/direct IO, for buffered writeback only the inode hint
is available.
This is in preparation for utilizing these hints in the block layer,
to guide on-media data placement.
/*
* writehint.c: get or set an inode write hint
*/
#include <stdio.h>
#include <fcntl.h>
#include <stdlib.h>
#include <unistd.h>
#include <stdbool.h>
#include <inttypes.h>
#ifndef F_GET_RW_HINT
#define F_LINUX_SPECIFIC_BASE 1024
#define F_GET_RW_HINT (F_LINUX_SPECIFIC_BASE + 11)
#define F_SET_RW_HINT (F_LINUX_SPECIFIC_BASE + 12)
#endif
static char *str[] = { "RWF_WRITE_LIFE_NOT_SET", "RWH_WRITE_LIFE_NONE",
"RWH_WRITE_LIFE_SHORT", "RWH_WRITE_LIFE_MEDIUM",
"RWH_WRITE_LIFE_LONG", "RWH_WRITE_LIFE_EXTREME" };
int main(int argc, char *argv[])
{
uint64_t hint;
int fd, ret;
if (argc < 2) {
fprintf(stderr, "%s: file <hint>\n", argv[0]);
return 1;
}
fd = open(argv[1], O_RDONLY);
if (fd < 0) {
perror("open");
return 2;
}
if (argc > 2) {
hint = atoi(argv[2]);
ret = fcntl(fd, F_SET_RW_HINT, &hint);
if (ret < 0) {
perror("fcntl: F_SET_RW_HINT");
return 4;
}
}
ret = fcntl(fd, F_GET_RW_HINT, &hint);
if (ret < 0) {
perror("fcntl: F_GET_RW_HINT");
return 3;
}
printf("%s: hint %s\n", argv[1], str[hint]);
close(fd);
return 0;
}
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2017-06-27 20:47:04 +03:00
inode - > i_write_hint = WRITE_LIFE_NOT_SET ;
2008-10-30 09:32:23 +03:00
inode - > i_blocks = 0 ;
inode - > i_bytes = 0 ;
inode - > i_generation = 0 ;
inode - > i_pipe = NULL ;
inode - > i_cdev = NULL ;
2015-05-02 16:54:06 +03:00
inode - > i_link = NULL ;
parallel lookups machinery, part 2
We'll need to verify that there's neither a hashed nor in-lookup
dentry with desired parent/name before adding to in-lookup set.
One possible solution would be to hold the parent's ->d_lock through
both checks, but while the in-lookup set is relatively small at any
time, dcache is not. And holding the parent's ->d_lock through
something like __d_lookup_rcu() would suck too badly.
So we leave the parent's ->d_lock alone, which means that we watch
out for the following scenario:
* we verify that there's no hashed match
* existing in-lookup match gets hashed by another process
* we verify that there's no in-lookup matches and decide
that everything's fine.
Solution: per-directory kinda-sorta seqlock, bumped around the times
we hash something that used to be in-lookup or move (and hash)
something in place of in-lookup. Then the above would turn into
* read the counter
* do dcache lookup
* if no matches found, check for in-lookup matches
* if there had been none of those either, check if the
counter has changed; repeat if it has.
The "kinda-sorta" part is due to the fact that we don't have much spare
space in inode. There is a spare word (shared with i_bdev/i_cdev/i_pipe),
so the counter part is not a problem, but spinlock is a different story.
We could use the parent's ->d_lock, and it would be less painful in
terms of contention, for __d_add() it would be rather inconvenient to
grab; we could do that (using lock_parent()), but...
Fortunately, we can get serialization on the counter itself, and it
might be a good idea in general; we can use cmpxchg() in a loop to
get from even to odd and smp_store_release() from odd to even.
This commit adds the counter and updating logics; the readers will be
added in the next commit.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-04-15 07:58:55 +03:00
inode - > i_dir_seq = 0 ;
2008-10-30 09:32:23 +03:00
inode - > i_rdev = 0 ;
inode - > dirtied_when = 0 ;
2009-02-04 17:06:57 +03:00
2016-02-17 00:34:39 +03:00
# ifdef CONFIG_CGROUP_WRITEBACK
inode - > i_wb_frn_winner = 0 ;
inode - > i_wb_frn_avg_time = 0 ;
inode - > i_wb_frn_history = 0 ;
# endif
2008-10-30 09:32:23 +03:00
spin_lock_init ( & inode - > i_lock ) ;
lockdep_set_class ( & inode - > i_lock , & sb - > s_type - > i_lock_key ) ;
2016-04-15 22:08:36 +03:00
init_rwsem ( & inode - > i_rwsem ) ;
lockdep_set_class ( & inode - > i_rwsem , & sb - > s_type - > i_mutex_key ) ;
2008-10-30 09:32:23 +03:00
2011-06-24 22:29:43 +04:00
atomic_set ( & inode - > i_dio_count , 0 ) ;
2008-10-30 09:32:23 +03:00
mapping - > a_ops = & empty_aops ;
mapping - > host = inode ;
mapping - > flags = 0 ;
2018-05-31 05:43:53 +03:00
mapping - > wb_err = 0 ;
2014-08-09 01:25:25 +04:00
atomic_set ( & mapping - > i_mmap_writable , 0 ) ;
2019-09-24 01:38:03 +03:00
# ifdef CONFIG_READ_ONLY_THP_FOR_FS
atomic_set ( & mapping - > nr_thps , 0 ) ;
# endif
2009-01-07 01:39:23 +03:00
mapping_set_gfp_mask ( mapping , GFP_HIGHUSER_MOVABLE ) ;
2012-12-12 04:02:35 +04:00
mapping - > private_data = NULL ;
2008-10-30 09:32:23 +03:00
mapping - > writeback_index = 0 ;
2021-09-01 11:44:03 +03:00
init_rwsem ( & mapping - > invalidate_lock ) ;
lockdep_set_class_and_name ( & mapping - > invalidate_lock ,
& sb - > s_type - > invalidate_lock_key ,
" mapping.invalidate_lock " ) ;
2008-10-30 09:32:23 +03:00
inode - > i_private = NULL ;
inode - > i_mapping = mapping ;
2012-06-09 21:51:19 +04:00
INIT_HLIST_HEAD ( & inode - > i_dentry ) ; /* buggered by rcu freeing */
2009-06-09 03:50:45 +04:00
# ifdef CONFIG_FS_POSIX_ACL
inode - > i_acl = inode - > i_default_acl = ACL_NOT_CACHED ;
# endif
2008-10-30 09:32:23 +03:00
2009-05-22 01:01:26 +04:00
# ifdef CONFIG_FSNOTIFY
inode - > i_fsnotify_mask = 0 ;
# endif
2015-01-16 23:05:54 +03:00
inode - > i_flctx = NULL ;
2022-08-16 07:08:58 +03:00
if ( unlikely ( security_inode_alloc ( inode ) ) )
return - ENOMEM ;
fs: use fast counters for vfs caches
percpu_counter library generates quite nasty code, so unless you need
to dynamically allocate counters or take fast approximate value, a
simple per cpu set of counters is much better.
The percpu_counter can never be made to work as well, because it has an
indirection from pointer to percpu memory, and it can't use direct
this_cpu_inc interfaces because it doesn't use static PER_CPU data, so
code will always be worse.
In the fastpath, it is the difference between this:
incl %gs:nr_dentry # nr_dentry
and this:
movl percpu_counter_batch(%rip), %edx # percpu_counter_batch,
movl $1, %esi #,
movq $nr_dentry, %rdi #,
call __percpu_counter_add # (plus I clobber registers)
__percpu_counter_add:
pushq %rbp #
movq %rsp, %rbp #,
subq $32, %rsp #,
movq %rbx, -24(%rbp) #,
movq %r12, -16(%rbp) #,
movq %r13, -8(%rbp) #,
movq %rdi, %rbx # fbc, fbc
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
incl -8124(%rax) # <variable>.preempt_count
movq 32(%rdi), %r12 # <variable>.counters, tcp_ptr__
#APP
# 78 "lib/percpu_counter.c" 1
add %gs:this_cpu_off, %r12 # this_cpu_off, tcp_ptr__
# 0 "" 2
#NO_APP
movslq (%r12),%r13 #* tcp_ptr__, tmp73
movslq %edx,%rax # batch, batch
addq %rsi, %r13 # amount, count
cmpq %rax, %r13 # batch, count
jge .L27 #,
negl %edx # tmp76
movslq %edx,%rdx # tmp76, tmp77
cmpq %rdx, %r13 # tmp77, count
jg .L28 #,
.L27:
movq %rbx, %rdi # fbc,
call _raw_spin_lock #
addq %r13, 8(%rbx) # count, <variable>.count
movq %rbx, %rdi # fbc,
movl $0, (%r12) #,* tcp_ptr__
call _raw_spin_unlock #
.L29:
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
decl -8124(%rax) # <variable>.preempt_count
movq -8136(%rax), %rax #, D.14625
testb $8, %al #, D.14625
jne .L32 #,
.L31:
movq -24(%rbp), %rbx #,
movq -16(%rbp), %r12 #,
movq -8(%rbp), %r13 #,
leave
ret
.p2align 4,,10
.p2align 3
.L28:
movl %r13d, (%r12) # count,*
jmp .L29 #
.L32:
call preempt_schedule #
.p2align 4,,6
jmp .L31 #
.size __percpu_counter_add, .-__percpu_counter_add
.p2align 4,,15
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 09:49:19 +03:00
this_cpu_inc ( nr_inodes ) ;
2010-10-23 13:03:02 +04:00
2009-08-07 21:38:25 +04:00
return 0 ;
2005-04-17 02:20:36 +04:00
}
2008-10-30 09:32:23 +03:00
EXPORT_SYMBOL ( inode_init_always ) ;
2019-04-10 21:43:44 +03:00
void free_inode_nonrcu ( struct inode * inode )
{
kmem_cache_free ( inode_cachep , inode ) ;
}
EXPORT_SYMBOL ( free_inode_nonrcu ) ;
static void i_callback ( struct rcu_head * head )
{
struct inode * inode = container_of ( head , struct inode , i_rcu ) ;
if ( inode - > free_inode )
inode - > free_inode ( inode ) ;
else
free_inode_nonrcu ( inode ) ;
}
2008-10-30 09:32:23 +03:00
static struct inode * alloc_inode ( struct super_block * sb )
{
2019-04-10 21:43:44 +03:00
const struct super_operations * ops = sb - > s_op ;
2008-10-30 09:32:23 +03:00
struct inode * inode ;
2019-04-10 21:43:44 +03:00
if ( ops - > alloc_inode )
inode = ops - > alloc_inode ( sb ) ;
2008-10-30 09:32:23 +03:00
else
2022-03-23 00:41:00 +03:00
inode = alloc_inode_sb ( sb , inode_cachep , GFP_KERNEL ) ;
2008-10-30 09:32:23 +03:00
2009-08-07 21:38:25 +04:00
if ( ! inode )
return NULL ;
if ( unlikely ( inode_init_always ( sb , inode ) ) ) {
2019-04-10 21:43:44 +03:00
if ( ops - > destroy_inode ) {
ops - > destroy_inode ( inode ) ;
if ( ! ops - > free_inode )
return NULL ;
}
inode - > free_inode = ops - > free_inode ;
i_callback ( & inode - > i_rcu ) ;
2009-08-07 21:38:25 +04:00
return NULL ;
}
return inode ;
2008-10-30 09:32:23 +03:00
}
2005-04-17 02:20:36 +04:00
2009-08-07 21:38:29 +04:00
void __destroy_inode ( struct inode * inode )
2005-04-17 02:20:36 +04:00
{
2006-04-02 15:38:18 +04:00
BUG_ON ( inode_has_buffers ( inode ) ) ;
2015-05-23 00:13:37 +03:00
inode_detach_wb ( inode ) ;
2005-04-17 02:20:36 +04:00
security_inode_free ( inode ) ;
2009-05-22 01:01:26 +04:00
fsnotify_inode_delete ( inode ) ;
2016-01-07 23:08:51 +03:00
locks_free_lock_context ( inode ) ;
2011-11-21 15:11:32 +04:00
if ( ! inode - > i_nlink ) {
WARN_ON ( atomic_long_read ( & inode - > i_sb - > s_remove_count ) = = 0 ) ;
atomic_long_dec ( & inode - > i_sb - > s_remove_count ) ;
}
2009-06-09 03:50:45 +04:00
# ifdef CONFIG_FS_POSIX_ACL
2016-03-24 16:38:37 +03:00
if ( inode - > i_acl & & ! is_uncached_acl ( inode - > i_acl ) )
2009-06-09 03:50:45 +04:00
posix_acl_release ( inode - > i_acl ) ;
2016-03-24 16:38:37 +03:00
if ( inode - > i_default_acl & & ! is_uncached_acl ( inode - > i_default_acl ) )
2009-06-09 03:50:45 +04:00
posix_acl_release ( inode - > i_default_acl ) ;
# endif
fs: use fast counters for vfs caches
percpu_counter library generates quite nasty code, so unless you need
to dynamically allocate counters or take fast approximate value, a
simple per cpu set of counters is much better.
The percpu_counter can never be made to work as well, because it has an
indirection from pointer to percpu memory, and it can't use direct
this_cpu_inc interfaces because it doesn't use static PER_CPU data, so
code will always be worse.
In the fastpath, it is the difference between this:
incl %gs:nr_dentry # nr_dentry
and this:
movl percpu_counter_batch(%rip), %edx # percpu_counter_batch,
movl $1, %esi #,
movq $nr_dentry, %rdi #,
call __percpu_counter_add # (plus I clobber registers)
__percpu_counter_add:
pushq %rbp #
movq %rsp, %rbp #,
subq $32, %rsp #,
movq %rbx, -24(%rbp) #,
movq %r12, -16(%rbp) #,
movq %r13, -8(%rbp) #,
movq %rdi, %rbx # fbc, fbc
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
incl -8124(%rax) # <variable>.preempt_count
movq 32(%rdi), %r12 # <variable>.counters, tcp_ptr__
#APP
# 78 "lib/percpu_counter.c" 1
add %gs:this_cpu_off, %r12 # this_cpu_off, tcp_ptr__
# 0 "" 2
#NO_APP
movslq (%r12),%r13 #* tcp_ptr__, tmp73
movslq %edx,%rax # batch, batch
addq %rsi, %r13 # amount, count
cmpq %rax, %r13 # batch, count
jge .L27 #,
negl %edx # tmp76
movslq %edx,%rdx # tmp76, tmp77
cmpq %rdx, %r13 # tmp77, count
jg .L28 #,
.L27:
movq %rbx, %rdi # fbc,
call _raw_spin_lock #
addq %r13, 8(%rbx) # count, <variable>.count
movq %rbx, %rdi # fbc,
movl $0, (%r12) #,* tcp_ptr__
call _raw_spin_unlock #
.L29:
#APP
# 216 "/home/npiggin/usr/src/linux-2.6/arch/x86/include/asm/thread_info.h" 1
movq %gs:kernel_stack,%rax #, pfo_ret__
# 0 "" 2
#NO_APP
decl -8124(%rax) # <variable>.preempt_count
movq -8136(%rax), %rax #, D.14625
testb $8, %al #, D.14625
jne .L32 #,
.L31:
movq -24(%rbp), %rbx #,
movq -16(%rbp), %r12 #,
movq -8(%rbp), %r13 #,
leave
ret
.p2align 4,,10
.p2align 3
.L28:
movl %r13d, (%r12) # count,*
jmp .L29 #
.L32:
call preempt_schedule #
.p2align 4,,6
jmp .L31 #
.size __percpu_counter_add, .-__percpu_counter_add
.p2align 4,,15
Signed-off-by: Nick Piggin <npiggin@kernel.dk>
2011-01-07 09:49:19 +03:00
this_cpu_dec ( nr_inodes ) ;
2009-08-07 21:38:29 +04:00
}
EXPORT_SYMBOL ( __destroy_inode ) ;
2010-10-06 12:48:55 +04:00
static void destroy_inode ( struct inode * inode )
2009-08-07 21:38:29 +04:00
{
2019-04-10 21:43:44 +03:00
const struct super_operations * ops = inode - > i_sb - > s_op ;
2010-10-21 04:49:30 +04:00
BUG_ON ( ! list_empty ( & inode - > i_lru ) ) ;
2009-08-07 21:38:29 +04:00
__destroy_inode ( inode ) ;
2019-04-10 21:43:44 +03:00
if ( ops - > destroy_inode ) {
ops - > destroy_inode ( inode ) ;
if ( ! ops - > free_inode )
return ;
}
inode - > free_inode = ops - > free_inode ;
call_rcu ( & inode - > i_rcu , i_callback ) ;
2005-04-17 02:20:36 +04:00
}
2011-11-21 15:11:32 +04:00
/**
* drop_nlink - directly drop an inode ' s link count
* @ inode : inode
*
* This is a low - level filesystem helper to replace any
* direct filesystem manipulation of i_nlink . In cases
* where we are attempting to track writes to the
* filesystem , a decrement to zero means an imminent
* write when the file is truncated and actually unlinked
* on the filesystem .
*/
void drop_nlink ( struct inode * inode )
{
WARN_ON ( inode - > i_nlink = = 0 ) ;
inode - > __i_nlink - - ;
if ( ! inode - > i_nlink )
atomic_long_inc ( & inode - > i_sb - > s_remove_count ) ;
}
EXPORT_SYMBOL ( drop_nlink ) ;
/**
* clear_nlink - directly zero an inode ' s link count
* @ inode : inode
*
* This is a low - level filesystem helper to replace any
* direct filesystem manipulation of i_nlink . See
* drop_nlink ( ) for why we care about i_nlink hitting zero .
*/
void clear_nlink ( struct inode * inode )
{
if ( inode - > i_nlink ) {
inode - > __i_nlink = 0 ;
atomic_long_inc ( & inode - > i_sb - > s_remove_count ) ;
}
}
EXPORT_SYMBOL ( clear_nlink ) ;
/**
* set_nlink - directly set an inode ' s link count
* @ inode : inode
* @ nlink : new nlink ( should be non - zero )
*
* This is a low - level filesystem helper to replace any
* direct filesystem manipulation of i_nlink .
*/
void set_nlink ( struct inode * inode , unsigned int nlink )
{
if ( ! nlink ) {
clear_nlink ( inode ) ;
} else {
/* Yes, some filesystems do change nlink from zero to one */
if ( inode - > i_nlink = = 0 )
atomic_long_dec ( & inode - > i_sb - > s_remove_count ) ;
inode - > __i_nlink = nlink ;
}
}
EXPORT_SYMBOL ( set_nlink ) ;
/**
* inc_nlink - directly increment an inode ' s link count
* @ inode : inode
*
* This is a low - level filesystem helper to replace any
* direct filesystem manipulation of i_nlink . Currently ,
* it is only here for parity with dec_nlink ( ) .
*/
void inc_nlink ( struct inode * inode )
{
2013-06-11 08:34:36 +04:00
if ( unlikely ( inode - > i_nlink = = 0 ) ) {
WARN_ON ( ! ( inode - > i_state & I_LINKABLE ) ) ;
2011-11-21 15:11:32 +04:00
atomic_long_dec ( & inode - > i_sb - > s_remove_count ) ;
2013-06-11 08:34:36 +04:00
}
2011-11-21 15:11:32 +04:00
inode - > __i_nlink + + ;
}
EXPORT_SYMBOL ( inc_nlink ) ;
2018-03-07 04:30:34 +03:00
static void __address_space_init_once ( struct address_space * mapping )
2011-02-23 15:49:47 +03:00
{
mm: fix page cache convergence regression
Since a28334862993 ("page cache: Finish XArray conversion"), on most
major Linux distributions, the page cache doesn't correctly transition
when the hot data set is changing, and leaves the new pages thrashing
indefinitely instead of kicking out the cold ones.
On a freshly booted, freshly ssh'd into virtual machine with 1G RAM
running stock Arch Linux:
[root@ham ~]# ./reclaimtest.sh
+ dd of=workingset-a bs=1M count=0 seek=600
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ ./mincore workingset-a
153600/153600 workingset-a
+ dd of=workingset-b bs=1M count=0 seek=600
+ cat workingset-b
+ cat workingset-b
+ cat workingset-b
+ cat workingset-b
+ ./mincore workingset-a workingset-b
104029/153600 workingset-a
120086/153600 workingset-b
+ cat workingset-b
+ cat workingset-b
+ cat workingset-b
+ cat workingset-b
+ ./mincore workingset-a workingset-b
104029/153600 workingset-a
120268/153600 workingset-b
workingset-b is a 600M file on a 1G host that is otherwise entirely
idle. No matter how often it's being accessed, it won't get cached.
While investigating, I noticed that the non-resident information gets
aggressively reclaimed - /proc/vmstat::workingset_nodereclaim. This is
a problem because a workingset transition like this relies on the
non-resident information tracked in the page cache tree of evicted
file ranges: when the cache faults are refaults of recently evicted
cache, we challenge the existing active set, and that allows a new
workingset to establish itself.
Tracing the shrinker that maintains this memory revealed that all page
cache tree nodes were allocated to the root cgroup. This is a problem,
because 1) the shrinker sizes the amount of non-resident information
it keeps to the size of the cgroup's other memory and 2) on most major
Linux distributions, only kernel threads live in the root cgroup and
everything else gets put into services or session groups:
[root@ham ~]# cat /proc/self/cgroup
0::/user.slice/user-0.slice/session-c1.scope
As a result, we basically maintain no non-resident information for the
workloads running on the system, thus breaking the caching algorithm.
Looking through the code, I found the culprit in the above-mentioned
patch: when switching from the radix tree to xarray, it dropped the
__GFP_ACCOUNT flag from the tree node allocations - the flag that
makes sure the allocated memory gets charged to and tracked by the
cgroup of the calling process - in this case, the one doing the fault.
To fix this, allow xarray users to specify per-tree flag that makes
xarray allocate nodes using __GFP_ACCOUNT. Then restore the page cache
tree annotation to request such cgroup tracking for the cache nodes.
With this patch applied, the page cache correctly converges on new
workingsets again after just a few iterations:
[root@ham ~]# ./reclaimtest.sh
+ dd of=workingset-a bs=1M count=0 seek=600
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ cat workingset-a
+ ./mincore workingset-a
153600/153600 workingset-a
+ dd of=workingset-b bs=1M count=0 seek=600
+ cat workingset-b
+ ./mincore workingset-a workingset-b
124607/153600 workingset-a
87876/153600 workingset-b
+ cat workingset-b
+ ./mincore workingset-a workingset-b
81313/153600 workingset-a
133321/153600 workingset-b
+ cat workingset-b
+ ./mincore workingset-a workingset-b
63036/153600 workingset-a
153600/153600 workingset-b
Cc: stable@vger.kernel.org # 4.20+
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Reviewed-by: Shakeel Butt <shakeelb@google.com>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
2019-05-24 17:12:46 +03:00
xa_init_flags ( & mapping - > i_pages , XA_FLAGS_LOCK_IRQ | XA_FLAGS_ACCOUNT ) ;
2014-12-13 03:54:24 +03:00
init_rwsem ( & mapping - > i_mmap_rwsem ) ;
2011-02-23 15:49:47 +03:00
INIT_LIST_HEAD ( & mapping - > private_list ) ;
spin_lock_init ( & mapping - > private_lock ) ;
2017-09-09 02:15:08 +03:00
mapping - > i_mmap = RB_ROOT_CACHED ;
2011-02-23 15:49:47 +03:00
}
2018-03-07 04:30:34 +03:00
void address_space_init_once ( struct address_space * mapping )
{
memset ( mapping , 0 , sizeof ( * mapping ) ) ;
__address_space_init_once ( mapping ) ;
}
2011-02-23 15:49:47 +03:00
EXPORT_SYMBOL ( address_space_init_once ) ;
2005-04-17 02:20:36 +04:00
/*
* These are initializations that only need to be done
* once , because the fields are idempotent across use
* of the inode , so let the slab aware of that .
*/
void inode_init_once ( struct inode * inode )
{
memset ( inode , 0 , sizeof ( * inode ) ) ;
INIT_HLIST_NODE ( & inode - > i_hash ) ;
INIT_LIST_HEAD ( & inode - > i_devices ) ;
2015-03-04 22:07:22 +03:00
INIT_LIST_HEAD ( & inode - > i_io_list ) ;
2016-07-27 01:21:50 +03:00
INIT_LIST_HEAD ( & inode - > i_wb_list ) ;
2010-10-21 04:49:30 +04:00
INIT_LIST_HEAD ( & inode - > i_lru ) ;
2022-03-31 23:29:00 +03:00
INIT_LIST_HEAD ( & inode - > i_sb_list ) ;
2018-03-07 04:30:34 +03:00
__address_space_init_once ( & inode - > i_data ) ;
2005-04-17 02:20:36 +04:00
i_size_ordered_init ( inode ) ;
}
EXPORT_SYMBOL ( inode_init_once ) ;
2008-07-26 06:45:34 +04:00
static void init_once ( void * foo )
2005-04-17 02:20:36 +04:00
{
2009-03-31 18:05:54 +04:00
struct inode * inode = ( struct inode * ) foo ;
2005-04-17 02:20:36 +04:00
2007-05-17 09:10:57 +04:00
inode_init_once ( inode ) ;
2005-04-17 02:20:36 +04:00
}
/*
2011-03-22 14:23:36 +03:00
* inode - > i_lock must be held
2005-04-17 02:20:36 +04:00
*/
2009-03-31 18:05:54 +04:00
void __iget ( struct inode * inode )
2005-04-17 02:20:36 +04:00
{
2010-10-23 14:55:17 +04:00
atomic_inc ( & inode - > i_count ) ;
}
2010-05-14 13:49:22 +04:00
2010-10-23 19:11:40 +04:00
/*
* get additional reference to inode ; caller must already hold one .
*/
void ihold ( struct inode * inode )
{
WARN_ON ( atomic_inc_return ( & inode - > i_count ) < 2 ) ;
}
EXPORT_SYMBOL ( ihold ) ;
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
static void __inode_add_lru ( struct inode * inode , bool rotate )
2010-10-23 14:55:17 +04:00
{
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
if ( inode - > i_state & ( I_DIRTY_ALL | I_SYNC | I_FREEING | I_WILL_FREE ) )
return ;
if ( atomic_read ( & inode - > i_count ) )
return ;
if ( ! ( inode - > i_sb - > s_flags & SB_ACTIVE ) )
return ;
if ( ! mapping_shrinkable ( & inode - > i_data ) )
return ;
2013-08-28 04:17:58 +04:00
if ( list_lru_add ( & inode - > i_sb - > s_inode_lru , & inode - > i_lru ) )
2011-07-08 08:14:38 +04:00
this_cpu_inc ( nr_unused ) ;
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
else if ( rotate )
2017-04-18 23:04:17 +03:00
inode - > i_state | = I_REFERENCED ;
2010-10-23 14:55:17 +04:00
}
2010-05-14 13:49:22 +04:00
2012-11-27 04:29:51 +04:00
/*
* Add inode to LRU if needed ( inode is unused and clean ) .
*
* Needs inode - > i_lock held .
