linux-stable/fs/xfs/xfs_buf.c

2244 lines
53 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (c) 2000-2006 Silicon Graphics, Inc.
* All Rights Reserved.
*/
#include "xfs.h"
#include <linux/stddef.h>
#include <linux/errno.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 08:04:11 +00:00
#include <linux/gfp.h>
#include <linux/pagemap.h>
#include <linux/init.h>
#include <linux/vmalloc.h>
#include <linux/bio.h>
#include <linux/sysctl.h>
#include <linux/proc_fs.h>
#include <linux/workqueue.h>
#include <linux/percpu.h>
#include <linux/blkdev.h>
#include <linux/hash.h>
#include <linux/kthread.h>
#include <linux/migrate.h>
#include <linux/backing-dev.h>
#include <linux/freezer.h>
#include "xfs_format.h"
#include "xfs_log_format.h"
#include "xfs_trans_resv.h"
#include "xfs_sb.h"
#include "xfs_mount.h"
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
#include "xfs_trace.h"
#include "xfs_log.h"
#include "xfs_errortag.h"
#include "xfs_error.h"
static kmem_zone_t *xfs_buf_zone;
#define xb_to_gfp(flags) \
((((flags) & XBF_READ_AHEAD) ? __GFP_NORETRY : GFP_NOFS) | __GFP_NOWARN)
xfs: fix use-after-free race in xfs_buf_rele When looking at a 4.18 based KASAN use after free report, I noticed that racing xfs_buf_rele() may race on dropping the last reference to the buffer and taking the buffer lock. This was the symptom displayed by the KASAN report, but the actual issue that was reported had already been fixed in 4.19-rc1 by commit e339dd8d8b04 ("xfs: use sync buffer I/O for sync delwri queue submission"). Despite this, I think there is still an issue with xfs_buf_rele() in this code: release = atomic_dec_and_lock(&bp->b_hold, &pag->pag_buf_lock); spin_lock(&bp->b_lock); if (!release) { ..... If two threads race on the b_lock after both dropping a reference and one getting dropping the last reference so release = true, we end up with: CPU 0 CPU 1 atomic_dec_and_lock() atomic_dec_and_lock() spin_lock(&bp->b_lock) spin_lock(&bp->b_lock) <spins> <release = true bp->b_lru_ref = 0> <remove from lists> freebuf = true spin_unlock(&bp->b_lock) xfs_buf_free(bp) <gets lock, reading and writing freed memory> <accesses freed memory> spin_unlock(&bp->b_lock) <reads/writes freed memory> IOWs, we can't safely take bp->b_lock after dropping the hold reference because the buffer may go away at any time after we drop that reference. However, this can be fixed simply by taking the bp->b_lock before we drop the reference. It is safe to nest the pag_buf_lock inside bp->b_lock as the pag_buf_lock is only used to serialise against lookup in xfs_buf_find() and no other locks are held over or under the pag_buf_lock there. Make this clear by documenting the buffer lock orders at the top of the file. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com Signed-off-by: Dave Chinner <david@fromorbit.com>
2018-10-18 06:21:29 +00:00
/*
* Locking orders
*
* xfs_buf_ioacct_inc:
* xfs_buf_ioacct_dec:
* b_sema (caller holds)
* b_lock
*
* xfs_buf_stale:
* b_sema (caller holds)
* b_lock
* lru_lock
*
* xfs_buf_rele:
* b_lock
* pag_buf_lock
* lru_lock
*
* xfs_buftarg_wait_rele
* lru_lock
* b_lock (trylock due to inversion)
*
* xfs_buftarg_isolate
* lru_lock
* b_lock (trylock due to inversion)
*/
static inline int
xfs_buf_is_vmapped(
struct xfs_buf *bp)
{
/*
* Return true if the buffer is vmapped.
*
* b_addr is null if the buffer is not mapped, but the code is clever
* enough to know it doesn't have to map a single page, so the check has
* to be both for b_addr and bp->b_page_count > 1.
*/
return bp->b_addr && bp->b_page_count > 1;
}
static inline int
xfs_buf_vmap_len(
struct xfs_buf *bp)
{
return (bp->b_page_count * PAGE_SIZE) - bp->b_offset;
}
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
/*
* Bump the I/O in flight count on the buftarg if we haven't yet done so for
* this buffer. The count is incremented once per buffer (per hold cycle)
* because the corresponding decrement is deferred to buffer release. Buffers
* can undergo I/O multiple times in a hold-release cycle and per buffer I/O
* tracking adds unnecessary overhead. This is used for sychronization purposes
* with unmount (see xfs_wait_buftarg()), so all we really need is a count of
* in-flight buffers.
*
* Buffers that are never released (e.g., superblock, iclog buffers) must set
* the XBF_NO_IOACCT flag before I/O submission. Otherwise, the buftarg count
* never reaches zero and unmount hangs indefinitely.
*/
static inline void
xfs_buf_ioacct_inc(
struct xfs_buf *bp)
{
xfs: use ->b_state to fix buffer I/O accounting release race We've had user reports of unmount hangs in xfs_wait_buftarg() that analysis shows is due to btp->bt_io_count == -1. bt_io_count represents the count of in-flight asynchronous buffers and thus should always be >= 0. xfs_wait_buftarg() waits for this value to stabilize to zero in order to ensure that all untracked (with respect to the lru) buffers have completed I/O processing before unmount proceeds to tear down in-core data structures. The value of -1 implies an I/O accounting decrement race. Indeed, the fact that xfs_buf_ioacct_dec() is called from xfs_buf_rele() (where the buffer lock is no longer held) means that bp->b_flags can be updated from an unsafe context. While a user-level reproducer is currently not available, some intrusive hacks to run racing buffer lookups/ioacct/releases from multiple threads was used to successfully manufacture this problem. Existing callers do not expect to acquire the buffer lock from xfs_buf_rele(). Therefore, we can not safely update ->b_flags from this context. It turns out that we already have separate buffer state bits and associated serialization for dealing with buffer LRU state in the form of ->b_state and ->b_lock. Therefore, replace the _XBF_IN_FLIGHT flag with a ->b_state variant, update the I/O accounting wrappers appropriately and make sure they are used with the correct locking. This ensures that buffer in-flight state can be modified at buffer release time without racing with modifications from a buffer lock holder. Fixes: 9c7504aa72b6 ("xfs: track and serialize in-flight async buffers against unmount") Cc: <stable@vger.kernel.org> # v4.8+ Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Nikolay Borisov <nborisov@suse.com> Tested-by: Libor Pechacek <lpechacek@suse.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-05-31 15:22:52 +00:00
if (bp->b_flags & XBF_NO_IOACCT)
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
return;
ASSERT(bp->b_flags & XBF_ASYNC);
xfs: use ->b_state to fix buffer I/O accounting release race We've had user reports of unmount hangs in xfs_wait_buftarg() that analysis shows is due to btp->bt_io_count == -1. bt_io_count represents the count of in-flight asynchronous buffers and thus should always be >= 0. xfs_wait_buftarg() waits for this value to stabilize to zero in order to ensure that all untracked (with respect to the lru) buffers have completed I/O processing before unmount proceeds to tear down in-core data structures. The value of -1 implies an I/O accounting decrement race. Indeed, the fact that xfs_buf_ioacct_dec() is called from xfs_buf_rele() (where the buffer lock is no longer held) means that bp->b_flags can be updated from an unsafe context. While a user-level reproducer is currently not available, some intrusive hacks to run racing buffer lookups/ioacct/releases from multiple threads was used to successfully manufacture this problem. Existing callers do not expect to acquire the buffer lock from xfs_buf_rele(). Therefore, we can not safely update ->b_flags from this context. It turns out that we already have separate buffer state bits and associated serialization for dealing with buffer LRU state in the form of ->b_state and ->b_lock. Therefore, replace the _XBF_IN_FLIGHT flag with a ->b_state variant, update the I/O accounting wrappers appropriately and make sure they are used with the correct locking. This ensures that buffer in-flight state can be modified at buffer release time without racing with modifications from a buffer lock holder. Fixes: 9c7504aa72b6 ("xfs: track and serialize in-flight async buffers against unmount") Cc: <stable@vger.kernel.org> # v4.8+ Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Nikolay Borisov <nborisov@suse.com> Tested-by: Libor Pechacek <lpechacek@suse.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-05-31 15:22:52 +00:00
spin_lock(&bp->b_lock);
if (!(bp->b_state & XFS_BSTATE_IN_FLIGHT)) {
bp->b_state |= XFS_BSTATE_IN_FLIGHT;
percpu_counter_inc(&bp->b_target->bt_io_count);
}
spin_unlock(&bp->b_lock);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
}
/*
* Clear the in-flight state on a buffer about to be released to the LRU or
* freed and unaccount from the buftarg.
*/
static inline void
xfs: use ->b_state to fix buffer I/O accounting release race We've had user reports of unmount hangs in xfs_wait_buftarg() that analysis shows is due to btp->bt_io_count == -1. bt_io_count represents the count of in-flight asynchronous buffers and thus should always be >= 0. xfs_wait_buftarg() waits for this value to stabilize to zero in order to ensure that all untracked (with respect to the lru) buffers have completed I/O processing before unmount proceeds to tear down in-core data structures. The value of -1 implies an I/O accounting decrement race. Indeed, the fact that xfs_buf_ioacct_dec() is called from xfs_buf_rele() (where the buffer lock is no longer held) means that bp->b_flags can be updated from an unsafe context. While a user-level reproducer is currently not available, some intrusive hacks to run racing buffer lookups/ioacct/releases from multiple threads was used to successfully manufacture this problem. Existing callers do not expect to acquire the buffer lock from xfs_buf_rele(). Therefore, we can not safely update ->b_flags from this context. It turns out that we already have separate buffer state bits and associated serialization for dealing with buffer LRU state in the form of ->b_state and ->b_lock. Therefore, replace the _XBF_IN_FLIGHT flag with a ->b_state variant, update the I/O accounting wrappers appropriately and make sure they are used with the correct locking. This ensures that buffer in-flight state can be modified at buffer release time without racing with modifications from a buffer lock holder. Fixes: 9c7504aa72b6 ("xfs: track and serialize in-flight async buffers against unmount") Cc: <stable@vger.kernel.org> # v4.8+ Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Nikolay Borisov <nborisov@suse.com> Tested-by: Libor Pechacek <lpechacek@suse.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-05-31 15:22:52 +00:00
__xfs_buf_ioacct_dec(
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
struct xfs_buf *bp)
{
lockdep_assert_held(&bp->b_lock);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
xfs: use ->b_state to fix buffer I/O accounting release race We've had user reports of unmount hangs in xfs_wait_buftarg() that analysis shows is due to btp->bt_io_count == -1. bt_io_count represents the count of in-flight asynchronous buffers and thus should always be >= 0. xfs_wait_buftarg() waits for this value to stabilize to zero in order to ensure that all untracked (with respect to the lru) buffers have completed I/O processing before unmount proceeds to tear down in-core data structures. The value of -1 implies an I/O accounting decrement race. Indeed, the fact that xfs_buf_ioacct_dec() is called from xfs_buf_rele() (where the buffer lock is no longer held) means that bp->b_flags can be updated from an unsafe context. While a user-level reproducer is currently not available, some intrusive hacks to run racing buffer lookups/ioacct/releases from multiple threads was used to successfully manufacture this problem. Existing callers do not expect to acquire the buffer lock from xfs_buf_rele(). Therefore, we can not safely update ->b_flags from this context. It turns out that we already have separate buffer state bits and associated serialization for dealing with buffer LRU state in the form of ->b_state and ->b_lock. Therefore, replace the _XBF_IN_FLIGHT flag with a ->b_state variant, update the I/O accounting wrappers appropriately and make sure they are used with the correct locking. This ensures that buffer in-flight state can be modified at buffer release time without racing with modifications from a buffer lock holder. Fixes: 9c7504aa72b6 ("xfs: track and serialize in-flight async buffers against unmount") Cc: <stable@vger.kernel.org> # v4.8+ Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Nikolay Borisov <nborisov@suse.com> Tested-by: Libor Pechacek <lpechacek@suse.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-05-31 15:22:52 +00:00
if (bp->b_state & XFS_BSTATE_IN_FLIGHT) {
bp->b_state &= ~XFS_BSTATE_IN_FLIGHT;
percpu_counter_dec(&bp->b_target->bt_io_count);
}
}
static inline void
xfs_buf_ioacct_dec(
struct xfs_buf *bp)
{
spin_lock(&bp->b_lock);
__xfs_buf_ioacct_dec(bp);
spin_unlock(&bp->b_lock);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
}
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
/*
* When we mark a buffer stale, we remove the buffer from the LRU and clear the
* b_lru_ref count so that the buffer is freed immediately when the buffer
* reference count falls to zero. If the buffer is already on the LRU, we need
* to remove the reference that LRU holds on the buffer.
*
* This prevents build-up of stale buffers on the LRU.
*/
void
xfs_buf_stale(
struct xfs_buf *bp)
{
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
ASSERT(xfs_buf_islocked(bp));
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
bp->b_flags |= XBF_STALE;
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
/*
* Clear the delwri status so that a delwri queue walker will not
* flush this buffer to disk now that it is stale. The delwri queue has
* a reference to the buffer, so this is safe to do.
*/
bp->b_flags &= ~_XBF_DELWRI_Q;
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
/*
* Once the buffer is marked stale and unlocked, a subsequent lookup
* could reset b_flags. There is no guarantee that the buffer is
* unaccounted (released to LRU) before that occurs. Drop in-flight
* status now to preserve accounting consistency.
*/
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
spin_lock(&bp->b_lock);
xfs: use ->b_state to fix buffer I/O accounting release race We've had user reports of unmount hangs in xfs_wait_buftarg() that analysis shows is due to btp->bt_io_count == -1. bt_io_count represents the count of in-flight asynchronous buffers and thus should always be >= 0. xfs_wait_buftarg() waits for this value to stabilize to zero in order to ensure that all untracked (with respect to the lru) buffers have completed I/O processing before unmount proceeds to tear down in-core data structures. The value of -1 implies an I/O accounting decrement race. Indeed, the fact that xfs_buf_ioacct_dec() is called from xfs_buf_rele() (where the buffer lock is no longer held) means that bp->b_flags can be updated from an unsafe context. While a user-level reproducer is currently not available, some intrusive hacks to run racing buffer lookups/ioacct/releases from multiple threads was used to successfully manufacture this problem. Existing callers do not expect to acquire the buffer lock from xfs_buf_rele(). Therefore, we can not safely update ->b_flags from this context. It turns out that we already have separate buffer state bits and associated serialization for dealing with buffer LRU state in the form of ->b_state and ->b_lock. Therefore, replace the _XBF_IN_FLIGHT flag with a ->b_state variant, update the I/O accounting wrappers appropriately and make sure they are used with the correct locking. This ensures that buffer in-flight state can be modified at buffer release time without racing with modifications from a buffer lock holder. Fixes: 9c7504aa72b6 ("xfs: track and serialize in-flight async buffers against unmount") Cc: <stable@vger.kernel.org> # v4.8+ Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Nikolay Borisov <nborisov@suse.com> Tested-by: Libor Pechacek <lpechacek@suse.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-05-31 15:22:52 +00:00
__xfs_buf_ioacct_dec(bp);
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
atomic_set(&bp->b_lru_ref, 0);
if (!(bp->b_state & XFS_BSTATE_DISPOSE) &&
(list_lru_del(&bp->b_target->bt_lru, &bp->b_lru)))
atomic_dec(&bp->b_hold);
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
ASSERT(atomic_read(&bp->b_hold) >= 1);
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
spin_unlock(&bp->b_lock);
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
}
static int
xfs_buf_get_maps(
struct xfs_buf *bp,
int map_count)
{
ASSERT(bp->b_maps == NULL);
bp->b_map_count = map_count;
if (map_count == 1) {
bp->b_maps = &bp->__b_map;
return 0;
}
bp->b_maps = kmem_zalloc(map_count * sizeof(struct xfs_buf_map),
KM_NOFS);
if (!bp->b_maps)
return -ENOMEM;
return 0;
}
/*
* Frees b_pages if it was allocated.
*/
static void
xfs_buf_free_maps(
struct xfs_buf *bp)
{
if (bp->b_maps != &bp->__b_map) {
kmem_free(bp->b_maps);
bp->b_maps = NULL;
}
}
struct xfs_buf *
_xfs_buf_alloc(
struct xfs_buftarg *target,
struct xfs_buf_map *map,
int nmaps,
xfs_buf_flags_t flags)
{
struct xfs_buf *bp;
int error;
int i;
bp = kmem_zone_zalloc(xfs_buf_zone, KM_NOFS);
if (unlikely(!bp))
return NULL;
/*
* We don't want certain flags to appear in b_flags unless they are
* specifically set by later operations on the buffer.
*/
flags &= ~(XBF_UNMAPPED | XBF_TRYLOCK | XBF_ASYNC | XBF_READ_AHEAD);
atomic_set(&bp->b_hold, 1);
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
atomic_set(&bp->b_lru_ref, 1);
init_completion(&bp->b_iowait);
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
INIT_LIST_HEAD(&bp->b_lru);
INIT_LIST_HEAD(&bp->b_list);
INIT_LIST_HEAD(&bp->b_li_list);
sema_init(&bp->b_sema, 0); /* held, no waiters */
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
spin_lock_init(&bp->b_lock);
bp->b_target = target;
bp->b_flags = flags;
/*
* Set length and io_length to the same value initially.
* I/O routines should use io_length, which will be the same in
* most cases but may be reset (e.g. XFS recovery).
*/
error = xfs_buf_get_maps(bp, nmaps);
if (error) {
kmem_zone_free(xfs_buf_zone, bp);
return NULL;
}
bp->b_bn = map[0].bm_bn;
bp->b_length = 0;
for (i = 0; i < nmaps; i++) {
bp->b_maps[i].bm_bn = map[i].bm_bn;
bp->b_maps[i].bm_len = map[i].bm_len;
bp->b_length += map[i].bm_len;
}
bp->b_io_length = bp->b_length;
atomic_set(&bp->b_pin_count, 0);
init_waitqueue_head(&bp->b_waiters);
XFS_STATS_INC(target->bt_mount, xb_create);
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_init(bp, _RET_IP_);
return bp;
}
/*
* Allocate a page array capable of holding a specified number
* of pages, and point the page buf at it.
*/
STATIC int
_xfs_buf_get_pages(
xfs_buf_t *bp,
int page_count)
{
/* Make sure that we have a page list */
if (bp->b_pages == NULL) {
bp->b_page_count = page_count;
if (page_count <= XB_PAGES) {
bp->b_pages = bp->b_page_array;
} else {
bp->b_pages = kmem_alloc(sizeof(struct page *) *
page_count, KM_NOFS);
if (bp->b_pages == NULL)
return -ENOMEM;
}
memset(bp->b_pages, 0, sizeof(struct page *) * page_count);
}
return 0;
}
/*
* Frees b_pages if it was allocated.
*/
STATIC void
_xfs_buf_free_pages(
xfs_buf_t *bp)
{
if (bp->b_pages != bp->b_page_array) {
kmem_free(bp->b_pages);
bp->b_pages = NULL;
}
}
/*
* Releases the specified buffer.
*
* The modification state of any associated pages is left unchanged.
* The buffer must not be on any hash - use xfs_buf_rele instead for
* hashed and refcounted buffers
*/
void
xfs_buf_free(
xfs_buf_t *bp)
{
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_free(bp, _RET_IP_);
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
ASSERT(list_empty(&bp->b_lru));
2011-03-25 22:16:45 +00:00
if (bp->b_flags & _XBF_PAGES) {
uint i;
if (xfs_buf_is_vmapped(bp))
vm_unmap_ram(bp->b_addr - bp->b_offset,
bp->b_page_count);
for (i = 0; i < bp->b_page_count; i++) {
struct page *page = bp->b_pages[i];
2011-03-25 22:16:45 +00:00
__free_page(page);
}
2011-03-25 22:16:45 +00:00
} else if (bp->b_flags & _XBF_KMEM)
kmem_free(bp->b_addr);
_xfs_buf_free_pages(bp);
xfs_buf_free_maps(bp);
kmem_zone_free(xfs_buf_zone, bp);
}
/*
2011-03-25 22:16:45 +00:00
* Allocates all the pages for buffer in question and builds it's page list.
*/
STATIC int
2011-03-25 22:16:45 +00:00
xfs_buf_allocate_memory(
xfs_buf_t *bp,
uint flags)
{
size_t size;
size_t nbytes, offset;
gfp_t gfp_mask = xb_to_gfp(flags);
unsigned short page_count, i;
xfs_off_t start, end;
int error;
2011-03-25 22:16:45 +00:00
/*
* for buffers that are contained within a single page, just allocate
* the memory from the heap - there's no need for the complexity of
* page arrays to keep allocation down to order 0.
*/
size = BBTOB(bp->b_length);
if (size < PAGE_SIZE) {
bp->b_addr = kmem_alloc(size, KM_NOFS);
2011-03-25 22:16:45 +00:00
if (!bp->b_addr) {
/* low memory - use alloc_page loop instead */
goto use_alloc_page;
}
if (((unsigned long)(bp->b_addr + size - 1) & PAGE_MASK) !=
2011-03-25 22:16:45 +00:00
((unsigned long)bp->b_addr & PAGE_MASK)) {
/* b_addr spans two pages - use alloc_page instead */
kmem_free(bp->b_addr);
bp->b_addr = NULL;
goto use_alloc_page;
}
bp->b_offset = offset_in_page(bp->b_addr);
bp->b_pages = bp->b_page_array;
bp->b_pages[0] = virt_to_page(bp->b_addr);
bp->b_page_count = 1;
bp->b_flags |= _XBF_KMEM;
2011-03-25 22:16:45 +00:00
return 0;
}
use_alloc_page:
start = BBTOB(bp->b_maps[0].bm_bn) >> PAGE_SHIFT;
end = (BBTOB(bp->b_maps[0].bm_bn + bp->b_length) + PAGE_SIZE - 1)
>> PAGE_SHIFT;
page_count = end - start;
error = _xfs_buf_get_pages(bp, page_count);
if (unlikely(error))
return error;
offset = bp->b_offset;
2011-03-25 22:16:45 +00:00
bp->b_flags |= _XBF_PAGES;
for (i = 0; i < bp->b_page_count; i++) {
struct page *page;
uint retries = 0;
2011-03-25 22:16:45 +00:00
retry:
page = alloc_page(gfp_mask);
if (unlikely(page == NULL)) {
if (flags & XBF_READ_AHEAD) {
bp->b_page_count = i;
error = -ENOMEM;
2011-03-25 22:16:45 +00:00
goto out_free_pages;
}
/*
* This could deadlock.
*
* But until all the XFS lowlevel code is revamped to
* handle buffer allocation failures we can't do much.
*/
if (!(++retries % 100))
xfs_err(NULL,
"%s(%u) possible memory allocation deadlock in %s (mode:0x%x)",
current->comm, current->pid,
__func__, gfp_mask);
XFS_STATS_INC(bp->b_target->bt_mount, xb_page_retries);
congestion_wait(BLK_RW_ASYNC, HZ/50);
goto retry;
}
XFS_STATS_INC(bp->b_target->bt_mount, xb_page_found);
2011-03-25 22:16:45 +00:00
nbytes = min_t(size_t, size, PAGE_SIZE - offset);
size -= nbytes;
bp->b_pages[i] = page;
offset = 0;
}
2011-03-25 22:16:45 +00:00
return 0;
2011-03-25 22:16:45 +00:00
out_free_pages:
for (i = 0; i < bp->b_page_count; i++)
__free_page(bp->b_pages[i]);
bp->b_flags &= ~_XBF_PAGES;
return error;
}
/*
* Map buffer into kernel address-space if necessary.
*/
STATIC int
_xfs_buf_map_pages(
xfs_buf_t *bp,
uint flags)
{
2011-03-25 22:16:45 +00:00
ASSERT(bp->b_flags & _XBF_PAGES);
if (bp->b_page_count == 1) {
2011-03-25 22:16:45 +00:00
/* A single page buffer is always mappable */
bp->b_addr = page_address(bp->b_pages[0]) + bp->b_offset;
} else if (flags & XBF_UNMAPPED) {
bp->b_addr = NULL;
} else {
int retried = 0;
unsigned nofs_flag;
xfs: use NOIO contexts for vm_map_ram When we map pages in the buffer cache, we can do so in GFP_NOFS contexts. However, the vmap interfaces do not provide any method of communicating this information to memory reclaim, and hence we get lockdep complaining about it regularly and occassionally see hangs that may be vmap related reclaim deadlocks. We can also see these same problems from anywhere where we use vmalloc for a large buffer (e.g. attribute code) inside a transaction context. A typical lockdep report shows up as a reclaim state warning like so: [14046.101458] ================================= [14046.102850] [ INFO: inconsistent lock state ] [14046.102850] 3.14.0-rc4+ #2 Not tainted [14046.102850] --------------------------------- [14046.102850] inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage. [14046.102850] kswapd0/14 [HC0[0]:SC0[0]:HE1:SE1] takes: [14046.102850] (&xfs_dir_ilock_class){++++?+}, at: [<791a04bb>] xfs_ilock+0xff/0x16a [14046.102850] {RECLAIM_FS-ON-W} state was registered at: [14046.102850] [<7904cdb1>] mark_held_locks+0x81/0xe7 [14046.102850] [<7904d390>] lockdep_trace_alloc+0x5c/0xb4 [14046.102850] [<790c2c28>] kmem_cache_alloc_trace+0x2b/0x11e [14046.102850] [<790ba7f4>] vm_map_ram+0x119/0x3e6 [14046.102850] [<7914e124>] _xfs_buf_map_pages+0x5b/0xcf [14046.102850] [<7914ed74>] xfs_buf_get_map+0x67/0x13f [14046.102850] [<7917506f>] xfs_attr_rmtval_set+0x396/0x4d5 [14046.102850] [<7916e8bb>] xfs_attr_leaf_addname+0x18f/0x37d [14046.102850] [<7916ed9e>] xfs_attr_set_int+0x2f5/0x3e8 [14046.102850] [<7916eefc>] xfs_attr_set+0x6b/0x74 [14046.102850] [<79168355>] xfs_xattr_set+0x61/0x81 [14046.102850] [<790e5b10>] generic_setxattr+0x59/0x68 [14046.102850] [<790e4c06>] __vfs_setxattr_noperm+0x58/0xce [14046.102850] [<790e4d0a>] vfs_setxattr+0x8e/0x92 [14046.102850] [<790e4ddd>] setxattr+0xcf/0x159 [14046.102850] [<790e5423>] SyS_lsetxattr+0x88/0xbb [14046.102850] [<79268438>] sysenter_do_call+0x12/0x36 Now, we can't completely remove these traces - mainly because vm_map_ram() will do GFP_KERNEL allocation and that generates the above warning before we get into the reclaim code, but we can turn them all into false positive warnings. To do that, use the method that DM and other IO context code uses to avoid this problem: there is a process flag to tell memory reclaim not to do IO that we can set appropriately. That prevents GFP_KERNEL context reclaim being done from deep inside the vmalloc code in places we can't directly pass a GFP_NOFS context to. That interface has a pair of wrapper functions: memalloc_noio_save() and memalloc_noio_restore(). Adding them around vm_map_ram and the vzalloc call in kmem_alloc_large() will prevent deadlocks and most lockdep reports for this issue. Also, convert the vzalloc() call in kmem_alloc_large() to use __vmalloc() so that we can pass the correct gfp context to the data page allocation routine inside __vmalloc() so that it is clear that GFP_NOFS context is important to this vmalloc call. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Dave Chinner <david@fromorbit.com>
2014-03-07 05:19:14 +00:00
/*
* vm_map_ram() will allocate auxillary structures (e.g.
* pagetables) with GFP_KERNEL, yet we are likely to be under
* GFP_NOFS context here. Hence we need to tell memory reclaim
* that we are in such a context via PF_MEMALLOC_NOFS to prevent
xfs: use NOIO contexts for vm_map_ram When we map pages in the buffer cache, we can do so in GFP_NOFS contexts. However, the vmap interfaces do not provide any method of communicating this information to memory reclaim, and hence we get lockdep complaining about it regularly and occassionally see hangs that may be vmap related reclaim deadlocks. We can also see these same problems from anywhere where we use vmalloc for a large buffer (e.g. attribute code) inside a transaction context. A typical lockdep report shows up as a reclaim state warning like so: [14046.101458] ================================= [14046.102850] [ INFO: inconsistent lock state ] [14046.102850] 3.14.0-rc4+ #2 Not tainted [14046.102850] --------------------------------- [14046.102850] inconsistent {RECLAIM_FS-ON-W} -> {IN-RECLAIM_FS-W} usage. [14046.102850] kswapd0/14 [HC0[0]:SC0[0]:HE1:SE1] takes: [14046.102850] (&xfs_dir_ilock_class){++++?+}, at: [<791a04bb>] xfs_ilock+0xff/0x16a [14046.102850] {RECLAIM_FS-ON-W} state was registered at: [14046.102850] [<7904cdb1>] mark_held_locks+0x81/0xe7 [14046.102850] [<7904d390>] lockdep_trace_alloc+0x5c/0xb4 [14046.102850] [<790c2c28>] kmem_cache_alloc_trace+0x2b/0x11e [14046.102850] [<790ba7f4>] vm_map_ram+0x119/0x3e6 [14046.102850] [<7914e124>] _xfs_buf_map_pages+0x5b/0xcf [14046.102850] [<7914ed74>] xfs_buf_get_map+0x67/0x13f [14046.102850] [<7917506f>] xfs_attr_rmtval_set+0x396/0x4d5 [14046.102850] [<7916e8bb>] xfs_attr_leaf_addname+0x18f/0x37d [14046.102850] [<7916ed9e>] xfs_attr_set_int+0x2f5/0x3e8 [14046.102850] [<7916eefc>] xfs_attr_set+0x6b/0x74 [14046.102850] [<79168355>] xfs_xattr_set+0x61/0x81 [14046.102850] [<790e5b10>] generic_setxattr+0x59/0x68 [14046.102850] [<790e4c06>] __vfs_setxattr_noperm+0x58/0xce [14046.102850] [<790e4d0a>] vfs_setxattr+0x8e/0x92 [14046.102850] [<790e4ddd>] setxattr+0xcf/0x159 [14046.102850] [<790e5423>] SyS_lsetxattr+0x88/0xbb [14046.102850] [<79268438>] sysenter_do_call+0x12/0x36 Now, we can't completely remove these traces - mainly because vm_map_ram() will do GFP_KERNEL allocation and that generates the above warning before we get into the reclaim code, but we can turn them all into false positive warnings. To do that, use the method that DM and other IO context code uses to avoid this problem: there is a process flag to tell memory reclaim not to do IO that we can set appropriately. That prevents GFP_KERNEL context reclaim being done from deep inside the vmalloc code in places we can't directly pass a GFP_NOFS context to. That interface has a pair of wrapper functions: memalloc_noio_save() and memalloc_noio_restore(). Adding them around vm_map_ram and the vzalloc call in kmem_alloc_large() will prevent deadlocks and most lockdep reports for this issue. Also, convert the vzalloc() call in kmem_alloc_large() to use __vmalloc() so that we can pass the correct gfp context to the data page allocation routine inside __vmalloc() so that it is clear that GFP_NOFS context is important to this vmalloc call. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Dave Chinner <david@fromorbit.com>
2014-03-07 05:19:14 +00:00
* memory reclaim re-entering the filesystem here and
* potentially deadlocking.
*/
nofs_flag = memalloc_nofs_save();
do {
bp->b_addr = vm_map_ram(bp->b_pages, bp->b_page_count,
-1, PAGE_KERNEL);
if (bp->b_addr)
break;
vm_unmap_aliases();
} while (retried++ <= 1);
memalloc_nofs_restore(nofs_flag);
if (!bp->b_addr)
return -ENOMEM;
bp->b_addr += bp->b_offset;
}
return 0;
}
/*
* Finding and Reading Buffers
*/
static int
_xfs_buf_obj_cmp(
struct rhashtable_compare_arg *arg,
const void *obj)
{
const struct xfs_buf_map *map = arg->key;
const struct xfs_buf *bp = obj;
/*
* The key hashing in the lookup path depends on the key being the
* first element of the compare_arg, make sure to assert this.
*/
BUILD_BUG_ON(offsetof(struct xfs_buf_map, bm_bn) != 0);
if (bp->b_bn != map->bm_bn)
return 1;
if (unlikely(bp->b_length != map->bm_len)) {
/*
* found a block number match. If the range doesn't
* match, the only way this is allowed is if the buffer
* in the cache is stale and the transaction that made
* it stale has not yet committed. i.e. we are
* reallocating a busy extent. Skip this buffer and
* continue searching for an exact match.
*/
ASSERT(bp->b_flags & XBF_STALE);
return 1;
}
return 0;
}
static const struct rhashtable_params xfs_buf_hash_params = {
.min_size = 32, /* empty AGs have minimal footprint */
.nelem_hint = 16,
.key_len = sizeof(xfs_daddr_t),
.key_offset = offsetof(struct xfs_buf, b_bn),
.head_offset = offsetof(struct xfs_buf, b_rhash_head),
.automatic_shrinking = true,
.obj_cmpfn = _xfs_buf_obj_cmp,
};
int
xfs_buf_hash_init(
struct xfs_perag *pag)
{
spin_lock_init(&pag->pag_buf_lock);
return rhashtable_init(&pag->pag_buf_hash, &xfs_buf_hash_params);
}
void
xfs_buf_hash_destroy(
struct xfs_perag *pag)
{
rhashtable_destroy(&pag->pag_buf_hash);
}
/*
* Look up a buffer in the buffer cache and return it referenced and locked
* in @found_bp.
*
* If @new_bp is supplied and we have a lookup miss, insert @new_bp into the
* cache.
*
* If XBF_TRYLOCK is set in @flags, only try to lock the buffer and return
* -EAGAIN if we fail to lock it.
*
* Return values are:
* -EFSCORRUPTED if have been supplied with an invalid address
* -EAGAIN on trylock failure
* -ENOENT if we fail to find a match and @new_bp was NULL
* 0, with @found_bp:
* - @new_bp if we inserted it into the cache
* - the buffer we found and locked.