*/
void inode_add_lru ( struct inode * inode )
{
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
__inode_add_lru ( inode , false ) ;
2012-11-27 04:29:51 +04:00
}
2010-10-23 14:55:17 +04:00
static void inode_lru_list_del ( struct inode * inode )
{
2013-08-28 04:17:58 +04:00
if ( list_lru_del ( & inode - > i_sb - > s_inode_lru , & inode - > i_lru ) )
2011-07-08 08:14:38 +04:00
this_cpu_dec ( nr_unused ) ;
2005-04-17 02:20:36 +04:00
}
2010-10-23 15:15:32 +04:00
/**
* inode_sb_list_add - add inode to the superblock list of inodes
* @ inode : inode to add
*/
void inode_sb_list_add ( struct inode * inode )
{
2015-03-04 20:37:22 +03:00
spin_lock ( & inode - > i_sb - > s_inode_list_lock ) ;
2011-03-22 14:23:40 +03:00
list_add ( & inode - > i_sb_list , & inode - > i_sb - > s_inodes ) ;
2015-03-04 20:37:22 +03:00
spin_unlock ( & inode - > i_sb - > s_inode_list_lock ) ;
2010-10-23 15:15:32 +04:00
}
EXPORT_SYMBOL_GPL ( inode_sb_list_add ) ;
2011-03-22 14:23:40 +03:00
static inline void inode_sb_list_del ( struct inode * inode )
2010-10-23 15:15:32 +04:00
{
2011-07-26 13:36:34 +04:00
if ( ! list_empty ( & inode - > i_sb_list ) ) {
2015-03-04 20:37:22 +03:00
spin_lock ( & inode - > i_sb - > s_inode_list_lock ) ;
2011-07-26 13:36:34 +04:00
list_del_init ( & inode - > i_sb_list ) ;
2015-03-04 20:37:22 +03:00
spin_unlock ( & inode - > i_sb - > s_inode_list_lock ) ;
2011-07-26 13:36:34 +04:00
}
2010-10-23 15:15:32 +04:00
}
2010-10-23 14:58:09 +04:00
static unsigned long hash ( struct super_block * sb , unsigned long hashval )
{
unsigned long tmp ;
tmp = ( hashval * ( unsigned long ) sb ) ^ ( GOLDEN_RATIO_PRIME + hashval ) /
L1_CACHE_BYTES ;
2011-05-27 17:28:01 +04:00
tmp = tmp ^ ( ( tmp ^ GOLDEN_RATIO_PRIME ) > > i_hash_shift ) ;
return tmp & i_hash_mask ;
2010-10-23 14:58:09 +04:00
}
/**
* __insert_inode_hash - hash an inode
* @ inode : unhashed inode
* @ hashval : unsigned long value used to locate this object in the
* inode_hashtable .
*
* Add an inode to the inode hash for this superblock .
*/
void __insert_inode_hash ( struct inode * inode , unsigned long hashval )
{
2010-10-23 15:15:32 +04:00
struct hlist_head * b = inode_hashtable + hash ( inode - > i_sb , hashval ) ;
2011-03-22 14:23:42 +03:00
spin_lock ( & inode_hash_lock ) ;
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
2017-12-01 14:40:16 +03:00
hlist_add_head_rcu ( & inode - > i_hash , b ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2011-03-22 14:23:42 +03:00
spin_unlock ( & inode_hash_lock ) ;
2010-10-23 14:58:09 +04:00
}
EXPORT_SYMBOL ( __insert_inode_hash ) ;
/**
2011-07-28 08:41:09 +04:00
* __remove_inode_hash - remove an inode from the hash
2010-10-23 14:58:09 +04:00
* @ inode : inode to unhash
*
* Remove an inode from the superblock .
*/
2011-07-28 08:41:09 +04:00
void __remove_inode_hash ( struct inode * inode )
2010-10-23 14:58:09 +04:00
{
2011-03-22 14:23:42 +03:00
spin_lock ( & inode_hash_lock ) ;
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
2017-12-01 14:40:16 +03:00
hlist_del_init_rcu ( & inode - > i_hash ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2011-03-22 14:23:42 +03:00
spin_unlock ( & inode_hash_lock ) ;
2010-10-23 14:58:09 +04:00
}
2011-07-28 08:41:09 +04:00
EXPORT_SYMBOL ( __remove_inode_hash ) ;
2010-10-23 14:58:09 +04:00
2022-01-15 01:05:04 +03:00
void dump_mapping ( const struct address_space * mapping )
{
struct inode * host ;
const struct address_space_operations * a_ops ;
struct hlist_node * dentry_first ;
struct dentry * dentry_ptr ;
struct dentry dentry ;
unsigned long ino ;
/*
* If mapping is an invalid pointer , we don ' t want to crash
* accessing it , so probe everything depending on it carefully .
*/
if ( get_kernel_nofault ( host , & mapping - > host ) | |
get_kernel_nofault ( a_ops , & mapping - > a_ops ) ) {
pr_warn ( " invalid mapping:%px \n " , mapping ) ;
return ;
}
if ( ! host ) {
pr_warn ( " aops:%ps \n " , a_ops ) ;
return ;
}
if ( get_kernel_nofault ( dentry_first , & host - > i_dentry . first ) | |
get_kernel_nofault ( ino , & host - > i_ino ) ) {
pr_warn ( " aops:%ps invalid inode:%px \n " , a_ops , host ) ;
return ;
}
if ( ! dentry_first ) {
pr_warn ( " aops:%ps ino:%lx \n " , a_ops , ino ) ;
return ;
}
dentry_ptr = container_of ( dentry_first , struct dentry , d_u . d_alias ) ;
if ( get_kernel_nofault ( dentry , dentry_ptr ) ) {
pr_warn ( " aops:%ps ino:%lx invalid dentry:%px \n " ,
a_ops , ino , dentry_ptr ) ;
return ;
}
/*
* if dentry is corrupted , the % pd handler may still crash ,
* but it ' s unlikely that we reach here with a corrupt mapping
*/
pr_warn ( " aops:%ps ino:%lx dentry name: \" %pd \" \n " , a_ops , ino , & dentry ) ;
}
2012-05-03 16:48:02 +04:00
void clear_inode ( struct inode * inode )
2010-06-05 04:55:25 +04:00
{
2011-06-28 03:18:10 +04:00
/*
2018-04-11 02:36:56 +03:00
* We have to cycle the i_pages lock here because reclaim can be in the
2022-06-29 03:41:40 +03:00
* process of removing the last page ( in __filemap_remove_folio ( ) )
2018-04-11 02:36:56 +03:00
* and we must not free the mapping under it .
2011-06-28 03:18:10 +04:00
*/
2018-04-11 02:36:56 +03:00
xa_lock_irq ( & inode - > i_data . i_pages ) ;
2010-06-05 04:55:25 +04:00
BUG_ON ( inode - > i_data . nrpages ) ;
2021-05-05 04:32:57 +03:00
/*
* Almost always , mapping_empty ( & inode - > i_data ) here ; but there are
* two known and long - standing ways in which nodes may get left behind
* ( when deep radix - tree node allocation failed partway ; or when THP
* collapse_file ( ) failed ) . Until those two known cases are cleaned up ,
* or a cleanup function is called here , do not BUG_ON ( ! mapping_empty ) ,
* nor even WARN_ON ( ! mapping_empty ) .
*/
2018-04-11 02:36:56 +03:00
xa_unlock_irq ( & inode - > i_data . i_pages ) ;
2010-06-05 04:55:25 +04:00
BUG_ON ( ! list_empty ( & inode - > i_data . private_list ) ) ;
BUG_ON ( ! ( inode - > i_state & I_FREEING ) ) ;
BUG_ON ( inode - > i_state & I_CLEAR ) ;
2016-07-27 01:21:50 +03:00
BUG_ON ( ! list_empty ( & inode - > i_wb_list ) ) ;
2011-01-07 09:49:49 +03:00
/* don't need i_lock here, no concurrent mods to i_state */
2010-06-05 04:55:25 +04:00
inode - > i_state = I_FREEING | I_CLEAR ;
}
2012-05-03 16:48:02 +04:00
EXPORT_SYMBOL ( clear_inode ) ;
2010-06-05 04:55:25 +04:00
2011-03-22 14:23:37 +03:00
/*
* Free the inode passed in , removing it from the lists it is still connected
* to . We remove any pages still attached to the inode and wait for any IO that
* is still in progress before finally destroying the inode .
*
* An inode must already be marked I_FREEING so that we avoid the inode being
* moved back onto lists if we race with other code that manipulates the lists
* ( e . g . writeback_single_inode ) . The caller is responsible for setting this .
*
* An inode must already be removed from the LRU list before being evicted from
* the cache . This should occur atomically with setting the I_FREEING state
* flag , so no inodes here should ever be on the LRU when being evicted .
*/
2010-06-07 21:21:05 +04:00
static void evict ( struct inode * inode )
2010-06-05 03:33:20 +04:00
{
const struct super_operations * op = inode - > i_sb - > s_op ;
2011-03-22 14:23:37 +03:00
BUG_ON ( ! ( inode - > i_state & I_FREEING ) ) ;
BUG_ON ( ! list_empty ( & inode - > i_lru ) ) ;
2015-03-04 22:07:22 +03:00
if ( ! list_empty ( & inode - > i_io_list ) )
inode_io_list_del ( inode ) ;
2011-07-28 08:11:47 +04:00
2011-03-22 14:23:40 +03:00
inode_sb_list_del ( inode ) ;
2012-05-03 16:48:03 +04:00
/*
* Wait for flusher thread to be done with the inode so that filesystem
* does not start destroying it while writeback is still running . Since
* the inode has I_FREEING set , flusher thread won ' t start new work on
* the inode . We just have to wait for running writeback to finish .
*/
inode_wait_for_writeback ( inode ) ;
2012-05-03 16:48:01 +04:00
2010-06-05 03:40:39 +04:00
if ( op - > evict_inode ) {
op - > evict_inode ( inode ) ;
2010-06-05 03:33:20 +04:00
} else {
2014-04-04 01:47:49 +04:00
truncate_inode_pages_final ( & inode - > i_data ) ;
2012-05-03 16:48:02 +04:00
clear_inode ( inode ) ;
2010-06-05 03:33:20 +04:00
}
2010-06-05 04:19:55 +04:00
if ( S_ISCHR ( inode - > i_mode ) & & inode - > i_cdev )
cd_forget ( inode ) ;
2011-03-22 14:23:37 +03:00
remove_inode_hash ( inode ) ;
spin_lock ( & inode - > i_lock ) ;
wake_up_bit ( & inode - > i_state , __I_NEW ) ;
BUG_ON ( inode - > i_state ! = ( I_FREEING | I_CLEAR ) ) ;
spin_unlock ( & inode - > i_lock ) ;
destroy_inode ( inode ) ;
2010-06-05 03:33:20 +04:00
}
2005-04-17 02:20:36 +04:00
/*
* dispose_list - dispose of the contents of a local list
* @ head : the head of the list to free
*
* Dispose - list gets a local list with local inodes in it , so it doesn ' t
* need to worry about list corruption and SMP locks .
*/
static void dispose_list ( struct list_head * head )
{
while ( ! list_empty ( head ) ) {
struct inode * inode ;
2010-10-21 04:49:30 +04:00
inode = list_first_entry ( head , struct inode , i_lru ) ;
list_del_init ( & inode - > i_lru ) ;
2005-04-17 02:20:36 +04:00
2010-06-07 21:21:05 +04:00
evict ( inode ) ;
2015-03-05 00:52:52 +03:00
cond_resched ( ) ;
2005-04-17 02:20:36 +04:00
}
}
2010-10-26 04:49:35 +04:00
/**
* evict_inodes - evict all evictable inodes for a superblock
* @ sb : superblock to operate on
*
* Make sure that no inodes with zero refcount are retained . This is
2017-11-28 00:05:09 +03:00
* called by superblock shutdown after having SB_ACTIVE flag removed ,
2010-10-26 04:49:35 +04:00
* so any inode reaching zero refcount during or after that call will
* be immediately evicted .
2005-04-17 02:20:36 +04:00
*/
2010-10-26 04:49:35 +04:00
void evict_inodes ( struct super_block * sb )
2005-04-17 02:20:36 +04:00
{
2010-10-26 04:49:35 +04:00
struct inode * inode , * next ;
LIST_HEAD ( dispose ) ;
2005-04-17 02:20:36 +04:00
2015-03-05 00:52:52 +03:00
again :
2015-03-04 20:37:22 +03:00
spin_lock ( & sb - > s_inode_list_lock ) ;
2010-10-26 04:49:35 +04:00
list_for_each_entry_safe ( inode , next , & sb - > s_inodes , i_sb_list ) {
if ( atomic_read ( & inode - > i_count ) )
2009-03-11 23:17:36 +03:00
continue ;
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
if ( inode - > i_state & ( I_NEW | I_FREEING | I_WILL_FREE ) ) {
spin_unlock ( & inode - > i_lock ) ;
2005-04-17 02:20:36 +04:00
continue ;
2011-03-22 14:23:36 +03:00
}
2010-10-26 04:49:35 +04:00
inode - > i_state | = I_FREEING ;
2011-03-22 14:23:38 +03:00
inode_lru_list_del ( inode ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2011-03-22 14:23:38 +03:00
list_add ( & inode - > i_lru , & dispose ) ;
2015-03-05 00:52:52 +03:00
/*
* We can have a ton of inodes to evict at unmount time given
* enough memory , check to see if we need to go to sleep for a
* bit so we don ' t livelock .
*/
if ( need_resched ( ) ) {
spin_unlock ( & sb - > s_inode_list_lock ) ;
cond_resched ( ) ;
dispose_list ( & dispose ) ;
goto again ;
}
2005-04-17 02:20:36 +04:00
}
2015-03-04 20:37:22 +03:00
spin_unlock ( & sb - > s_inode_list_lock ) ;
2010-10-26 04:49:35 +04:00
dispose_list ( & dispose ) ;
2005-04-17 02:20:36 +04:00
}
2017-08-19 04:08:25 +03:00
EXPORT_SYMBOL_GPL ( evict_inodes ) ;
2005-04-17 02:20:36 +04:00
/**
2010-10-24 21:40:33 +04:00
* invalidate_inodes - attempt to free all inodes on a superblock
* @ sb : superblock to operate on
2011-02-24 09:25:47 +03:00
* @ kill_dirty : flag to guide handling of dirty inodes
2005-04-17 02:20:36 +04:00
*
2010-10-24 21:40:33 +04:00
* Attempts to free all inodes for a given superblock . If there were any
* busy inodes return a non - zero value , else zero .
2011-02-24 09:25:47 +03:00
* If @ kill_dirty is set , discard dirty inodes too , otherwise treat
* them as busy .
2005-04-17 02:20:36 +04:00
*/
2011-02-24 09:25:47 +03:00
int invalidate_inodes ( struct super_block * sb , bool kill_dirty )
2005-04-17 02:20:36 +04:00
{
2010-10-23 13:03:02 +04:00
int busy = 0 ;
2010-10-24 21:40:33 +04:00
struct inode * inode , * next ;
LIST_HEAD ( dispose ) ;
2005-04-17 02:20:36 +04:00
2019-12-06 19:54:23 +03:00
again :
2015-03-04 20:37:22 +03:00
spin_lock ( & sb - > s_inode_list_lock ) ;
2010-10-24 21:40:33 +04:00
list_for_each_entry_safe ( inode , next , & sb - > s_inodes , i_sb_list ) {
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
if ( inode - > i_state & ( I_NEW | I_FREEING | I_WILL_FREE ) ) {
spin_unlock ( & inode - > i_lock ) ;
2009-03-11 23:17:36 +03:00
continue ;
2011-03-22 14:23:36 +03:00
}
2015-02-02 08:37:00 +03:00
if ( inode - > i_state & I_DIRTY_ALL & & ! kill_dirty ) {
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2011-02-24 09:25:47 +03:00
busy = 1 ;
continue ;
}
2010-10-23 21:07:20 +04:00
if ( atomic_read ( & inode - > i_count ) ) {
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2010-10-23 21:07:20 +04:00
busy = 1 ;
2005-04-17 02:20:36 +04:00
continue ;
}
2010-10-23 21:07:20 +04:00
inode - > i_state | = I_FREEING ;
2011-03-22 14:23:38 +03:00
inode_lru_list_del ( inode ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2011-03-22 14:23:38 +03:00
list_add ( & inode - > i_lru , & dispose ) ;
2019-12-06 19:54:23 +03:00
if ( need_resched ( ) ) {
spin_unlock ( & sb - > s_inode_list_lock ) ;
cond_resched ( ) ;
dispose_list ( & dispose ) ;
goto again ;
}
2005-04-17 02:20:36 +04:00
}
2015-03-04 20:37:22 +03:00
spin_unlock ( & sb - > s_inode_list_lock ) ;
2005-04-17 02:20:36 +04:00
2010-10-24 21:40:33 +04:00
dispose_list ( & dispose ) ;
2005-04-17 02:20:36 +04:00
return busy ;
}
/*
2013-08-28 04:17:58 +04:00
* Isolate the inode from the LRU in preparation for freeing it .
2005-04-17 02:20:36 +04:00
*
2010-10-23 14:55:17 +04:00
* If the inode has the I_REFERENCED flag set , then it means that it has been
* used recently - the flag is set in iput_final ( ) . When we encounter such an
* inode , clear the flag and move it to the back of the LRU so it gets another
* pass through the LRU before it gets reclaimed . This is necessary because of
* the fact we are doing lazy LRU updates to minimise lock contention so the
* LRU does not have strict ordering . Hence we don ' t want to reclaim inodes
* with this flag set because they are the inodes that are out of order .
2005-04-17 02:20:36 +04:00
*/
2015-02-13 01:59:35 +03:00
static enum lru_status inode_lru_isolate ( struct list_head * item ,
struct list_lru_one * lru , spinlock_t * lru_lock , void * arg )
2005-04-17 02:20:36 +04:00
{
2013-08-28 04:17:58 +04:00
struct list_head * freeable = arg ;
struct inode * inode = container_of ( item , struct inode , i_lru ) ;
2005-04-17 02:20:36 +04:00
2013-08-28 04:17:58 +04:00
/*
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
* We are inverting the lru lock / inode - > i_lock here , so use a
* trylock . If we fail to get the lock , just skip it .
2013-08-28 04:17:58 +04:00
*/
if ( ! spin_trylock ( & inode - > i_lock ) )
return LRU_SKIP ;
2005-04-17 02:20:36 +04:00
2013-08-28 04:17:58 +04:00
/*
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
* Inodes can get referenced , redirtied , or repopulated while
* they ' re already on the LRU , and this can make them
* unreclaimable for a while . Remove them lazily here ; iput ,
* sync , or the last page cache deletion will requeue them .
2013-08-28 04:17:58 +04:00
*/
if ( atomic_read ( & inode - > i_count ) | |
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
( inode - > i_state & ~ I_REFERENCED ) | |
! mapping_shrinkable ( & inode - > i_data ) ) {
2015-02-13 01:59:35 +03:00
list_lru_isolate ( lru , & inode - > i_lru ) ;
2013-08-28 04:17:58 +04:00
spin_unlock ( & inode - > i_lock ) ;
this_cpu_dec ( nr_unused ) ;
return LRU_REMOVED ;
}
2005-04-17 02:20:36 +04:00
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
/* Recently referenced inodes get one more pass */
2019-02-13 02:35:51 +03:00
if ( inode - > i_state & I_REFERENCED ) {
2013-08-28 04:17:58 +04:00
inode - > i_state & = ~ I_REFERENCED ;
spin_unlock ( & inode - > i_lock ) ;
return LRU_ROTATE ;
}
2005-04-17 02:20:36 +04:00
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
/*
* On highmem systems , mapping_shrinkable ( ) permits dropping
* page cache in order to free up struct inodes : lowmem might
* be under pressure before the cache inside the highmem zone .
*/
2021-09-03 00:53:24 +03:00
if ( inode_has_buffers ( inode ) | | ! mapping_empty ( & inode - > i_data ) ) {
2013-08-28 04:17:58 +04:00
__iget ( inode ) ;
spin_unlock ( & inode - > i_lock ) ;
spin_unlock ( lru_lock ) ;
if ( remove_inode_buffers ( inode ) ) {
unsigned long reap ;
reap = invalidate_mapping_pages ( & inode - > i_data , 0 , - 1 ) ;
if ( current_is_kswapd ( ) )
__count_vm_events ( KSWAPD_INODESTEAL , reap ) ;
else
__count_vm_events ( PGINODESTEAL , reap ) ;
2023-04-13 13:40:34 +03:00
mm_account_reclaimed_pages ( reap ) ;
2011-03-22 14:23:38 +03:00
}
2013-08-28 04:17:58 +04:00
iput ( inode ) ;
spin_lock ( lru_lock ) ;
return LRU_RETRY ;
}
2011-03-22 14:23:38 +03:00
2013-08-28 04:17:58 +04:00
WARN_ON ( inode - > i_state & I_NEW ) ;
inode - > i_state | = I_FREEING ;
2015-02-13 01:59:35 +03:00
list_lru_isolate_move ( lru , & inode - > i_lru , freeable ) ;
2013-08-28 04:17:58 +04:00
spin_unlock ( & inode - > i_lock ) ;
2010-10-23 14:55:17 +04:00
2013-08-28 04:17:58 +04:00
this_cpu_dec ( nr_unused ) ;
return LRU_REMOVED ;
}
2010-10-21 04:49:30 +04:00
2013-08-28 04:17:58 +04:00
/*
* Walk the superblock inode LRU for freeable inodes and attempt to free them .
* This is called from the superblock shrinker function with a number of inodes
* to trim from the LRU . Inodes to be freed are moved to a temporary list and
* then are freed outside inode_lock by dispose_list ( ) .
*/
list_lru: introduce list_lru_shrink_{count,walk}
Kmem accounting of memcg is unusable now, because it lacks slab shrinker
support. That means when we hit the limit we will get ENOMEM w/o any
chance to recover. What we should do then is to call shrink_slab, which
would reclaim old inode/dentry caches from this cgroup. This is what
this patch set is intended to do.
Basically, it does two things. First, it introduces the notion of
per-memcg slab shrinker. A shrinker that wants to reclaim objects per
cgroup should mark itself as SHRINKER_MEMCG_AWARE. Then it will be
passed the memory cgroup to scan from in shrink_control->memcg. For
such shrinkers shrink_slab iterates over the whole cgroup subtree under
the target cgroup and calls the shrinker for each kmem-active memory
cgroup.
Secondly, this patch set makes the list_lru structure per-memcg. It's
done transparently to list_lru users - everything they have to do is to
tell list_lru_init that they want memcg-aware list_lru. Then the
list_lru will automatically distribute objects among per-memcg lists
basing on which cgroup the object is accounted to. This way to make FS
shrinkers (icache, dcache) memcg-aware we only need to make them use
memcg-aware list_lru, and this is what this patch set does.
As before, this patch set only enables per-memcg kmem reclaim when the
pressure goes from memory.limit, not from memory.kmem.limit. Handling
memory.kmem.limit is going to be tricky due to GFP_NOFS allocations, and
it is still unclear whether we will have this knob in the unified
hierarchy.
This patch (of 9):
NUMA aware slab shrinkers use the list_lru structure to distribute
objects coming from different NUMA nodes to different lists. Whenever
such a shrinker needs to count or scan objects from a particular node,
it issues commands like this:
count = list_lru_count_node(lru, sc->nid);
freed = list_lru_walk_node(lru, sc->nid, isolate_func,
isolate_arg, &sc->nr_to_scan);
where sc is an instance of the shrink_control structure passed to it
from vmscan.
To simplify this, let's add special list_lru functions to be used by
shrinkers, list_lru_shrink_count() and list_lru_shrink_walk(), which
consolidate the nid and nr_to_scan arguments in the shrink_control
structure.
This will also allow us to avoid patching shrinkers that use list_lru
when we make shrink_slab() per-memcg - all we will have to do is extend
the shrink_control structure to include the target memcg and make
list_lru_shrink_{count,walk} handle this appropriately.
Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Suggested-by: Dave Chinner <david@fromorbit.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Greg Thelen <gthelen@google.com>
Cc: Glauber Costa <glommer@gmail.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-13 01:58:47 +03:00
long prune_icache_sb ( struct super_block * sb , struct shrink_control * sc )
2013-08-28 04:17:58 +04:00
{
LIST_HEAD ( freeable ) ;
long freed ;
2005-04-17 02:20:36 +04:00
list_lru: introduce list_lru_shrink_{count,walk}
Kmem accounting of memcg is unusable now, because it lacks slab shrinker
support. That means when we hit the limit we will get ENOMEM w/o any
chance to recover. What we should do then is to call shrink_slab, which
would reclaim old inode/dentry caches from this cgroup. This is what
this patch set is intended to do.
Basically, it does two things. First, it introduces the notion of
per-memcg slab shrinker. A shrinker that wants to reclaim objects per
cgroup should mark itself as SHRINKER_MEMCG_AWARE. Then it will be
passed the memory cgroup to scan from in shrink_control->memcg. For
such shrinkers shrink_slab iterates over the whole cgroup subtree under
the target cgroup and calls the shrinker for each kmem-active memory
cgroup.
Secondly, this patch set makes the list_lru structure per-memcg. It's
done transparently to list_lru users - everything they have to do is to
tell list_lru_init that they want memcg-aware list_lru. Then the
list_lru will automatically distribute objects among per-memcg lists
basing on which cgroup the object is accounted to. This way to make FS
shrinkers (icache, dcache) memcg-aware we only need to make them use
memcg-aware list_lru, and this is what this patch set does.
As before, this patch set only enables per-memcg kmem reclaim when the
pressure goes from memory.limit, not from memory.kmem.limit. Handling
memory.kmem.limit is going to be tricky due to GFP_NOFS allocations, and
it is still unclear whether we will have this knob in the unified
hierarchy.
This patch (of 9):
NUMA aware slab shrinkers use the list_lru structure to distribute
objects coming from different NUMA nodes to different lists. Whenever
such a shrinker needs to count or scan objects from a particular node,
it issues commands like this:
count = list_lru_count_node(lru, sc->nid);
freed = list_lru_walk_node(lru, sc->nid, isolate_func,
isolate_arg, &sc->nr_to_scan);
where sc is an instance of the shrink_control structure passed to it
from vmscan.
To simplify this, let's add special list_lru functions to be used by
shrinkers, list_lru_shrink_count() and list_lru_shrink_walk(), which
consolidate the nid and nr_to_scan arguments in the shrink_control
structure.
This will also allow us to avoid patching shrinkers that use list_lru
when we make shrink_slab() per-memcg - all we will have to do is extend
the shrink_control structure to include the target memcg and make
list_lru_shrink_{count,walk} handle this appropriately.
Signed-off-by: Vladimir Davydov <vdavydov@parallels.com>
Suggested-by: Dave Chinner <david@fromorbit.com>
Cc: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@suse.cz>
Cc: Greg Thelen <gthelen@google.com>
Cc: Glauber Costa <glommer@gmail.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-13 01:58:47 +03:00
freed = list_lru_shrink_walk ( & sb - > s_inode_lru , sc ,
inode_lru_isolate , & freeable ) ;
2005-04-17 02:20:36 +04:00
dispose_list ( & freeable ) ;
2013-08-28 04:17:57 +04:00
return freed ;
2005-04-17 02:20:36 +04:00
}
static void __wait_on_freeing_inode ( struct inode * inode ) ;
/*
* Called with the inode lock held .