*/
static int
xfs_buf_find(
struct xfs_buftarg *btp,
struct xfs_buf_map *map,
int nmaps,
xfs_buf_flags_t flags,
struct xfs_buf *new_bp,
struct xfs_buf **found_bp)
{
struct xfs_perag *pag;
xfs_buf_t *bp;
struct xfs_buf_map cmap = { .bm_bn = map[0].bm_bn };
xfs_daddr_t eofs;
int i;
*found_bp = NULL;
for (i = 0; i < nmaps; i++)
cmap.bm_len += map[i].bm_len;
/* Check for IOs smaller than the sector size / not sector aligned */
ASSERT(!(BBTOB(cmap.bm_len) < btp->bt_meta_sectorsize));
ASSERT(!(BBTOB(cmap.bm_bn) & (xfs_off_t)btp->bt_meta_sectormask));
/*
* Corrupted block numbers can get through to here, unfortunately, so we
* have to check that the buffer falls within the filesystem bounds.
*/
eofs = XFS_FSB_TO_BB(btp->bt_mount, btp->bt_mount->m_sb.sb_dblocks);
if (cmap.bm_bn < 0 || cmap.bm_bn >= eofs) {
xfs_alert(btp->bt_mount,
"%s: daddr 0x%llx out of range, EOFS 0x%llx",
__func__, cmap.bm_bn, eofs);
xfs: rework remote attr CRCs Note: this changes the on-disk remote attribute format. I assert that this is OK to do as CRCs are marked experimental and the first kernel it is included in has not yet reached release yet. Further, the userspace utilities are still evolving and so anyone using this stuff right now is a developer or tester using volatile filesystems for testing this feature. Hence changing the format right now to save longer term pain is the right thing to do. The fundamental change is to move from a header per extent in the attribute to a header per filesytem block in the attribute. This means there are more header blocks and the parsing of the attribute data is slightly more complex, but it has the advantage that we always know the size of the attribute on disk based on the length of the data it contains. This is where the header-per-extent method has problems. We don't know the size of the attribute on disk without first knowing how many extents are used to hold it. And we can't tell from a mapping lookup, either, because remote attributes can be allocated contiguously with other attribute blocks and so there is no obvious way of determining the actual size of the atribute on disk short of walking and mapping buffers. The problem with this approach is that if we map a buffer incorrectly (e.g. we make the last buffer for the attribute data too long), we then get buffer cache lookup failure when we map it correctly. i.e. we get a size mismatch on lookup. This is not necessarily fatal, but it's a cache coherency problem that can lead to returning the wrong data to userspace or writing the wrong data to disk. And debug kernels will assert fail if this occurs. I found lots of niggly little problems trying to fix this issue on a 4k block size filesystem, finally getting it to pass with lots of fixes. The thing is, 1024 byte filesystems still failed, and it was getting really complex handling all the corner cases that were showing up. And there were clearly more that I hadn't found yet. It is complex, fragile code, and if we don't fix it now, it will be complex, fragile code forever more. Hence the simple fix is to add a header to each filesystem block. This gives us the same relationship between the attribute data length and the number of blocks on disk as we have without CRCs - it's a linear mapping and doesn't require us to guess anything. It is simple to implement, too - the remote block count calculated at lookup time can be used by the remote attribute set/get/remove code without modification for both CRC and non-CRC filesystems. The world becomes sane again. Because the copy-in and copy-out now need to iterate over each filesystem block, I moved them into helper functions so we separate the block mapping and buffer manupulations from the attribute data and CRC header manipulations. The code becomes much clearer as a result, and it is a lot easier to understand and debug. It also appears to be much more robust - once it worked on 4k block size filesystems, it has worked without failure on 1k block size filesystems, too. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Ben Myers <bpm@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com> (cherry picked from commit ad1858d77771172e08016890f0eb2faedec3ecee)
2013-05-21 08:02:08 +00:00
WARN_ON(1);
return -EFSCORRUPTED;
}
pag = xfs_perag_get(btp->bt_mount,
xfs_daddr_to_agno(btp->bt_mount, cmap.bm_bn));
spin_lock(&pag->pag_buf_lock);
bp = rhashtable_lookup_fast(&pag->pag_buf_hash, &cmap,
xfs_buf_hash_params);
if (bp) {
atomic_inc(&bp->b_hold);
goto found;
}
/* No match found */
if (!new_bp) {
XFS_STATS_INC(btp->bt_mount, xb_miss_locked);
spin_unlock(&pag->pag_buf_lock);
xfs_perag_put(pag);
return -ENOENT;
}
/* the buffer keeps the perag reference until it is freed */
new_bp->b_pag = pag;
rhashtable_insert_fast(&pag->pag_buf_hash, &new_bp->b_rhash_head,
xfs_buf_hash_params);
spin_unlock(&pag->pag_buf_lock);
*found_bp = new_bp;
return 0;
found:
spin_unlock(&pag->pag_buf_lock);
xfs_perag_put(pag);
if (!xfs_buf_trylock(bp)) {
if (flags & XBF_TRYLOCK) {
xfs_buf_rele(bp);
XFS_STATS_INC(btp->bt_mount, xb_busy_locked);
return -EAGAIN;
}
xfs_buf_lock(bp);
XFS_STATS_INC(btp->bt_mount, xb_get_locked_waited);
}
2011-03-25 22:16:45 +00:00
/*
* if the buffer is stale, clear all the external state associated with
* it. We need to keep flags such as how we allocated the buffer memory
* intact here.
*/
if (bp->b_flags & XBF_STALE) {
ASSERT((bp->b_flags & _XBF_DELWRI_Q) == 0);
ASSERT(bp->b_iodone == NULL);
bp->b_flags &= _XBF_KMEM | _XBF_PAGES;
bp->b_ops = NULL;
}
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_find(bp, flags, _RET_IP_);
XFS_STATS_INC(btp->bt_mount, xb_get_locked);
*found_bp = bp;
return 0;
}
struct xfs_buf *
xfs_buf_incore(
struct xfs_buftarg *target,
xfs_daddr_t blkno,
size_t numblks,
xfs_buf_flags_t flags)
{
struct xfs_buf *bp;
int error;
DEFINE_SINGLE_BUF_MAP(map, blkno, numblks);
error = xfs_buf_find(target, &map, 1, flags, NULL, &bp);
if (error)
return NULL;
return bp;
}
/*
xfs: Don't allocate new buffers on every call to _xfs_buf_find Stats show that for an 8-way unlink @ ~80,000 unlinks/s we are doing ~1 million cache hit lookups to ~3000 buffer creates. That's almost 3 orders of magnitude more cahce hits than misses, so optimising for cache hits is quite important. In the cache hit case, we do not need to allocate a new buffer in case of a cache miss, so we are effectively hitting the allocator for no good reason for vast the majority of calls to _xfs_buf_find. 8-way create workloads are showing similar cache hit/miss ratios. The result is profiles that look like this: samples pcnt function DSO _______ _____ _______________________________ _________________ 1036.00 10.0% _xfs_buf_find [kernel.kallsyms] 582.00 5.6% kmem_cache_alloc [kernel.kallsyms] 519.00 5.0% __memcpy [kernel.kallsyms] 468.00 4.5% __ticket_spin_lock [kernel.kallsyms] 388.00 3.7% kmem_cache_free [kernel.kallsyms] 331.00 3.2% xfs_log_commit_cil [kernel.kallsyms] Further, there is a fair bit of work involved in initialising a new buffer once a cache miss has occurred and we currently do that under the rbtree spinlock. That increases spinlock hold time on what are heavily used trees. To fix this, remove the initialisation of the buffer from _xfs_buf_find() and only allocate the new buffer once we've had a cache miss. Initialise the buffer immediately after allocating it in xfs_buf_get, too, so that is it ready for insert if we get another cache miss after allocation. This minimises lock hold time and avoids unnecessary allocator churn. The resulting profiles look like: samples pcnt function DSO _______ _____ ___________________________ _________________ 8111.00 9.1% _xfs_buf_find [kernel.kallsyms] 4380.00 4.9% __memcpy [kernel.kallsyms] 4341.00 4.8% __ticket_spin_lock [kernel.kallsyms] 3401.00 3.8% kmem_cache_alloc [kernel.kallsyms] 2856.00 3.2% xfs_log_commit_cil [kernel.kallsyms] 2625.00 2.9% __kmalloc [kernel.kallsyms] 2380.00 2.7% kfree [kernel.kallsyms] 2016.00 2.3% kmem_cache_free [kernel.kallsyms] Showing a significant reduction in time spent doing allocation and freeing from slabs (kmem_cache_alloc and kmem_cache_free). Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2011-09-30 04:45:02 +00:00
* Assembles a buffer covering the specified range. The code is optimised for
* cache hits, as metadata intensive workloads will see 3 orders of magnitude
* more hits than misses.
*/
xfs: Don't allocate new buffers on every call to _xfs_buf_find Stats show that for an 8-way unlink @ ~80,000 unlinks/s we are doing ~1 million cache hit lookups to ~3000 buffer creates. That's almost 3 orders of magnitude more cahce hits than misses, so optimising for cache hits is quite important. In the cache hit case, we do not need to allocate a new buffer in case of a cache miss, so we are effectively hitting the allocator for no good reason for vast the majority of calls to _xfs_buf_find. 8-way create workloads are showing similar cache hit/miss ratios. The result is profiles that look like this: samples pcnt function DSO _______ _____ _______________________________ _________________ 1036.00 10.0% _xfs_buf_find [kernel.kallsyms] 582.00 5.6% kmem_cache_alloc [kernel.kallsyms] 519.00 5.0% __memcpy [kernel.kallsyms] 468.00 4.5% __ticket_spin_lock [kernel.kallsyms] 388.00 3.7% kmem_cache_free [kernel.kallsyms] 331.00 3.2% xfs_log_commit_cil [kernel.kallsyms] Further, there is a fair bit of work involved in initialising a new buffer once a cache miss has occurred and we currently do that under the rbtree spinlock. That increases spinlock hold time on what are heavily used trees. To fix this, remove the initialisation of the buffer from _xfs_buf_find() and only allocate the new buffer once we've had a cache miss. Initialise the buffer immediately after allocating it in xfs_buf_get, too, so that is it ready for insert if we get another cache miss after allocation. This minimises lock hold time and avoids unnecessary allocator churn. The resulting profiles look like: samples pcnt function DSO _______ _____ ___________________________ _________________ 8111.00 9.1% _xfs_buf_find [kernel.kallsyms] 4380.00 4.9% __memcpy [kernel.kallsyms] 4341.00 4.8% __ticket_spin_lock [kernel.kallsyms] 3401.00 3.8% kmem_cache_alloc [kernel.kallsyms] 2856.00 3.2% xfs_log_commit_cil [kernel.kallsyms] 2625.00 2.9% __kmalloc [kernel.kallsyms] 2380.00 2.7% kfree [kernel.kallsyms] 2016.00 2.3% kmem_cache_free [kernel.kallsyms] Showing a significant reduction in time spent doing allocation and freeing from slabs (kmem_cache_alloc and kmem_cache_free). Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2011-09-30 04:45:02 +00:00
struct xfs_buf *
xfs_buf_get_map(
struct xfs_buftarg *target,
struct xfs_buf_map *map,
int nmaps,
xfs_buf_flags_t flags)
{
xfs: Don't allocate new buffers on every call to _xfs_buf_find Stats show that for an 8-way unlink @ ~80,000 unlinks/s we are doing ~1 million cache hit lookups to ~3000 buffer creates. That's almost 3 orders of magnitude more cahce hits than misses, so optimising for cache hits is quite important. In the cache hit case, we do not need to allocate a new buffer in case of a cache miss, so we are effectively hitting the allocator for no good reason for vast the majority of calls to _xfs_buf_find. 8-way create workloads are showing similar cache hit/miss ratios. The result is profiles that look like this: samples pcnt function DSO _______ _____ _______________________________ _________________ 1036.00 10.0% _xfs_buf_find [kernel.kallsyms] 582.00 5.6% kmem_cache_alloc [kernel.kallsyms] 519.00 5.0% __memcpy [kernel.kallsyms] 468.00 4.5% __ticket_spin_lock [kernel.kallsyms] 388.00 3.7% kmem_cache_free [kernel.kallsyms] 331.00 3.2% xfs_log_commit_cil [kernel.kallsyms] Further, there is a fair bit of work involved in initialising a new buffer once a cache miss has occurred and we currently do that under the rbtree spinlock. That increases spinlock hold time on what are heavily used trees. To fix this, remove the initialisation of the buffer from _xfs_buf_find() and only allocate the new buffer once we've had a cache miss. Initialise the buffer immediately after allocating it in xfs_buf_get, too, so that is it ready for insert if we get another cache miss after allocation. This minimises lock hold time and avoids unnecessary allocator churn. The resulting profiles look like: samples pcnt function DSO _______ _____ ___________________________ _________________ 8111.00 9.1% _xfs_buf_find [kernel.kallsyms] 4380.00 4.9% __memcpy [kernel.kallsyms] 4341.00 4.8% __ticket_spin_lock [kernel.kallsyms] 3401.00 3.8% kmem_cache_alloc [kernel.kallsyms] 2856.00 3.2% xfs_log_commit_cil [kernel.kallsyms] 2625.00 2.9% __kmalloc [kernel.kallsyms] 2380.00 2.7% kfree [kernel.kallsyms] 2016.00 2.3% kmem_cache_free [kernel.kallsyms] Showing a significant reduction in time spent doing allocation and freeing from slabs (kmem_cache_alloc and kmem_cache_free). Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2011-09-30 04:45:02 +00:00
struct xfs_buf *bp;
struct xfs_buf *new_bp;
2011-03-25 22:16:45 +00:00
int error = 0;
error = xfs_buf_find(target, map, nmaps, flags, NULL, &bp);
switch (error) {
case 0:
/* cache hit */
xfs: Don't allocate new buffers on every call to _xfs_buf_find Stats show that for an 8-way unlink @ ~80,000 unlinks/s we are doing ~1 million cache hit lookups to ~3000 buffer creates. That's almost 3 orders of magnitude more cahce hits than misses, so optimising for cache hits is quite important. In the cache hit case, we do not need to allocate a new buffer in case of a cache miss, so we are effectively hitting the allocator for no good reason for vast the majority of calls to _xfs_buf_find. 8-way create workloads are showing similar cache hit/miss ratios. The result is profiles that look like this: samples pcnt function DSO _______ _____ _______________________________ _________________ 1036.00 10.0% _xfs_buf_find [kernel.kallsyms] 582.00 5.6% kmem_cache_alloc [kernel.kallsyms] 519.00 5.0% __memcpy [kernel.kallsyms] 468.00 4.5% __ticket_spin_lock [kernel.kallsyms] 388.00 3.7% kmem_cache_free [kernel.kallsyms] 331.00 3.2% xfs_log_commit_cil [kernel.kallsyms] Further, there is a fair bit of work involved in initialising a new buffer once a cache miss has occurred and we currently do that under the rbtree spinlock. That increases spinlock hold time on what are heavily used trees. To fix this, remove the initialisation of the buffer from _xfs_buf_find() and only allocate the new buffer once we've had a cache miss. Initialise the buffer immediately after allocating it in xfs_buf_get, too, so that is it ready for insert if we get another cache miss after allocation. This minimises lock hold time and avoids unnecessary allocator churn. The resulting profiles look like: samples pcnt function DSO _______ _____ ___________________________ _________________ 8111.00 9.1% _xfs_buf_find [kernel.kallsyms] 4380.00 4.9% __memcpy [kernel.kallsyms] 4341.00 4.8% __ticket_spin_lock [kernel.kallsyms] 3401.00 3.8% kmem_cache_alloc [kernel.kallsyms] 2856.00 3.2% xfs_log_commit_cil [kernel.kallsyms] 2625.00 2.9% __kmalloc [kernel.kallsyms] 2380.00 2.7% kfree [kernel.kallsyms] 2016.00 2.3% kmem_cache_free [kernel.kallsyms] Showing a significant reduction in time spent doing allocation and freeing from slabs (kmem_cache_alloc and kmem_cache_free). Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2011-09-30 04:45:02 +00:00
goto found;
case -EAGAIN:
/* cache hit, trylock failure, caller handles failure */
ASSERT(flags & XBF_TRYLOCK);
return NULL;
case -ENOENT:
/* cache miss, go for insert */
break;
case -EFSCORRUPTED:
default:
/*
* None of the higher layers understand failure types
* yet, so return NULL to signal a fatal lookup error.
*/
return NULL;
}
xfs: Don't allocate new buffers on every call to _xfs_buf_find Stats show that for an 8-way unlink @ ~80,000 unlinks/s we are doing ~1 million cache hit lookups to ~3000 buffer creates. That's almost 3 orders of magnitude more cahce hits than misses, so optimising for cache hits is quite important. In the cache hit case, we do not need to allocate a new buffer in case of a cache miss, so we are effectively hitting the allocator for no good reason for vast the majority of calls to _xfs_buf_find. 8-way create workloads are showing similar cache hit/miss ratios. The result is profiles that look like this: samples pcnt function DSO _______ _____ _______________________________ _________________ 1036.00 10.0% _xfs_buf_find [kernel.kallsyms] 582.00 5.6% kmem_cache_alloc [kernel.kallsyms] 519.00 5.0% __memcpy [kernel.kallsyms] 468.00 4.5% __ticket_spin_lock [kernel.kallsyms] 388.00 3.7% kmem_cache_free [kernel.kallsyms] 331.00 3.2% xfs_log_commit_cil [kernel.kallsyms] Further, there is a fair bit of work involved in initialising a new buffer once a cache miss has occurred and we currently do that under the rbtree spinlock. That increases spinlock hold time on what are heavily used trees. To fix this, remove the initialisation of the buffer from _xfs_buf_find() and only allocate the new buffer once we've had a cache miss. Initialise the buffer immediately after allocating it in xfs_buf_get, too, so that is it ready for insert if we get another cache miss after allocation. This minimises lock hold time and avoids unnecessary allocator churn. The resulting profiles look like: samples pcnt function DSO _______ _____ ___________________________ _________________ 8111.00 9.1% _xfs_buf_find [kernel.kallsyms] 4380.00 4.9% __memcpy [kernel.kallsyms] 4341.00 4.8% __ticket_spin_lock [kernel.kallsyms] 3401.00 3.8% kmem_cache_alloc [kernel.kallsyms] 2856.00 3.2% xfs_log_commit_cil [kernel.kallsyms] 2625.00 2.9% __kmalloc [kernel.kallsyms] 2380.00 2.7% kfree [kernel.kallsyms] 2016.00 2.3% kmem_cache_free [kernel.kallsyms] Showing a significant reduction in time spent doing allocation and freeing from slabs (kmem_cache_alloc and kmem_cache_free). Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2011-09-30 04:45:02 +00:00
new_bp = _xfs_buf_alloc(target, map, nmaps, flags);
if (unlikely(!new_bp))
return NULL;
error = xfs_buf_allocate_memory(new_bp, flags);
if (error) {
xfs_buf_free(new_bp);
return NULL;
}
error = xfs_buf_find(target, map, nmaps, flags, new_bp, &bp);
if (error) {
xfs_buf_free(new_bp);
xfs: Don't allocate new buffers on every call to _xfs_buf_find Stats show that for an 8-way unlink @ ~80,000 unlinks/s we are doing ~1 million cache hit lookups to ~3000 buffer creates. That's almost 3 orders of magnitude more cahce hits than misses, so optimising for cache hits is quite important. In the cache hit case, we do not need to allocate a new buffer in case of a cache miss, so we are effectively hitting the allocator for no good reason for vast the majority of calls to _xfs_buf_find. 8-way create workloads are showing similar cache hit/miss ratios. The result is profiles that look like this: samples pcnt function DSO _______ _____ _______________________________ _________________ 1036.00 10.0% _xfs_buf_find [kernel.kallsyms] 582.00 5.6% kmem_cache_alloc [kernel.kallsyms] 519.00 5.0% __memcpy [kernel.kallsyms] 468.00 4.5% __ticket_spin_lock [kernel.kallsyms] 388.00 3.7% kmem_cache_free [kernel.kallsyms] 331.00 3.2% xfs_log_commit_cil [kernel.kallsyms] Further, there is a fair bit of work involved in initialising a new buffer once a cache miss has occurred and we currently do that under the rbtree spinlock. That increases spinlock hold time on what are heavily used trees. To fix this, remove the initialisation of the buffer from _xfs_buf_find() and only allocate the new buffer once we've had a cache miss. Initialise the buffer immediately after allocating it in xfs_buf_get, too, so that is it ready for insert if we get another cache miss after allocation. This minimises lock hold time and avoids unnecessary allocator churn. The resulting profiles look like: samples pcnt function DSO _______ _____ ___________________________ _________________ 8111.00 9.1% _xfs_buf_find [kernel.kallsyms] 4380.00 4.9% __memcpy [kernel.kallsyms] 4341.00 4.8% __ticket_spin_lock [kernel.kallsyms] 3401.00 3.8% kmem_cache_alloc [kernel.kallsyms] 2856.00 3.2% xfs_log_commit_cil [kernel.kallsyms] 2625.00 2.9% __kmalloc [kernel.kallsyms] 2380.00 2.7% kfree [kernel.kallsyms] 2016.00 2.3% kmem_cache_free [kernel.kallsyms] Showing a significant reduction in time spent doing allocation and freeing from slabs (kmem_cache_alloc and kmem_cache_free). Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2011-09-30 04:45:02 +00:00
return NULL;
}
if (bp != new_bp)
xfs_buf_free(new_bp);
xfs: Don't allocate new buffers on every call to _xfs_buf_find Stats show that for an 8-way unlink @ ~80,000 unlinks/s we are doing ~1 million cache hit lookups to ~3000 buffer creates. That's almost 3 orders of magnitude more cahce hits than misses, so optimising for cache hits is quite important. In the cache hit case, we do not need to allocate a new buffer in case of a cache miss, so we are effectively hitting the allocator for no good reason for vast the majority of calls to _xfs_buf_find. 8-way create workloads are showing similar cache hit/miss ratios. The result is profiles that look like this: samples pcnt function DSO _______ _____ _______________________________ _________________ 1036.00 10.0% _xfs_buf_find [kernel.kallsyms] 582.00 5.6% kmem_cache_alloc [kernel.kallsyms] 519.00 5.0% __memcpy [kernel.kallsyms] 468.00 4.5% __ticket_spin_lock [kernel.kallsyms] 388.00 3.7% kmem_cache_free [kernel.kallsyms] 331.00 3.2% xfs_log_commit_cil [kernel.kallsyms] Further, there is a fair bit of work involved in initialising a new buffer once a cache miss has occurred and we currently do that under the rbtree spinlock. That increases spinlock hold time on what are heavily used trees. To fix this, remove the initialisation of the buffer from _xfs_buf_find() and only allocate the new buffer once we've had a cache miss. Initialise the buffer immediately after allocating it in xfs_buf_get, too, so that is it ready for insert if we get another cache miss after allocation. This minimises lock hold time and avoids unnecessary allocator churn. The resulting profiles look like: samples pcnt function DSO _______ _____ ___________________________ _________________ 8111.00 9.1% _xfs_buf_find [kernel.kallsyms] 4380.00 4.9% __memcpy [kernel.kallsyms] 4341.00 4.8% __ticket_spin_lock [kernel.kallsyms] 3401.00 3.8% kmem_cache_alloc [kernel.kallsyms] 2856.00 3.2% xfs_log_commit_cil [kernel.kallsyms] 2625.00 2.9% __kmalloc [kernel.kallsyms] 2380.00 2.7% kfree [kernel.kallsyms] 2016.00 2.3% kmem_cache_free [kernel.kallsyms] Showing a significant reduction in time spent doing allocation and freeing from slabs (kmem_cache_alloc and kmem_cache_free). Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2011-09-30 04:45:02 +00:00
found:
if (!bp->b_addr) {
error = _xfs_buf_map_pages(bp, flags);
if (unlikely(error)) {
xfs_warn(target->bt_mount,
"%s: failed to map pagesn", __func__);
xfs_buf_relse(bp);
return NULL;
}
}
xfs: inode recovery readahead can race with inode buffer creation When we do inode readahead in log recovery, we do can do the readahead before we've replayed the icreate transaction that stamps the buffer with inode cores. The inode readahead verifier catches this and marks the buffer as !done to indicate that it doesn't yet contain valid inodes. In adding buffer error notification (i.e. setting b_error = -EIO at the same time as as we clear the done flag) to such a readahead verifier failure, we can then get subsequent inode recovery failing with this error: XFS (dm-0): metadata I/O error: block 0xa00060 ("xlog_recover_do..(read#2)") error 5 numblks 32 This occurs when readahead completion races with icreate item replay such as: inode readahead find buffer lock buffer submit RA io .... icreate recovery xfs_trans_get_buffer find buffer lock buffer <blocks on RA completion> ..... <ra completion> fails verifier clear XBF_DONE set bp->b_error = -EIO release and unlock buffer <icreate gains lock> icreate initialises buffer marks buffer as done adds buffer to delayed write queue releases buffer At this point, we have an initialised inode buffer that is up to date but has an -EIO state registered against it. When we finally get to recovering an inode in that buffer: inode item recovery xfs_trans_read_buffer find buffer lock buffer sees XBF_DONE is set, returns buffer sees bp->b_error is set fail log recovery! Essentially, we need xfs_trans_get_buf_map() to clear the error status of the buffer when doing a lookup. This function returns uninitialised buffers, so the buffer returned can not be in an error state and none of the code that uses this function expects b_error to be set on return. Indeed, there is an ASSERT(!bp->b_error); in the transaction case in xfs_trans_get_buf_map() that would have caught this if log recovery used transactions.... This patch firstly changes the inode readahead failure to set -EIO on the buffer, and secondly changes xfs_buf_get_map() to never return a buffer with an error state set so this first change doesn't cause unexpected log recovery failures. cc: <stable@vger.kernel.org> # 3.12 - current Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-01-11 20:03:44 +00:00
/*
* Clear b_error if this is a lookup from a caller that doesn't expect
* valid data to be found in the buffer.
*/
if (!(flags & XBF_READ))
xfs_buf_ioerror(bp, 0);
XFS_STATS_INC(target->bt_mount, xb_get);
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_get(bp, flags, _RET_IP_);
return bp;
}
STATIC int
_xfs_buf_read(
xfs_buf_t *bp,
xfs_buf_flags_t flags)
{
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
ASSERT(!(flags & XBF_WRITE));
ASSERT(bp->b_maps[0].bm_bn != XFS_BUF_DADDR_NULL);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
bp->b_flags &= ~(XBF_WRITE | XBF_ASYNC | XBF_READ_AHEAD);
bp->b_flags |= flags & (XBF_READ | XBF_ASYNC | XBF_READ_AHEAD);
return xfs_buf_submit(bp);
}
/*
* Reverify a buffer found in cache without an attached ->b_ops.
*
* If the caller passed an ops structure and the buffer doesn't have ops
* assigned, set the ops and use it to verify the contents. If verification
* fails, clear XBF_DONE. We assume the buffer has no recorded errors and is
* already in XBF_DONE state on entry.
*
* Under normal operations, every in-core buffer is verified on read I/O
* completion. There are two scenarios that can lead to in-core buffers without
* an assigned ->b_ops. The first is during log recovery of buffers on a V4
* filesystem, though these buffers are purged at the end of recovery. The
* other is online repair, which intentionally reads with a NULL buffer ops to
* run several verifiers across an in-core buffer in order to establish buffer
* type. If repair can't establish that, the buffer will be left in memory
* with NULL buffer ops.
*/
int
xfs_buf_reverify(
struct xfs_buf *bp,
const struct xfs_buf_ops *ops)
{
ASSERT(bp->b_flags & XBF_DONE);
ASSERT(bp->b_error == 0);
if (!ops || bp->b_ops)
return 0;
bp->b_ops = ops;
bp->b_ops->verify_read(bp);
if (bp->b_error)
bp->b_flags &= ~XBF_DONE;
return bp->b_error;
}
xfs_buf_t *
xfs_buf_read_map(
struct xfs_buftarg *target,
struct xfs_buf_map *map,
int nmaps,
xfs_buf_flags_t flags,
const struct xfs_buf_ops *ops)
{
struct xfs_buf *bp;
flags |= XBF_READ;
bp = xfs_buf_get_map(target, map, nmaps, flags);
if (!bp)
return NULL;
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_read(bp, flags, _RET_IP_);
if (!(bp->b_flags & XBF_DONE)) {
XFS_STATS_INC(target->bt_mount, xb_get_read);
bp->b_ops = ops;
_xfs_buf_read(bp, flags);
return bp;
}
xfs_buf_reverify(bp, ops);
if (flags & XBF_ASYNC) {
/*
* Read ahead call which is already satisfied,
* drop the buffer
*/
xfs_buf_relse(bp);
return NULL;
}
/* We do not want read in the flags */
bp->b_flags &= ~XBF_READ;
ASSERT(bp->b_ops != NULL || ops == NULL);
return bp;
}
/*
* If we are not low on memory then do the readahead in a deadlock
* safe manner.
*/
void
xfs_buf_readahead_map(
struct xfs_buftarg *target,
struct xfs_buf_map *map,
int nmaps,
const struct xfs_buf_ops *ops)
{
if (bdi_read_congested(target->bt_bdev->bd_bdi))
return;
xfs_buf_read_map(target, map, nmaps,
XBF_TRYLOCK|XBF_ASYNC|XBF_READ_AHEAD, ops);
}
/*
* Read an uncached buffer from disk. Allocates and returns a locked
* buffer containing the disk contents or nothing.
*/
int
xfs_buf_read_uncached(
struct xfs_buftarg *target,
xfs_daddr_t daddr,
size_t numblks,
int flags,
struct xfs_buf **bpp,
const struct xfs_buf_ops *ops)
{
struct xfs_buf *bp;
*bpp = NULL;
bp = xfs_buf_get_uncached(target, numblks, flags);
if (!bp)
return -ENOMEM;
/* set up the buffer for a read IO */
ASSERT(bp->b_map_count == 1);
bp->b_bn = XFS_BUF_DADDR_NULL; /* always null for uncached buffers */
bp->b_maps[0].bm_bn = daddr;
bp->b_flags |= XBF_READ;
bp->b_ops = ops;
xfs_buf_submit(bp);
if (bp->b_error) {
int error = bp->b_error;
xfs_buf_relse(bp);
return error;
}
*bpp = bp;
return 0;
}
/*
* Return a buffer allocated as an empty buffer and associated to external
* memory via xfs_buf_associate_memory() back to it's empty state.
*/
void
xfs_buf_set_empty(
struct xfs_buf *bp,
size_t numblks)
{
if (bp->b_pages)
_xfs_buf_free_pages(bp);
bp->b_pages = NULL;
bp->b_page_count = 0;
bp->b_addr = NULL;
bp->b_length = numblks;
bp->b_io_length = numblks;
ASSERT(bp->b_map_count == 1);
bp->b_bn = XFS_BUF_DADDR_NULL;
bp->b_maps[0].bm_bn = XFS_BUF_DADDR_NULL;
bp->b_maps[0].bm_len = bp->b_length;
}
static inline struct page *
mem_to_page(
void *addr)
{
if ((!is_vmalloc_addr(addr))) {
return virt_to_page(addr);
} else {
return vmalloc_to_page(addr);
}
}
int
xfs_buf_associate_memory(
xfs_buf_t *bp,
void *mem,
size_t len)
{
int rval;
int i = 0;
unsigned long pageaddr;
unsigned long offset;
size_t buflen;
int page_count;
2011-03-25 22:16:45 +00:00
pageaddr = (unsigned long)mem & PAGE_MASK;
offset = (unsigned long)mem - pageaddr;
2011-03-25 22:16:45 +00:00
buflen = PAGE_ALIGN(len + offset);
page_count = buflen >> PAGE_SHIFT;
/* Free any previous set of page pointers */
if (bp->b_pages)
_xfs_buf_free_pages(bp);
bp->b_pages = NULL;
bp->b_addr = mem;
rval = _xfs_buf_get_pages(bp, page_count);
if (rval)
return rval;
bp->b_offset = offset;
for (i = 0; i < bp->b_page_count; i++) {
bp->b_pages[i] = mem_to_page((void *)pageaddr);
2011-03-25 22:16:45 +00:00
pageaddr += PAGE_SIZE;
}
bp->b_io_length = BTOBB(len);
bp->b_length = BTOBB(buflen);
return 0;
}
xfs_buf_t *
xfs_buf_get_uncached(
struct xfs_buftarg *target,
size_t numblks,
int flags)
{
unsigned long page_count;
int error, i;
struct xfs_buf *bp;
DEFINE_SINGLE_BUF_MAP(map, XFS_BUF_DADDR_NULL, numblks);
/* flags might contain irrelevant bits, pass only what we care about */
bp = _xfs_buf_alloc(target, &map, 1, flags & XBF_NO_IOACCT);
if (unlikely(bp == NULL))
goto fail;
page_count = PAGE_ALIGN(numblks << BBSHIFT) >> PAGE_SHIFT;
error = _xfs_buf_get_pages(bp, page_count);
if (error)
goto fail_free_buf;
for (i = 0; i < page_count; i++) {
bp->b_pages[i] = alloc_page(xb_to_gfp(flags));
if (!bp->b_pages[i])
goto fail_free_mem;
}
bp->b_flags |= _XBF_PAGES;
error = _xfs_buf_map_pages(bp, 0);
if (unlikely(error)) {
xfs_warn(target->bt_mount,
"%s: failed to map pages", __func__);
goto fail_free_mem;
}
trace_xfs_buf_get_uncached(bp, _RET_IP_);
return bp;
fail_free_mem:
while (--i >= 0)
__free_page(bp->b_pages[i]);
_xfs_buf_free_pages(bp);
fail_free_buf:
xfs_buf_free_maps(bp);
kmem_zone_free(xfs_buf_zone, bp);
fail:
return NULL;
}
/*
* Increment reference count on buffer, to hold the buffer concurrently
* with another thread which may release (free) the buffer asynchronously.
* Must hold the buffer already to call this function.
*/
void
xfs_buf_hold(
xfs_buf_t *bp)
{
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_hold(bp, _RET_IP_);
atomic_inc(&bp->b_hold);
}
/*
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
* Release a hold on the specified buffer. If the hold count is 1, the buffer is
* placed on LRU or freed (depending on b_lru_ref).