*/
2009-03-31 18:05:54 +04:00
static struct inode * find_inode ( struct super_block * sb ,
struct hlist_head * head ,
int ( * test ) ( struct inode * , void * ) ,
void * data )
2005-04-17 02:20:36 +04:00
{
2009-03-31 18:05:54 +04:00
struct inode * inode = NULL ;
2005-04-17 02:20:36 +04:00
repeat :
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 05:06:00 +04:00
hlist_for_each_entry ( inode , head , i_hash ) {
2013-11-06 18:54:52 +04:00
if ( inode - > i_sb ! = sb )
2005-04-17 02:20:36 +04:00
continue ;
2013-11-06 18:54:52 +04:00
if ( ! test ( inode , data ) )
2005-04-17 02:20:36 +04:00
continue ;
2013-11-06 18:54:52 +04:00
spin_lock ( & inode - > i_lock ) ;
2010-06-03 01:38:30 +04:00
if ( inode - > i_state & ( I_FREEING | I_WILL_FREE ) ) {
2005-04-17 02:20:36 +04:00
__wait_on_freeing_inode ( inode ) ;
goto repeat ;
}
2018-06-28 22:53:17 +03:00
if ( unlikely ( inode - > i_state & I_CREATING ) ) {
spin_unlock ( & inode - > i_lock ) ;
return ERR_PTR ( - ESTALE ) ;
}
2010-10-23 15:09:06 +04:00
__iget ( inode ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2010-10-23 15:09:06 +04:00
return inode ;
2005-04-17 02:20:36 +04:00
}
2010-10-23 15:09:06 +04:00
return NULL ;
2005-04-17 02:20:36 +04:00
}
/*
* find_inode_fast is the fast path version of find_inode , see the comment at
* iget_locked for details .
*/
2009-03-31 18:05:54 +04:00
static struct inode * find_inode_fast ( struct super_block * sb ,
struct hlist_head * head , unsigned long ino )
2005-04-17 02:20:36 +04:00
{
2009-03-31 18:05:54 +04:00
struct inode * inode = NULL ;
2005-04-17 02:20:36 +04:00
repeat :
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 05:06:00 +04:00
hlist_for_each_entry ( inode , head , i_hash ) {
2013-11-06 18:54:52 +04:00
if ( inode - > i_ino ! = ino )
2005-04-17 02:20:36 +04:00
continue ;
2013-11-06 18:54:52 +04:00
if ( inode - > i_sb ! = sb )
2005-04-17 02:20:36 +04:00
continue ;
2013-11-06 18:54:52 +04:00
spin_lock ( & inode - > i_lock ) ;
2010-06-03 01:38:30 +04:00
if ( inode - > i_state & ( I_FREEING | I_WILL_FREE ) ) {
2005-04-17 02:20:36 +04:00
__wait_on_freeing_inode ( inode ) ;
goto repeat ;
}
2018-06-28 22:53:17 +03:00
if ( unlikely ( inode - > i_state & I_CREATING ) ) {
spin_unlock ( & inode - > i_lock ) ;
return ERR_PTR ( - ESTALE ) ;
}
2010-10-23 15:09:06 +04:00
__iget ( inode ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2010-10-23 15:09:06 +04:00
return inode ;
2005-04-17 02:20:36 +04:00
}
2010-10-23 15:09:06 +04:00
return NULL ;
2008-10-30 09:35:24 +03:00
}
2010-10-23 19:18:01 +04:00
/*
* Each cpu owns a range of LAST_INO_BATCH numbers .
* ' shared_last_ino ' is dirtied only once out of LAST_INO_BATCH allocations ,
* to renew the exhausted range .
2008-10-30 09:35:24 +03:00
*
2010-10-23 19:18:01 +04:00
* This does not significantly increase overflow rate because every CPU can
* consume at most LAST_INO_BATCH - 1 unused inode numbers . So there is
* NR_CPUS * ( LAST_INO_BATCH - 1 ) wastage . At 4096 and 1024 , this is ~ 0.1 % of the
* 2 ^ 32 range , and is a worst - case . Even a 50 % wastage would only increase
* overflow rate by 2 x , which does not seem too significant .
*
* On a 32 bit , non LFS stat ( ) call , glibc will generate an EOVERFLOW
* error if st_ino won ' t fit in target struct field . Use 32 bit counter
* here to attempt to avoid that .
2008-10-30 09:35:24 +03:00
*/
2010-10-23 19:18:01 +04:00
# define LAST_INO_BATCH 1024
static DEFINE_PER_CPU ( unsigned int , last_ino ) ;
2010-10-23 19:19:54 +04:00
unsigned int get_next_ino ( void )
2008-10-30 09:35:24 +03:00
{
2010-10-23 19:18:01 +04:00
unsigned int * p = & get_cpu_var ( last_ino ) ;
unsigned int res = * p ;
2008-10-30 09:35:24 +03:00
2010-10-23 19:18:01 +04:00
# ifdef CONFIG_SMP
if ( unlikely ( ( res & ( LAST_INO_BATCH - 1 ) ) = = 0 ) ) {
static atomic_t shared_last_ino ;
int next = atomic_add_return ( LAST_INO_BATCH , & shared_last_ino ) ;
res = next - LAST_INO_BATCH ;
}
# endif
vfs: avoid creation of inode number 0 in get_next_ino
currently, get_next_ino() is able to create inodes with inode number = 0.
This have a bad impact in the filesystems relying in this function to generate
inode numbers.
While there is no problem at all in having inodes with number 0, userspace tools
which handle file management tasks can have problems handling these files, like
for example, the impossiblity of users to delete these files, since glibc will
ignore them. So, I believe the best way is kernel to avoid creating them.
This problem has been raised previously, but the old thread didn't have any
other update for a year+, and I've seen too many users hitting the same issue
regarding the impossibility to delete files while using filesystems relying on
this function. So, I'm starting the thread again, with the same patch
that I believe is enough to address this problem.
Signed-off-by: Carlos Maiolino <cmaiolino@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2015-06-25 18:25:58 +03:00
res + + ;
/* get_next_ino should not provide a 0 inode number */
if ( unlikely ( ! res ) )
res + + ;
* p = res ;
2010-10-23 19:18:01 +04:00
put_cpu_var ( last_ino ) ;
return res ;
2008-10-30 09:35:24 +03:00
}
2010-10-23 19:19:54 +04:00
EXPORT_SYMBOL ( get_next_ino ) ;
2008-10-30 09:35:24 +03:00
2011-07-26 13:36:34 +04:00
/**
* new_inode_pseudo - obtain an inode
* @ sb : superblock
*
* Allocates a new inode for given superblock .
* Inode wont be chained in superblock s_inodes list
* This means :
* - fs can ' t be unmount
* - quotas , fsnotify , writeback can ' t work
*/
struct inode * new_inode_pseudo ( struct super_block * sb )
{
struct inode * inode = alloc_inode ( sb ) ;
if ( inode ) {
spin_lock ( & inode - > i_lock ) ;
inode - > i_state = 0 ;
spin_unlock ( & inode - > i_lock ) ;
}
return inode ;
}
2005-04-17 02:20:36 +04:00
/**
* new_inode - obtain an inode
* @ sb : superblock
*
2007-07-17 15:03:05 +04:00
* Allocates a new inode for given superblock . The default gfp_mask
2009-01-07 01:39:23 +03:00
* for allocations related to inode - > i_mapping is GFP_HIGHUSER_MOVABLE .
2007-07-17 15:03:05 +04:00
* If HIGHMEM pages are unsuitable or it is known that pages allocated
* for the page cache are not reclaimable or migratable ,
* mapping_set_gfp_mask ( ) must be called with suitable flags on the
* newly created inode ' s mapping
*
2005-04-17 02:20:36 +04:00
*/
struct inode * new_inode ( struct super_block * sb )
{
2009-03-31 18:05:54 +04:00
struct inode * inode ;
2005-04-17 02:20:36 +04:00
2011-07-26 13:36:34 +04:00
inode = new_inode_pseudo ( sb ) ;
if ( inode )
2011-03-22 14:23:40 +03:00
inode_sb_list_add ( inode ) ;
2005-04-17 02:20:36 +04:00
return inode ;
}
EXPORT_SYMBOL ( new_inode ) ;
2007-10-14 03:38:33 +04:00
# ifdef CONFIG_DEBUG_LOCK_ALLOC
lockdep: Add helper function for dir vs file i_mutex annotation
Purely in-memory filesystems do not use the inode hash as the dcache
tells us if an entry already exists. As a result, they do not call
unlock_new_inode, and thus directory inodes do not get put into a
different lockdep class for i_sem.
We need the different lockdep classes, because the locking order for
i_mutex is different for directory inodes and regular inodes. Directory
inodes can do "readdir()", which takes i_mutex *before* possibly taking
mm->mmap_sem (due to a page fault while copying the directory entry to
user space).
In contrast, regular inodes can be mmap'ed, which takes mm->mmap_sem
before accessing i_mutex.
The two cases can never happen for the same inode, so no real deadlock
can occur, but without the different lockdep classes, lockdep cannot
understand that. As a result, if CONFIG_DEBUG_LOCK_ALLOC is set, this
can lead to false positives from lockdep like below:
find/645 is trying to acquire lock:
(&mm->mmap_sem){++++++}, at: [<ffffffff81109514>] might_fault+0x5c/0xac
but task is already holding lock:
(&sb->s_type->i_mutex_key#15){+.+.+.}, at: [<ffffffff81149f34>]
vfs_readdir+0x5b/0xb4
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&sb->s_type->i_mutex_key#15){+.+.+.}:
[<ffffffff8108ac26>] lock_acquire+0xbf/0x103
[<ffffffff814db822>] __mutex_lock_common+0x4c/0x361
[<ffffffff814dbc46>] mutex_lock_nested+0x40/0x45
[<ffffffff811daa87>] hugetlbfs_file_mmap+0x82/0x110
[<ffffffff81111557>] mmap_region+0x258/0x432
[<ffffffff811119dd>] do_mmap_pgoff+0x2ac/0x306
[<ffffffff81111b4f>] sys_mmap_pgoff+0x118/0x16a
[<ffffffff8100c858>] sys_mmap+0x22/0x24
[<ffffffff814e3ec2>] system_call_fastpath+0x16/0x1b
-> #0 (&mm->mmap_sem){++++++}:
[<ffffffff8108a4bc>] __lock_acquire+0xa1a/0xcf7
[<ffffffff8108ac26>] lock_acquire+0xbf/0x103
[<ffffffff81109541>] might_fault+0x89/0xac
[<ffffffff81149cff>] filldir+0x6f/0xc7
[<ffffffff811586ea>] dcache_readdir+0x67/0x205
[<ffffffff81149f54>] vfs_readdir+0x7b/0xb4
[<ffffffff8114a073>] sys_getdents+0x7e/0xd1
[<ffffffff814e3ec2>] system_call_fastpath+0x16/0x1b
This patch moves the directory vs file lockdep annotation into a helper
function that can be called by in-memory filesystems and has hugetlbfs
call it.
Signed-off-by: Josh Boyer <jwboyer@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-25 15:48:12 +04:00
void lockdep_annotate_inode_mutex_key ( struct inode * inode )
{
2010-10-11 17:38:00 +04:00
if ( S_ISDIR ( inode - > i_mode ) ) {
2007-10-16 08:47:54 +04:00
struct file_system_type * type = inode - > i_sb - > s_type ;
2009-06-04 17:26:49 +04:00
/* Set new key only if filesystem hasn't already changed it */
2016-04-15 22:08:36 +03:00
if ( lockdep_match_class ( & inode - > i_rwsem , & type - > i_mutex_key ) ) {
2009-06-04 17:26:49 +04:00
/*
* ensure nobody is actually holding i_mutex
*/
2016-04-15 22:08:36 +03:00
// mutex_destroy(&inode->i_mutex);
init_rwsem ( & inode - > i_rwsem ) ;
lockdep_set_class ( & inode - > i_rwsem ,
2009-06-04 17:26:49 +04:00
& type - > i_mutex_dir_key ) ;
}
2007-10-16 08:47:54 +04:00
}
lockdep: Add helper function for dir vs file i_mutex annotation
Purely in-memory filesystems do not use the inode hash as the dcache
tells us if an entry already exists. As a result, they do not call
unlock_new_inode, and thus directory inodes do not get put into a
different lockdep class for i_sem.
We need the different lockdep classes, because the locking order for
i_mutex is different for directory inodes and regular inodes. Directory
inodes can do "readdir()", which takes i_mutex *before* possibly taking
mm->mmap_sem (due to a page fault while copying the directory entry to
user space).
In contrast, regular inodes can be mmap'ed, which takes mm->mmap_sem
before accessing i_mutex.
The two cases can never happen for the same inode, so no real deadlock
can occur, but without the different lockdep classes, lockdep cannot
understand that. As a result, if CONFIG_DEBUG_LOCK_ALLOC is set, this
can lead to false positives from lockdep like below:
find/645 is trying to acquire lock:
(&mm->mmap_sem){++++++}, at: [<ffffffff81109514>] might_fault+0x5c/0xac
but task is already holding lock:
(&sb->s_type->i_mutex_key#15){+.+.+.}, at: [<ffffffff81149f34>]
vfs_readdir+0x5b/0xb4
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&sb->s_type->i_mutex_key#15){+.+.+.}:
[<ffffffff8108ac26>] lock_acquire+0xbf/0x103
[<ffffffff814db822>] __mutex_lock_common+0x4c/0x361
[<ffffffff814dbc46>] mutex_lock_nested+0x40/0x45
[<ffffffff811daa87>] hugetlbfs_file_mmap+0x82/0x110
[<ffffffff81111557>] mmap_region+0x258/0x432
[<ffffffff811119dd>] do_mmap_pgoff+0x2ac/0x306
[<ffffffff81111b4f>] sys_mmap_pgoff+0x118/0x16a
[<ffffffff8100c858>] sys_mmap+0x22/0x24
[<ffffffff814e3ec2>] system_call_fastpath+0x16/0x1b
-> #0 (&mm->mmap_sem){++++++}:
[<ffffffff8108a4bc>] __lock_acquire+0xa1a/0xcf7
[<ffffffff8108ac26>] lock_acquire+0xbf/0x103
[<ffffffff81109541>] might_fault+0x89/0xac
[<ffffffff81149cff>] filldir+0x6f/0xc7
[<ffffffff811586ea>] dcache_readdir+0x67/0x205
[<ffffffff81149f54>] vfs_readdir+0x7b/0xb4
[<ffffffff8114a073>] sys_getdents+0x7e/0xd1
[<ffffffff814e3ec2>] system_call_fastpath+0x16/0x1b
This patch moves the directory vs file lockdep annotation into a helper
function that can be called by in-memory filesystems and has hugetlbfs
call it.
Signed-off-by: Josh Boyer <jwboyer@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-25 15:48:12 +04:00
}
EXPORT_SYMBOL ( lockdep_annotate_inode_mutex_key ) ;
2007-10-14 03:38:33 +04:00
# endif
lockdep: Add helper function for dir vs file i_mutex annotation
Purely in-memory filesystems do not use the inode hash as the dcache
tells us if an entry already exists. As a result, they do not call
unlock_new_inode, and thus directory inodes do not get put into a
different lockdep class for i_sem.
We need the different lockdep classes, because the locking order for
i_mutex is different for directory inodes and regular inodes. Directory
inodes can do "readdir()", which takes i_mutex *before* possibly taking
mm->mmap_sem (due to a page fault while copying the directory entry to
user space).
In contrast, regular inodes can be mmap'ed, which takes mm->mmap_sem
before accessing i_mutex.
The two cases can never happen for the same inode, so no real deadlock
can occur, but without the different lockdep classes, lockdep cannot
understand that. As a result, if CONFIG_DEBUG_LOCK_ALLOC is set, this
can lead to false positives from lockdep like below:
find/645 is trying to acquire lock:
(&mm->mmap_sem){++++++}, at: [<ffffffff81109514>] might_fault+0x5c/0xac
but task is already holding lock:
(&sb->s_type->i_mutex_key#15){+.+.+.}, at: [<ffffffff81149f34>]
vfs_readdir+0x5b/0xb4
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&sb->s_type->i_mutex_key#15){+.+.+.}:
[<ffffffff8108ac26>] lock_acquire+0xbf/0x103
[<ffffffff814db822>] __mutex_lock_common+0x4c/0x361
[<ffffffff814dbc46>] mutex_lock_nested+0x40/0x45
[<ffffffff811daa87>] hugetlbfs_file_mmap+0x82/0x110
[<ffffffff81111557>] mmap_region+0x258/0x432
[<ffffffff811119dd>] do_mmap_pgoff+0x2ac/0x306
[<ffffffff81111b4f>] sys_mmap_pgoff+0x118/0x16a
[<ffffffff8100c858>] sys_mmap+0x22/0x24
[<ffffffff814e3ec2>] system_call_fastpath+0x16/0x1b
-> #0 (&mm->mmap_sem){++++++}:
[<ffffffff8108a4bc>] __lock_acquire+0xa1a/0xcf7
[<ffffffff8108ac26>] lock_acquire+0xbf/0x103
[<ffffffff81109541>] might_fault+0x89/0xac
[<ffffffff81149cff>] filldir+0x6f/0xc7
[<ffffffff811586ea>] dcache_readdir+0x67/0x205
[<ffffffff81149f54>] vfs_readdir+0x7b/0xb4
[<ffffffff8114a073>] sys_getdents+0x7e/0xd1
[<ffffffff814e3ec2>] system_call_fastpath+0x16/0x1b
This patch moves the directory vs file lockdep annotation into a helper
function that can be called by in-memory filesystems and has hugetlbfs
call it.
Signed-off-by: Josh Boyer <jwboyer@redhat.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2011-08-25 15:48:12 +04:00
/**
* unlock_new_inode - clear the I_NEW state and wake up any waiters
* @ inode : new inode to unlock
*
* Called when the inode is fully initialised to clear the new state of the
* inode and wake up anyone waiting for the inode to finish initialisation .
*/
void unlock_new_inode ( struct inode * inode )
{
lockdep_annotate_inode_mutex_key ( inode ) ;
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
2009-12-17 16:25:01 +03:00
WARN_ON ( ! ( inode - > i_state & I_NEW ) ) ;
2018-06-28 22:53:17 +03:00
inode - > i_state & = ~ I_NEW & ~ I_CREATING ;
2012-03-11 02:07:28 +04:00
smp_mb ( ) ;
2011-03-22 14:23:36 +03:00
wake_up_bit ( & inode - > i_state , __I_NEW ) ;
spin_unlock ( & inode - > i_lock ) ;
2005-04-17 02:20:36 +04:00
}
EXPORT_SYMBOL ( unlock_new_inode ) ;
2018-06-28 22:53:17 +03:00
void discard_new_inode ( struct inode * inode )
{
lockdep_annotate_inode_mutex_key ( inode ) ;
spin_lock ( & inode - > i_lock ) ;
WARN_ON ( ! ( inode - > i_state & I_NEW ) ) ;
inode - > i_state & = ~ I_NEW ;
smp_mb ( ) ;
wake_up_bit ( & inode - > i_state , __I_NEW ) ;
spin_unlock ( & inode - > i_lock ) ;
iput ( inode ) ;
}
EXPORT_SYMBOL ( discard_new_inode ) ;
2023-06-01 13:58:24 +03:00
/**
* lock_two_inodes - lock two inodes ( may be regular files but also dirs )
*
* Lock any non - NULL argument . The caller must make sure that if he is passing
* in two directories , one is not ancestor of the other . Zero , one or two
* objects may be locked by this function .
*
* @ inode1 : first inode to lock
* @ inode2 : second inode to lock
* @ subclass1 : inode lock subclass for the first lock obtained
* @ subclass2 : inode lock subclass for the second lock obtained
*/
void lock_two_inodes ( struct inode * inode1 , struct inode * inode2 ,
unsigned subclass1 , unsigned subclass2 )
{
if ( ! inode1 | | ! inode2 ) {
/*
* Make sure @ subclass1 will be used for the acquired lock .
* This is not strictly necessary ( no current caller cares ) but
* let ' s keep things consistent .
*/
if ( ! inode1 )
swap ( inode1 , inode2 ) ;
goto lock ;
}
/*
* If one object is directory and the other is not , we must make sure
* to lock directory first as the other object may be its child .
*/
if ( S_ISDIR ( inode2 - > i_mode ) = = S_ISDIR ( inode1 - > i_mode ) ) {
if ( inode1 > inode2 )
swap ( inode1 , inode2 ) ;
} else if ( ! S_ISDIR ( inode1 - > i_mode ) )
swap ( inode1 , inode2 ) ;
lock :
if ( inode1 )
inode_lock_nested ( inode1 , subclass1 ) ;
if ( inode2 & & inode2 ! = inode1 )
inode_lock_nested ( inode2 , subclass2 ) ;
}
2012-04-18 23:16:33 +04:00
/**
* lock_two_nondirectories - take two i_mutexes on non - directory objects
2014-04-01 19:08:43 +04:00
*
2023-06-01 13:58:26 +03:00
* Lock any non - NULL argument . Passed objects must not be directories .
2014-04-01 19:08:43 +04:00
* Zero , one or two objects may be locked by this function .
*
2012-04-18 23:16:33 +04:00
* @ inode1 : first inode to lock
* @ inode2 : second inode to lock
*/
void lock_two_nondirectories ( struct inode * inode1 , struct inode * inode2 )
{
2023-07-03 17:49:12 +03:00
if ( inode1 )
WARN_ON_ONCE ( S_ISDIR ( inode1 - > i_mode ) ) ;
if ( inode2 )
WARN_ON_ONCE ( S_ISDIR ( inode2 - > i_mode ) ) ;
2023-06-01 13:58:26 +03:00
lock_two_inodes ( inode1 , inode2 , I_MUTEX_NORMAL , I_MUTEX_NONDIR2 ) ;
2012-04-18 23:16:33 +04:00
}
EXPORT_SYMBOL ( lock_two_nondirectories ) ;
/**
* unlock_two_nondirectories - release locks from lock_two_nondirectories ( )
* @ inode1 : first inode to unlock
* @ inode2 : second inode to unlock
*/
void unlock_two_nondirectories ( struct inode * inode1 , struct inode * inode2 )
{
2023-06-01 13:58:26 +03:00
if ( inode1 ) {
WARN_ON_ONCE ( S_ISDIR ( inode1 - > i_mode ) ) ;
2016-01-22 23:40:57 +03:00
inode_unlock ( inode1 ) ;
2023-06-01 13:58:26 +03:00
}
if ( inode2 & & inode2 ! = inode1 ) {
WARN_ON_ONCE ( S_ISDIR ( inode2 - > i_mode ) ) ;
2016-01-22 23:40:57 +03:00
inode_unlock ( inode2 ) ;
2023-06-01 13:58:26 +03:00
}
2012-04-18 23:16:33 +04:00
}
EXPORT_SYMBOL ( unlock_two_nondirectories ) ;
2018-05-17 11:53:05 +03:00
/**
* inode_insert5 - obtain an inode from a mounted file system
* @ inode : pre - allocated inode to use for insert to cache
* @ hashval : hash value ( usually inode number ) to get
* @ test : callback used for comparisons between inodes
* @ set : callback used to initialize a new struct inode
* @ data : opaque data pointer to pass to @ test and @ set
*
* Search for the inode specified by @ hashval and @ data in the inode cache ,
* and if present it is return it with an increased reference count . This is
* a variant of iget5_locked ( ) for callers that don ' t want to fail on memory
* allocation of inode .
*
* If the inode is not in cache , insert the pre - allocated inode to cache and
* return it locked , hashed , and with the I_NEW flag set . The file system gets
* to fill it in before unlocking it via unlock_new_inode ( ) .
*
* Note both @ test and @ set are called with the inode_hash_lock held , so can ' t
* sleep .
*/
struct inode * inode_insert5 ( struct inode * inode , unsigned long hashval ,
int ( * test ) ( struct inode * , void * ) ,
int ( * set ) ( struct inode * , void * ) , void * data )
{
struct hlist_head * head = inode_hashtable + hash ( inode - > i_sb , hashval ) ;
struct inode * old ;
again :
spin_lock ( & inode_hash_lock ) ;
old = find_inode ( inode - > i_sb , head , test , data ) ;
if ( unlikely ( old ) ) {
/*
* Uhhuh , somebody else created the same inode under us .
* Use the old inode instead of the preallocated one .
*/
spin_unlock ( & inode_hash_lock ) ;
2018-06-28 22:53:17 +03:00
if ( IS_ERR ( old ) )
return NULL ;
2018-05-17 11:53:05 +03:00
wait_on_inode ( old ) ;
if ( unlikely ( inode_unhashed ( old ) ) ) {
iput ( old ) ;
goto again ;
}
return old ;
}
if ( set & & unlikely ( set ( inode , data ) ) ) {
inode = NULL ;
goto unlock ;
}
/*
* Return the locked inode with I_NEW set , the
* caller is responsible for filling in the contents
*/
spin_lock ( & inode - > i_lock ) ;
inode - > i_state | = I_NEW ;
2017-12-01 14:40:16 +03:00
hlist_add_head_rcu ( & inode - > i_hash , head ) ;
2018-05-17 11:53:05 +03:00
spin_unlock ( & inode - > i_lock ) ;
2022-03-31 23:29:00 +03:00
/*
* Add inode to the sb list if it ' s not already . It has I_NEW at this
* point , so it should be safe to test i_sb_list locklessly .
*/
if ( list_empty ( & inode - > i_sb_list ) )
2018-07-24 16:01:55 +03:00
inode_sb_list_add ( inode ) ;
2018-05-17 11:53:05 +03:00
unlock :
spin_unlock ( & inode_hash_lock ) ;
return inode ;
}
EXPORT_SYMBOL ( inode_insert5 ) ;
2011-03-23 22:03:28 +03:00
/**
* iget5_locked - obtain an inode from a mounted file system
* @ sb : super block of file system
* @ hashval : hash value ( usually inode number ) to get
* @ test : callback used for comparisons between inodes
* @ set : callback used to initialize a new struct inode
* @ data : opaque data pointer to pass to @ test and @ set
*
* Search for the inode specified by @ hashval and @ data in the inode cache ,
* and if present it is return it with an increased reference count . This is
* a generalized version of iget_locked ( ) for file systems where the inode
* number is not sufficient for unique identification of an inode .
*
* If the inode is not in cache , allocate a new inode and return it locked ,
* hashed , and with the I_NEW flag set . The file system gets to fill it in
* before unlocking it via unlock_new_inode ( ) .
2005-04-17 02:20:36 +04:00
*
2011-03-23 22:03:28 +03:00
* Note both @ test and @ set are called with the inode_hash_lock held , so can ' t
* sleep .
2005-04-17 02:20:36 +04:00
*/
2011-03-23 22:03:28 +03:00
struct inode * iget5_locked ( struct super_block * sb , unsigned long hashval ,
int ( * test ) ( struct inode * , void * ) ,
int ( * set ) ( struct inode * , void * ) , void * data )
2005-04-17 02:20:36 +04:00
{
2018-05-17 11:53:05 +03:00
struct inode * inode = ilookup5 ( sb , hashval , test , data ) ;
2011-03-23 22:03:28 +03:00
2018-05-17 11:53:05 +03:00
if ( ! inode ) {
2018-07-24 16:01:55 +03:00
struct inode * new = alloc_inode ( sb ) ;
2011-03-23 22:03:28 +03:00
2018-05-17 11:53:05 +03:00
if ( new ) {
2018-07-24 16:01:55 +03:00
new - > i_state = 0 ;
2018-05-17 11:53:05 +03:00
inode = inode_insert5 ( new , hashval , test , set , data ) ;
if ( unlikely ( inode ! = new ) )
2018-07-24 16:01:55 +03:00
destroy_inode ( new ) ;
2016-07-04 06:15:21 +03:00
}
2005-04-17 02:20:36 +04:00
}
return inode ;
}
2011-03-23 22:03:28 +03:00
EXPORT_SYMBOL ( iget5_locked ) ;
2005-04-17 02:20:36 +04:00
2011-03-23 22:03:28 +03:00
/**
* iget_locked - obtain an inode from a mounted file system
* @ sb : super block of file system
* @ ino : inode number to get
*
* Search for the inode specified by @ ino in the inode cache and if present
* return it with an increased reference count . This is for file systems
* where the inode number is sufficient for unique identification of an inode .
*
* If the inode is not in cache , allocate a new inode and return it locked ,
* hashed , and with the I_NEW flag set . The file system gets to fill it in
* before unlocking it via unlock_new_inode ( ) .
2005-04-17 02:20:36 +04:00
*/
2011-03-23 22:03:28 +03:00
struct inode * iget_locked ( struct super_block * sb , unsigned long ino )
2005-04-17 02:20:36 +04:00
{
2011-03-23 22:03:28 +03:00
struct hlist_head * head = inode_hashtable + hash ( sb , ino ) ;
2009-03-31 18:05:54 +04:00
struct inode * inode ;
2016-07-04 06:15:21 +03:00
again :
2011-03-23 22:03:28 +03:00
spin_lock ( & inode_hash_lock ) ;
inode = find_inode_fast ( sb , head , ino ) ;
spin_unlock ( & inode_hash_lock ) ;
if ( inode ) {
2018-06-28 22:53:17 +03:00
if ( IS_ERR ( inode ) )
return NULL ;
2011-03-23 22:03:28 +03:00
wait_on_inode ( inode ) ;
2016-07-04 06:15:21 +03:00
if ( unlikely ( inode_unhashed ( inode ) ) ) {
iput ( inode ) ;
goto again ;
}
2011-03-23 22:03:28 +03:00
return inode ;
}
2005-04-17 02:20:36 +04:00
inode = alloc_inode ( sb ) ;
if ( inode ) {
2009-03-31 18:05:54 +04:00
struct inode * old ;
2005-04-17 02:20:36 +04:00
2011-03-22 14:23:42 +03:00
spin_lock ( & inode_hash_lock ) ;
2005-04-17 02:20:36 +04:00
/* We released the lock, so.. */
old = find_inode_fast ( sb , head , ino ) ;
if ( ! old ) {
inode - > i_ino = ino ;
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
inode - > i_state = I_NEW ;
2017-12-01 14:40:16 +03:00
hlist_add_head_rcu ( & inode - > i_hash , head ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2011-03-22 14:23:40 +03:00
inode_sb_list_add ( inode ) ;
2011-03-22 14:23:42 +03:00
spin_unlock ( & inode_hash_lock ) ;
2005-04-17 02:20:36 +04:00
/* Return the locked inode with I_NEW set, the
* caller is responsible for filling in the contents
*/
return inode ;
}
/*
* Uhhuh , somebody else created the same inode under
* us . Use the old inode instead of the one we just
* allocated .
*/
2011-03-22 14:23:42 +03:00
spin_unlock ( & inode_hash_lock ) ;
2005-04-17 02:20:36 +04:00
destroy_inode ( inode ) ;
2018-06-28 22:53:17 +03:00
if ( IS_ERR ( old ) )
return NULL ;
2005-04-17 02:20:36 +04:00
inode = old ;
wait_on_inode ( inode ) ;
2016-07-04 06:15:21 +03:00
if ( unlikely ( inode_unhashed ( inode ) ) ) {
iput ( inode ) ;
goto again ;
}
2005-04-17 02:20:36 +04:00
}
return inode ;
}
2011-03-23 22:03:28 +03:00
EXPORT_SYMBOL ( iget_locked ) ;
2005-04-17 02:20:36 +04:00
2010-10-23 15:00:16 +04:00
/*
* search the inode cache for a matching inode number .
* If we find one , then the inode number we are trying to
* allocate is not unique and so we should not use it .
*
* Returns 1 if the inode number is unique , 0 if it is not .
*/
static int test_inode_iunique ( struct super_block * sb , unsigned long ino )
{
struct hlist_head * b = inode_hashtable + hash ( sb , ino ) ;
struct inode * inode ;
2017-12-01 14:40:16 +03:00
hlist_for_each_entry_rcu ( inode , b , i_hash ) {
if ( inode - > i_ino = = ino & & inode - > i_sb = = sb )
2010-10-23 15:00:16 +04:00
return 0 ;
}
return 1 ;
}
2005-04-17 02:20:36 +04:00
/**
* iunique - get a unique inode number
* @ sb : superblock
* @ max_reserved : highest reserved inode number
*
* Obtain an inode number that is unique on the system for a given
* superblock . This is used by file systems that have no natural
* permanent inode numbering system . An inode number is returned that
* is higher than the reserved limit but unique .
*
* BUGS :
* With a large number of inodes live on the file system this function
* currently becomes quite slow .
*/
ino_t iunique ( struct super_block * sb , ino_t max_reserved )
{
inode numbering: make static counters in new_inode and iunique be 32 bits
The problems are:
- on filesystems w/o permanent inode numbers, i_ino values can be larger
than 32 bits, which can cause problems for some 32 bit userspace programs on
a 64 bit kernel. We can't do anything for filesystems that have actual
>32-bit inode numbers, but on filesystems that generate i_ino values on the
fly, we should try to have them fit in 32 bits. We could trivially fix this
by making the static counters in new_inode and iunique 32 bits, but...
- many filesystems call new_inode and assume that the i_ino values they are
given are unique. They are not guaranteed to be so, since the static
counter can wrap. This problem is exacerbated by the fix for #1.
- after allocating a new inode, some filesystems call iunique to try to get
a unique i_ino value, but they don't actually add their inodes to the
hashtable, and so they're still not guaranteed to be unique if that counter
wraps.
This patch set takes the simpler approach of simply using iunique and hashing
the inodes afterward. Christoph H. previously mentioned that he thought that
this approach may slow down lookups for filesystems that currently hash their
inodes.
The questions are:
1) how much would this slow down lookups for these filesystems?
2) is it enough to justify adding more infrastructure to avoid it?
What might be best is to start with this approach and then only move to using
IDR or some other scheme if these extra inodes in the hashtable prove to be
problematic.
I've done some cursory testing with this patch and the overhead of hashing and
unhashing the inodes with pipefs is pretty low -- just a few seconds of system
time added on to the creation and destruction of 10 million pipes (very
similar to the overhead that the IDR approach would add).
The hard thing to measure is what effect this has on other filesystems. I'm
open to ways to try and gauge this.
Again, I've only converted pipefs as an example. If this approach is
acceptable then I'll start work on patches to convert other filesystems.
With a pretty-much-worst-case microbenchmark provided by Eric Dumazet
<dada1@cosmosbay.com>:
hashing patch (pipebench):
sys 1m15.329s
sys 1m16.249s
sys 1m17.169s
unpatched (pipebench):
sys 1m9.836s
sys 1m12.541s
sys 1m14.153s
Which works out to 1.05642174294555027017. So ~5-6% slowdown.
This patch:
When a 32-bit program that was not compiled with large file offsets does a
stat and gets a st_ino value back that won't fit in the 32 bit field, glibc
(correctly) generates an EOVERFLOW error. We can't do anything about fs's
with larger permanent inode numbers, but when we generate them on the fly, we
ought to try and have them fit within a 32 bit field.
This patch takes the first step toward this by making the static counters in
these two functions be 32 bits.
[jlayton@redhat.com: mention that it's only the case for 32bit, non-LFS stat]
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-05-08 11:32:29 +04:00
/*
* On a 32 bit , non LFS stat ( ) call , glibc will generate an EOVERFLOW
* error if st_ino won ' t fit in target struct field . Use 32 bit counter
* here to attempt to avoid that .
*/
2010-10-23 15:00:16 +04:00
static DEFINE_SPINLOCK ( iunique_lock ) ;
inode numbering: make static counters in new_inode and iunique be 32 bits
The problems are:
- on filesystems w/o permanent inode numbers, i_ino values can be larger
than 32 bits, which can cause problems for some 32 bit userspace programs on
a 64 bit kernel. We can't do anything for filesystems that have actual
>32-bit inode numbers, but on filesystems that generate i_ino values on the
fly, we should try to have them fit in 32 bits. We could trivially fix this
by making the static counters in new_inode and iunique 32 bits, but...
- many filesystems call new_inode and assume that the i_ino values they are
given are unique. They are not guaranteed to be so, since the static
counter can wrap. This problem is exacerbated by the fix for #1.
- after allocating a new inode, some filesystems call iunique to try to get
a unique i_ino value, but they don't actually add their inodes to the
hashtable, and so they're still not guaranteed to be unique if that counter
wraps.
This patch set takes the simpler approach of simply using iunique and hashing
the inodes afterward. Christoph H. previously mentioned that he thought that
this approach may slow down lookups for filesystems that currently hash their
inodes.
The questions are:
1) how much would this slow down lookups for these filesystems?
2) is it enough to justify adding more infrastructure to avoid it?
What might be best is to start with this approach and then only move to using
IDR or some other scheme if these extra inodes in the hashtable prove to be
problematic.
I've done some cursory testing with this patch and the overhead of hashing and
unhashing the inodes with pipefs is pretty low -- just a few seconds of system
time added on to the creation and destruction of 10 million pipes (very
similar to the overhead that the IDR approach would add).
The hard thing to measure is what effect this has on other filesystems. I'm
open to ways to try and gauge this.
Again, I've only converted pipefs as an example. If this approach is
acceptable then I'll start work on patches to convert other filesystems.
With a pretty-much-worst-case microbenchmark provided by Eric Dumazet
<dada1@cosmosbay.com>:
hashing patch (pipebench):
sys 1m15.329s
sys 1m16.249s
sys 1m17.169s
unpatched (pipebench):
sys 1m9.836s
sys 1m12.541s
sys 1m14.153s
Which works out to 1.05642174294555027017. So ~5-6% slowdown.
This patch:
When a 32-bit program that was not compiled with large file offsets does a
stat and gets a st_ino value back that won't fit in the 32 bit field, glibc
(correctly) generates an EOVERFLOW error. We can't do anything about fs's
with larger permanent inode numbers, but when we generate them on the fly, we
ought to try and have them fit within a 32 bit field.
This patch takes the first step toward this by making the static counters in
these two functions be 32 bits.
[jlayton@redhat.com: mention that it's only the case for 32bit, non-LFS stat]
Signed-off-by: Jeff Layton <jlayton@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-05-08 11:32:29 +04:00
static unsigned int counter ;
2005-04-17 02:20:36 +04:00
ino_t res ;
2007-05-08 11:29:48 +04:00
2017-12-01 14:40:16 +03:00
rcu_read_lock ( ) ;
2010-10-23 15:00:16 +04:00
spin_lock ( & iunique_lock ) ;
2007-05-08 11:29:48 +04:00
do {
if ( counter < = max_reserved )
counter = max_reserved + 1 ;
2005-04-17 02:20:36 +04:00
res = counter + + ;
2010-10-23 15:00:16 +04:00
} while ( ! test_inode_iunique ( sb , res ) ) ;
spin_unlock ( & iunique_lock ) ;
2017-12-01 14:40:16 +03:00
rcu_read_unlock ( ) ;
2005-04-17 02:20:36 +04:00
2007-05-08 11:29:48 +04:00
return res ;
}
2005-04-17 02:20:36 +04:00
EXPORT_SYMBOL ( iunique ) ;
struct inode * igrab ( struct inode * inode )
{
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
if ( ! ( inode - > i_state & ( I_FREEING | I_WILL_FREE ) ) ) {
2005-04-17 02:20:36 +04:00
__iget ( inode ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
} else {
spin_unlock ( & inode - > i_lock ) ;
2005-04-17 02:20:36 +04:00
/*
* Handle the case where s_op - > clear_inode is not been
* called yet , and somebody is calling igrab
* while the inode is getting freed .
*/
inode = NULL ;
2011-03-22 14:23:36 +03:00
}
2005-04-17 02:20:36 +04:00
return inode ;
}
EXPORT_SYMBOL ( igrab ) ;
/**
2011-03-23 22:03:28 +03:00
* ilookup5_nowait - search for an inode in the inode cache
2005-04-17 02:20:36 +04:00
* @ sb : super block of file system to search
2011-03-23 22:03:28 +03:00
* @ hashval : hash value ( usually inode number ) to search for
2005-04-17 02:20:36 +04:00
* @ test : callback used for comparisons between inodes
* @ data : opaque data pointer to pass to @ test
*
2011-03-23 22:03:28 +03:00
* Search for the inode specified by @ hashval and @ data in the inode cache .
2005-04-17 02:20:36 +04:00
* If the inode is in the cache , the inode is returned with an incremented
* reference count .
*
2011-03-23 22:03:28 +03:00
* Note : I_NEW is not waited upon so you have to be very careful what you do
* with the returned inode . You probably should be using ilookup5 ( ) instead .
2005-04-17 02:20:36 +04:00
*
2011-03-26 23:27:47 +03:00
* Note2 : @ test is called with the inode_hash_lock held , so can ' t sleep .
2005-04-17 02:20:36 +04:00
*/
2011-03-23 22:03:28 +03:00
struct inode * ilookup5_nowait ( struct super_block * sb , unsigned long hashval ,
int ( * test ) ( struct inode * , void * ) , void * data )
2005-04-17 02:20:36 +04:00
{
2011-03-23 22:03:28 +03:00
struct hlist_head * head = inode_hashtable + hash ( sb , hashval ) ;
2005-04-17 02:20:36 +04:00
struct inode * inode ;
2011-03-22 14:23:42 +03:00
spin_lock ( & inode_hash_lock ) ;
2005-04-17 02:20:36 +04:00
inode = find_inode ( sb , head , test , data ) ;
2011-03-22 14:23:42 +03:00
spin_unlock ( & inode_hash_lock ) ;
2005-07-13 12:10:44 +04:00
2018-06-28 22:53:17 +03:00
return IS_ERR ( inode ) ? NULL : inode ;
2005-07-13 12:10:44 +04:00
}
EXPORT_SYMBOL ( ilookup5_nowait ) ;
/**
* ilookup5 - search for an inode in the inode cache
* @ sb : super block of file system to search
* @ hashval : hash value ( usually inode number ) to search for
* @ test : callback used for comparisons between inodes
* @ data : opaque data pointer to pass to @ test
*
2011-03-23 22:03:28 +03:00
* Search for the inode specified by @ hashval and @ data in the inode cache ,
* and if the inode is in the cache , return the inode with an incremented
* reference count . Waits on I_NEW before returning the inode .
2005-07-13 12:10:44 +04:00
* returned with an incremented reference count .
2005-04-17 02:20:36 +04:00
*
2011-03-23 22:03:28 +03:00
* This is a generalized version of ilookup ( ) for file systems where the
* inode number is not sufficient for unique identification of an inode .
2005-04-17 02:20:36 +04:00
*
2011-03-23 22:03:28 +03:00
* Note : @ test is called with the inode_hash_lock held , so can ' t sleep .
2005-04-17 02:20:36 +04:00
*/
struct inode * ilookup5 ( struct super_block * sb , unsigned long hashval ,
int ( * test ) ( struct inode * , void * ) , void * data )
{
2016-07-04 06:15:21 +03:00
struct inode * inode ;
again :
inode = ilookup5_nowait ( sb , hashval , test , data ) ;
if ( inode ) {
2011-03-23 22:03:28 +03:00
wait_on_inode ( inode ) ;
2016-07-04 06:15:21 +03:00
if ( unlikely ( inode_unhashed ( inode ) ) ) {
iput ( inode ) ;
goto again ;
}
}
2011-03-23 22:03:28 +03:00
return inode ;
2005-04-17 02:20:36 +04:00
}
EXPORT_SYMBOL ( ilookup5 ) ;
/**
* ilookup - search for an inode in the inode cache
* @ sb : super block of file system to search
* @ ino : inode number to search for
*
2011-03-23 22:03:28 +03:00
* Search for the inode @ ino in the inode cache , and if the inode is in the
* cache , the inode is returned with an incremented reference count .
2005-04-17 02:20:36 +04:00
*/
struct inode * ilookup ( struct super_block * sb , unsigned long ino )
{
struct hlist_head * head = inode_hashtable + hash ( sb , ino ) ;
struct inode * inode ;
2016-07-04 06:15:21 +03:00
again :
2011-03-23 22:03:28 +03:00
spin_lock ( & inode_hash_lock ) ;
inode = find_inode_fast ( sb , head , ino ) ;
spin_unlock ( & inode_hash_lock ) ;
2005-04-17 02:20:36 +04:00
2016-07-04 06:15:21 +03:00
if ( inode ) {
2018-06-28 22:53:17 +03:00
if ( IS_ERR ( inode ) )
return NULL ;
2011-03-23 22:03:28 +03:00
wait_on_inode ( inode ) ;
2016-07-04 06:15:21 +03:00
if ( unlikely ( inode_unhashed ( inode ) ) ) {
iput ( inode ) ;
goto again ;
}
}
2011-03-23 22:03:28 +03:00
return inode ;
2005-04-17 02:20:36 +04:00
}
2011-03-23 22:03:28 +03:00
EXPORT_SYMBOL ( ilookup ) ;
2005-04-17 02:20:36 +04:00
2015-02-02 08:37:01 +03:00
/**
* find_inode_nowait - find an inode in the inode cache
* @ sb : super block of file system to search
* @ hashval : hash value ( usually inode number ) to search for
* @ match : callback used for comparisons between inodes
* @ data : opaque data pointer to pass to @ match
*
* Search for the inode specified by @ hashval and @ data in the inode
* cache , where the helper function @ match will return 0 if the inode
* does not match , 1 if the inode does match , and - 1 if the search
* should be stopped . The @ match function must be responsible for
* taking the i_lock spin_lock and checking i_state for an inode being
* freed or being initialized , and incrementing the reference count
* before returning 1. It also must not sleep , since it is called with
* the inode_hash_lock spinlock held .
*
* This is a even more generalized version of ilookup5 ( ) when the
* function must never block - - - find_inode ( ) can block in
* __wait_on_freeing_inode ( ) - - - or when the caller can not increment
* the reference count because the resulting iput ( ) might cause an
* inode eviction . The tradeoff is that the @ match funtion must be
* very carefully implemented .
*/
struct inode * find_inode_nowait ( struct super_block * sb ,
unsigned long hashval ,
int ( * match ) ( struct inode * , unsigned long ,
void * ) ,
void * data )
{
struct hlist_head * head = inode_hashtable + hash ( sb , hashval ) ;
struct inode * inode , * ret_inode = NULL ;
int mval ;
spin_lock ( & inode_hash_lock ) ;
hlist_for_each_entry ( inode , head , i_hash ) {
if ( inode - > i_sb ! = sb )
continue ;
mval = match ( inode , hashval , data ) ;
if ( mval = = 0 )
continue ;
if ( mval = = 1 )
ret_inode = inode ;
goto out ;
}
out :
spin_unlock ( & inode_hash_lock ) ;
return ret_inode ;
}
EXPORT_SYMBOL ( find_inode_nowait ) ;
2017-12-01 14:40:16 +03:00
/**
* find_inode_rcu - find an inode in the inode cache
* @ sb : Super block of file system to search
* @ hashval : Key to hash
* @ test : Function to test match on an inode
* @ data : Data for test function
*
* Search for the inode specified by @ hashval and @ data in the inode cache ,
* where the helper function @ test will return 0 if the inode does not match
* and 1 if it does . The @ test function must be responsible for taking the
* i_lock spin_lock and checking i_state for an inode being freed or being
* initialized .
*
* If successful , this will return the inode for which the @ test function
* returned 1 and NULL otherwise .
*
* The @ test function is not permitted to take a ref on any inode presented .
* It is also not permitted to sleep .
*
* The caller must hold the RCU read lock .
*/
struct inode * find_inode_rcu ( struct super_block * sb , unsigned long hashval ,
int ( * test ) ( struct inode * , void * ) , void * data )
{
struct hlist_head * head = inode_hashtable + hash ( sb , hashval ) ;
struct inode * inode ;
RCU_LOCKDEP_WARN ( ! rcu_read_lock_held ( ) ,
" suspicious find_inode_rcu() usage " ) ;
hlist_for_each_entry_rcu ( inode , head , i_hash ) {
if ( inode - > i_sb = = sb & &
! ( READ_ONCE ( inode - > i_state ) & ( I_FREEING | I_WILL_FREE ) ) & &
test ( inode , data ) )
return inode ;
}
return NULL ;
}
EXPORT_SYMBOL ( find_inode_rcu ) ;
/**
2021-01-14 11:04:39 +03:00
* find_inode_by_ino_rcu - Find an inode in the inode cache
2017-12-01 14:40:16 +03:00
* @ sb : Super block of file system to search
* @ ino : The inode number to match
*
* Search for the inode specified by @ hashval and @ data in the inode cache ,
* where the helper function @ test will return 0 if the inode does not match
* and 1 if it does . The @ test function must be responsible for taking the
* i_lock spin_lock and checking i_state for an inode being freed or being
* initialized .
*
* If successful , this will return the inode for which the @ test function
* returned 1 and NULL otherwise .
*
* The @ test function is not permitted to take a ref on any inode presented .
* It is also not permitted to sleep .
*
* The caller must hold the RCU read lock .
*/
struct inode * find_inode_by_ino_rcu ( struct super_block * sb ,
unsigned long ino )
{
struct hlist_head * head = inode_hashtable + hash ( sb , ino ) ;
struct inode * inode ;
RCU_LOCKDEP_WARN ( ! rcu_read_lock_held ( ) ,
" suspicious find_inode_by_ino_rcu() usage " ) ;
hlist_for_each_entry_rcu ( inode , head , i_hash ) {
if ( inode - > i_ino = = ino & &
inode - > i_sb = = sb & &
! ( READ_ONCE ( inode - > i_state ) & ( I_FREEING | I_WILL_FREE ) ) )
return inode ;
}
return NULL ;
}
EXPORT_SYMBOL ( find_inode_by_ino_rcu ) ;
2008-12-30 09:48:21 +03:00
int insert_inode_locked ( struct inode * inode )
{
struct super_block * sb = inode - > i_sb ;
ino_t ino = inode - > i_ino ;
struct hlist_head * head = inode_hashtable + hash ( sb , ino ) ;
while ( 1 ) {
2009-05-13 22:13:40 +04:00
struct inode * old = NULL ;
2011-03-22 14:23:42 +03:00
spin_lock ( & inode_hash_lock ) ;
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 05:06:00 +04:00
hlist_for_each_entry ( old , head , i_hash ) {
2009-05-13 22:13:40 +04:00
if ( old - > i_ino ! = ino )
continue ;
if ( old - > i_sb ! = sb )
continue ;
2011-03-22 14:23:36 +03:00
spin_lock ( & old - > i_lock ) ;
if ( old - > i_state & ( I_FREEING | I_WILL_FREE ) ) {
spin_unlock ( & old - > i_lock ) ;
2009-05-13 22:13:40 +04:00
continue ;
2011-03-22 14:23:36 +03:00
}
2009-05-13 22:13:40 +04:00
break ;
}
hlist: drop the node parameter from iterators
I'm not sure why, but the hlist for each entry iterators were conceived
list_for_each_entry(pos, head, member)
The hlist ones were greedy and wanted an extra parameter:
hlist_for_each_entry(tpos, pos, head, member)
Why did they need an extra pos parameter? I'm not quite sure. Not only
they don't really need it, it also prevents the iterator from looking
exactly like the list iterator, which is unfortunate.
Besides the semantic patch, there was some manual work required:
- Fix up the actual hlist iterators in linux/list.h
- Fix up the declaration of other iterators based on the hlist ones.
- A very small amount of places were using the 'node' parameter, this
was modified to use 'obj->member' instead.
- Coccinelle didn't handle the hlist_for_each_entry_safe iterator
properly, so those had to be fixed up manually.