*/
void
xfs_buf_rele(
xfs_buf_t *bp)
{
struct xfs_perag *pag = bp->b_pag;
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
bool release;
bool freebuf = false;
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_rele(bp, _RET_IP_);
if (!pag) {
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
ASSERT(list_empty(&bp->b_lru));
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
if (atomic_dec_and_test(&bp->b_hold)) {
xfs_buf_ioacct_dec(bp);
xfs_buf_free(bp);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
}
return;
}
ASSERT(atomic_read(&bp->b_hold) > 0);
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
xfs: fix use-after-free race in xfs_buf_rele When looking at a 4.18 based KASAN use after free report, I noticed that racing xfs_buf_rele() may race on dropping the last reference to the buffer and taking the buffer lock. This was the symptom displayed by the KASAN report, but the actual issue that was reported had already been fixed in 4.19-rc1 by commit e339dd8d8b04 ("xfs: use sync buffer I/O for sync delwri queue submission"). Despite this, I think there is still an issue with xfs_buf_rele() in this code: release = atomic_dec_and_lock(&bp->b_hold, &pag->pag_buf_lock); spin_lock(&bp->b_lock); if (!release) { ..... If two threads race on the b_lock after both dropping a reference and one getting dropping the last reference so release = true, we end up with: CPU 0 CPU 1 atomic_dec_and_lock() atomic_dec_and_lock() spin_lock(&bp->b_lock) spin_lock(&bp->b_lock) <spins> <release = true bp->b_lru_ref = 0> <remove from lists> freebuf = true spin_unlock(&bp->b_lock) xfs_buf_free(bp) <gets lock, reading and writing freed memory> <accesses freed memory> spin_unlock(&bp->b_lock) <reads/writes freed memory> IOWs, we can't safely take bp->b_lock after dropping the hold reference because the buffer may go away at any time after we drop that reference. However, this can be fixed simply by taking the bp->b_lock before we drop the reference. It is safe to nest the pag_buf_lock inside bp->b_lock as the pag_buf_lock is only used to serialise against lookup in xfs_buf_find() and no other locks are held over or under the pag_buf_lock there. Make this clear by documenting the buffer lock orders at the top of the file. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com Signed-off-by: Dave Chinner <david@fromorbit.com>
2018-10-18 06:21:29 +00:00
/*
* We grab the b_lock here first to serialise racing xfs_buf_rele()
* calls. The pag_buf_lock being taken on the last reference only
* serialises against racing lookups in xfs_buf_find(). IOWs, the second
* to last reference we drop here is not serialised against the last
* reference until we take bp->b_lock. Hence if we don't grab b_lock
* first, the last "release" reference can win the race to the lock and
* free the buffer before the second-to-last reference is processed,
* leading to a use-after-free scenario.
*/
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
spin_lock(&bp->b_lock);
xfs: fix use-after-free race in xfs_buf_rele When looking at a 4.18 based KASAN use after free report, I noticed that racing xfs_buf_rele() may race on dropping the last reference to the buffer and taking the buffer lock. This was the symptom displayed by the KASAN report, but the actual issue that was reported had already been fixed in 4.19-rc1 by commit e339dd8d8b04 ("xfs: use sync buffer I/O for sync delwri queue submission"). Despite this, I think there is still an issue with xfs_buf_rele() in this code: release = atomic_dec_and_lock(&bp->b_hold, &pag->pag_buf_lock); spin_lock(&bp->b_lock); if (!release) { ..... If two threads race on the b_lock after both dropping a reference and one getting dropping the last reference so release = true, we end up with: CPU 0 CPU 1 atomic_dec_and_lock() atomic_dec_and_lock() spin_lock(&bp->b_lock) spin_lock(&bp->b_lock) <spins> <release = true bp->b_lru_ref = 0> <remove from lists> freebuf = true spin_unlock(&bp->b_lock) xfs_buf_free(bp) <gets lock, reading and writing freed memory> <accesses freed memory> spin_unlock(&bp->b_lock) <reads/writes freed memory> IOWs, we can't safely take bp->b_lock after dropping the hold reference because the buffer may go away at any time after we drop that reference. However, this can be fixed simply by taking the bp->b_lock before we drop the reference. It is safe to nest the pag_buf_lock inside bp->b_lock as the pag_buf_lock is only used to serialise against lookup in xfs_buf_find() and no other locks are held over or under the pag_buf_lock there. Make this clear by documenting the buffer lock orders at the top of the file. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Carlos Maiolino <cmaiolino@redhat.com Signed-off-by: Dave Chinner <david@fromorbit.com>
2018-10-18 06:21:29 +00:00
release = atomic_dec_and_lock(&bp->b_hold, &pag->pag_buf_lock);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
if (!release) {
/*
* Drop the in-flight state if the buffer is already on the LRU
* and it holds the only reference. This is racy because we
* haven't acquired the pag lock, but the use of _XBF_IN_FLIGHT
* ensures the decrement occurs only once per-buf.
*/
if ((atomic_read(&bp->b_hold) == 1) && !list_empty(&bp->b_lru))
xfs: use ->b_state to fix buffer I/O accounting release race We've had user reports of unmount hangs in xfs_wait_buftarg() that analysis shows is due to btp->bt_io_count == -1. bt_io_count represents the count of in-flight asynchronous buffers and thus should always be >= 0. xfs_wait_buftarg() waits for this value to stabilize to zero in order to ensure that all untracked (with respect to the lru) buffers have completed I/O processing before unmount proceeds to tear down in-core data structures. The value of -1 implies an I/O accounting decrement race. Indeed, the fact that xfs_buf_ioacct_dec() is called from xfs_buf_rele() (where the buffer lock is no longer held) means that bp->b_flags can be updated from an unsafe context. While a user-level reproducer is currently not available, some intrusive hacks to run racing buffer lookups/ioacct/releases from multiple threads was used to successfully manufacture this problem. Existing callers do not expect to acquire the buffer lock from xfs_buf_rele(). Therefore, we can not safely update ->b_flags from this context. It turns out that we already have separate buffer state bits and associated serialization for dealing with buffer LRU state in the form of ->b_state and ->b_lock. Therefore, replace the _XBF_IN_FLIGHT flag with a ->b_state variant, update the I/O accounting wrappers appropriately and make sure they are used with the correct locking. This ensures that buffer in-flight state can be modified at buffer release time without racing with modifications from a buffer lock holder. Fixes: 9c7504aa72b6 ("xfs: track and serialize in-flight async buffers against unmount") Cc: <stable@vger.kernel.org> # v4.8+ Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Nikolay Borisov <nborisov@suse.com> Tested-by: Libor Pechacek <lpechacek@suse.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-05-31 15:22:52 +00:00
__xfs_buf_ioacct_dec(bp);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
goto out_unlock;
}
/* the last reference has been dropped ... */
xfs: use ->b_state to fix buffer I/O accounting release race We've had user reports of unmount hangs in xfs_wait_buftarg() that analysis shows is due to btp->bt_io_count == -1. bt_io_count represents the count of in-flight asynchronous buffers and thus should always be >= 0. xfs_wait_buftarg() waits for this value to stabilize to zero in order to ensure that all untracked (with respect to the lru) buffers have completed I/O processing before unmount proceeds to tear down in-core data structures. The value of -1 implies an I/O accounting decrement race. Indeed, the fact that xfs_buf_ioacct_dec() is called from xfs_buf_rele() (where the buffer lock is no longer held) means that bp->b_flags can be updated from an unsafe context. While a user-level reproducer is currently not available, some intrusive hacks to run racing buffer lookups/ioacct/releases from multiple threads was used to successfully manufacture this problem. Existing callers do not expect to acquire the buffer lock from xfs_buf_rele(). Therefore, we can not safely update ->b_flags from this context. It turns out that we already have separate buffer state bits and associated serialization for dealing with buffer LRU state in the form of ->b_state and ->b_lock. Therefore, replace the _XBF_IN_FLIGHT flag with a ->b_state variant, update the I/O accounting wrappers appropriately and make sure they are used with the correct locking. This ensures that buffer in-flight state can be modified at buffer release time without racing with modifications from a buffer lock holder. Fixes: 9c7504aa72b6 ("xfs: track and serialize in-flight async buffers against unmount") Cc: <stable@vger.kernel.org> # v4.8+ Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Nikolay Borisov <nborisov@suse.com> Tested-by: Libor Pechacek <lpechacek@suse.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-05-31 15:22:52 +00:00
__xfs_buf_ioacct_dec(bp);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
if (!(bp->b_flags & XBF_STALE) && atomic_read(&bp->b_lru_ref)) {
/*
* If the buffer is added to the LRU take a new reference to the
* buffer for the LRU and clear the (now stale) dispose list
* state flag
*/
if (list_lru_add(&bp->b_target->bt_lru, &bp->b_lru)) {
bp->b_state &= ~XFS_BSTATE_DISPOSE;
atomic_inc(&bp->b_hold);
}
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
spin_unlock(&pag->pag_buf_lock);
} else {
/*
* most of the time buffers will already be removed from the
* LRU, so optimise that case by checking for the
* XFS_BSTATE_DISPOSE flag indicating the last list the buffer
* was on was the disposal list
*/
if (!(bp->b_state & XFS_BSTATE_DISPOSE)) {
list_lru_del(&bp->b_target->bt_lru, &bp->b_lru);
} else {
ASSERT(list_empty(&bp->b_lru));
}
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
ASSERT(!(bp->b_flags & _XBF_DELWRI_Q));
rhashtable_remove_fast(&pag->pag_buf_hash, &bp->b_rhash_head,
xfs_buf_hash_params);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
spin_unlock(&pag->pag_buf_lock);
xfs_perag_put(pag);
freebuf = true;
}
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
out_unlock:
spin_unlock(&bp->b_lock);
if (freebuf)
xfs_buf_free(bp);
}
/*
2011-03-25 22:16:45 +00:00
* Lock a buffer object, if it is not already locked.
*
* If we come across a stale, pinned, locked buffer, we know that we are
* being asked to lock a buffer that has been reallocated. Because it is
* pinned, we know that the log has not been pushed to disk and hence it
* will still be locked. Rather than continuing to have trylock attempts
* fail until someone else pushes the log, push it ourselves before
* returning. This means that the xfsaild will not get stuck trying
* to push on stale inode buffers.
*/
int
xfs_buf_trylock(
struct xfs_buf *bp)
{
int locked;
locked = down_trylock(&bp->b_sema) == 0;
if (locked)
trace_xfs_buf_trylock(bp, _RET_IP_);
else
trace_xfs_buf_trylock_fail(bp, _RET_IP_);
return locked;
}
/*
2011-03-25 22:16:45 +00:00
* Lock a buffer object.
xfs: Improve scalability of busy extent tracking When we free a metadata extent, we record it in the per-AG busy extent array so that it is not re-used before the freeing transaction hits the disk. This array is fixed size, so when it overflows we make further allocation transactions synchronous because we cannot track more freed extents until those transactions hit the disk and are completed. Under heavy mixed allocation and freeing workloads with large log buffers, we can overflow this array quite easily. Further, the array is sparsely populated, which means that inserts need to search for a free slot, and array searches often have to search many more slots that are actually used to check all the busy extents. Quite inefficient, really. To enable this aspect of extent freeing to scale better, we need a structure that can grow dynamically. While in other areas of XFS we have used radix trees, the extents being freed are at random locations on disk so are better suited to being indexed by an rbtree. So, use a per-AG rbtree indexed by block number to track busy extents. This incures a memory allocation when marking an extent busy, but should not occur too often in low memory situations. This should scale to an arbitrary number of extents so should not be a limitation for features such as in-memory aggregation of transactions. However, there are still situations where we can't avoid allocating busy extents (such as allocation from the AGFL). To minimise the overhead of such occurences, we need to avoid doing a synchronous log force while holding the AGF locked to ensure that the previous transactions are safely on disk before we use the extent. We can do this by marking the transaction doing the allocation as synchronous rather issuing a log force. Because of the locking involved and the ordering of transactions, the synchronous transaction provides the same guarantees as a synchronous log force because it ensures that all the prior transactions are already on disk when the synchronous transaction hits the disk. i.e. it preserves the free->allocate order of the extent correctly in recovery. By doing this, we avoid holding the AGF locked while log writes are in progress, hence reducing the length of time the lock is held and therefore we increase the rate at which we can allocate and free from the allocation group, thereby increasing overall throughput. The only problem with this approach is that when a metadata buffer is marked stale (e.g. a directory block is removed), then buffer remains pinned and locked until the log goes to disk. The issue here is that if that stale buffer is reallocated in a subsequent transaction, the attempt to lock that buffer in the transaction will hang waiting the log to go to disk to unlock and unpin the buffer. Hence if someone tries to lock a pinned, stale, locked buffer we need to push on the log to get it unlocked ASAP. Effectively we are trading off a guaranteed log force for a much less common trigger for log force to occur. Ideally we should not reallocate busy extents. That is a much more complex fix to the problem as it involves direct intervention in the allocation btree searches in many places. This is left to a future set of modifications. Finally, now that we track busy extents in allocated memory, we don't need the descriptors in the transaction structure to point to them. We can replace the complex busy chunk infrastructure with a simple linked list of busy extents. This allows us to remove a large chunk of code, making the overall change a net reduction in code size. Signed-off-by: Dave Chinner <david@fromorbit.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2010-05-21 02:07:08 +00:00
*
* If we come across a stale, pinned, locked buffer, we know that we
* are being asked to lock a buffer that has been reallocated. Because
* it is pinned, we know that the log has not been pushed to disk and
* hence it will still be locked. Rather than sleeping until someone
* else pushes the log, push it ourselves before trying to get the lock.
*/
void
xfs_buf_lock(
struct xfs_buf *bp)
{
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_lock(bp, _RET_IP_);
xfs: Improve scalability of busy extent tracking When we free a metadata extent, we record it in the per-AG busy extent array so that it is not re-used before the freeing transaction hits the disk. This array is fixed size, so when it overflows we make further allocation transactions synchronous because we cannot track more freed extents until those transactions hit the disk and are completed. Under heavy mixed allocation and freeing workloads with large log buffers, we can overflow this array quite easily. Further, the array is sparsely populated, which means that inserts need to search for a free slot, and array searches often have to search many more slots that are actually used to check all the busy extents. Quite inefficient, really. To enable this aspect of extent freeing to scale better, we need a structure that can grow dynamically. While in other areas of XFS we have used radix trees, the extents being freed are at random locations on disk so are better suited to being indexed by an rbtree. So, use a per-AG rbtree indexed by block number to track busy extents. This incures a memory allocation when marking an extent busy, but should not occur too often in low memory situations. This should scale to an arbitrary number of extents so should not be a limitation for features such as in-memory aggregation of transactions. However, there are still situations where we can't avoid allocating busy extents (such as allocation from the AGFL). To minimise the overhead of such occurences, we need to avoid doing a synchronous log force while holding the AGF locked to ensure that the previous transactions are safely on disk before we use the extent. We can do this by marking the transaction doing the allocation as synchronous rather issuing a log force. Because of the locking involved and the ordering of transactions, the synchronous transaction provides the same guarantees as a synchronous log force because it ensures that all the prior transactions are already on disk when the synchronous transaction hits the disk. i.e. it preserves the free->allocate order of the extent correctly in recovery. By doing this, we avoid holding the AGF locked while log writes are in progress, hence reducing the length of time the lock is held and therefore we increase the rate at which we can allocate and free from the allocation group, thereby increasing overall throughput. The only problem with this approach is that when a metadata buffer is marked stale (e.g. a directory block is removed), then buffer remains pinned and locked until the log goes to disk. The issue here is that if that stale buffer is reallocated in a subsequent transaction, the attempt to lock that buffer in the transaction will hang waiting the log to go to disk to unlock and unpin the buffer. Hence if someone tries to lock a pinned, stale, locked buffer we need to push on the log to get it unlocked ASAP. Effectively we are trading off a guaranteed log force for a much less common trigger for log force to occur. Ideally we should not reallocate busy extents. That is a much more complex fix to the problem as it involves direct intervention in the allocation btree searches in many places. This is left to a future set of modifications. Finally, now that we track busy extents in allocated memory, we don't need the descriptors in the transaction structure to point to them. We can replace the complex busy chunk infrastructure with a simple linked list of busy extents. This allows us to remove a large chunk of code, making the overall change a net reduction in code size. Signed-off-by: Dave Chinner <david@fromorbit.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2010-05-21 02:07:08 +00:00
if (atomic_read(&bp->b_pin_count) && (bp->b_flags & XBF_STALE))
xfs_log_force(bp->b_target->bt_mount, 0);
down(&bp->b_sema);
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_lock_done(bp, _RET_IP_);
}
void
xfs_buf_unlock(
struct xfs_buf *bp)
{
ASSERT(xfs_buf_islocked(bp));
up(&bp->b_sema);
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
trace_xfs_buf_unlock(bp, _RET_IP_);
}
STATIC void
xfs_buf_wait_unpin(
xfs_buf_t *bp)
{
DECLARE_WAITQUEUE (wait, current);
if (atomic_read(&bp->b_pin_count) == 0)
return;
add_wait_queue(&bp->b_waiters, &wait);
for (;;) {
set_current_state(TASK_UNINTERRUPTIBLE);
if (atomic_read(&bp->b_pin_count) == 0)
break;
io_schedule();
}
remove_wait_queue(&bp->b_waiters, &wait);
set_current_state(TASK_RUNNING);
}
/*
* Buffer Utility Routines
*/
void
xfs_buf_ioend(
struct xfs_buf *bp)
{
bool read = bp->b_flags & XBF_READ;
trace_xfs_buf_iodone(bp, _RET_IP_);
bp->b_flags &= ~(XBF_READ | XBF_WRITE | XBF_READ_AHEAD);
/*
* Pull in IO completion errors now. We are guaranteed to be running
* single threaded, so we don't need the lock to read b_io_error.