The semantic patch which is mostly the work of Peter Senna Tschudin is here:
@@
iterator name hlist_for_each_entry, hlist_for_each_entry_continue, hlist_for_each_entry_from, hlist_for_each_entry_rcu, hlist_for_each_entry_rcu_bh, hlist_for_each_entry_continue_rcu_bh, for_each_busy_worker, ax25_uid_for_each, ax25_for_each, inet_bind_bucket_for_each, sctp_for_each_hentry, sk_for_each, sk_for_each_rcu, sk_for_each_from, sk_for_each_safe, sk_for_each_bound, hlist_for_each_entry_safe, hlist_for_each_entry_continue_rcu, nr_neigh_for_each, nr_neigh_for_each_safe, nr_node_for_each, nr_node_for_each_safe, for_each_gfn_indirect_valid_sp, for_each_gfn_sp, for_each_host;
type T;
expression a,c,d,e;
identifier b;
statement S;
@@
-T b;
<+... when != b
(
hlist_for_each_entry(a,
- b,
c, d) S
|
hlist_for_each_entry_continue(a,
- b,
c) S
|
hlist_for_each_entry_from(a,
- b,
c) S
|
hlist_for_each_entry_rcu(a,
- b,
c, d) S
|
hlist_for_each_entry_rcu_bh(a,
- b,
c, d) S
|
hlist_for_each_entry_continue_rcu_bh(a,
- b,
c) S
|
for_each_busy_worker(a, c,
- b,
d) S
|
ax25_uid_for_each(a,
- b,
c) S
|
ax25_for_each(a,
- b,
c) S
|
inet_bind_bucket_for_each(a,
- b,
c) S
|
sctp_for_each_hentry(a,
- b,
c) S
|
sk_for_each(a,
- b,
c) S
|
sk_for_each_rcu(a,
- b,
c) S
|
sk_for_each_from
-(a, b)
+(a)
S
+ sk_for_each_from(a) S
|
sk_for_each_safe(a,
- b,
c, d) S
|
sk_for_each_bound(a,
- b,
c) S
|
hlist_for_each_entry_safe(a,
- b,
c, d, e) S
|
hlist_for_each_entry_continue_rcu(a,
- b,
c) S
|
nr_neigh_for_each(a,
- b,
c) S
|
nr_neigh_for_each_safe(a,
- b,
c, d) S
|
nr_node_for_each(a,
- b,
c) S
|
nr_node_for_each_safe(a,
- b,
c, d) S
|
- for_each_gfn_sp(a, c, d, b) S
+ for_each_gfn_sp(a, c, d) S
|
- for_each_gfn_indirect_valid_sp(a, c, d, b) S
+ for_each_gfn_indirect_valid_sp(a, c, d) S
|
for_each_host(a,
- b,
c) S
|
for_each_host_safe(a,
- b,
c, d) S
|
for_each_mesh_entry(a,
- b,
c, d) S
)
...+>
[akpm@linux-foundation.org: drop bogus change from net/ipv4/raw.c]
[akpm@linux-foundation.org: drop bogus hunk from net/ipv6/raw.c]
[akpm@linux-foundation.org: checkpatch fixes]
[akpm@linux-foundation.org: fix warnings]
[akpm@linux-foudnation.org: redo intrusive kvm changes]
Tested-by: Peter Senna Tschudin <peter.senna@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Sasha Levin <sasha.levin@oracle.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2013-02-28 05:06:00 +04:00
if ( likely ( ! old ) ) {
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
2018-06-28 22:53:17 +03:00
inode - > i_state | = I_NEW | I_CREATING ;
2017-12-01 14:40:16 +03:00
hlist_add_head_rcu ( & inode - > i_hash , head ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2011-03-22 14:23:42 +03:00
spin_unlock ( & inode_hash_lock ) ;
2008-12-30 09:48:21 +03:00
return 0 ;
}
2018-06-28 22:53:17 +03:00
if ( unlikely ( old - > i_state & I_CREATING ) ) {
spin_unlock ( & old - > i_lock ) ;
spin_unlock ( & inode_hash_lock ) ;
return - EBUSY ;
}
2008-12-30 09:48:21 +03:00
__iget ( old ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & old - > i_lock ) ;
2011-03-22 14:23:42 +03:00
spin_unlock ( & inode_hash_lock ) ;
2008-12-30 09:48:21 +03:00
wait_on_inode ( old ) ;
2010-10-23 23:19:20 +04:00
if ( unlikely ( ! inode_unhashed ( old ) ) ) {
2008-12-30 09:48:21 +03:00
iput ( old ) ;
return - EBUSY ;
}
iput ( old ) ;
}
}
EXPORT_SYMBOL ( insert_inode_locked ) ;
int insert_inode_locked4 ( struct inode * inode , unsigned long hashval ,
int ( * test ) ( struct inode * , void * ) , void * data )
{
2018-06-28 22:53:17 +03:00
struct inode * old ;
inode - > i_state | = I_CREATING ;
old = inode_insert5 ( inode , hashval , test , NULL , data ) ;
2008-12-30 09:48:21 +03:00
2018-05-17 11:53:05 +03:00
if ( old ! = inode ) {
2008-12-30 09:48:21 +03:00
iput ( old ) ;
2018-05-17 11:53:05 +03:00
return - EBUSY ;
2008-12-30 09:48:21 +03:00
}
2018-05-17 11:53:05 +03:00
return 0 ;
2008-12-30 09:48:21 +03:00
}
EXPORT_SYMBOL ( insert_inode_locked4 ) ;
2005-04-17 02:20:36 +04:00
2010-06-07 21:43:19 +04:00
int generic_delete_inode ( struct inode * inode )
{
return 1 ;
}
EXPORT_SYMBOL ( generic_delete_inode ) ;
/*
* Called when we ' re dropping the last reference
* to an inode .
2009-09-19 00:05:44 +04:00
*
2010-06-07 21:43:19 +04:00
* Call the FS " drop_inode() " function , defaulting to
* the legacy UNIX filesystem behaviour . If it tells
* us to evict inode , do so . Otherwise , retain inode
* in cache if fs is alive , sync and evict if fs is
* shutting down .
2009-09-19 00:05:44 +04:00
*/
2010-06-07 21:43:19 +04:00
static void iput_final ( struct inode * inode )
2005-04-17 02:20:36 +04:00
{
struct super_block * sb = inode - > i_sb ;
2010-06-07 21:43:19 +04:00
const struct super_operations * op = inode - > i_sb - > s_op ;
2017-12-01 14:40:16 +03:00
unsigned long state ;
2010-06-07 21:43:19 +04:00
int drop ;
2011-03-22 14:23:36 +03:00
WARN_ON ( inode - > i_state & I_NEW ) ;
2011-07-07 23:45:59 +04:00
if ( op - > drop_inode )
2010-06-07 21:43:19 +04:00
drop = op - > drop_inode ( inode ) ;
else
drop = generic_drop_inode ( inode ) ;
2005-04-17 02:20:36 +04:00
2020-12-08 05:08:43 +03:00
if ( ! drop & &
! ( inode - > i_state & I_DONTCACHE ) & &
( sb - > s_flags & SB_ACTIVE ) ) {
vfs: keep inodes with page cache off the inode shrinker LRU
Historically (pre-2.5), the inode shrinker used to reclaim only empty
inodes and skip over those that still contained page cache. This caused
problems on highmem hosts: struct inode could put fill lowmem zones
before the cache was getting reclaimed in the highmem zones.
To address this, the inode shrinker started to strip page cache to
facilitate reclaiming lowmem. However, this comes with its own set of
problems: the shrinkers may drop actively used page cache just because
the inodes are not currently open or dirty - think working with a large
git tree. It further doesn't respect cgroup memory protection settings
and can cause priority inversions between containers.
Nowadays, the page cache also holds non-resident info for evicted cache
pages in order to detect refaults. We've come to rely heavily on this
data inside reclaim for protecting the cache workingset and driving swap
behavior. We also use it to quantify and report workload health through
psi. The latter in turn is used for fleet health monitoring, as well as
driving automated memory sizing of workloads and containers, proactive
reclaim and memory offloading schemes.
The consequences of dropping page cache prematurely is that we're seeing
subtle and not-so-subtle failures in all of the above-mentioned
scenarios, with the workload generally entering unexpected thrashing
states while losing the ability to reliably detect it.
To fix this on non-highmem systems at least, going back to rotating
inodes on the LRU isn't feasible. We've tried (commit a76cf1a474d7
("mm: don't reclaim inodes with many attached pages")) and failed
(commit 69056ee6a8a3 ("Revert "mm: don't reclaim inodes with many
attached pages"")).
The issue is mostly that shrinker pools attract pressure based on their
size, and when objects get skipped the shrinkers remember this as
deferred reclaim work. This accumulates excessive pressure on the
remaining inodes, and we can quickly eat into heavily used ones, or
dirty ones that require IO to reclaim, when there potentially is plenty
of cold, clean cache around still.
Instead, this patch keeps populated inodes off the inode LRU in the
first place - just like an open file or dirty state would. An otherwise
clean and unused inode then gets queued when the last cache entry
disappears. This solves the problem without reintroducing the reclaim
issues, and generally is a bit more scalable than having to wade through
potentially hundreds of thousands of busy inodes.
Locking is a bit tricky because the locks protecting the inode state
(i_lock) and the inode LRU (lru_list.lock) don't nest inside the
irq-safe page cache lock (i_pages.xa_lock). Page cache deletions are
serialized through i_lock, taken before the i_pages lock, to make sure
depopulated inodes are queued reliably. Additions may race with
deletions, but we'll check again in the shrinker. If additions race
with the shrinker itself, we're protected by the i_lock: if find_inode()
or iput() win, the shrinker will bail on the elevated i_count or
I_REFERENCED; if the shrinker wins and goes ahead with the inode, it
will set I_FREEING and inhibit further igets(), which will cause the
other side to create a new instance of the inode instead.
Link: https://lkml.kernel.org/r/20210614211904.14420-4-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Roman Gushchin <guro@fb.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Dave Chinner <david@fromorbit.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2021-11-09 05:31:24 +03:00
__inode_add_lru ( inode , true ) ;
2011-03-22 14:23:37 +03:00
spin_unlock ( & inode - > i_lock ) ;
return ;
}
2017-12-01 14:40:16 +03:00
state = inode - > i_state ;
2010-06-07 21:43:19 +04:00
if ( ! drop ) {
2017-12-01 14:40:16 +03:00
WRITE_ONCE ( inode - > i_state , state | I_WILL_FREE ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2017-12-01 14:40:16 +03:00
2005-04-17 02:20:36 +04:00
write_inode_now ( inode , 1 ) ;
2017-12-01 14:40:16 +03:00
2011-03-22 14:23:36 +03:00
spin_lock ( & inode - > i_lock ) ;
2017-12-01 14:40:16 +03:00
state = inode - > i_state ;
WARN_ON ( state & I_NEW ) ;
state & = ~ I_WILL_FREE ;
2005-04-17 02:20:36 +04:00
}
2010-10-21 04:49:30 +04:00
2017-12-01 14:40:16 +03:00
WRITE_ONCE ( inode - > i_state , state | I_FREEING ) ;
2011-07-28 08:55:13 +04:00
if ( ! list_empty ( & inode - > i_lru ) )
inode_lru_list_del ( inode ) ;
2011-03-22 14:23:37 +03:00
spin_unlock ( & inode - > i_lock ) ;
2010-06-07 21:21:05 +04:00
evict ( inode ) ;
2005-04-17 02:20:36 +04:00
}
/**
2009-03-31 18:05:54 +04:00
* iput - put an inode
2005-04-17 02:20:36 +04:00
* @ inode : inode to put
*
* Puts an inode , dropping its usage count . If the inode use count hits
* zero , the inode is then freed and may also be destroyed .
*
* Consequently , iput ( ) can sleep .
*/
void iput ( struct inode * inode )
{
2015-02-02 08:37:00 +03:00
if ( ! inode )
return ;
BUG_ON ( inode - > i_state & I_CLEAR ) ;
retry :
if ( atomic_dec_and_lock ( & inode - > i_count , & inode - > i_lock ) ) {
if ( inode - > i_nlink & & ( inode - > i_state & I_DIRTY_TIME ) ) {
atomic_inc ( & inode - > i_count ) ;
spin_unlock ( & inode - > i_lock ) ;
trace_writeback_lazytime_iput ( inode ) ;
mark_inode_dirty_sync ( inode ) ;
goto retry ;
}
iput_final ( inode ) ;
2005-04-17 02:20:36 +04:00
}
}
EXPORT_SYMBOL ( iput ) ;
2020-01-09 16:30:41 +03:00
# ifdef CONFIG_BLOCK
2005-04-17 02:20:36 +04:00
/**
* bmap - find a block number in a file
2020-01-09 16:30:41 +03:00
* @ inode : inode owning the block number being requested
* @ block : pointer containing the block to find
2005-04-17 02:20:36 +04:00
*
2020-04-14 19:48:57 +03:00
* Replaces the value in ` ` * block ` ` with the block number on the device holding
2020-01-09 16:30:41 +03:00
* corresponding to the requested block number in the file .
* That is , asked for block 4 of inode 1 the function will replace the
2020-04-14 19:48:57 +03:00
* 4 in ` ` * block ` ` , with disk block relative to the disk start that holds that
2020-01-09 16:30:41 +03:00
* block of the file .
*
* Returns - EINVAL in case of error , 0 otherwise . If mapping falls into a
2020-04-14 19:48:57 +03:00
* hole , returns 0 and ` ` * block ` ` is also set to 0.
2005-04-17 02:20:36 +04:00
*/
2020-01-09 16:30:41 +03:00
int bmap ( struct inode * inode , sector_t * block )
2005-04-17 02:20:36 +04:00
{
2020-01-09 16:30:41 +03:00
if ( ! inode - > i_mapping - > a_ops - > bmap )
return - EINVAL ;
* block = inode - > i_mapping - > a_ops - > bmap ( inode - > i_mapping , * block ) ;
return 0 ;
2005-04-17 02:20:36 +04:00
}
EXPORT_SYMBOL ( bmap ) ;
2020-01-09 16:30:41 +03:00
# endif
2005-04-17 02:20:36 +04:00
2009-03-26 20:32:14 +03:00
/*
* With relative atime , only update atime if the previous atime is
2023-03-25 11:22:32 +03:00
* earlier than or equal to either the ctime or mtime ,
* or if at least a day has passed since the last atime update .
2009-03-26 20:32:14 +03:00
*/
2018-07-18 16:44:43 +03:00
static int relatime_need_update ( struct vfsmount * mnt , struct inode * inode ,
2019-04-26 17:50:41 +03:00
struct timespec64 now )
2009-03-26 20:32:14 +03:00
{
2018-07-18 16:44:43 +03:00
if ( ! ( mnt - > mnt_flags & MNT_RELATIME ) )
2009-03-26 20:32:14 +03:00
return 1 ;
/*
2023-03-25 11:22:32 +03:00
* Is mtime younger than or equal to atime ? If yes , update atime :
2009-03-26 20:32:14 +03:00
*/
vfs: change inode times to use struct timespec64
struct timespec is not y2038 safe. Transition vfs to use
y2038 safe struct timespec64 instead.
The change was made with the help of the following cocinelle
script. This catches about 80% of the changes.
All the header file and logic changes are included in the
first 5 rules. The rest are trivial substitutions.
I avoid changing any of the function signatures or any other
filesystem specific data structures to keep the patch simple
for review.
The script can be a little shorter by combining different cases.
But, this version was sufficient for my usecase.
virtual patch
@ depends on patch @
identifier now;
@@
- struct timespec
+ struct timespec64
current_time ( ... )
{
- struct timespec now = current_kernel_time();
+ struct timespec64 now = current_kernel_time64();
...
- return timespec_trunc(
+ return timespec64_trunc(
... );
}
@ depends on patch @
identifier xtime;
@@
struct \( iattr \| inode \| kstat \) {
...
- struct timespec xtime;
+ struct timespec64 xtime;
...
}
@ depends on patch @
identifier t;
@@
struct inode_operations {
...
int (*update_time) (...,
- struct timespec t,
+ struct timespec64 t,
...);
...
}
@ depends on patch @
identifier t;
identifier fn_update_time =~ "update_time$";
@@
fn_update_time (...,
- struct timespec *t,
+ struct timespec64 *t,
...) { ... }
@ depends on patch @
identifier t;
@@
lease_get_mtime( ... ,
- struct timespec *t
+ struct timespec64 *t
) { ... }
@te depends on patch forall@
identifier ts;
local idexpression struct inode *inode_node;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn_update_time =~ "update_time$";
identifier fn;
expression e, E3;
local idexpression struct inode *node1;
local idexpression struct inode *node2;
local idexpression struct iattr *attr1;
local idexpression struct iattr *attr2;
local idexpression struct iattr attr;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
@@
(
(
- struct timespec ts;
+ struct timespec64 ts;
|
- struct timespec ts = current_time(inode_node);
+ struct timespec64 ts = current_time(inode_node);
)
<+... when != ts
(
- timespec_equal(&inode_node->i_xtime, &ts)
+ timespec64_equal(&inode_node->i_xtime, &ts)
|
- timespec_equal(&ts, &inode_node->i_xtime)
+ timespec64_equal(&ts, &inode_node->i_xtime)
|
- timespec_compare(&inode_node->i_xtime, &ts)
+ timespec64_compare(&inode_node->i_xtime, &ts)
|
- timespec_compare(&ts, &inode_node->i_xtime)
+ timespec64_compare(&ts, &inode_node->i_xtime)
|
ts = current_time(e)
|
fn_update_time(..., &ts,...)
|
inode_node->i_xtime = ts
|
node1->i_xtime = ts
|
ts = inode_node->i_xtime
|
<+... attr1->ia_xtime ...+> = ts
|
ts = attr1->ia_xtime
|
ts.tv_sec
|
ts.tv_nsec
|
btrfs_set_stack_timespec_sec(..., ts.tv_sec)
|
btrfs_set_stack_timespec_nsec(..., ts.tv_nsec)
|
- ts = timespec64_to_timespec(
+ ts =
...
-)
|
- ts = ktime_to_timespec(
+ ts = ktime_to_timespec64(
...)
|
- ts = E3
+ ts = timespec_to_timespec64(E3)
|
- ktime_get_real_ts(&ts)
+ ktime_get_real_ts64(&ts)
|
fn(...,
- ts
+ timespec64_to_timespec(ts)
,...)
)
...+>
(
<... when != ts
- return ts;
+ return timespec64_to_timespec(ts);
...>
)
|
- timespec_equal(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_equal(&node1->i_xtime2, &node2->i_xtime2)
|
- timespec_equal(&node1->i_xtime1, &attr2->ia_xtime2)
+ timespec64_equal(&node1->i_xtime2, &attr2->ia_xtime2)
|
- timespec_compare(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_compare(&node1->i_xtime1, &node2->i_xtime2)
|
node1->i_xtime1 =
- timespec_trunc(attr1->ia_xtime1,
+ timespec64_trunc(attr1->ia_xtime1,
...)
|
- attr1->ia_xtime1 = timespec_trunc(attr2->ia_xtime2,
+ attr1->ia_xtime1 = timespec64_trunc(attr2->ia_xtime2,
...)
|
- ktime_get_real_ts(&attr1->ia_xtime1)
+ ktime_get_real_ts64(&attr1->ia_xtime1)
|
- ktime_get_real_ts(&attr.ia_xtime1)
+ ktime_get_real_ts64(&attr.ia_xtime1)
)
@ depends on patch @
struct inode *node;
struct iattr *attr;
identifier fn;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
expression e;
@@
(
- fn(node->i_xtime);
+ fn(timespec64_to_timespec(node->i_xtime));
|
fn(...,
- node->i_xtime);
+ timespec64_to_timespec(node->i_xtime));
|
- e = fn(attr->ia_xtime);
+ e = fn(timespec64_to_timespec(attr->ia_xtime));
)
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
fn (...,
- &attr->ia_xtime,
+ &ts,
...);
)
...+>
}
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
struct kstat *stat;
identifier ia_xtime =~ "^ia_[acm]time$";
identifier i_xtime =~ "^i_[acm]time$";
identifier xtime =~ "^[acm]time$";
identifier fn, ret;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(stat->xtime);
ret = fn (...,
- &stat->xtime);
+ &ts);
)
...+>
}
@ depends on patch @
struct inode *node;
struct inode *node2;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier i_xtime3 =~ "^i_[acm]time$";
struct iattr *attrp;
struct iattr *attrp2;
struct iattr attr ;
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
struct kstat *stat;
struct kstat stat1;
struct timespec64 ts;
identifier xtime =~ "^[acmb]time$";
expression e;
@@
(
( node->i_xtime2 \| attrp->ia_xtime2 \| attr.ia_xtime2 \) = node->i_xtime1 ;
|
node->i_xtime2 = \( node2->i_xtime1 \| timespec64_trunc(...) \);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
stat->xtime = node2->i_xtime1;
|
stat1.xtime = node2->i_xtime1;
|
( node->i_xtime2 \| attrp->ia_xtime2 \) = attrp->ia_xtime1 ;
|
( attrp->ia_xtime1 \| attr.ia_xtime1 \) = attrp2->ia_xtime2;
|
- e = node->i_xtime1;
+ e = timespec64_to_timespec( node->i_xtime1 );
|
- e = attrp->ia_xtime1;
+ e = timespec64_to_timespec( attrp->ia_xtime1 );
|
node->i_xtime1 = current_time(...);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
- node->i_xtime1 = e;
+ node->i_xtime1 = timespec_to_timespec64(e);
)
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: <anton@tuxera.com>
Cc: <balbi@kernel.org>
Cc: <bfields@fieldses.org>
Cc: <darrick.wong@oracle.com>
Cc: <dhowells@redhat.com>
Cc: <dsterba@suse.com>
Cc: <dwmw2@infradead.org>
Cc: <hch@lst.de>
Cc: <hirofumi@mail.parknet.co.jp>
Cc: <hubcap@omnibond.com>
Cc: <jack@suse.com>
Cc: <jaegeuk@kernel.org>
Cc: <jaharkes@cs.cmu.edu>
Cc: <jslaby@suse.com>
Cc: <keescook@chromium.org>
Cc: <mark@fasheh.com>
Cc: <miklos@szeredi.hu>
Cc: <nico@linaro.org>
Cc: <reiserfs-devel@vger.kernel.org>
Cc: <richard@nod.at>
Cc: <sage@redhat.com>
Cc: <sfrench@samba.org>
Cc: <swhiteho@redhat.com>
Cc: <tj@kernel.org>
Cc: <trond.myklebust@primarydata.com>
Cc: <tytso@mit.edu>
Cc: <viro@zeniv.linux.org.uk>
2018-05-09 05:36:02 +03:00
if ( timespec64_compare ( & inode - > i_mtime , & inode - > i_atime ) > = 0 )
2009-03-26 20:32:14 +03:00
return 1 ;
/*
2023-03-25 11:22:32 +03:00
* Is ctime younger than or equal to atime ? If yes , update atime :
2009-03-26 20:32:14 +03:00
*/
vfs: change inode times to use struct timespec64
struct timespec is not y2038 safe. Transition vfs to use
y2038 safe struct timespec64 instead.
The change was made with the help of the following cocinelle
script. This catches about 80% of the changes.
All the header file and logic changes are included in the
first 5 rules. The rest are trivial substitutions.
I avoid changing any of the function signatures or any other
filesystem specific data structures to keep the patch simple
for review.
The script can be a little shorter by combining different cases.
But, this version was sufficient for my usecase.
virtual patch
@ depends on patch @
identifier now;
@@
- struct timespec
+ struct timespec64
current_time ( ... )
{
- struct timespec now = current_kernel_time();
+ struct timespec64 now = current_kernel_time64();
...
- return timespec_trunc(
+ return timespec64_trunc(
... );
}
@ depends on patch @
identifier xtime;
@@
struct \( iattr \| inode \| kstat \) {
...
- struct timespec xtime;
+ struct timespec64 xtime;
...
}
@ depends on patch @
identifier t;
@@
struct inode_operations {
...
int (*update_time) (...,
- struct timespec t,
+ struct timespec64 t,
...);
...
}
@ depends on patch @
identifier t;
identifier fn_update_time =~ "update_time$";
@@
fn_update_time (...,
- struct timespec *t,
+ struct timespec64 *t,
...) { ... }
@ depends on patch @
identifier t;
@@
lease_get_mtime( ... ,
- struct timespec *t
+ struct timespec64 *t
) { ... }
@te depends on patch forall@
identifier ts;
local idexpression struct inode *inode_node;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn_update_time =~ "update_time$";
identifier fn;
expression e, E3;
local idexpression struct inode *node1;
local idexpression struct inode *node2;
local idexpression struct iattr *attr1;
local idexpression struct iattr *attr2;
local idexpression struct iattr attr;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
@@
(
(
- struct timespec ts;
+ struct timespec64 ts;
|
- struct timespec ts = current_time(inode_node);
+ struct timespec64 ts = current_time(inode_node);
)
<+... when != ts
(
- timespec_equal(&inode_node->i_xtime, &ts)
+ timespec64_equal(&inode_node->i_xtime, &ts)
|
- timespec_equal(&ts, &inode_node->i_xtime)
+ timespec64_equal(&ts, &inode_node->i_xtime)
|
- timespec_compare(&inode_node->i_xtime, &ts)
+ timespec64_compare(&inode_node->i_xtime, &ts)
|
- timespec_compare(&ts, &inode_node->i_xtime)
+ timespec64_compare(&ts, &inode_node->i_xtime)
|
ts = current_time(e)
|
fn_update_time(..., &ts,...)
|
inode_node->i_xtime = ts
|
node1->i_xtime = ts
|
ts = inode_node->i_xtime
|
<+... attr1->ia_xtime ...+> = ts
|
ts = attr1->ia_xtime
|
ts.tv_sec
|
ts.tv_nsec
|
btrfs_set_stack_timespec_sec(..., ts.tv_sec)
|
btrfs_set_stack_timespec_nsec(..., ts.tv_nsec)
|
- ts = timespec64_to_timespec(
+ ts =
...
-)
|
- ts = ktime_to_timespec(
+ ts = ktime_to_timespec64(
...)
|
- ts = E3
+ ts = timespec_to_timespec64(E3)
|
- ktime_get_real_ts(&ts)
+ ktime_get_real_ts64(&ts)
|
fn(...,
- ts
+ timespec64_to_timespec(ts)
,...)
)
...+>
(
<... when != ts
- return ts;
+ return timespec64_to_timespec(ts);
...>
)
|
- timespec_equal(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_equal(&node1->i_xtime2, &node2->i_xtime2)
|
- timespec_equal(&node1->i_xtime1, &attr2->ia_xtime2)
+ timespec64_equal(&node1->i_xtime2, &attr2->ia_xtime2)
|
- timespec_compare(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_compare(&node1->i_xtime1, &node2->i_xtime2)
|
node1->i_xtime1 =
- timespec_trunc(attr1->ia_xtime1,
+ timespec64_trunc(attr1->ia_xtime1,
...)
|
- attr1->ia_xtime1 = timespec_trunc(attr2->ia_xtime2,
+ attr1->ia_xtime1 = timespec64_trunc(attr2->ia_xtime2,
...)
|
- ktime_get_real_ts(&attr1->ia_xtime1)
+ ktime_get_real_ts64(&attr1->ia_xtime1)
|
- ktime_get_real_ts(&attr.ia_xtime1)
+ ktime_get_real_ts64(&attr.ia_xtime1)
)
@ depends on patch @
struct inode *node;
struct iattr *attr;
identifier fn;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
expression e;
@@
(
- fn(node->i_xtime);
+ fn(timespec64_to_timespec(node->i_xtime));
|
fn(...,
- node->i_xtime);
+ timespec64_to_timespec(node->i_xtime));
|
- e = fn(attr->ia_xtime);
+ e = fn(timespec64_to_timespec(attr->ia_xtime));
)
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
fn (...,
- &attr->ia_xtime,
+ &ts,
...);
)
...+>
}
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
struct kstat *stat;
identifier ia_xtime =~ "^ia_[acm]time$";
identifier i_xtime =~ "^i_[acm]time$";
identifier xtime =~ "^[acm]time$";
identifier fn, ret;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(stat->xtime);
ret = fn (...,
- &stat->xtime);
+ &ts);
)
...+>
}
@ depends on patch @
struct inode *node;
struct inode *node2;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier i_xtime3 =~ "^i_[acm]time$";
struct iattr *attrp;
struct iattr *attrp2;
struct iattr attr ;
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
struct kstat *stat;
struct kstat stat1;
struct timespec64 ts;
identifier xtime =~ "^[acmb]time$";
expression e;
@@
(
( node->i_xtime2 \| attrp->ia_xtime2 \| attr.ia_xtime2 \) = node->i_xtime1 ;
|
node->i_xtime2 = \( node2->i_xtime1 \| timespec64_trunc(...) \);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
stat->xtime = node2->i_xtime1;
|
stat1.xtime = node2->i_xtime1;
|
( node->i_xtime2 \| attrp->ia_xtime2 \) = attrp->ia_xtime1 ;
|
( attrp->ia_xtime1 \| attr.ia_xtime1 \) = attrp2->ia_xtime2;
|
- e = node->i_xtime1;
+ e = timespec64_to_timespec( node->i_xtime1 );
|
- e = attrp->ia_xtime1;
+ e = timespec64_to_timespec( attrp->ia_xtime1 );
|
node->i_xtime1 = current_time(...);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
- node->i_xtime1 = e;
+ node->i_xtime1 = timespec_to_timespec64(e);
)
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: <anton@tuxera.com>
Cc: <balbi@kernel.org>
Cc: <bfields@fieldses.org>
Cc: <darrick.wong@oracle.com>
Cc: <dhowells@redhat.com>
Cc: <dsterba@suse.com>
Cc: <dwmw2@infradead.org>
Cc: <hch@lst.de>
Cc: <hirofumi@mail.parknet.co.jp>
Cc: <hubcap@omnibond.com>
Cc: <jack@suse.com>
Cc: <jaegeuk@kernel.org>
Cc: <jaharkes@cs.cmu.edu>
Cc: <jslaby@suse.com>
Cc: <keescook@chromium.org>
Cc: <mark@fasheh.com>
Cc: <miklos@szeredi.hu>
Cc: <nico@linaro.org>
Cc: <reiserfs-devel@vger.kernel.org>
Cc: <richard@nod.at>
Cc: <sage@redhat.com>
Cc: <sfrench@samba.org>
Cc: <swhiteho@redhat.com>
Cc: <tj@kernel.org>
Cc: <trond.myklebust@primarydata.com>
Cc: <tytso@mit.edu>
Cc: <viro@zeniv.linux.org.uk>
2018-05-09 05:36:02 +03:00
if ( timespec64_compare ( & inode - > i_ctime , & inode - > i_atime ) > = 0 )
2009-03-26 20:32:14 +03:00
return 1 ;
/*
* Is the previous atime value older than a day ? If yes ,
* update atime :
*/
if ( ( long ) ( now . tv_sec - inode - > i_atime . tv_sec ) > = 24 * 60 * 60 )
return 1 ;
/*
* Good , we can skip the atime update :
*/
return 0 ;
}
vfs: change inode times to use struct timespec64
struct timespec is not y2038 safe. Transition vfs to use
y2038 safe struct timespec64 instead.
The change was made with the help of the following cocinelle
script. This catches about 80% of the changes.
All the header file and logic changes are included in the
first 5 rules. The rest are trivial substitutions.
I avoid changing any of the function signatures or any other
filesystem specific data structures to keep the patch simple
for review.
The script can be a little shorter by combining different cases.
But, this version was sufficient for my usecase.
virtual patch
@ depends on patch @
identifier now;
@@
- struct timespec
+ struct timespec64
current_time ( ... )
{
- struct timespec now = current_kernel_time();
+ struct timespec64 now = current_kernel_time64();
...
- return timespec_trunc(
+ return timespec64_trunc(
... );
}
@ depends on patch @
identifier xtime;
@@
struct \( iattr \| inode \| kstat \) {
...
- struct timespec xtime;
+ struct timespec64 xtime;
...
}
@ depends on patch @
identifier t;
@@
struct inode_operations {
...
int (*update_time) (...,
- struct timespec t,
+ struct timespec64 t,
...);
...
}
@ depends on patch @
identifier t;
identifier fn_update_time =~ "update_time$";
@@
fn_update_time (...,
- struct timespec *t,
+ struct timespec64 *t,
...) { ... }
@ depends on patch @
identifier t;
@@
lease_get_mtime( ... ,
- struct timespec *t
+ struct timespec64 *t
) { ... }
@te depends on patch forall@
identifier ts;
local idexpression struct inode *inode_node;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn_update_time =~ "update_time$";
identifier fn;
expression e, E3;
local idexpression struct inode *node1;
local idexpression struct inode *node2;
local idexpression struct iattr *attr1;
local idexpression struct iattr *attr2;
local idexpression struct iattr attr;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
@@
(
(
- struct timespec ts;
+ struct timespec64 ts;
|
- struct timespec ts = current_time(inode_node);
+ struct timespec64 ts = current_time(inode_node);
)
<+... when != ts
(
- timespec_equal(&inode_node->i_xtime, &ts)
+ timespec64_equal(&inode_node->i_xtime, &ts)
|
- timespec_equal(&ts, &inode_node->i_xtime)
+ timespec64_equal(&ts, &inode_node->i_xtime)
|
- timespec_compare(&inode_node->i_xtime, &ts)
+ timespec64_compare(&inode_node->i_xtime, &ts)
|
- timespec_compare(&ts, &inode_node->i_xtime)
+ timespec64_compare(&ts, &inode_node->i_xtime)
|
ts = current_time(e)
|
fn_update_time(..., &ts,...)
|
inode_node->i_xtime = ts
|
node1->i_xtime = ts
|
ts = inode_node->i_xtime
|
<+... attr1->ia_xtime ...+> = ts
|
ts = attr1->ia_xtime
|
ts.tv_sec
|
ts.tv_nsec
|
btrfs_set_stack_timespec_sec(..., ts.tv_sec)
|
btrfs_set_stack_timespec_nsec(..., ts.tv_nsec)
|
- ts = timespec64_to_timespec(
+ ts =
...
-)
|
- ts = ktime_to_timespec(
+ ts = ktime_to_timespec64(
...)
|
- ts = E3
+ ts = timespec_to_timespec64(E3)
|
- ktime_get_real_ts(&ts)
+ ktime_get_real_ts64(&ts)
|
fn(...,
- ts
+ timespec64_to_timespec(ts)
,...)
)
...+>
(
<... when != ts
- return ts;
+ return timespec64_to_timespec(ts);
...>
)
|
- timespec_equal(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_equal(&node1->i_xtime2, &node2->i_xtime2)
|
- timespec_equal(&node1->i_xtime1, &attr2->ia_xtime2)
+ timespec64_equal(&node1->i_xtime2, &attr2->ia_xtime2)
|
- timespec_compare(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_compare(&node1->i_xtime1, &node2->i_xtime2)
|
node1->i_xtime1 =
- timespec_trunc(attr1->ia_xtime1,
+ timespec64_trunc(attr1->ia_xtime1,
...)
|
- attr1->ia_xtime1 = timespec_trunc(attr2->ia_xtime2,
+ attr1->ia_xtime1 = timespec64_trunc(attr2->ia_xtime2,
...)
|
- ktime_get_real_ts(&attr1->ia_xtime1)
+ ktime_get_real_ts64(&attr1->ia_xtime1)
|
- ktime_get_real_ts(&attr.ia_xtime1)
+ ktime_get_real_ts64(&attr.ia_xtime1)
)
@ depends on patch @
struct inode *node;
struct iattr *attr;
identifier fn;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
expression e;
@@
(
- fn(node->i_xtime);
+ fn(timespec64_to_timespec(node->i_xtime));
|
fn(...,
- node->i_xtime);
+ timespec64_to_timespec(node->i_xtime));
|
- e = fn(attr->ia_xtime);
+ e = fn(timespec64_to_timespec(attr->ia_xtime));
)
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
fn (...,
- &attr->ia_xtime,
+ &ts,
...);
)
...+>
}
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
struct kstat *stat;
identifier ia_xtime =~ "^ia_[acm]time$";
identifier i_xtime =~ "^i_[acm]time$";
identifier xtime =~ "^[acm]time$";
identifier fn, ret;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(stat->xtime);
ret = fn (...,
- &stat->xtime);
+ &ts);
)
...+>
}
@ depends on patch @
struct inode *node;
struct inode *node2;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier i_xtime3 =~ "^i_[acm]time$";
struct iattr *attrp;
struct iattr *attrp2;
struct iattr attr ;
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
struct kstat *stat;
struct kstat stat1;
struct timespec64 ts;
identifier xtime =~ "^[acmb]time$";
expression e;
@@
(
( node->i_xtime2 \| attrp->ia_xtime2 \| attr.ia_xtime2 \) = node->i_xtime1 ;
|
node->i_xtime2 = \( node2->i_xtime1 \| timespec64_trunc(...) \);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
stat->xtime = node2->i_xtime1;
|
stat1.xtime = node2->i_xtime1;
|
( node->i_xtime2 \| attrp->ia_xtime2 \) = attrp->ia_xtime1 ;
|
( attrp->ia_xtime1 \| attr.ia_xtime1 \) = attrp2->ia_xtime2;
|
- e = node->i_xtime1;
+ e = timespec64_to_timespec( node->i_xtime1 );
|
- e = attrp->ia_xtime1;
+ e = timespec64_to_timespec( attrp->ia_xtime1 );
|
node->i_xtime1 = current_time(...);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
- node->i_xtime1 = e;
+ node->i_xtime1 = timespec_to_timespec64(e);
)
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: <anton@tuxera.com>
Cc: <balbi@kernel.org>
Cc: <bfields@fieldses.org>
Cc: <darrick.wong@oracle.com>
Cc: <dhowells@redhat.com>
Cc: <dsterba@suse.com>
Cc: <dwmw2@infradead.org>
Cc: <hch@lst.de>
Cc: <hirofumi@mail.parknet.co.jp>
Cc: <hubcap@omnibond.com>
Cc: <jack@suse.com>
Cc: <jaegeuk@kernel.org>
Cc: <jaharkes@cs.cmu.edu>
Cc: <jslaby@suse.com>
Cc: <keescook@chromium.org>
Cc: <mark@fasheh.com>
Cc: <miklos@szeredi.hu>
Cc: <nico@linaro.org>
Cc: <reiserfs-devel@vger.kernel.org>
Cc: <richard@nod.at>
Cc: <sage@redhat.com>
Cc: <sfrench@samba.org>
Cc: <swhiteho@redhat.com>
Cc: <tj@kernel.org>
Cc: <trond.myklebust@primarydata.com>
Cc: <tytso@mit.edu>
Cc: <viro@zeniv.linux.org.uk>
2018-05-09 05:36:02 +03:00
int generic_update_time ( struct inode * inode , struct timespec64 * time , int flags )
2012-03-26 17:59:21 +04:00
{
2021-01-12 22:02:45 +03:00
int dirty_flags = 0 ;
if ( flags & ( S_ATIME | S_CTIME | S_MTIME ) ) {
if ( flags & S_ATIME )
inode - > i_atime = * time ;
if ( flags & S_CTIME )
inode - > i_ctime = * time ;
if ( flags & S_MTIME )
inode - > i_mtime = * time ;
if ( inode - > i_sb - > s_flags & SB_LAZYTIME )
dirty_flags | = I_DIRTY_TIME ;
else
dirty_flags | = I_DIRTY_SYNC ;
}
if ( ( flags & S_VERSION ) & & inode_maybe_inc_iversion ( inode , false ) )
dirty_flags | = I_DIRTY_SYNC ;
__mark_inode_dirty ( inode , dirty_flags ) ;
2012-03-26 17:59:21 +04:00
return 0 ;
}
2015-02-02 08:37:00 +03:00
EXPORT_SYMBOL ( generic_update_time ) ;
/*
* This does the actual work of updating an inodes time or version . Must have
* had called mnt_want_write ( ) before calling this .
*/
2021-10-14 20:11:00 +03:00
int inode_update_time ( struct inode * inode , struct timespec64 * time , int flags )
2015-02-02 08:37:00 +03:00
{
2019-12-03 08:19:45 +03:00
if ( inode - > i_op - > update_time )
return inode - > i_op - > update_time ( inode , time , flags ) ;
return generic_update_time ( inode , time , flags ) ;
2015-02-02 08:37:00 +03:00
}
2021-10-14 20:11:00 +03:00
EXPORT_SYMBOL ( inode_update_time ) ;
2012-03-26 17:59:21 +04:00
2005-04-17 02:20:36 +04:00
/**
2021-01-14 11:04:39 +03:00
* atime_needs_update - update the access time
2012-04-18 04:03:25 +04:00
* @ path : the & struct path to update
2015-11-10 01:57:58 +03:00
* @ inode : inode to update
2005-04-17 02:20:36 +04:00
*
* Update the accessed time on an inode and mark it for writeback .
* This function automatically handles read only file systems and media ,
* as well as the " noatime " flag and inode specific " noatime " markers .
*/
2018-07-18 16:44:43 +03:00
bool atime_needs_update ( const struct path * path , struct inode * inode )
2005-04-17 02:20:36 +04:00
{
2012-03-15 16:21:57 +04:00
struct vfsmount * mnt = path - > mnt ;
vfs: change inode times to use struct timespec64
struct timespec is not y2038 safe. Transition vfs to use
y2038 safe struct timespec64 instead.
The change was made with the help of the following cocinelle
script. This catches about 80% of the changes.
All the header file and logic changes are included in the
first 5 rules. The rest are trivial substitutions.
I avoid changing any of the function signatures or any other
filesystem specific data structures to keep the patch simple
for review.
The script can be a little shorter by combining different cases.
But, this version was sufficient for my usecase.
virtual patch
@ depends on patch @
identifier now;
@@
- struct timespec
+ struct timespec64
current_time ( ... )
{
- struct timespec now = current_kernel_time();
+ struct timespec64 now = current_kernel_time64();
...
- return timespec_trunc(
+ return timespec64_trunc(
... );
}
@ depends on patch @
identifier xtime;
@@
struct \( iattr \| inode \| kstat \) {
...
- struct timespec xtime;
+ struct timespec64 xtime;
...
}
@ depends on patch @
identifier t;
@@
struct inode_operations {
...
int (*update_time) (...,
- struct timespec t,
+ struct timespec64 t,
...);
...
}
@ depends on patch @
identifier t;
identifier fn_update_time =~ "update_time$";
@@
fn_update_time (...,
- struct timespec *t,
+ struct timespec64 *t,
...) { ... }
@ depends on patch @
identifier t;
@@
lease_get_mtime( ... ,
- struct timespec *t
+ struct timespec64 *t
) { ... }
@te depends on patch forall@
identifier ts;
local idexpression struct inode *inode_node;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn_update_time =~ "update_time$";
identifier fn;
expression e, E3;
local idexpression struct inode *node1;
local idexpression struct inode *node2;
local idexpression struct iattr *attr1;
local idexpression struct iattr *attr2;
local idexpression struct iattr attr;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
@@
(
(
- struct timespec ts;
+ struct timespec64 ts;
|
- struct timespec ts = current_time(inode_node);
+ struct timespec64 ts = current_time(inode_node);
)
<+... when != ts
(
- timespec_equal(&inode_node->i_xtime, &ts)
+ timespec64_equal(&inode_node->i_xtime, &ts)
|
- timespec_equal(&ts, &inode_node->i_xtime)
+ timespec64_equal(&ts, &inode_node->i_xtime)
|
- timespec_compare(&inode_node->i_xtime, &ts)
+ timespec64_compare(&inode_node->i_xtime, &ts)
|
- timespec_compare(&ts, &inode_node->i_xtime)
+ timespec64_compare(&ts, &inode_node->i_xtime)
|
ts = current_time(e)
|
fn_update_time(..., &ts,...)
|
inode_node->i_xtime = ts
|
node1->i_xtime = ts
|
ts = inode_node->i_xtime
|
<+... attr1->ia_xtime ...+> = ts
|
ts = attr1->ia_xtime
|
ts.tv_sec
|
ts.tv_nsec
|
btrfs_set_stack_timespec_sec(..., ts.tv_sec)
|
btrfs_set_stack_timespec_nsec(..., ts.tv_nsec)
|
- ts = timespec64_to_timespec(
+ ts =
...
-)
|
- ts = ktime_to_timespec(
+ ts = ktime_to_timespec64(
...)
|
- ts = E3
+ ts = timespec_to_timespec64(E3)
|
- ktime_get_real_ts(&ts)
+ ktime_get_real_ts64(&ts)
|
fn(...,
- ts
+ timespec64_to_timespec(ts)
,...)
)
...+>
(
<... when != ts
- return ts;
+ return timespec64_to_timespec(ts);
...>
)
|
- timespec_equal(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_equal(&node1->i_xtime2, &node2->i_xtime2)
|
- timespec_equal(&node1->i_xtime1, &attr2->ia_xtime2)
+ timespec64_equal(&node1->i_xtime2, &attr2->ia_xtime2)
|
- timespec_compare(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_compare(&node1->i_xtime1, &node2->i_xtime2)
|
node1->i_xtime1 =
- timespec_trunc(attr1->ia_xtime1,
+ timespec64_trunc(attr1->ia_xtime1,
...)
|
- attr1->ia_xtime1 = timespec_trunc(attr2->ia_xtime2,
+ attr1->ia_xtime1 = timespec64_trunc(attr2->ia_xtime2,
...)
|
- ktime_get_real_ts(&attr1->ia_xtime1)
+ ktime_get_real_ts64(&attr1->ia_xtime1)
|
- ktime_get_real_ts(&attr.ia_xtime1)
+ ktime_get_real_ts64(&attr.ia_xtime1)
)
@ depends on patch @
struct inode *node;
struct iattr *attr;
identifier fn;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
expression e;
@@
(
- fn(node->i_xtime);
+ fn(timespec64_to_timespec(node->i_xtime));
|
fn(...,
- node->i_xtime);
+ timespec64_to_timespec(node->i_xtime));
|
- e = fn(attr->ia_xtime);
+ e = fn(timespec64_to_timespec(attr->ia_xtime));
)
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
fn (...,
- &attr->ia_xtime,
+ &ts,
...);
)
...+>
}
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
struct kstat *stat;
identifier ia_xtime =~ "^ia_[acm]time$";
identifier i_xtime =~ "^i_[acm]time$";
identifier xtime =~ "^[acm]time$";
identifier fn, ret;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(stat->xtime);
ret = fn (...,
- &stat->xtime);
+ &ts);
)
...+>
}
@ depends on patch @
struct inode *node;
struct inode *node2;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier i_xtime3 =~ "^i_[acm]time$";
struct iattr *attrp;
struct iattr *attrp2;
struct iattr attr ;
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
struct kstat *stat;
struct kstat stat1;
struct timespec64 ts;
identifier xtime =~ "^[acmb]time$";
expression e;
@@
(
( node->i_xtime2 \| attrp->ia_xtime2 \| attr.ia_xtime2 \) = node->i_xtime1 ;
|
node->i_xtime2 = \( node2->i_xtime1 \| timespec64_trunc(...) \);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
stat->xtime = node2->i_xtime1;
|
stat1.xtime = node2->i_xtime1;
|
( node->i_xtime2 \| attrp->ia_xtime2 \) = attrp->ia_xtime1 ;
|
( attrp->ia_xtime1 \| attr.ia_xtime1 \) = attrp2->ia_xtime2;
|
- e = node->i_xtime1;
+ e = timespec64_to_timespec( node->i_xtime1 );
|
- e = attrp->ia_xtime1;
+ e = timespec64_to_timespec( attrp->ia_xtime1 );
|
node->i_xtime1 = current_time(...);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
- node->i_xtime1 = e;
+ node->i_xtime1 = timespec_to_timespec64(e);
)
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: <anton@tuxera.com>
Cc: <balbi@kernel.org>
Cc: <bfields@fieldses.org>
Cc: <darrick.wong@oracle.com>
Cc: <dhowells@redhat.com>
Cc: <dsterba@suse.com>
Cc: <dwmw2@infradead.org>
Cc: <hch@lst.de>
Cc: <hirofumi@mail.parknet.co.jp>
Cc: <hubcap@omnibond.com>
Cc: <jack@suse.com>
Cc: <jaegeuk@kernel.org>
Cc: <jaharkes@cs.cmu.edu>
Cc: <jslaby@suse.com>
Cc: <keescook@chromium.org>
Cc: <mark@fasheh.com>
Cc: <miklos@szeredi.hu>
Cc: <nico@linaro.org>
Cc: <reiserfs-devel@vger.kernel.org>
Cc: <richard@nod.at>
Cc: <sage@redhat.com>
Cc: <sfrench@samba.org>
Cc: <swhiteho@redhat.com>
Cc: <tj@kernel.org>
Cc: <trond.myklebust@primarydata.com>
Cc: <tytso@mit.edu>
Cc: <viro@zeniv.linux.org.uk>
2018-05-09 05:36:02 +03:00
struct timespec64 now ;
2005-04-17 02:20:36 +04:00
2008-02-16 01:37:41 +03:00
if ( inode - > i_flags & S_NOATIME )
2015-03-23 05:37:40 +03:00
return false ;
2016-06-29 22:54:46 +03:00
/* Atime updates will likely cause i_uid and i_gid to be written
* back improprely if their true value is unknown to the vfs .
*/
2023-01-13 14:49:22 +03:00
if ( HAS_UNMAPPED_ID ( mnt_idmap ( mnt ) , inode ) )
2016-06-29 22:54:46 +03:00
return false ;
2007-02-10 12:44:49 +03:00
if ( IS_NOATIME ( inode ) )
2015-03-23 05:37:40 +03:00
return false ;
2017-11-28 00:05:09 +03:00
if ( ( inode - > i_sb - > s_flags & SB_NODIRATIME ) & & S_ISDIR ( inode - > i_mode ) )
2015-03-23 05:37:40 +03:00
return false ;
2006-12-13 11:34:34 +03:00
2008-02-16 01:37:41 +03:00
if ( mnt - > mnt_flags & MNT_NOATIME )
2015-03-23 05:37:40 +03:00
return false ;
2008-02-16 01:37:41 +03:00
if ( ( mnt - > mnt_flags & MNT_NODIRATIME ) & & S_ISDIR ( inode - > i_mode ) )
2015-03-23 05:37:40 +03:00
return false ;
2005-04-17 02:20:36 +04:00
2016-09-14 17:48:06 +03:00
now = current_time ( inode ) ;
2009-03-26 20:32:14 +03:00
2019-04-26 17:50:41 +03:00
if ( ! relatime_need_update ( mnt , inode , now ) )
2015-03-23 05:37:40 +03:00
return false ;
2009-03-26 20:32:14 +03:00
vfs: change inode times to use struct timespec64
struct timespec is not y2038 safe. Transition vfs to use
y2038 safe struct timespec64 instead.
The change was made with the help of the following cocinelle
script. This catches about 80% of the changes.
All the header file and logic changes are included in the
first 5 rules. The rest are trivial substitutions.
I avoid changing any of the function signatures or any other
filesystem specific data structures to keep the patch simple
for review.
The script can be a little shorter by combining different cases.
But, this version was sufficient for my usecase.
virtual patch
@ depends on patch @
identifier now;
@@
- struct timespec
+ struct timespec64
current_time ( ... )
{
- struct timespec now = current_kernel_time();
+ struct timespec64 now = current_kernel_time64();
...
- return timespec_trunc(
+ return timespec64_trunc(
... );
}
@ depends on patch @
identifier xtime;
@@
struct \( iattr \| inode \| kstat \) {
...
- struct timespec xtime;
+ struct timespec64 xtime;
...
}
@ depends on patch @
identifier t;
@@
struct inode_operations {
...
int (*update_time) (...,
- struct timespec t,
+ struct timespec64 t,
...);
...
}
@ depends on patch @
identifier t;
identifier fn_update_time =~ "update_time$";
@@
fn_update_time (...,
- struct timespec *t,
+ struct timespec64 *t,
...) { ... }
@ depends on patch @
identifier t;
@@
lease_get_mtime( ... ,
- struct timespec *t
+ struct timespec64 *t
) { ... }
@te depends on patch forall@
identifier ts;
local idexpression struct inode *inode_node;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn_update_time =~ "update_time$";
identifier fn;
expression e, E3;
local idexpression struct inode *node1;
local idexpression struct inode *node2;
local idexpression struct iattr *attr1;
local idexpression struct iattr *attr2;
local idexpression struct iattr attr;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
@@
(
(
- struct timespec ts;
+ struct timespec64 ts;
|
- struct timespec ts = current_time(inode_node);
+ struct timespec64 ts = current_time(inode_node);
)
<+... when != ts
(
- timespec_equal(&inode_node->i_xtime, &ts)
+ timespec64_equal(&inode_node->i_xtime, &ts)
|
- timespec_equal(&ts, &inode_node->i_xtime)
+ timespec64_equal(&ts, &inode_node->i_xtime)
|
- timespec_compare(&inode_node->i_xtime, &ts)
+ timespec64_compare(&inode_node->i_xtime, &ts)
|
- timespec_compare(&ts, &inode_node->i_xtime)
+ timespec64_compare(&ts, &inode_node->i_xtime)
|
ts = current_time(e)
|
fn_update_time(..., &ts,...)
|
inode_node->i_xtime = ts
|
node1->i_xtime = ts
|
ts = inode_node->i_xtime
|
<+... attr1->ia_xtime ...+> = ts
|
ts = attr1->ia_xtime
|
ts.tv_sec
|
ts.tv_nsec
|
btrfs_set_stack_timespec_sec(..., ts.tv_sec)
|
btrfs_set_stack_timespec_nsec(..., ts.tv_nsec)
|
- ts = timespec64_to_timespec(
+ ts =
...
-)
|
- ts = ktime_to_timespec(
+ ts = ktime_to_timespec64(
...)
|
- ts = E3
+ ts = timespec_to_timespec64(E3)
|
- ktime_get_real_ts(&ts)
+ ktime_get_real_ts64(&ts)
|
fn(...,
- ts
+ timespec64_to_timespec(ts)
,...)