*/
if (!bp->b_error && bp->b_io_error)
xfs_buf_ioerror(bp, bp->b_io_error);
/* Only validate buffers that were read without errors */
if (read && !bp->b_error && bp->b_ops) {
ASSERT(!bp->b_iodone);
bp->b_ops->verify_read(bp);
}
if (!bp->b_error)
bp->b_flags |= XBF_DONE;
if (bp->b_iodone)
(*(bp->b_iodone))(bp);
else if (bp->b_flags & XBF_ASYNC)
xfs_buf_relse(bp);
else
complete(&bp->b_iowait);
}
static void
xfs_buf_ioend_work(
struct work_struct *work)
{
struct xfs_buf *bp =
container_of(work, xfs_buf_t, b_ioend_work);
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
xfs_buf_ioend(bp);
}
static void
xfs_buf_ioend_async(
struct xfs_buf *bp)
{
INIT_WORK(&bp->b_ioend_work, xfs_buf_ioend_work);
queue_work(bp->b_ioend_wq, &bp->b_ioend_work);
}
void
__xfs_buf_ioerror(
xfs_buf_t *bp,
int error,
xfs_failaddr_t failaddr)
{
ASSERT(error <= 0 && error >= -1000);
bp->b_error = error;
trace_xfs_buf_ioerror(bp, error, failaddr);
}
void
xfs_buf_ioerror_alert(
struct xfs_buf *bp,
const char *func)
{
xfs_alert(bp->b_target->bt_mount,
"metadata I/O error in \"%s\" at daddr 0x%llx len %d error %d",
func, (uint64_t)XFS_BUF_ADDR(bp), bp->b_length,
-bp->b_error);
}
int
xfs_bwrite(
struct xfs_buf *bp)
{
int error;
ASSERT(xfs_buf_islocked(bp));
bp->b_flags |= XBF_WRITE;
bp->b_flags &= ~(XBF_ASYNC | XBF_READ | _XBF_DELWRI_Q |
XBF_WRITE_FAIL | XBF_DONE);
error = xfs_buf_submit(bp);
if (error) {
xfs_force_shutdown(bp->b_target->bt_mount,
SHUTDOWN_META_IO_ERROR);
}
return error;
}
static void
xfs_buf_bio_end_io(
struct bio *bio)
{
struct xfs_buf *bp = (struct xfs_buf *)bio->bi_private;
/*
* don't overwrite existing errors - otherwise we can lose errors on
* buffers that require multiple bios to complete.
*/
if (bio->bi_status) {
int error = blk_status_to_errno(bio->bi_status);
cmpxchg(&bp->b_io_error, 0, error);
}
if (!bp->b_error && xfs_buf_is_vmapped(bp) && (bp->b_flags & XBF_READ))
invalidate_kernel_vmap_range(bp->b_addr, xfs_buf_vmap_len(bp));
if (atomic_dec_and_test(&bp->b_io_remaining) == 1)
xfs_buf_ioend_async(bp);
bio_put(bio);
}
static void
xfs_buf_ioapply_map(
struct xfs_buf *bp,
int map,
int *buf_offset,
int *count,
int op,
int op_flags)
{
int page_index;
int total_nr_pages = bp->b_page_count;
int nr_pages;
struct bio *bio;
sector_t sector = bp->b_maps[map].bm_bn;
int size;
int offset;
/* skip the pages in the buffer before the start offset */
page_index = 0;
offset = *buf_offset;
while (offset >= PAGE_SIZE) {
page_index++;
offset -= PAGE_SIZE;
}
/*
* Limit the IO size to the length of the current vector, and update the
* remaining IO count for the next time around.
*/
size = min_t(int, BBTOB(bp->b_maps[map].bm_len), *count);
*count -= size;
*buf_offset += size;
next_chunk:
atomic_inc(&bp->b_io_remaining);
nr_pages = min(total_nr_pages, BIO_MAX_PAGES);
bio = bio_alloc(GFP_NOIO, nr_pages);
bio_set_dev(bio, bp->b_target->bt_bdev);
block: Abstract out bvec iterator Immutable biovecs are going to require an explicit iterator. To implement immutable bvecs, a later patch is going to add a bi_bvec_done member to this struct; for now, this patch effectively just renames things. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: "Ed L. Cashin" <ecashin@coraid.com> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Lars Ellenberg <drbd-dev@lists.linbit.com> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Matthew Wilcox <willy@linux.intel.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Yehuda Sadeh <yehuda@inktank.com> Cc: Sage Weil <sage@inktank.com> Cc: Alex Elder <elder@inktank.com> Cc: ceph-devel@vger.kernel.org Cc: Joshua Morris <josh.h.morris@us.ibm.com> Cc: Philip Kelleher <pjk1939@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Neil Brown <neilb@suse.de> Cc: Alasdair Kergon <agk@redhat.com> Cc: Mike Snitzer <snitzer@redhat.com> Cc: dm-devel@redhat.com Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: linux390@de.ibm.com Cc: Boaz Harrosh <bharrosh@panasas.com> Cc: Benny Halevy <bhalevy@tonian.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Chris Mason <chris.mason@fusionio.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Andreas Dilger <adilger.kernel@dilger.ca> Cc: Jaegeuk Kim <jaegeuk.kim@samsung.com> Cc: Steven Whitehouse <swhiteho@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Joern Engel <joern@logfs.org> Cc: Prasad Joshi <prasadjoshi.linux@gmail.com> Cc: Trond Myklebust <Trond.Myklebust@netapp.com> Cc: KONISHI Ryusuke <konishi.ryusuke@lab.ntt.co.jp> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Ben Myers <bpm@sgi.com> Cc: xfs@oss.sgi.com Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Len Brown <len.brown@intel.com> Cc: Pavel Machek <pavel@ucw.cz> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Guo Chao <yan@linux.vnet.ibm.com> Cc: Tejun Heo <tj@kernel.org> Cc: Asai Thambi S P <asamymuthupa@micron.com> Cc: Selvan Mani <smani@micron.com> Cc: Sam Bradshaw <sbradshaw@micron.com> Cc: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Cc: "Roger Pau Monné" <roger.pau@citrix.com> Cc: Jan Beulich <jbeulich@suse.com> Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Cc: Ian Campbell <Ian.Campbell@citrix.com> Cc: Sebastian Ott <sebott@linux.vnet.ibm.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Jiang Liu <jiang.liu@huawei.com> Cc: Nitin Gupta <ngupta@vflare.org> Cc: Jerome Marchand <jmarchand@redhat.com> Cc: Joe Perches <joe@perches.com> Cc: Peng Tao <tao.peng@emc.com> Cc: Andy Adamson <andros@netapp.com> Cc: fanchaoting <fanchaoting@cn.fujitsu.com> Cc: Jie Liu <jeff.liu@oracle.com> Cc: Sunil Mushran <sunil.mushran@gmail.com> Cc: "Martin K. Petersen" <martin.petersen@oracle.com> Cc: Namjae Jeon <namjae.jeon@samsung.com> Cc: Pankaj Kumar <pankaj.km@samsung.com> Cc: Dan Magenheimer <dan.magenheimer@oracle.com> Cc: Mel Gorman <mgorman@suse.de>6
2013-10-11 22:44:27 +00:00
bio->bi_iter.bi_sector = sector;
bio->bi_end_io = xfs_buf_bio_end_io;
bio->bi_private = bp;
bio_set_op_attrs(bio, op, op_flags);
2011-03-25 22:16:45 +00:00
for (; size && nr_pages; nr_pages--, page_index++) {
2011-03-25 22:16:45 +00:00
int rbytes, nbytes = PAGE_SIZE - offset;
if (nbytes > size)
nbytes = size;
rbytes = bio_add_page(bio, bp->b_pages[page_index], nbytes,
offset);
if (rbytes < nbytes)
break;
offset = 0;
sector += BTOBB(nbytes);
size -= nbytes;
total_nr_pages--;
}
block: Abstract out bvec iterator Immutable biovecs are going to require an explicit iterator. To implement immutable bvecs, a later patch is going to add a bi_bvec_done member to this struct; for now, this patch effectively just renames things. Signed-off-by: Kent Overstreet <kmo@daterainc.com> Cc: Jens Axboe <axboe@kernel.dk> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Paul Mackerras <paulus@samba.org> Cc: "Ed L. Cashin" <ecashin@coraid.com> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Lars Ellenberg <drbd-dev@lists.linbit.com> Cc: Jiri Kosina <jkosina@suse.cz> Cc: Matthew Wilcox <willy@linux.intel.com> Cc: Geoff Levand <geoff@infradead.org> Cc: Yehuda Sadeh <yehuda@inktank.com> Cc: Sage Weil <sage@inktank.com> Cc: Alex Elder <elder@inktank.com> Cc: ceph-devel@vger.kernel.org Cc: Joshua Morris <josh.h.morris@us.ibm.com> Cc: Philip Kelleher <pjk1939@linux.vnet.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Cc: Neil Brown <neilb@suse.de> Cc: Alasdair Kergon <agk@redhat.com> Cc: Mike Snitzer <snitzer@redhat.com> Cc: dm-devel@redhat.com Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: linux390@de.ibm.com Cc: Boaz Harrosh <bharrosh@panasas.com> Cc: Benny Halevy <bhalevy@tonian.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Cc: "Nicholas A. Bellinger" <nab@linux-iscsi.org> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Chris Mason <chris.mason@fusionio.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Andreas Dilger <adilger.kernel@dilger.ca> Cc: Jaegeuk Kim <jaegeuk.kim@samsung.com> Cc: Steven Whitehouse <swhiteho@redhat.com> Cc: Dave Kleikamp <shaggy@kernel.org> Cc: Joern Engel <joern@logfs.org> Cc: Prasad Joshi <prasadjoshi.linux@gmail.com> Cc: Trond Myklebust <Trond.Myklebust@netapp.com> Cc: KONISHI Ryusuke <konishi.ryusuke@lab.ntt.co.jp> Cc: Mark Fasheh <mfasheh@suse.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Ben Myers <bpm@sgi.com> Cc: xfs@oss.sgi.com Cc: Steven Rostedt <rostedt@goodmis.org> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Len Brown <len.brown@intel.com> Cc: Pavel Machek <pavel@ucw.cz> Cc: "Rafael J. Wysocki" <rjw@sisk.pl> Cc: Herton Ronaldo Krzesinski <herton.krzesinski@canonical.com> Cc: Ben Hutchings <ben@decadent.org.uk> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Guo Chao <yan@linux.vnet.ibm.com> Cc: Tejun Heo <tj@kernel.org> Cc: Asai Thambi S P <asamymuthupa@micron.com> Cc: Selvan Mani <smani@micron.com> Cc: Sam Bradshaw <sbradshaw@micron.com> Cc: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Cc: "Roger Pau Monné" <roger.pau@citrix.com> Cc: Jan Beulich <jbeulich@suse.com> Cc: Stefano Stabellini <stefano.stabellini@eu.citrix.com> Cc: Ian Campbell <Ian.Campbell@citrix.com> Cc: Sebastian Ott <sebott@linux.vnet.ibm.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Minchan Kim <minchan@kernel.org> Cc: Jiang Liu <jiang.liu@huawei.com> Cc: Nitin Gupta <ngupta@vflare.org> Cc: Jerome Marchand <jmarchand@redhat.com> Cc: Joe Perches <joe@perches.com> Cc: Peng Tao <tao.peng@emc.com> Cc: Andy Adamson <andros@netapp.com> Cc: fanchaoting <fanchaoting@cn.fujitsu.com> Cc: Jie Liu <jeff.liu@oracle.com> Cc: Sunil Mushran <sunil.mushran@gmail.com> Cc: "Martin K. Petersen" <martin.petersen@oracle.com> Cc: Namjae Jeon <namjae.jeon@samsung.com> Cc: Pankaj Kumar <pankaj.km@samsung.com> Cc: Dan Magenheimer <dan.magenheimer@oracle.com> Cc: Mel Gorman <mgorman@suse.de>6
2013-10-11 22:44:27 +00:00
if (likely(bio->bi_iter.bi_size)) {
if (xfs_buf_is_vmapped(bp)) {
flush_kernel_vmap_range(bp->b_addr,
xfs_buf_vmap_len(bp));
}
submit_bio(bio);
if (size)
goto next_chunk;
} else {
/*
* This is guaranteed not to be the last io reference count
* because the caller (xfs_buf_submit) holds a count itself.
*/
atomic_dec(&bp->b_io_remaining);
xfs_buf_ioerror(bp, -EIO);
bio_put(bio);
}
}
STATIC void
_xfs_buf_ioapply(
struct xfs_buf *bp)
{
struct blk_plug plug;
int op;
int op_flags = 0;
int offset;
int size;
int i;
/*
* Make sure we capture only current IO errors rather than stale errors
* left over from previous use of the buffer (e.g. failed readahead).
*/
bp->b_error = 0;
/*
* Initialize the I/O completion workqueue if we haven't yet or the
* submitter has not opted to specify a custom one.
*/
if (!bp->b_ioend_wq)
bp->b_ioend_wq = bp->b_target->bt_mount->m_buf_workqueue;
if (bp->b_flags & XBF_WRITE) {
op = REQ_OP_WRITE;
if (bp->b_flags & XBF_SYNCIO)
op_flags = REQ_SYNC;
if (bp->b_flags & XBF_FUA)
op_flags |= REQ_FUA;
if (bp->b_flags & XBF_FLUSH)
op_flags |= REQ_PREFLUSH;
/*
* Run the write verifier callback function if it exists. If
* this function fails it will mark the buffer with an error and
* the IO should not be dispatched.
*/
if (bp->b_ops) {
bp->b_ops->verify_write(bp);
if (bp->b_error) {
xfs_force_shutdown(bp->b_target->bt_mount,
SHUTDOWN_CORRUPT_INCORE);
return;
}
} else if (bp->b_bn != XFS_BUF_DADDR_NULL) {
struct xfs_mount *mp = bp->b_target->bt_mount;
/*
* non-crc filesystems don't attach verifiers during
* log recovery, so don't warn for such filesystems.
*/
if (xfs_sb_version_hascrc(&mp->m_sb)) {
xfs_warn(mp,
"%s: no buf ops on daddr 0x%llx len %d",
__func__, bp->b_bn, bp->b_length);
xfs_hex_dump(bp->b_addr,
XFS_CORRUPTION_DUMP_LEN);
dump_stack();
}
}
} else if (bp->b_flags & XBF_READ_AHEAD) {
op = REQ_OP_READ;
op_flags = REQ_RAHEAD;
} else {
op = REQ_OP_READ;
}
/* we only use the buffer cache for meta-data */
op_flags |= REQ_META;
/*
* Walk all the vectors issuing IO on them. Set up the initial offset
* into the buffer and the desired IO size before we start -
* _xfs_buf_ioapply_vec() will modify them appropriately for each
* subsequent call.
*/
offset = bp->b_offset;
size = BBTOB(bp->b_io_length);
blk_start_plug(&plug);
for (i = 0; i < bp->b_map_count; i++) {
xfs_buf_ioapply_map(bp, i, &offset, &size, op, op_flags);
if (bp->b_error)
break;
if (size <= 0)
break; /* all done */
}
blk_finish_plug(&plug);
}
/*
* Wait for I/O completion of a sync buffer and return the I/O error code.
*/
static int
xfs_buf_iowait(
struct xfs_buf *bp)
{
ASSERT(!(bp->b_flags & XBF_ASYNC));
trace_xfs_buf_iowait(bp, _RET_IP_);
wait_for_completion(&bp->b_iowait);
trace_xfs_buf_iowait_done(bp, _RET_IP_);
return bp->b_error;
}
/*
* Buffer I/O submission path, read or write. Asynchronous submission transfers
* the buffer lock ownership and the current reference to the IO. It is not
* safe to reference the buffer after a call to this function unless the caller
* holds an additional reference itself.
*/
int
__xfs_buf_submit(
struct xfs_buf *bp,
bool wait)
{
int error = 0;
trace_xfs_buf_submit(bp, _RET_IP_);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
ASSERT(!(bp->b_flags & _XBF_DELWRI_Q));
/* on shutdown we stale and complete the buffer immediately */
if (XFS_FORCED_SHUTDOWN(bp->b_target->bt_mount)) {
xfs_buf_ioerror(bp, -EIO);
bp->b_flags &= ~XBF_DONE;
xfs_buf_stale(bp);
xfs_buf_ioend(bp);
return -EIO;
}
/*
* Grab a reference so the buffer does not go away underneath us. For
* async buffers, I/O completion drops the callers reference, which
* could occur before submission returns.
*/
xfs_buf_hold(bp);
if (bp->b_flags & XBF_WRITE)
xfs_buf_wait_unpin(bp);
/* clear the internal error state to avoid spurious errors */
bp->b_io_error = 0;
/*
* Set the count to 1 initially, this will stop an I/O completion
* callout which happens before we have started all the I/O from calling
* xfs_buf_ioend too early.
*/
atomic_set(&bp->b_io_remaining, 1);
if (bp->b_flags & XBF_ASYNC)
xfs_buf_ioacct_inc(bp);
_xfs_buf_ioapply(bp);
/*
* If _xfs_buf_ioapply failed, we can get back here with only the IO
* reference we took above. If we drop it to zero, run completion so
* that we don't return to the caller with completion still pending.
*/
if (atomic_dec_and_test(&bp->b_io_remaining) == 1) {
if (bp->b_error || !(bp->b_flags & XBF_ASYNC))
xfs_buf_ioend(bp);
else
xfs_buf_ioend_async(bp);
}
if (wait)
error = xfs_buf_iowait(bp);
/*
* Release the hold that keeps the buffer referenced for the entire
* I/O. Note that if the buffer is async, it is not safe to reference
* after this release.