)
...+>
(
<... when != ts
- return ts;
+ return timespec64_to_timespec(ts);
...>
)
|
- timespec_equal(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_equal(&node1->i_xtime2, &node2->i_xtime2)
|
- timespec_equal(&node1->i_xtime1, &attr2->ia_xtime2)
+ timespec64_equal(&node1->i_xtime2, &attr2->ia_xtime2)
|
- timespec_compare(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_compare(&node1->i_xtime1, &node2->i_xtime2)
|
node1->i_xtime1 =
- timespec_trunc(attr1->ia_xtime1,
+ timespec64_trunc(attr1->ia_xtime1,
...)
|
- attr1->ia_xtime1 = timespec_trunc(attr2->ia_xtime2,
+ attr1->ia_xtime1 = timespec64_trunc(attr2->ia_xtime2,
...)
|
- ktime_get_real_ts(&attr1->ia_xtime1)
+ ktime_get_real_ts64(&attr1->ia_xtime1)
|
- ktime_get_real_ts(&attr.ia_xtime1)
+ ktime_get_real_ts64(&attr.ia_xtime1)
)
@ depends on patch @
struct inode *node;
struct iattr *attr;
identifier fn;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
expression e;
@@
(
- fn(node->i_xtime);
+ fn(timespec64_to_timespec(node->i_xtime));
|
fn(...,
- node->i_xtime);
+ timespec64_to_timespec(node->i_xtime));
|
- e = fn(attr->ia_xtime);
+ e = fn(timespec64_to_timespec(attr->ia_xtime));
)
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
fn (...,
- &attr->ia_xtime,
+ &ts,
...);
)
...+>
}
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
struct kstat *stat;
identifier ia_xtime =~ "^ia_[acm]time$";
identifier i_xtime =~ "^i_[acm]time$";
identifier xtime =~ "^[acm]time$";
identifier fn, ret;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(stat->xtime);
ret = fn (...,
- &stat->xtime);
+ &ts);
)
...+>
}
@ depends on patch @
struct inode *node;
struct inode *node2;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier i_xtime3 =~ "^i_[acm]time$";
struct iattr *attrp;
struct iattr *attrp2;
struct iattr attr ;
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
struct kstat *stat;
struct kstat stat1;
struct timespec64 ts;
identifier xtime =~ "^[acmb]time$";
expression e;
@@
(
( node->i_xtime2 \| attrp->ia_xtime2 \| attr.ia_xtime2 \) = node->i_xtime1 ;
|
node->i_xtime2 = \( node2->i_xtime1 \| timespec64_trunc(...) \);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
stat->xtime = node2->i_xtime1;
|
stat1.xtime = node2->i_xtime1;
|
( node->i_xtime2 \| attrp->ia_xtime2 \) = attrp->ia_xtime1 ;
|
( attrp->ia_xtime1 \| attr.ia_xtime1 \) = attrp2->ia_xtime2;
|
- e = node->i_xtime1;
+ e = timespec64_to_timespec( node->i_xtime1 );
|
- e = attrp->ia_xtime1;
+ e = timespec64_to_timespec( attrp->ia_xtime1 );
|
node->i_xtime1 = current_time(...);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
- node->i_xtime1 = e;
+ node->i_xtime1 = timespec_to_timespec64(e);
)
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: <anton@tuxera.com>
Cc: <balbi@kernel.org>
Cc: <bfields@fieldses.org>
Cc: <darrick.wong@oracle.com>
Cc: <dhowells@redhat.com>
Cc: <dsterba@suse.com>
Cc: <dwmw2@infradead.org>
Cc: <hch@lst.de>
Cc: <hirofumi@mail.parknet.co.jp>
Cc: <hubcap@omnibond.com>
Cc: <jack@suse.com>
Cc: <jaegeuk@kernel.org>
Cc: <jaharkes@cs.cmu.edu>
Cc: <jslaby@suse.com>
Cc: <keescook@chromium.org>
Cc: <mark@fasheh.com>
Cc: <miklos@szeredi.hu>
Cc: <nico@linaro.org>
Cc: <reiserfs-devel@vger.kernel.org>
Cc: <richard@nod.at>
Cc: <sage@redhat.com>
Cc: <sfrench@samba.org>
Cc: <swhiteho@redhat.com>
Cc: <tj@kernel.org>
Cc: <trond.myklebust@primarydata.com>
Cc: <tytso@mit.edu>
Cc: <viro@zeniv.linux.org.uk>
2018-05-09 05:36:02 +03:00
if ( timespec64_equal ( & inode - > i_atime , & now ) )
2015-03-23 05:37:40 +03:00
return false ;
return true ;
}
void touch_atime ( const struct path * path )
{
struct vfsmount * mnt = path - > mnt ;
struct inode * inode = d_inode ( path - > dentry ) ;
vfs: change inode times to use struct timespec64
struct timespec is not y2038 safe. Transition vfs to use
y2038 safe struct timespec64 instead.
The change was made with the help of the following cocinelle
script. This catches about 80% of the changes.
All the header file and logic changes are included in the
first 5 rules. The rest are trivial substitutions.
I avoid changing any of the function signatures or any other
filesystem specific data structures to keep the patch simple
for review.
The script can be a little shorter by combining different cases.
But, this version was sufficient for my usecase.
virtual patch
@ depends on patch @
identifier now;
@@
- struct timespec
+ struct timespec64
current_time ( ... )
{
- struct timespec now = current_kernel_time();
+ struct timespec64 now = current_kernel_time64();
...
- return timespec_trunc(
+ return timespec64_trunc(
... );
}
@ depends on patch @
identifier xtime;
@@
struct \( iattr \| inode \| kstat \) {
...
- struct timespec xtime;
+ struct timespec64 xtime;
...
}
@ depends on patch @
identifier t;
@@
struct inode_operations {
...
int (*update_time) (...,
- struct timespec t,
+ struct timespec64 t,
...);
...
}
@ depends on patch @
identifier t;
identifier fn_update_time =~ "update_time$";
@@
fn_update_time (...,
- struct timespec *t,
+ struct timespec64 *t,
...) { ... }
@ depends on patch @
identifier t;
@@
lease_get_mtime( ... ,
- struct timespec *t
+ struct timespec64 *t
) { ... }
@te depends on patch forall@
identifier ts;
local idexpression struct inode *inode_node;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn_update_time =~ "update_time$";
identifier fn;
expression e, E3;
local idexpression struct inode *node1;
local idexpression struct inode *node2;
local idexpression struct iattr *attr1;
local idexpression struct iattr *attr2;
local idexpression struct iattr attr;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
@@
(
(
- struct timespec ts;
+ struct timespec64 ts;
|
- struct timespec ts = current_time(inode_node);
+ struct timespec64 ts = current_time(inode_node);
)
<+... when != ts
(
- timespec_equal(&inode_node->i_xtime, &ts)
+ timespec64_equal(&inode_node->i_xtime, &ts)
|
- timespec_equal(&ts, &inode_node->i_xtime)
+ timespec64_equal(&ts, &inode_node->i_xtime)
|
- timespec_compare(&inode_node->i_xtime, &ts)
+ timespec64_compare(&inode_node->i_xtime, &ts)
|
- timespec_compare(&ts, &inode_node->i_xtime)
+ timespec64_compare(&ts, &inode_node->i_xtime)
|
ts = current_time(e)
|
fn_update_time(..., &ts,...)
|
inode_node->i_xtime = ts
|
node1->i_xtime = ts
|
ts = inode_node->i_xtime
|
<+... attr1->ia_xtime ...+> = ts
|
ts = attr1->ia_xtime
|
ts.tv_sec
|
ts.tv_nsec
|
btrfs_set_stack_timespec_sec(..., ts.tv_sec)
|
btrfs_set_stack_timespec_nsec(..., ts.tv_nsec)
|
- ts = timespec64_to_timespec(
+ ts =
...
-)
|
- ts = ktime_to_timespec(
+ ts = ktime_to_timespec64(
...)
|
- ts = E3
+ ts = timespec_to_timespec64(E3)
|
- ktime_get_real_ts(&ts)
+ ktime_get_real_ts64(&ts)
|
fn(...,
- ts
+ timespec64_to_timespec(ts)
,...)
)
...+>
(
<... when != ts
- return ts;
+ return timespec64_to_timespec(ts);
...>
)
|
- timespec_equal(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_equal(&node1->i_xtime2, &node2->i_xtime2)
|
- timespec_equal(&node1->i_xtime1, &attr2->ia_xtime2)
+ timespec64_equal(&node1->i_xtime2, &attr2->ia_xtime2)
|
- timespec_compare(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_compare(&node1->i_xtime1, &node2->i_xtime2)
|
node1->i_xtime1 =
- timespec_trunc(attr1->ia_xtime1,
+ timespec64_trunc(attr1->ia_xtime1,
...)
|
- attr1->ia_xtime1 = timespec_trunc(attr2->ia_xtime2,
+ attr1->ia_xtime1 = timespec64_trunc(attr2->ia_xtime2,
...)
|
- ktime_get_real_ts(&attr1->ia_xtime1)
+ ktime_get_real_ts64(&attr1->ia_xtime1)
|
- ktime_get_real_ts(&attr.ia_xtime1)
+ ktime_get_real_ts64(&attr.ia_xtime1)
)
@ depends on patch @
struct inode *node;
struct iattr *attr;
identifier fn;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
expression e;
@@
(
- fn(node->i_xtime);
+ fn(timespec64_to_timespec(node->i_xtime));
|
fn(...,
- node->i_xtime);
+ timespec64_to_timespec(node->i_xtime));
|
- e = fn(attr->ia_xtime);
+ e = fn(timespec64_to_timespec(attr->ia_xtime));
)
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
fn (...,
- &attr->ia_xtime,
+ &ts,
...);
)
...+>
}
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
struct kstat *stat;
identifier ia_xtime =~ "^ia_[acm]time$";
identifier i_xtime =~ "^i_[acm]time$";
identifier xtime =~ "^[acm]time$";
identifier fn, ret;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(stat->xtime);
ret = fn (...,
- &stat->xtime);
+ &ts);
)
...+>
}
@ depends on patch @
struct inode *node;
struct inode *node2;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier i_xtime3 =~ "^i_[acm]time$";
struct iattr *attrp;
struct iattr *attrp2;
struct iattr attr ;
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
struct kstat *stat;
struct kstat stat1;
struct timespec64 ts;
identifier xtime =~ "^[acmb]time$";
expression e;
@@
(
( node->i_xtime2 \| attrp->ia_xtime2 \| attr.ia_xtime2 \) = node->i_xtime1 ;
|
node->i_xtime2 = \( node2->i_xtime1 \| timespec64_trunc(...) \);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
stat->xtime = node2->i_xtime1;
|
stat1.xtime = node2->i_xtime1;
|
( node->i_xtime2 \| attrp->ia_xtime2 \) = attrp->ia_xtime1 ;
|
( attrp->ia_xtime1 \| attr.ia_xtime1 \) = attrp2->ia_xtime2;
|
- e = node->i_xtime1;
+ e = timespec64_to_timespec( node->i_xtime1 );
|
- e = attrp->ia_xtime1;
+ e = timespec64_to_timespec( attrp->ia_xtime1 );
|
node->i_xtime1 = current_time(...);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
- node->i_xtime1 = e;
+ node->i_xtime1 = timespec_to_timespec64(e);
)
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: <anton@tuxera.com>
Cc: <balbi@kernel.org>
Cc: <bfields@fieldses.org>
Cc: <darrick.wong@oracle.com>
Cc: <dhowells@redhat.com>
Cc: <dsterba@suse.com>
Cc: <dwmw2@infradead.org>
Cc: <hch@lst.de>
Cc: <hirofumi@mail.parknet.co.jp>
Cc: <hubcap@omnibond.com>
Cc: <jack@suse.com>
Cc: <jaegeuk@kernel.org>
Cc: <jaharkes@cs.cmu.edu>
Cc: <jslaby@suse.com>
Cc: <keescook@chromium.org>
Cc: <mark@fasheh.com>
Cc: <miklos@szeredi.hu>
Cc: <nico@linaro.org>
Cc: <reiserfs-devel@vger.kernel.org>
Cc: <richard@nod.at>
Cc: <sage@redhat.com>
Cc: <sfrench@samba.org>
Cc: <swhiteho@redhat.com>
Cc: <tj@kernel.org>
Cc: <trond.myklebust@primarydata.com>
Cc: <tytso@mit.edu>
Cc: <viro@zeniv.linux.org.uk>
2018-05-09 05:36:02 +03:00
struct timespec64 now ;
2015-03-23 05:37:40 +03:00
2018-07-18 16:44:43 +03:00
if ( ! atime_needs_update ( path , inode ) )
2009-09-19 00:05:47 +04:00
return ;
2012-06-12 18:20:36 +04:00
if ( ! sb_start_write_trylock ( inode - > i_sb ) )
2009-09-19 00:05:47 +04:00
return ;
2006-12-13 11:34:34 +03:00
2015-03-23 05:37:40 +03:00
if ( __mnt_want_write ( mnt ) ! = 0 )
2012-06-12 18:20:36 +04:00
goto skip_update ;
2012-03-26 17:59:21 +04:00
/*
* File systems can error out when updating inodes if they need to
* allocate new space to modify an inode ( such is the case for
* Btrfs ) , but since we touch atime while walking down the path we
* really don ' t care if we failed to update the atime of the file ,
* so just ignore the return value .
2012-06-15 11:49:33 +04:00
* We may also fail on filesystems that have the ability to make parts
* of the fs read only , e . g . subvolumes in Btrfs .
2012-03-26 17:59:21 +04:00
*/
2016-09-14 17:48:06 +03:00
now = current_time ( inode ) ;
2021-10-14 20:11:00 +03:00
inode_update_time ( inode , & now , S_ATIME ) ;
2012-06-12 18:20:36 +04:00
__mnt_drop_write ( mnt ) ;
skip_update :
sb_end_write ( inode - > i_sb ) ;
2005-04-17 02:20:36 +04:00
}
2006-01-10 07:52:03 +03:00
EXPORT_SYMBOL ( touch_atime ) ;
2005-04-17 02:20:36 +04:00
2015-05-21 17:05:54 +03:00
/*
* Return mask of changes for notify_change ( ) that need to be done as a
* response to write or truncate . Return 0 if nothing has to be changed .
* Negative value on error ( change should be denied ) .
*/
2023-01-13 14:49:27 +03:00
int dentry_needs_remove_privs ( struct mnt_idmap * idmap ,
attr: use consistent sgid stripping checks
Currently setgid stripping in file_remove_privs()'s should_remove_suid()
helper is inconsistent with other parts of the vfs. Specifically, it only
raises ATTR_KILL_SGID if the inode is S_ISGID and S_IXGRP but not if the
inode isn't in the caller's groups and the caller isn't privileged over the
inode although we require this already in setattr_prepare() and
setattr_copy() and so all filesystem implement this requirement implicitly
because they have to use setattr_{prepare,copy}() anyway.
But the inconsistency shows up in setgid stripping bugs for overlayfs in
xfstests (e.g., generic/673, generic/683, generic/685, generic/686,
generic/687). For example, we test whether suid and setgid stripping works
correctly when performing various write-like operations as an unprivileged
user (fallocate, reflink, write, etc.):
echo "Test 1 - qa_user, non-exec file $verb"
setup_testfile
chmod a+rws $junk_file
commit_and_check "$qa_user" "$verb" 64k 64k
The test basically creates a file with 6666 permissions. While the file has
the S_ISUID and S_ISGID bits set it does not have the S_IXGRP set. On a
regular filesystem like xfs what will happen is:
sys_fallocate()
-> vfs_fallocate()
-> xfs_file_fallocate()
-> file_modified()
-> __file_remove_privs()
-> dentry_needs_remove_privs()
-> should_remove_suid()
-> __remove_privs()
newattrs.ia_valid = ATTR_FORCE | kill;
-> notify_change()
-> setattr_copy()
In should_remove_suid() we can see that ATTR_KILL_SUID is raised
unconditionally because the file in the test has S_ISUID set.
But we also see that ATTR_KILL_SGID won't be set because while the file
is S_ISGID it is not S_IXGRP (see above) which is a condition for
ATTR_KILL_SGID being raised.
So by the time we call notify_change() we have attr->ia_valid set to
ATTR_KILL_SUID | ATTR_FORCE. Now notify_change() sees that
ATTR_KILL_SUID is set and does:
ia_valid = attr->ia_valid |= ATTR_MODE
attr->ia_mode = (inode->i_mode & ~S_ISUID);
which means that when we call setattr_copy() later we will definitely
update inode->i_mode. Note that attr->ia_mode still contains S_ISGID.
Now we call into the filesystem's ->setattr() inode operation which will
end up calling setattr_copy(). Since ATTR_MODE is set we will hit:
if (ia_valid & ATTR_MODE) {
umode_t mode = attr->ia_mode;
vfsgid_t vfsgid = i_gid_into_vfsgid(mnt_userns, inode);
if (!vfsgid_in_group_p(vfsgid) &&
!capable_wrt_inode_uidgid(mnt_userns, inode, CAP_FSETID))
mode &= ~S_ISGID;
inode->i_mode = mode;
}
and since the caller in the test is neither capable nor in the group of the
inode the S_ISGID bit is stripped.
But assume the file isn't suid then ATTR_KILL_SUID won't be raised which
has the consequence that neither the setgid nor the suid bits are stripped
even though it should be stripped because the inode isn't in the caller's
groups and the caller isn't privileged over the inode.
If overlayfs is in the mix things become a bit more complicated and the bug
shows up more clearly. When e.g., ovl_setattr() is hit from
ovl_fallocate()'s call to file_remove_privs() then ATTR_KILL_SUID and
ATTR_KILL_SGID might be raised but because the check in notify_change() is
questioning the ATTR_KILL_SGID flag again by requiring S_IXGRP for it to be
stripped the S_ISGID bit isn't removed even though it should be stripped:
sys_fallocate()
-> vfs_fallocate()
-> ovl_fallocate()
-> file_remove_privs()
-> dentry_needs_remove_privs()
-> should_remove_suid()
-> __remove_privs()
newattrs.ia_valid = ATTR_FORCE | kill;
-> notify_change()
-> ovl_setattr()
// TAKE ON MOUNTER'S CREDS
-> ovl_do_notify_change()
-> notify_change()
// GIVE UP MOUNTER'S CREDS
// TAKE ON MOUNTER'S CREDS
-> vfs_fallocate()
-> xfs_file_fallocate()
-> file_modified()
-> __file_remove_privs()
-> dentry_needs_remove_privs()
-> should_remove_suid()
-> __remove_privs()
newattrs.ia_valid = attr_force | kill;
-> notify_change()
The fix for all of this is to make file_remove_privs()'s
should_remove_suid() helper to perform the same checks as we already
require in setattr_prepare() and setattr_copy() and have notify_change()
not pointlessly requiring S_IXGRP again. It doesn't make any sense in the
first place because the caller must calculate the flags via
should_remove_suid() anyway which would raise ATTR_KILL_SGID.
While we're at it we move should_remove_suid() from inode.c to attr.c
where it belongs with the rest of the iattr helpers. Especially since it
returns ATTR_KILL_S{G,U}ID flags. We also rename it to
setattr_should_drop_suidgid() to better reflect that it indicates both
setuid and setgid bit removal and also that it returns attr flags.
Running xfstests with this doesn't report any regressions. We should really
try and use consistent checks.
Reviewed-by: Amir Goldstein <amir73il@gmail.com>
Signed-off-by: Christian Brauner (Microsoft) <brauner@kernel.org>
2022-10-17 18:06:37 +03:00
struct dentry * dentry )
2015-05-21 17:05:54 +03:00
{
struct inode * inode = d_inode ( dentry ) ;
int mask = 0 ;
int ret ;
if ( IS_NOSEC ( inode ) )
return 0 ;
2023-01-13 14:49:27 +03:00
mask = setattr_should_drop_suidgid ( idmap , inode ) ;
2015-05-21 17:05:54 +03:00
ret = security_inode_need_killpriv ( dentry ) ;
if ( ret < 0 )
return ret ;
if ( ret )
mask | = ATTR_KILL_PRIV ;
return mask ;
}
2023-01-13 14:49:10 +03:00
static int __remove_privs ( struct mnt_idmap * idmap ,
2021-01-21 16:19:34 +03:00
struct dentry * dentry , int kill )
2012-05-15 10:57:33 +04:00
{
struct iattr newattrs ;
newattrs . ia_valid = ATTR_FORCE | kill ;
2011-09-21 01:19:26 +04:00
/*
* Note we call this on write , so notify_change will not
* encounter any conflicting delegations :
*/
2023-01-13 14:49:10 +03:00
return notify_change ( idmap , dentry , & newattrs , NULL ) ;
2012-05-15 10:57:33 +04:00
}
2022-06-23 20:51:51 +03:00
static int __file_remove_privs ( struct file * file , unsigned int flags )
2012-05-15 10:57:33 +04:00
{
2016-08-03 14:44:27 +03:00
struct dentry * dentry = file_dentry ( file ) ;
struct inode * inode = file_inode ( file ) ;
2022-08-16 18:31:58 +03:00
int error = 0 ;
2015-05-21 17:05:54 +03:00
int kill ;
2012-05-15 10:57:33 +04:00
2018-12-14 13:55:52 +03:00
if ( IS_NOSEC ( inode ) | | ! S_ISREG ( inode - > i_mode ) )
2012-05-15 10:57:33 +04:00
return 0 ;
2023-01-13 14:49:27 +03:00
kill = dentry_needs_remove_privs ( file_mnt_idmap ( file ) , dentry ) ;
2022-08-16 18:31:58 +03:00
if ( kill < 0 )
2015-05-21 17:05:54 +03:00
return kill ;
2022-06-23 20:51:51 +03:00
2022-08-16 18:31:58 +03:00
if ( kill ) {
if ( flags & IOCB_NOWAIT )
return - EAGAIN ;
2023-01-13 14:49:10 +03:00
error = __remove_privs ( file_mnt_idmap ( file ) , dentry , kill ) ;
2022-08-16 18:31:58 +03:00
}
2022-06-23 20:51:51 +03:00
2015-05-21 17:05:52 +03:00
if ( ! error )
inode_has_no_xattr ( inode ) ;
2012-05-15 10:57:33 +04:00
return error ;
}
2022-06-23 20:51:51 +03:00
/**
* file_remove_privs - remove special file privileges ( suid , capabilities )
* @ file : file to remove privileges from
*
* When file is modified by a write or truncation ensure that special
* file privileges are removed .
*
* Return : 0 on success , negative errno on failure .
*/
int file_remove_privs ( struct file * file )
{
return __file_remove_privs ( file , 0 ) ;
}
2015-05-21 17:05:53 +03:00
EXPORT_SYMBOL ( file_remove_privs ) ;
2012-05-15 10:57:33 +04:00
2022-06-23 20:51:52 +03:00
static int inode_needs_update_time ( struct inode * inode , struct timespec64 * now )
2005-04-17 02:20:36 +04:00
{
2012-03-26 17:59:21 +04:00
int sync_it = 0 ;
2005-04-17 02:20:36 +04:00
2009-09-19 00:05:48 +04:00
/* First try to exhaust all avenues to not sync */
2005-04-17 02:20:36 +04:00
if ( IS_NOCMTIME ( inode ) )
2012-03-26 17:59:21 +04:00
return 0 ;
2008-02-16 01:37:43 +03:00
2022-06-23 20:51:52 +03:00
if ( ! timespec64_equal ( & inode - > i_mtime , now ) )
2009-09-19 00:05:48 +04:00
sync_it = S_MTIME ;
2005-04-17 02:20:36 +04:00
2022-06-23 20:51:52 +03:00
if ( ! timespec64_equal ( & inode - > i_ctime , now ) )
2009-09-19 00:05:48 +04:00
sync_it | = S_CTIME ;
2006-01-10 07:52:01 +03:00
2017-12-11 14:35:22 +03:00
if ( IS_I_VERSION ( inode ) & & inode_iversion_need_inc ( inode ) )
2009-09-19 00:05:48 +04:00
sync_it | = S_VERSION ;
2008-01-29 07:58:27 +03:00
2022-06-23 20:51:52 +03:00
return sync_it ;
}
static int __file_update_time ( struct file * file , struct timespec64 * now ,
int sync_mode )
{
int ret = 0 ;
struct inode * inode = file_inode ( file ) ;
2009-09-19 00:05:48 +04:00
2022-06-23 20:51:52 +03:00
/* try to update time settings */
if ( ! __mnt_want_write_file ( file ) ) {
ret = inode_update_time ( inode , now , sync_mode ) ;
__mnt_drop_write_file ( file ) ;
}
2012-03-26 17:59:21 +04:00
return ret ;
2005-04-17 02:20:36 +04:00
}
2022-06-23 20:51:52 +03:00
/**
* file_update_time - update mtime and ctime time
* @ file : file accessed
*
* Update the mtime and ctime members of an inode and mark the inode for
* writeback . Note that this function is meant exclusively for usage in
* the file write path of filesystems , and filesystems may choose to
* explicitly ignore updates via this function with the _NOCMTIME inode
* flag , e . g . for network filesystem where these imestamps are handled
* by the server . This can return an error for file systems who need to
* allocate space in order to update an inode .
*
* Return : 0 on success , negative errno on failure .
*/
int file_update_time ( struct file * file )
{
int ret ;
struct inode * inode = file_inode ( file ) ;
struct timespec64 now = current_time ( inode ) ;
ret = inode_needs_update_time ( inode , & now ) ;
if ( ret < = 0 )
return ret ;
return __file_update_time ( file , & now , ret ) ;
}
2006-01-10 07:52:01 +03:00
EXPORT_SYMBOL ( file_update_time ) ;
2005-04-17 02:20:36 +04:00
2022-06-23 20:51:51 +03:00
/**
2022-06-23 20:51:53 +03:00
* file_modified_flags - handle mandated vfs changes when modifying a file
2022-06-23 20:51:51 +03:00
* @ file : file that was modified
2022-06-23 20:51:53 +03:00
* @ flags : kiocb flags
2022-06-23 20:51:51 +03:00
*
* When file has been modified ensure that special
* file privileges are removed and time settings are updated .
*
2022-06-23 20:51:53 +03:00
* If IOCB_NOWAIT is set , special file privileges will not be removed and
* time settings will not be updated . It will return - EAGAIN .
*
2022-06-23 20:51:51 +03:00
* Context : Caller must hold the file ' s inode lock .
*
* Return : 0 on success , negative errno on failure .
*/
2022-06-23 20:51:53 +03:00
static int file_modified_flags ( struct file * file , int flags )
2019-06-05 18:04:49 +03:00
{
2022-06-23 20:51:51 +03:00
int ret ;
2022-06-23 20:51:52 +03:00
struct inode * inode = file_inode ( file ) ;
struct timespec64 now = current_time ( inode ) ;
2019-06-05 18:04:49 +03:00
/*
* Clear the security bits if the process is not being run by root .
* This keeps people from modifying setuid and setgid binaries .
*/
2022-06-23 20:51:53 +03:00
ret = __file_remove_privs ( file , flags ) ;
2022-06-23 20:51:51 +03:00
if ( ret )
return ret ;
2019-06-05 18:04:49 +03:00
if ( unlikely ( file - > f_mode & FMODE_NOCMTIME ) )
return 0 ;
2022-06-23 20:51:52 +03:00
ret = inode_needs_update_time ( inode , & now ) ;
if ( ret < = 0 )
return ret ;
2022-06-23 20:51:53 +03:00
if ( flags & IOCB_NOWAIT )
return - EAGAIN ;
2022-06-23 20:51:52 +03:00
return __file_update_time ( file , & now , ret ) ;
2019-06-05 18:04:49 +03:00
}
2022-06-23 20:51:53 +03:00
/**
* file_modified - handle mandated vfs changes when modifying a file
* @ file : file that was modified
*
* When file has been modified ensure that special
* file privileges are removed and time settings are updated .
*
* Context : Caller must hold the file ' s inode lock .
*
* Return : 0 on success , negative errno on failure .
*/
int file_modified ( struct file * file )
{
return file_modified_flags ( file , 0 ) ;
}
2019-06-05 18:04:49 +03:00
EXPORT_SYMBOL ( file_modified ) ;
2022-06-23 20:51:53 +03:00
/**
* kiocb_modified - handle mandated vfs changes when modifying a file
* @ iocb : iocb that was modified
*
* When file has been modified ensure that special
* file privileges are removed and time settings are updated .
*
* Context : Caller must hold the file ' s inode lock .
*
* Return : 0 on success , negative errno on failure .
*/
int kiocb_modified ( struct kiocb * iocb )
{
return file_modified_flags ( iocb - > ki_filp , iocb - > ki_flags ) ;
}
EXPORT_SYMBOL_GPL ( kiocb_modified ) ;
2005-04-17 02:20:36 +04:00
int inode_needs_sync ( struct inode * inode )
{
if ( IS_SYNC ( inode ) )
return 1 ;
if ( S_ISDIR ( inode - > i_mode ) & & IS_DIRSYNC ( inode ) )
return 1 ;
return 0 ;
}
EXPORT_SYMBOL ( inode_needs_sync ) ;
/*
2005-07-13 00:58:10 +04:00
* If we try to find an inode in the inode hash while it is being
* deleted , we have to wait until the filesystem completes its
* deletion before reporting that it isn ' t found . This function waits
* until the deletion _might_ have completed . Callers are responsible
* to recheck inode state .