*/
xfs_buf_rele(bp);
return error;
}
void *
xfs_buf_offset(
struct xfs_buf *bp,
size_t offset)
{
struct page *page;
if (bp->b_addr)
return bp->b_addr + offset;
offset += bp->b_offset;
2011-03-25 22:16:45 +00:00
page = bp->b_pages[offset >> PAGE_SHIFT];
return page_address(page) + (offset & (PAGE_SIZE-1));
}
/*
* Move data into or out of a buffer.
*/
void
xfs_buf_iomove(
xfs_buf_t *bp, /* buffer to process */
size_t boff, /* starting buffer offset */
size_t bsize, /* length to copy */
void *data, /* data address */
xfs_buf_rw_t mode) /* read/write/zero flag */
{
size_t bend;
bend = boff + bsize;
while (boff < bend) {
struct page *page;
int page_index, page_offset, csize;
page_index = (boff + bp->b_offset) >> PAGE_SHIFT;
page_offset = (boff + bp->b_offset) & ~PAGE_MASK;
page = bp->b_pages[page_index];
csize = min_t(size_t, PAGE_SIZE - page_offset,
BBTOB(bp->b_io_length) - boff);
ASSERT((csize + page_offset) <= PAGE_SIZE);
switch (mode) {
case XBRW_ZERO:
memset(page_address(page) + page_offset, 0, csize);
break;
case XBRW_READ:
memcpy(data, page_address(page) + page_offset, csize);
break;
case XBRW_WRITE:
memcpy(page_address(page) + page_offset, data, csize);
}
boff += csize;
data += csize;
}
}
/*
* Handling of buffer targets (buftargs).
*/
/*
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
* Wait for any bufs with callbacks that have been submitted but have not yet
* returned. These buffers will have an elevated hold count, so wait on those
* while freeing all the buffers only held by the LRU.
*/
static enum lru_status
xfs_buftarg_wait_rele(
struct list_head *item,
list_lru: add helpers to isolate items Currently, the isolate callback passed to the list_lru_walk family of functions is supposed to just delete an item from the list upon returning LRU_REMOVED or LRU_REMOVED_RETRY, while nr_items counter is fixed by __list_lru_walk_one after the callback returns. Since the callback is allowed to drop the lock after removing an item (it has to return LRU_REMOVED_RETRY then), the nr_items can be less than the actual number of elements on the list even if we check them under the lock. This makes it difficult to move items from one list_lru_one to another, which is required for per-memcg list_lru reparenting - we can't just splice the lists, we have to move entries one by one. This patch therefore introduces helpers that must be used by callback functions to isolate items instead of raw list_del/list_move. These are list_lru_isolate and list_lru_isolate_move. They not only remove the entry from the list, but also fix the nr_items counter, making sure nr_items always reflects the actual number of elements on the list if checked under the appropriate lock. Signed-off-by: Vladimir Davydov <vdavydov@parallels.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> 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: Dave Chinner <david@fromorbit.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-12 22:59:35 +00:00
struct list_lru_one *lru,
spinlock_t *lru_lock,
void *arg)
{
struct xfs_buf *bp = container_of(item, struct xfs_buf, b_lru);
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
struct list_head *dispose = arg;
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
if (atomic_read(&bp->b_hold) > 1) {
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
/* need to wait, so skip it this pass */
trace_xfs_buf_wait_buftarg(bp, _RET_IP_);
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
return LRU_SKIP;
}
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
if (!spin_trylock(&bp->b_lock))
return LRU_SKIP;
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
/*
* clear the LRU reference count so the buffer doesn't get
* ignored in xfs_buf_rele().
*/
atomic_set(&bp->b_lru_ref, 0);
bp->b_state |= XFS_BSTATE_DISPOSE;
list_lru: add helpers to isolate items Currently, the isolate callback passed to the list_lru_walk family of functions is supposed to just delete an item from the list upon returning LRU_REMOVED or LRU_REMOVED_RETRY, while nr_items counter is fixed by __list_lru_walk_one after the callback returns. Since the callback is allowed to drop the lock after removing an item (it has to return LRU_REMOVED_RETRY then), the nr_items can be less than the actual number of elements on the list even if we check them under the lock. This makes it difficult to move items from one list_lru_one to another, which is required for per-memcg list_lru reparenting - we can't just splice the lists, we have to move entries one by one. This patch therefore introduces helpers that must be used by callback functions to isolate items instead of raw list_del/list_move. These are list_lru_isolate and list_lru_isolate_move. They not only remove the entry from the list, but also fix the nr_items counter, making sure nr_items always reflects the actual number of elements on the list if checked under the appropriate lock. Signed-off-by: Vladimir Davydov <vdavydov@parallels.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> 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: Dave Chinner <david@fromorbit.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-12 22:59:35 +00:00
list_lru_isolate_move(lru, item, dispose);
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
spin_unlock(&bp->b_lock);
return LRU_REMOVED;
}
void
xfs_wait_buftarg(
struct xfs_buftarg *btp)
{
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
LIST_HEAD(dispose);
int loop = 0;
xfs: log mount failures don't wait for buffers to be released Recently I've been seeing xfs/051 fail on 1k block size filesystems. Trying to trace the events during the test lead to the problem going away, indicating that it was a race condition that lead to this ASSERT failure: XFS: Assertion failed: atomic_read(&pag->pag_ref) == 0, file: fs/xfs/xfs_mount.c, line: 156 ..... [<ffffffff814e1257>] xfs_free_perag+0x87/0xb0 [<ffffffff814e21b9>] xfs_mountfs+0x4d9/0x900 [<ffffffff814e5dff>] xfs_fs_fill_super+0x3bf/0x4d0 [<ffffffff811d8800>] mount_bdev+0x180/0x1b0 [<ffffffff814e3ff5>] xfs_fs_mount+0x15/0x20 [<ffffffff811d90a8>] mount_fs+0x38/0x170 [<ffffffff811f4347>] vfs_kern_mount+0x67/0x120 [<ffffffff811f7018>] do_mount+0x218/0xd60 [<ffffffff811f7e5b>] SyS_mount+0x8b/0xd0 When I finally caught it with tracing enabled, I saw that AG 2 had an elevated reference count and a buffer was responsible for it. I tracked down the specific buffer, and found that it was missing the final reference count release that would put it back on the LRU and hence be found by xfs_wait_buftarg() calls in the log mount failure handling. The last four traces for the buffer before the assert were (trimmed for relevance) kworker/0:1-5259 xfs_buf_iodone: hold 2 lock 0 flags ASYNC kworker/0:1-5259 xfs_buf_ioerror: hold 2 lock 0 error -5 mount-7163 xfs_buf_lock_done: hold 2 lock 0 flags ASYNC mount-7163 xfs_buf_unlock: hold 2 lock 1 flags ASYNC This is an async write that is completing, so there's nobody waiting for it directly. Hence we call xfs_buf_relse() once all the processing is complete. That does: static inline void xfs_buf_relse(xfs_buf_t *bp) { xfs_buf_unlock(bp); xfs_buf_rele(bp); } Now, it's clear that mount is waiting on the buffer lock, and that it has been released by xfs_buf_relse() and gained by mount. This is expected, because at this point the mount process is in xfs_buf_delwri_submit() waiting for all the IO it submitted to complete. The mount process, however, is waiting on the lock for the buffer because it is in xfs_buf_delwri_submit(). This waits for IO completion, but it doesn't wait for the buffer reference owned by the IO to go away. The mount process collects all the completions, fails the log recovery, and the higher level code then calls xfs_wait_buftarg() to free all the remaining buffers in the filesystem. The issue is that on unlocking the buffer, the scheduler has decided that the mount process has higher priority than the the kworker thread that is running the IO completion, and so immediately switched contexts to the mount process from the semaphore unlock code, hence preventing the kworker thread from finishing the IO completion and releasing the IO reference to the buffer. Hence by the time that xfs_wait_buftarg() is run, the buffer still has an active reference and so isn't on the LRU list that the function walks to free the remaining buffers. Hence we miss that buffer and continue onwards to tear down the mount structures, at which time we get find a stray reference count on the perag structure. On a non-debug kernel, this will be ignored and the structure torn down and freed. Hence when the kworker thread is then rescheduled and the buffer released and freed, it will access a freed perag structure. The problem here is that when the log mount fails, we still need to quiesce the log to ensure that the IO workqueues have returned to idle before we run xfs_wait_buftarg(). By synchronising the workqueues, we ensure that all IO completions are fully processed, not just to the point where buffers have been unlocked. This ensures we don't end up in the situation above. cc: <stable@vger.kernel.org> # 3.18 Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-01-18 21:28:10 +00:00
/*
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
* First wait on the buftarg I/O count for all in-flight buffers to be
* released. This is critical as new buffers do not make the LRU until
* they are released.
*
* Next, flush the buffer workqueue to ensure all completion processing
* has finished. Just waiting on buffer locks is not sufficient for
* async IO as the reference count held over IO is not released until
* after the buffer lock is dropped. Hence we need to ensure here that
* all reference counts have been dropped before we start walking the
* LRU list.
xfs: log mount failures don't wait for buffers to be released Recently I've been seeing xfs/051 fail on 1k block size filesystems. Trying to trace the events during the test lead to the problem going away, indicating that it was a race condition that lead to this ASSERT failure: XFS: Assertion failed: atomic_read(&pag->pag_ref) == 0, file: fs/xfs/xfs_mount.c, line: 156 ..... [<ffffffff814e1257>] xfs_free_perag+0x87/0xb0 [<ffffffff814e21b9>] xfs_mountfs+0x4d9/0x900 [<ffffffff814e5dff>] xfs_fs_fill_super+0x3bf/0x4d0 [<ffffffff811d8800>] mount_bdev+0x180/0x1b0 [<ffffffff814e3ff5>] xfs_fs_mount+0x15/0x20 [<ffffffff811d90a8>] mount_fs+0x38/0x170 [<ffffffff811f4347>] vfs_kern_mount+0x67/0x120 [<ffffffff811f7018>] do_mount+0x218/0xd60 [<ffffffff811f7e5b>] SyS_mount+0x8b/0xd0 When I finally caught it with tracing enabled, I saw that AG 2 had an elevated reference count and a buffer was responsible for it. I tracked down the specific buffer, and found that it was missing the final reference count release that would put it back on the LRU and hence be found by xfs_wait_buftarg() calls in the log mount failure handling. The last four traces for the buffer before the assert were (trimmed for relevance) kworker/0:1-5259 xfs_buf_iodone: hold 2 lock 0 flags ASYNC kworker/0:1-5259 xfs_buf_ioerror: hold 2 lock 0 error -5 mount-7163 xfs_buf_lock_done: hold 2 lock 0 flags ASYNC mount-7163 xfs_buf_unlock: hold 2 lock 1 flags ASYNC This is an async write that is completing, so there's nobody waiting for it directly. Hence we call xfs_buf_relse() once all the processing is complete. That does: static inline void xfs_buf_relse(xfs_buf_t *bp) { xfs_buf_unlock(bp); xfs_buf_rele(bp); } Now, it's clear that mount is waiting on the buffer lock, and that it has been released by xfs_buf_relse() and gained by mount. This is expected, because at this point the mount process is in xfs_buf_delwri_submit() waiting for all the IO it submitted to complete. The mount process, however, is waiting on the lock for the buffer because it is in xfs_buf_delwri_submit(). This waits for IO completion, but it doesn't wait for the buffer reference owned by the IO to go away. The mount process collects all the completions, fails the log recovery, and the higher level code then calls xfs_wait_buftarg() to free all the remaining buffers in the filesystem. The issue is that on unlocking the buffer, the scheduler has decided that the mount process has higher priority than the the kworker thread that is running the IO completion, and so immediately switched contexts to the mount process from the semaphore unlock code, hence preventing the kworker thread from finishing the IO completion and releasing the IO reference to the buffer. Hence by the time that xfs_wait_buftarg() is run, the buffer still has an active reference and so isn't on the LRU list that the function walks to free the remaining buffers. Hence we miss that buffer and continue onwards to tear down the mount structures, at which time we get find a stray reference count on the perag structure. On a non-debug kernel, this will be ignored and the structure torn down and freed. Hence when the kworker thread is then rescheduled and the buffer released and freed, it will access a freed perag structure. The problem here is that when the log mount fails, we still need to quiesce the log to ensure that the IO workqueues have returned to idle before we run xfs_wait_buftarg(). By synchronising the workqueues, we ensure that all IO completions are fully processed, not just to the point where buffers have been unlocked. This ensures we don't end up in the situation above. cc: <stable@vger.kernel.org> # 3.18 Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-01-18 21:28:10 +00:00
*/
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
while (percpu_counter_sum(&btp->bt_io_count))
delay(100);
xfs: prevent dropping ioend completions during buftarg wait xfs_wait_buftarg() waits for all pending I/O, drains the ioend completion workqueue and walks the LRU until all buffers in the cache have been released. This is traditionally an unmount operation` but the mechanism is also reused during filesystem freeze. xfs_wait_buftarg() invokes drain_workqueue() as part of the quiesce, which is intended more for a shutdown sequence in that it indicates to the queue that new operations are not expected once the drain has begun. New work jobs after this point result in a WARN_ON_ONCE() and are otherwise dropped. With filesystem freeze, however, read operations are allowed and can proceed during or after the workqueue drain. If such a read occurs during the drain sequence, the workqueue infrastructure complains about the queued ioend completion work item and drops it on the floor. As a result, the buffer remains on the LRU and the freeze never completes. Despite the fact that the overall buffer cache cleanup is not necessary during freeze, fix up this operation such that it is safe to invoke during non-unmount quiesce operations. Replace the drain_workqueue() call with flush_workqueue(), which runs a similar serialization on pending workqueue jobs without causing new jobs to be dropped. This is safe for unmount as unmount independently locks out new operations by the time xfs_wait_buftarg() is invoked. cc: <stable@vger.kernel.org> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-08-26 06:01:59 +00:00
flush_workqueue(btp->bt_mount->m_buf_workqueue);
xfs: log mount failures don't wait for buffers to be released Recently I've been seeing xfs/051 fail on 1k block size filesystems. Trying to trace the events during the test lead to the problem going away, indicating that it was a race condition that lead to this ASSERT failure: XFS: Assertion failed: atomic_read(&pag->pag_ref) == 0, file: fs/xfs/xfs_mount.c, line: 156 ..... [<ffffffff814e1257>] xfs_free_perag+0x87/0xb0 [<ffffffff814e21b9>] xfs_mountfs+0x4d9/0x900 [<ffffffff814e5dff>] xfs_fs_fill_super+0x3bf/0x4d0 [<ffffffff811d8800>] mount_bdev+0x180/0x1b0 [<ffffffff814e3ff5>] xfs_fs_mount+0x15/0x20 [<ffffffff811d90a8>] mount_fs+0x38/0x170 [<ffffffff811f4347>] vfs_kern_mount+0x67/0x120 [<ffffffff811f7018>] do_mount+0x218/0xd60 [<ffffffff811f7e5b>] SyS_mount+0x8b/0xd0 When I finally caught it with tracing enabled, I saw that AG 2 had an elevated reference count and a buffer was responsible for it. I tracked down the specific buffer, and found that it was missing the final reference count release that would put it back on the LRU and hence be found by xfs_wait_buftarg() calls in the log mount failure handling. The last four traces for the buffer before the assert were (trimmed for relevance) kworker/0:1-5259 xfs_buf_iodone: hold 2 lock 0 flags ASYNC kworker/0:1-5259 xfs_buf_ioerror: hold 2 lock 0 error -5 mount-7163 xfs_buf_lock_done: hold 2 lock 0 flags ASYNC mount-7163 xfs_buf_unlock: hold 2 lock 1 flags ASYNC This is an async write that is completing, so there's nobody waiting for it directly. Hence we call xfs_buf_relse() once all the processing is complete. That does: static inline void xfs_buf_relse(xfs_buf_t *bp) { xfs_buf_unlock(bp); xfs_buf_rele(bp); } Now, it's clear that mount is waiting on the buffer lock, and that it has been released by xfs_buf_relse() and gained by mount. This is expected, because at this point the mount process is in xfs_buf_delwri_submit() waiting for all the IO it submitted to complete. The mount process, however, is waiting on the lock for the buffer because it is in xfs_buf_delwri_submit(). This waits for IO completion, but it doesn't wait for the buffer reference owned by the IO to go away. The mount process collects all the completions, fails the log recovery, and the higher level code then calls xfs_wait_buftarg() to free all the remaining buffers in the filesystem. The issue is that on unlocking the buffer, the scheduler has decided that the mount process has higher priority than the the kworker thread that is running the IO completion, and so immediately switched contexts to the mount process from the semaphore unlock code, hence preventing the kworker thread from finishing the IO completion and releasing the IO reference to the buffer. Hence by the time that xfs_wait_buftarg() is run, the buffer still has an active reference and so isn't on the LRU list that the function walks to free the remaining buffers. Hence we miss that buffer and continue onwards to tear down the mount structures, at which time we get find a stray reference count on the perag structure. On a non-debug kernel, this will be ignored and the structure torn down and freed. Hence when the kworker thread is then rescheduled and the buffer released and freed, it will access a freed perag structure. The problem here is that when the log mount fails, we still need to quiesce the log to ensure that the IO workqueues have returned to idle before we run xfs_wait_buftarg(). By synchronising the workqueues, we ensure that all IO completions are fully processed, not just to the point where buffers have been unlocked. This ensures we don't end up in the situation above. cc: <stable@vger.kernel.org> # 3.18 Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-01-18 21:28:10 +00:00
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
/* loop until there is nothing left on the lru list. */
while (list_lru_count(&btp->bt_lru)) {
list_lru_walk(&btp->bt_lru, xfs_buftarg_wait_rele,
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
&dispose, LONG_MAX);
while (!list_empty(&dispose)) {
struct xfs_buf *bp;
bp = list_first_entry(&dispose, struct xfs_buf, b_lru);
list_del_init(&bp->b_lru);