*
2009-12-17 16:25:01 +03:00
* It doesn ' t matter if I_NEW is not set initially , a call to
2011-03-22 14:23:36 +03:00
* wake_up_bit ( & inode - > i_state , __I_NEW ) after removing from the hash list
* will DTRT .
2005-04-17 02:20:36 +04:00
*/
static void __wait_on_freeing_inode ( struct inode * inode )
{
wait_queue_head_t * wq ;
2009-12-17 16:25:01 +03:00
DEFINE_WAIT_BIT ( wait , & inode - > i_state , __I_NEW ) ;
wq = bit_waitqueue ( & inode - > i_state , __I_NEW ) ;
2017-03-05 13:25:39 +03:00
prepare_to_wait ( wq , & wait . wq_entry , TASK_UNINTERRUPTIBLE ) ;
2011-03-22 14:23:36 +03:00
spin_unlock ( & inode - > i_lock ) ;
2011-03-22 14:23:42 +03:00
spin_unlock ( & inode_hash_lock ) ;
2005-04-17 02:20:36 +04:00
schedule ( ) ;
2017-03-05 13:25:39 +03:00
finish_wait ( wq , & wait . wq_entry ) ;
2011-03-22 14:23:42 +03:00
spin_lock ( & inode_hash_lock ) ;
2005-04-17 02:20:36 +04:00
}
static __initdata unsigned long ihash_entries ;
static int __init set_ihash_entries ( char * str )
{
if ( ! str )
return 0 ;
ihash_entries = simple_strtoul ( str , & str , 0 ) ;
return 1 ;
}
__setup ( " ihash_entries= " , set_ihash_entries ) ;
/*
* Initialize the waitqueues and inode hash table .
*/
void __init inode_init_early ( void )
{
/* If hashes are distributed across NUMA nodes, defer
* hash allocation until vmalloc space is available .
*/
if ( hashdist )
return ;
inode_hashtable =
alloc_large_system_hash ( " Inode-cache " ,
sizeof ( struct hlist_head ) ,
ihash_entries ,
14 ,
2017-07-07 01:39:11 +03:00
HASH_EARLY | HASH_ZERO ,
2005-04-17 02:20:36 +04:00
& i_hash_shift ,
& i_hash_mask ,
2012-05-23 17:33:35 +04:00
0 ,
2005-04-17 02:20:36 +04:00
0 ) ;
}
2007-10-17 10:26:30 +04:00
void __init inode_init ( void )
2005-04-17 02:20:36 +04:00
{
/* inode slab cache */
2006-03-24 14:16:09 +03:00
inode_cachep = kmem_cache_create ( " inode_cache " ,
sizeof ( struct inode ) ,
0 ,
( SLAB_RECLAIM_ACCOUNT | SLAB_PANIC |
2016-01-15 02:18:21 +03:00
SLAB_MEM_SPREAD | SLAB_ACCOUNT ) ,
2007-07-20 05:11:58 +04:00
init_once ) ;
2005-04-17 02:20:36 +04:00
/* Hash may have been set up in inode_init_early */
if ( ! hashdist )
return ;
inode_hashtable =
alloc_large_system_hash ( " Inode-cache " ,
sizeof ( struct hlist_head ) ,
ihash_entries ,
14 ,
2017-07-07 01:39:11 +03:00
HASH_ZERO ,
2005-04-17 02:20:36 +04:00
& i_hash_shift ,
& i_hash_mask ,
2012-05-23 17:33:35 +04:00
0 ,
2005-04-17 02:20:36 +04:00
0 ) ;
}
void init_special_inode ( struct inode * inode , umode_t mode , dev_t rdev )
{
inode - > i_mode = mode ;
if ( S_ISCHR ( mode ) ) {
inode - > i_fop = & def_chr_fops ;
inode - > i_rdev = rdev ;
} else if ( S_ISBLK ( mode ) ) {
2023-05-08 17:44:05 +03:00
if ( IS_ENABLED ( CONFIG_BLOCK ) )
inode - > i_fop = & def_blk_fops ;
2005-04-17 02:20:36 +04:00
inode - > i_rdev = rdev ;
} else if ( S_ISFIFO ( mode ) )
2013-03-12 17:58:10 +04:00
inode - > i_fop = & pipefifo_fops ;
2005-04-17 02:20:36 +04:00
else if ( S_ISSOCK ( mode ) )
2014-11-19 07:38:21 +03:00
; /* leave it no_open_fops */
2005-04-17 02:20:36 +04:00
else
2009-09-19 00:05:43 +04:00
printk ( KERN_DEBUG " init_special_inode: bogus i_mode (%o) for "
" inode %s:%lu \n " , mode , inode - > i_sb - > s_id ,
inode - > i_ino ) ;
2005-04-17 02:20:36 +04:00
}
EXPORT_SYMBOL ( init_special_inode ) ;
2010-03-04 17:29:14 +03:00
/**
2011-02-15 15:48:09 +03:00
* inode_init_owner - Init uid , gid , mode for new inode according to posix standards
2023-01-13 14:49:25 +03:00
* @ idmap : idmap of the mount the inode was created from
2010-03-04 17:29:14 +03:00
* @ inode : New inode
* @ dir : Directory inode
* @ mode : mode of the new inode
2021-01-21 16:19:25 +03:00
*
2023-01-13 14:49:25 +03:00
* If the inode has been created through an idmapped mount the idmap of
* the vfsmount must be passed through @ idmap . This function will then take
* care to map the inode according to @ idmap before checking permissions
2021-01-21 16:19:25 +03:00
* and initializing i_uid and i_gid . On non - idmapped mounts or if permission
2023-01-13 14:49:25 +03:00
* checking is to be performed on the raw inode simply pass @ nop_mnt_idmap .
2010-03-04 17:29:14 +03:00
*/
2023-01-13 14:49:25 +03:00
void inode_init_owner ( struct mnt_idmap * idmap , struct inode * inode ,
2021-01-21 16:19:25 +03:00
const struct inode * dir , umode_t mode )
2010-03-04 17:29:14 +03:00
{
2023-01-13 14:49:31 +03:00
inode_fsuid_set ( inode , idmap ) ;
2010-03-04 17:29:14 +03:00
if ( dir & & dir - > i_mode & S_ISGID ) {
inode - > i_gid = dir - > i_gid ;
2018-07-04 03:10:19 +03:00
/* Directories are special, and always inherit S_ISGID */
2010-03-04 17:29:14 +03:00
if ( S_ISDIR ( mode ) )
mode | = S_ISGID ;
} else
2023-01-13 14:49:31 +03:00
inode_fsgid_set ( inode , idmap ) ;
2010-03-04 17:29:14 +03:00
inode - > i_mode = mode ;
}
EXPORT_SYMBOL ( inode_init_owner ) ;
2011-03-24 02:43:25 +03:00
2011-03-24 02:43:26 +03:00
/**
* inode_owner_or_capable - check current task permissions to inode
2023-01-13 14:49:26 +03:00
* @ idmap : idmap of the mount the inode was found from
2011-03-24 02:43:26 +03:00
* @ inode : inode being checked
*
2014-06-10 23:45:42 +04:00
* Return true if current either has CAP_FOWNER in a namespace with the
* inode owner uid mapped , or owns the file .
2021-01-21 16:19:25 +03:00
*
2023-01-13 14:49:26 +03:00
* If the inode has been found through an idmapped mount the idmap of
* the vfsmount must be passed through @ idmap . This function will then take
* care to map the inode according to @ idmap before checking permissions .
2021-01-21 16:19:25 +03:00
* On non - idmapped mounts or if permission checking is to be performed on the
2023-01-13 14:49:26 +03:00
* raw inode simply passs @ nop_mnt_idmap .
2011-03-24 02:43:25 +03:00
*/
2023-01-13 14:49:26 +03:00
bool inode_owner_or_capable ( struct mnt_idmap * idmap ,
2021-01-21 16:19:25 +03:00
const struct inode * inode )
2011-03-24 02:43:25 +03:00
{
2022-06-22 23:12:16 +03:00
vfsuid_t vfsuid ;
2014-06-10 23:45:42 +04:00
struct user_namespace * ns ;
2023-01-13 14:49:30 +03:00
vfsuid = i_uid_into_vfsuid ( idmap , inode ) ;
2022-06-22 23:12:16 +03:00
if ( vfsuid_eq_kuid ( vfsuid , current_fsuid ( ) ) )
2011-03-24 02:43:25 +03:00
return true ;
2014-06-10 23:45:42 +04:00
ns = current_user_ns ( ) ;
2022-06-22 23:12:16 +03:00
if ( vfsuid_has_mapping ( ns , vfsuid ) & & ns_capable ( ns , CAP_FOWNER ) )
2011-03-24 02:43:25 +03:00
return true ;
return false ;
}
2011-03-24 02:43:26 +03:00
EXPORT_SYMBOL ( inode_owner_or_capable ) ;
2012-05-31 20:22:33 +04:00
/*
* Direct i / o helper functions
*/
static void __inode_dio_wait ( struct inode * inode )
{
wait_queue_head_t * wq = bit_waitqueue ( & inode - > i_state , __I_DIO_WAKEUP ) ;
DEFINE_WAIT_BIT ( q , & inode - > i_state , __I_DIO_WAKEUP ) ;
do {
2017-03-05 13:25:39 +03:00
prepare_to_wait ( wq , & q . wq_entry , TASK_UNINTERRUPTIBLE ) ;
2012-05-31 20:22:33 +04:00
if ( atomic_read ( & inode - > i_dio_count ) )
schedule ( ) ;
} while ( atomic_read ( & inode - > i_dio_count ) ) ;
2017-03-05 13:25:39 +03:00
finish_wait ( wq , & q . wq_entry ) ;
2012-05-31 20:22:33 +04:00
}
/**
* inode_dio_wait - wait for outstanding DIO requests to finish
* @ inode : inode to wait for
*
* Waits for all pending direct I / O requests to finish so that we can
* proceed with a truncate or equivalent operation .
*
* Must be called under a lock that serializes taking new references
* to i_dio_count , usually by inode - > i_mutex .
*/
void inode_dio_wait ( struct inode * inode )
{
if ( atomic_read ( & inode - > i_dio_count ) )
__inode_dio_wait ( inode ) ;
}
EXPORT_SYMBOL ( inode_dio_wait ) ;
2014-03-24 22:43:12 +04:00
/*
* inode_set_flags - atomically set some inode flags
*
* Note : the caller should be holding i_mutex , or else be sure that
* they have exclusive access to the inode structure ( i . e . , while the
* inode is being instantiated ) . The reason for the cmpxchg ( ) loop
* - - - which wouldn ' t be necessary if all code paths which modify
* i_flags actually followed this rule , is that there is at least one
2015-05-21 17:05:53 +03:00
* code path which doesn ' t today so we use cmpxchg ( ) out of an abundance
* of caution .
2014-03-24 22:43:12 +04:00
*
* In the long run , i_mutex is overkill , and we should probably look
* at using the i_lock spinlock to protect i_flags , and then make sure
* it is so documented in include / linux / fs . h and that all code follows
* the locking convention ! !
*/
void inode_set_flags ( struct inode * inode , unsigned int flags ,
unsigned int mask )
{
WARN_ON_ONCE ( flags & ~ mask ) ;
2019-03-06 02:41:52 +03:00
set_mask_bits ( & inode - > i_flags , mask , flags ) ;
2014-03-24 22:43:12 +04:00
}
EXPORT_SYMBOL ( inode_set_flags ) ;
2015-11-17 09:07:57 +03:00
void inode_nohighmem ( struct inode * inode )
{
mapping_set_gfp_mask ( inode - > i_mapping , GFP_USER ) ;
}
EXPORT_SYMBOL ( inode_nohighmem ) ;
2016-09-14 17:48:02 +03:00
2018-01-22 05:04:25 +03:00
/**
* timestamp_truncate - Truncate timespec to a granularity
* @ t : Timespec
* @ inode : inode being updated
*
* Truncate a timespec to the granularity supported by the fs
* containing the inode . Always rounds down . gran must
* not be 0 nor greater than a second ( NSEC_PER_SEC , or 10 ^ 9 ns ) .
*/
struct timespec64 timestamp_truncate ( struct timespec64 t , struct inode * inode )
{
struct super_block * sb = inode - > i_sb ;
unsigned int gran = sb - > s_time_gran ;
t . tv_sec = clamp ( t . tv_sec , sb - > s_time_min , sb - > s_time_max ) ;
if ( unlikely ( t . tv_sec = = sb - > s_time_max | | t . tv_sec = = sb - > s_time_min ) )
t . tv_nsec = 0 ;
/* Avoid division in the common cases 1 ns and 1 s. */
if ( gran = = 1 )
; /* nothing */
else if ( gran = = NSEC_PER_SEC )
t . tv_nsec = 0 ;
else if ( gran > 1 & & gran < NSEC_PER_SEC )
t . tv_nsec - = t . tv_nsec % gran ;
else
WARN ( 1 , " invalid file time granularity: %u " , gran ) ;
return t ;
}
EXPORT_SYMBOL ( timestamp_truncate ) ;
2016-09-14 17:48:02 +03:00
/**
* current_time - Return FS time
* @ inode : inode .
*
* Return the current time truncated to the time granularity supported by
* the fs .
*
* Note that inode and inode - > sb cannot be NULL .
* Otherwise , the function warns and returns time without truncation .
*/
vfs: change inode times to use struct timespec64
struct timespec is not y2038 safe. Transition vfs to use
y2038 safe struct timespec64 instead.
The change was made with the help of the following cocinelle
script. This catches about 80% of the changes.
All the header file and logic changes are included in the
first 5 rules. The rest are trivial substitutions.
I avoid changing any of the function signatures or any other
filesystem specific data structures to keep the patch simple
for review.
The script can be a little shorter by combining different cases.
But, this version was sufficient for my usecase.
virtual patch
@ depends on patch @
identifier now;
@@
- struct timespec
+ struct timespec64
current_time ( ... )
{
- struct timespec now = current_kernel_time();
+ struct timespec64 now = current_kernel_time64();
...
- return timespec_trunc(
+ return timespec64_trunc(
... );
}
@ depends on patch @
identifier xtime;
@@
struct \( iattr \| inode \| kstat \) {
...
- struct timespec xtime;
+ struct timespec64 xtime;
...
}
@ depends on patch @
identifier t;
@@
struct inode_operations {
...
int (*update_time) (...,
- struct timespec t,
+ struct timespec64 t,
...);
...
}
@ depends on patch @
identifier t;
identifier fn_update_time =~ "update_time$";
@@
fn_update_time (...,
- struct timespec *t,
+ struct timespec64 *t,
...) { ... }
@ depends on patch @
identifier t;
@@
lease_get_mtime( ... ,
- struct timespec *t
+ struct timespec64 *t
) { ... }
@te depends on patch forall@
identifier ts;
local idexpression struct inode *inode_node;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn_update_time =~ "update_time$";
identifier fn;
expression e, E3;
local idexpression struct inode *node1;
local idexpression struct inode *node2;
local idexpression struct iattr *attr1;
local idexpression struct iattr *attr2;
local idexpression struct iattr attr;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
@@
(
(
- struct timespec ts;
+ struct timespec64 ts;
|
- struct timespec ts = current_time(inode_node);
+ struct timespec64 ts = current_time(inode_node);
)
<+... when != ts
(
- timespec_equal(&inode_node->i_xtime, &ts)
+ timespec64_equal(&inode_node->i_xtime, &ts)
|
- timespec_equal(&ts, &inode_node->i_xtime)
+ timespec64_equal(&ts, &inode_node->i_xtime)
|
- timespec_compare(&inode_node->i_xtime, &ts)
+ timespec64_compare(&inode_node->i_xtime, &ts)
|
- timespec_compare(&ts, &inode_node->i_xtime)
+ timespec64_compare(&ts, &inode_node->i_xtime)
|
ts = current_time(e)
|
fn_update_time(..., &ts,...)
|
inode_node->i_xtime = ts
|
node1->i_xtime = ts
|
ts = inode_node->i_xtime
|
<+... attr1->ia_xtime ...+> = ts
|
ts = attr1->ia_xtime
|
ts.tv_sec
|
ts.tv_nsec
|
btrfs_set_stack_timespec_sec(..., ts.tv_sec)
|
btrfs_set_stack_timespec_nsec(..., ts.tv_nsec)
|
- ts = timespec64_to_timespec(
+ ts =
...
-)
|
- ts = ktime_to_timespec(
+ ts = ktime_to_timespec64(
...)
|
- ts = E3
+ ts = timespec_to_timespec64(E3)
|
- ktime_get_real_ts(&ts)
+ ktime_get_real_ts64(&ts)
|
fn(...,
- ts
+ timespec64_to_timespec(ts)
,...)
)
...+>
(
<... when != ts
- return ts;
+ return timespec64_to_timespec(ts);
...>
)
|
- timespec_equal(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_equal(&node1->i_xtime2, &node2->i_xtime2)
|
- timespec_equal(&node1->i_xtime1, &attr2->ia_xtime2)
+ timespec64_equal(&node1->i_xtime2, &attr2->ia_xtime2)
|
- timespec_compare(&node1->i_xtime1, &node2->i_xtime2)
+ timespec64_compare(&node1->i_xtime1, &node2->i_xtime2)
|
node1->i_xtime1 =
- timespec_trunc(attr1->ia_xtime1,
+ timespec64_trunc(attr1->ia_xtime1,
...)
|
- attr1->ia_xtime1 = timespec_trunc(attr2->ia_xtime2,
+ attr1->ia_xtime1 = timespec64_trunc(attr2->ia_xtime2,
...)
|
- ktime_get_real_ts(&attr1->ia_xtime1)
+ ktime_get_real_ts64(&attr1->ia_xtime1)
|
- ktime_get_real_ts(&attr.ia_xtime1)
+ ktime_get_real_ts64(&attr.ia_xtime1)
)
@ depends on patch @
struct inode *node;
struct iattr *attr;
identifier fn;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
expression e;
@@
(
- fn(node->i_xtime);
+ fn(timespec64_to_timespec(node->i_xtime));
|
fn(...,
- node->i_xtime);
+ timespec64_to_timespec(node->i_xtime));
|
- e = fn(attr->ia_xtime);
+ e = fn(timespec64_to_timespec(attr->ia_xtime));
)
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
identifier i_xtime =~ "^i_[acm]time$";
identifier ia_xtime =~ "^ia_[acm]time$";
identifier fn;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
fn (...,
- &attr->ia_xtime,
+ &ts,
...);
)
...+>
}
@ depends on patch forall @
struct inode *node;
struct iattr *attr;
struct kstat *stat;
identifier ia_xtime =~ "^ia_[acm]time$";
identifier i_xtime =~ "^i_[acm]time$";
identifier xtime =~ "^[acm]time$";
identifier fn, ret;
@@
{
+ struct timespec ts;
<+...
(
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(node->i_xtime);
ret = fn (...,
- &node->i_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime,
+ &ts,
...);
|
+ ts = timespec64_to_timespec(attr->ia_xtime);
ret = fn (...,
- &attr->ia_xtime);
+ &ts);
|
+ ts = timespec64_to_timespec(stat->xtime);
ret = fn (...,
- &stat->xtime);
+ &ts);
)
...+>
}
@ depends on patch @
struct inode *node;
struct inode *node2;
identifier i_xtime1 =~ "^i_[acm]time$";
identifier i_xtime2 =~ "^i_[acm]time$";
identifier i_xtime3 =~ "^i_[acm]time$";
struct iattr *attrp;
struct iattr *attrp2;
struct iattr attr ;
identifier ia_xtime1 =~ "^ia_[acm]time$";
identifier ia_xtime2 =~ "^ia_[acm]time$";
struct kstat *stat;
struct kstat stat1;
struct timespec64 ts;
identifier xtime =~ "^[acmb]time$";
expression e;
@@
(
( node->i_xtime2 \| attrp->ia_xtime2 \| attr.ia_xtime2 \) = node->i_xtime1 ;
|
node->i_xtime2 = \( node2->i_xtime1 \| timespec64_trunc(...) \);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
node->i_xtime1 = node->i_xtime3 = \(ts \| current_time(...) \);
|
stat->xtime = node2->i_xtime1;
|
stat1.xtime = node2->i_xtime1;
|
( node->i_xtime2 \| attrp->ia_xtime2 \) = attrp->ia_xtime1 ;
|
( attrp->ia_xtime1 \| attr.ia_xtime1 \) = attrp2->ia_xtime2;
|
- e = node->i_xtime1;
+ e = timespec64_to_timespec( node->i_xtime1 );
|
- e = attrp->ia_xtime1;
+ e = timespec64_to_timespec( attrp->ia_xtime1 );
|
node->i_xtime1 = current_time(...);
|
node->i_xtime2 = node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
node->i_xtime1 = node->i_xtime3 =
- e;
+ timespec_to_timespec64(e);
|
- node->i_xtime1 = e;
+ node->i_xtime1 = timespec_to_timespec64(e);
)
Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
Cc: <anton@tuxera.com>
Cc: <balbi@kernel.org>
Cc: <bfields@fieldses.org>
Cc: <darrick.wong@oracle.com>
Cc: <dhowells@redhat.com>
Cc: <dsterba@suse.com>
Cc: <dwmw2@infradead.org>
Cc: <hch@lst.de>
Cc: <hirofumi@mail.parknet.co.jp>
Cc: <hubcap@omnibond.com>
Cc: <jack@suse.com>
Cc: <jaegeuk@kernel.org>
Cc: <jaharkes@cs.cmu.edu>
Cc: <jslaby@suse.com>
Cc: <keescook@chromium.org>
Cc: <mark@fasheh.com>
Cc: <miklos@szeredi.hu>
Cc: <nico@linaro.org>
Cc: <reiserfs-devel@vger.kernel.org>
Cc: <richard@nod.at>
Cc: <sage@redhat.com>
Cc: <sfrench@samba.org>
Cc: <swhiteho@redhat.com>
Cc: <tj@kernel.org>
Cc: <trond.myklebust@primarydata.com>
Cc: <tytso@mit.edu>
Cc: <viro@zeniv.linux.org.uk>
2018-05-09 05:36:02 +03:00
struct timespec64 current_time ( struct inode * inode )
2016-09-14 17:48:02 +03:00
{
vfs: replace current_kernel_time64 with ktime equivalent
current_time is the last remaining caller of current_kernel_time64(),
which is a wrapper around ktime_get_coarse_real_ts64(). This calls the
latter directly for consistency with the rest of the kernel that is moving
to the ktime_get_ family of time accessors, as now documented in
Documentation/core-api/timekeeping.rst.
An open questions is whether we may want to actually call the more
accurate ktime_get_real_ts64() for file systems that save high-resolution
timestamps in their on-disk format. This would add a small overhead to
each update of the inode stamps but lead to inode timestamps to actually
have a usable resolution better than one jiffy (1 to 10 milliseconds
normally). Experiments on a variety of hardware platforms show a typical
time of around 100 CPU cycles to read the cycle counter and calculate the
accurate time from that. On old platforms without a cycle counter, this
can be signiciantly higher, up to several microseconds to access a
hardware clock, but those have become very rare by now.
I traced the original addition of the current_kernel_time() call to set
the nanosecond fields back to linux-2.5.48, where Andi Kleen added a patch
with subject "nanosecond stat timefields". Andi explains that the
motivation was to introduce as little overhead as possible back then. At
this time, reading the clock hardware was also more expensive when most
architectures did not have a cycle counter.
One side effect of having more accurate inode timestamp would be having to
write out the inode every time that mtime/ctime/atime get touched on most
systems, whereas many file systems today only write it when the timestamps
have changed, i.e. at most once per jiffy unless something else changes
as well. That change would certainly be noticed in some workloads, which
is enough reason to not do it without a good reason, regardless of the
cost of reading the time.
One thing we could still consider however would be to round the timestamps
from current_time() to multiples of NSEC_PER_JIFFY, e.g. full
milliseconds rather than having six or seven meaningless but confusing
digits at the end of the timestamp.
Link: http://lkml.kernel.org/r/20180726130820.4174359-1-arnd@arndb.de
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
2018-12-05 03:14:27 +03:00
struct timespec64 now ;
ktime_get_coarse_real_ts64 ( & now ) ;
2016-09-14 17:48:02 +03:00
if ( unlikely ( ! inode - > i_sb ) ) {
WARN ( 1 , " current_time() called with uninitialized super_block in the inode " ) ;
return now ;
}
2018-01-22 05:04:25 +03:00
return timestamp_truncate ( now , inode ) ;
2016-09-14 17:48:02 +03:00
}
EXPORT_SYMBOL ( current_time ) ;
2022-07-14 09:11:25 +03:00
2022-10-17 18:06:34 +03:00
/**
* in_group_or_capable - check whether caller is CAP_FSETID privileged
2023-01-13 14:49:27 +03:00
* @ idmap : idmap of the mount @ inode was found from
2022-10-17 18:06:34 +03:00
* @ inode : inode to check
* @ vfsgid : the new / current vfsgid of @ inode
*
* Check wether @ vfsgid is in the caller ' s group list or if the caller is
* privileged with CAP_FSETID over @ inode . This can be used to determine
* whether the setgid bit can be kept or must be dropped .
*
* Return : true if the caller is sufficiently privileged , false if not .
*/
2023-01-13 14:49:27 +03:00
bool in_group_or_capable ( struct mnt_idmap * idmap ,
2022-10-17 18:06:34 +03:00
const struct inode * inode , vfsgid_t vfsgid )
{
if ( vfsgid_in_group_p ( vfsgid ) )
return true ;
2023-01-13 14:49:27 +03:00
if ( capable_wrt_inode_uidgid ( idmap , inode , CAP_FSETID ) )
2022-10-17 18:06:34 +03:00
return true ;
return false ;
}
2022-07-14 09:11:25 +03:00
/**
* mode_strip_sgid - handle the sgid bit for non - directories
2023-01-13 14:49:27 +03:00
* @ idmap : idmap of the mount the inode was created from
2022-07-14 09:11:25 +03:00
* @ dir : parent directory inode
* @ mode : mode of the file to be created in @ dir
*
* If the @ mode of the new file has both the S_ISGID and S_IXGRP bit
* raised and @ dir has the S_ISGID bit raised ensure that the caller is
* either in the group of the parent directory or they have CAP_FSETID
* in their user namespace and are privileged over the parent directory .
* In all other cases , strip the S_ISGID bit from @ mode .
*
* Return : the new mode to use for the file
*/
2023-01-13 14:49:27 +03:00
umode_t mode_strip_sgid ( struct mnt_idmap * idmap ,
2022-07-14 09:11:25 +03:00
const struct inode * dir , umode_t mode )
{
if ( ( mode & ( S_ISGID | S_IXGRP ) ) ! = ( S_ISGID | S_IXGRP ) )
return mode ;
if ( S_ISDIR ( mode ) | | ! dir | | ! ( dir - > i_mode & S_ISGID ) )
return mode ;
2023-01-13 14:49:30 +03:00
if ( in_group_or_capable ( idmap , dir , i_gid_into_vfsgid ( idmap , dir ) ) )
2022-07-14 09:11:25 +03:00
return mode ;
return mode & ~ S_ISGID ;
}
EXPORT_SYMBOL ( mode_strip_sgid ) ;