xfs: abort metadata writeback on permanent errors If we are doing aysnc writeback of metadata, we can get write errors but have nobody to report them to. At the moment, we simply attempt to reissue the write from io completion in the hope that it's a transient error. When it's not a transient error, the buffer is stuck forever in this loop, and we cannot break out of it. Eventually, unmount will hang because the AIL cannot be emptied and everything goes downhill from them. To solve this problem, only retry the write IO once before aborting it. We don't throw the buffer away because some transient errors can last minutes (e.g. FC path failover) or even hours (thin provisioned devices that have run out of backing space) before they go away. Hence we really want to keep trying until we can't try any more. Because the buffer was not cleaned, however, it does not get removed from the AIL and hence the next pass across the AIL will start IO on it again. As such, we still get the "retry forever" semantics that we currently have, but we allow other access to the buffer in the mean time. Meanwhile the filesystem can continue to modify the buffer and relog it, so the IO errors won't hang the log or the filesystem. Now when we are pushing the AIL, we can see all these "permanent IO error" buffers and we can issue a warning about failures before we retry the IO. We can also catch these buffers when unmounting an issue a corruption warning, too. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-12-12 05:34:38 +00:00
if (bp->b_flags & XBF_WRITE_FAIL) {
xfs_alert(btp->bt_mount,
"Corruption Alert: Buffer at daddr 0x%llx had permanent write failures!",
xfs: abort metadata writeback on permanent errors If we are doing aysnc writeback of metadata, we can get write errors but have nobody to report them to. At the moment, we simply attempt to reissue the write from io completion in the hope that it's a transient error. When it's not a transient error, the buffer is stuck forever in this loop, and we cannot break out of it. Eventually, unmount will hang because the AIL cannot be emptied and everything goes downhill from them. To solve this problem, only retry the write IO once before aborting it. We don't throw the buffer away because some transient errors can last minutes (e.g. FC path failover) or even hours (thin provisioned devices that have run out of backing space) before they go away. Hence we really want to keep trying until we can't try any more. Because the buffer was not cleaned, however, it does not get removed from the AIL and hence the next pass across the AIL will start IO on it again. As such, we still get the "retry forever" semantics that we currently have, but we allow other access to the buffer in the mean time. Meanwhile the filesystem can continue to modify the buffer and relog it, so the IO errors won't hang the log or the filesystem. Now when we are pushing the AIL, we can see all these "permanent IO error" buffers and we can issue a warning about failures before we retry the IO. We can also catch these buffers when unmounting an issue a corruption warning, too. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-12-12 05:34:38 +00:00
(long long)bp->b_bn);
xfs_alert(btp->bt_mount,
"Please run xfs_repair to determine the extent of the problem.");
xfs: abort metadata writeback on permanent errors If we are doing aysnc writeback of metadata, we can get write errors but have nobody to report them to. At the moment, we simply attempt to reissue the write from io completion in the hope that it's a transient error. When it's not a transient error, the buffer is stuck forever in this loop, and we cannot break out of it. Eventually, unmount will hang because the AIL cannot be emptied and everything goes downhill from them. To solve this problem, only retry the write IO once before aborting it. We don't throw the buffer away because some transient errors can last minutes (e.g. FC path failover) or even hours (thin provisioned devices that have run out of backing space) before they go away. Hence we really want to keep trying until we can't try any more. Because the buffer was not cleaned, however, it does not get removed from the AIL and hence the next pass across the AIL will start IO on it again. As such, we still get the "retry forever" semantics that we currently have, but we allow other access to the buffer in the mean time. Meanwhile the filesystem can continue to modify the buffer and relog it, so the IO errors won't hang the log or the filesystem. Now when we are pushing the AIL, we can see all these "permanent IO error" buffers and we can issue a warning about failures before we retry the IO. We can also catch these buffers when unmounting an issue a corruption warning, too. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Ben Myers <bpm@sgi.com>
2013-12-12 05:34:38 +00:00
}
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
xfs_buf_rele(bp);
}
if (loop++ != 0)
delay(100);
}
}
static enum lru_status
xfs_buftarg_isolate(
struct list_head *item,
list_lru: add helpers to isolate items Currently, the isolate callback passed to the list_lru_walk family of functions is supposed to just delete an item from the list upon returning LRU_REMOVED or LRU_REMOVED_RETRY, while nr_items counter is fixed by __list_lru_walk_one after the callback returns. Since the callback is allowed to drop the lock after removing an item (it has to return LRU_REMOVED_RETRY then), the nr_items can be less than the actual number of elements on the list even if we check them under the lock. This makes it difficult to move items from one list_lru_one to another, which is required for per-memcg list_lru reparenting - we can't just splice the lists, we have to move entries one by one. This patch therefore introduces helpers that must be used by callback functions to isolate items instead of raw list_del/list_move. These are list_lru_isolate and list_lru_isolate_move. They not only remove the entry from the list, but also fix the nr_items counter, making sure nr_items always reflects the actual number of elements on the list if checked under the appropriate lock. Signed-off-by: Vladimir Davydov <vdavydov@parallels.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> 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: Dave Chinner <david@fromorbit.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-12 22:59:35 +00:00
struct list_lru_one *lru,
spinlock_t *lru_lock,
void *arg)
{
struct xfs_buf *bp = container_of(item, struct xfs_buf, b_lru);
struct list_head *dispose = arg;
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
/*
* we are inverting the lru lock/bp->b_lock here, so use a trylock.
* If we fail to get the lock, just skip it.
*/
if (!spin_trylock(&bp->b_lock))
return LRU_SKIP;
/*
* Decrement the b_lru_ref count unless the value is already
* zero. If the value is already zero, we need to reclaim the
* buffer, otherwise it gets another trip through the LRU.
*/
if (atomic_add_unless(&bp->b_lru_ref, -1, 0)) {
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
spin_unlock(&bp->b_lock);
return LRU_ROTATE;
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
}
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
bp->b_state |= XFS_BSTATE_DISPOSE;
list_lru: add helpers to isolate items Currently, the isolate callback passed to the list_lru_walk family of functions is supposed to just delete an item from the list upon returning LRU_REMOVED or LRU_REMOVED_RETRY, while nr_items counter is fixed by __list_lru_walk_one after the callback returns. Since the callback is allowed to drop the lock after removing an item (it has to return LRU_REMOVED_RETRY then), the nr_items can be less than the actual number of elements on the list even if we check them under the lock. This makes it difficult to move items from one list_lru_one to another, which is required for per-memcg list_lru reparenting - we can't just splice the lists, we have to move entries one by one. This patch therefore introduces helpers that must be used by callback functions to isolate items instead of raw list_del/list_move. These are list_lru_isolate and list_lru_isolate_move. They not only remove the entry from the list, but also fix the nr_items counter, making sure nr_items always reflects the actual number of elements on the list if checked under the appropriate lock. Signed-off-by: Vladimir Davydov <vdavydov@parallels.com> Cc: Johannes Weiner <hannes@cmpxchg.org> Cc: Michal Hocko <mhocko@suse.cz> Cc: Tejun Heo <tj@kernel.org> 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: Dave Chinner <david@fromorbit.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-12 22:59:35 +00:00
list_lru_isolate_move(lru, item, dispose);
xfs: rework buffer dispose list tracking In converting the buffer lru lists to use the generic code, the locking for marking the buffers as on the dispose list was lost. This results in confusion in LRU buffer tracking and acocunting, resulting in reference counts being mucked up and filesystem beig unmountable. To fix this, introduce an internal buffer spinlock to protect the state field that holds the dispose list information. Because there is now locking needed around xfs_buf_lru_add/del, and they are used in exactly one place each two lines apart, get rid of the wrappers and code the logic directly in place. Further, the LRU emptying code used on unmount is less than optimal. Convert it to use a dispose list as per a normal shrinker walk, and repeat the walk that fills the dispose list until the LRU is empty. Thi avoids needing to drop and regain the LRU lock for every item being freed, and allows the same logic as the shrinker isolate call to be used. Simpler, easier to understand. Signed-off-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Glauber Costa <glommer@openvz.org> 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: 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 00:18:06 +00:00
spin_unlock(&bp->b_lock);
return LRU_REMOVED;
}
static unsigned long
xfs_buftarg_shrink_scan(
struct shrinker *shrink,
struct shrink_control *sc)
{
struct xfs_buftarg *btp = container_of(shrink,
struct xfs_buftarg, bt_shrinker);
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
LIST_HEAD(dispose);
unsigned long freed;
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
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>
2015-02-12 22:58:47 +00:00
freed = list_lru_shrink_walk(&btp->bt_lru, sc,
xfs_buftarg_isolate, &dispose);
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
while (!list_empty(&dispose)) {
struct xfs_buf *bp;
xfs: add a lru to the XFS buffer cache Introduce a per-buftarg LRU for memory reclaim to operate on. This is the last piece we need to put in place so that we can fully control the buffer lifecycle. This allows XFS to be responsibile for maintaining the working set of buffers under memory pressure instead of relying on the VM reclaim not to take pages we need out from underneath us. The implementation introduces a b_lru_ref counter into the buffer. This is currently set to 1 whenever the buffer is referenced and so is used to determine if the buffer should be added to the LRU or not when freed. Effectively it allows lazy LRU initialisation of the buffer so we do not need to touch the LRU list and locks in xfs_buf_find(). Instead, when the buffer is being released and we drop the last reference to it, we check the b_lru_ref count and if it is none zero we re-add the buffer reference and add the inode to the LRU. The b_lru_ref counter is decremented by the shrinker, and whenever the shrinker comes across a buffer with a zero b_lru_ref counter, if released the LRU reference on the buffer. In the absence of a lookup race, this will result in the buffer being freed. This counting mechanism is used instead of a reference flag so that it is simple to re-introduce buffer-type specific reclaim reference counts to prioritise reclaim more effectively. We still have all those hooks in the XFS code, so this will provide the infrastructure to re-implement that functionality. Signed-off-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Christoph Hellwig <hch@lst.de>
2010-12-02 05:30:55 +00:00
bp = list_first_entry(&dispose, struct xfs_buf, b_lru);
list_del_init(&bp->b_lru);
xfs_buf_rele(bp);
}
return freed;
}
static unsigned long
xfs_buftarg_shrink_count(
struct shrinker *shrink,
struct shrink_control *sc)
{
struct xfs_buftarg *btp = container_of(shrink,
struct xfs_buftarg, bt_shrinker);
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>
2015-02-12 22:58:47 +00:00
return list_lru_shrink_count(&btp->bt_lru, sc);
}
void
xfs_free_buftarg(
struct xfs_buftarg *btp)
{
unregister_shrinker(&btp->bt_shrinker);
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
ASSERT(percpu_counter_sum(&btp->bt_io_count) == 0);
percpu_counter_destroy(&btp->bt_io_count);
list_lru_destroy(&btp->bt_lru);
xfs_blkdev_issue_flush(btp);
kmem_free(btp);
}
int
xfs_setsize_buftarg(
xfs_buftarg_t *btp,
unsigned int sectorsize)
{
xfs: allow logical-sector sized O_DIRECT Some time ago, mkfs.xfs started picking the storage physical sector size as the default filesystem "sector size" in order to avoid RMW costs incurred by doing IOs at logical sector size alignments. However, this means that for a filesystem made with i.e. a 4k sector size on an "advanced format" 4k/512 disk, 512-byte direct IOs are no longer allowed. This means that XFS has essentially turned this AF drive into a hard 4K device, from the filesystem on up. XFS's mkfs-specified "sector size" is really just controlling the minimum size & alignment of filesystem metadata. There is no real need to tightly couple XFS's minimal metadata size to the minimum allowed direct IO size; XFS can continue doing metadata in optimal sizes, but still allow smaller DIOs for apps which issue them, for whatever reason. This patch adds a new field to the xfs_buftarg, so that we now track 2 sizes: 1) The metadata sector size, which is the minimum unit and alignment of IO which will be performed by metadata operations. 2) The device logical sector size The first is used internally by the file system for metadata alignment and IOs. The second is used for the minimum allowed direct IO alignment. This has passed xfstests on filesystems made with 4k sectors, including when run under the patch I sent to ignore XFS_IOC_DIOINFO, and issue 512 DIOs anyway. I also directly tested end of block behavior on preallocated, sparse, and existing files when we do a 512 IO into a 4k file on a 4k-sector filesystem, to be sure there were no unexpected behaviors. Signed-off-by: Eric Sandeen <sandeen@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2014-01-21 22:46:23 +00:00
/* Set up metadata sector size info */
btp->bt_meta_sectorsize = sectorsize;
btp->bt_meta_sectormask = sectorsize - 1;
if (set_blocksize(btp->bt_bdev, sectorsize)) {
xfs_warn(btp->bt_mount,
"Cannot set_blocksize to %u on device %pg",
sectorsize, btp->bt_bdev);
return -EINVAL;
}
xfs: allow logical-sector sized O_DIRECT Some time ago, mkfs.xfs started picking the storage physical sector size as the default filesystem "sector size" in order to avoid RMW costs incurred by doing IOs at logical sector size alignments. However, this means that for a filesystem made with i.e. a 4k sector size on an "advanced format" 4k/512 disk, 512-byte direct IOs are no longer allowed. This means that XFS has essentially turned this AF drive into a hard 4K device, from the filesystem on up. XFS's mkfs-specified "sector size" is really just controlling the minimum size & alignment of filesystem metadata. There is no real need to tightly couple XFS's minimal metadata size to the minimum allowed direct IO size; XFS can continue doing metadata in optimal sizes, but still allow smaller DIOs for apps which issue them, for whatever reason. This patch adds a new field to the xfs_buftarg, so that we now track 2 sizes: 1) The metadata sector size, which is the minimum unit and alignment of IO which will be performed by metadata operations. 2) The device logical sector size The first is used internally by the file system for metadata alignment and IOs. The second is used for the minimum allowed direct IO alignment. This has passed xfstests on filesystems made with 4k sectors, including when run under the patch I sent to ignore XFS_IOC_DIOINFO, and issue 512 DIOs anyway. I also directly tested end of block behavior on preallocated, sparse, and existing files when we do a 512 IO into a 4k file on a 4k-sector filesystem, to be sure there were no unexpected behaviors. Signed-off-by: Eric Sandeen <sandeen@redhat.com> Reviewed-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2014-01-21 22:46:23 +00:00
/* Set up device logical sector size mask */
btp->bt_logical_sectorsize = bdev_logical_block_size(btp->bt_bdev);
btp->bt_logical_sectormask = bdev_logical_block_size(btp->bt_bdev) - 1;
return 0;
}
/*
* When allocating the initial buffer target we have not yet
* read in the superblock, so don't know what sized sectors
* are being used at this early stage. Play safe.
*/
STATIC int
xfs_setsize_buftarg_early(
xfs_buftarg_t *btp,
struct block_device *bdev)
{
return xfs_setsize_buftarg(btp, bdev_logical_block_size(bdev));
}
xfs_buftarg_t *
xfs_alloc_buftarg(
struct xfs_mount *mp,
struct block_device *bdev,
struct dax_device *dax_dev)
{
xfs_buftarg_t *btp;
btp = kmem_zalloc(sizeof(*btp), KM_SLEEP | KM_NOFS);
btp->bt_mount = mp;
btp->bt_dev = bdev->bd_dev;
btp->bt_bdev = bdev;
btp->bt_daxdev = dax_dev;
2011-03-25 22:16:45 +00:00
if (xfs_setsize_buftarg_early(btp, bdev))
goto error_free;
list_lru: dynamically adjust node arrays We currently use a compile-time constant to size the node array for the list_lru structure. Due to this, we don't need to allocate any memory at initialization time. But as a consequence, the structures that contain embedded list_lru lists can become way too big (the superblock for instance contains two of them). This patch aims at ameliorating this situation by dynamically allocating the node arrays with the firmware provided nr_node_ids. Signed-off-by: Glauber Costa <glommer@openvz.org> Cc: Dave Chinner <dchinner@redhat.com> Cc: Mel Gorman <mgorman@suse.de> 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: 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 00:18:18 +00:00
if (list_lru_init(&btp->bt_lru))
goto error_free;
list_lru: dynamically adjust node arrays We currently use a compile-time constant to size the node array for the list_lru structure. Due to this, we don't need to allocate any memory at initialization time. But as a consequence, the structures that contain embedded list_lru lists can become way too big (the superblock for instance contains two of them). This patch aims at ameliorating this situation by dynamically allocating the node arrays with the firmware provided nr_node_ids. Signed-off-by: Glauber Costa <glommer@openvz.org> Cc: Dave Chinner <dchinner@redhat.com> Cc: Mel Gorman <mgorman@suse.de> 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: 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 00:18:18 +00:00
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
if (percpu_counter_init(&btp->bt_io_count, 0, GFP_KERNEL))
goto error_lru;
xfs: track and serialize in-flight async buffers against unmount Newly allocated XFS metadata buffers are added to the LRU once the hold count is released, which typically occurs after I/O completion. There is no other mechanism at current that tracks the existence or I/O state of a new buffer. Further, readahead I/O tends to be submitted asynchronously by nature, which means the I/O can remain in flight and actually complete long after the calling context is gone. This means that file descriptors or any other holds on the filesystem can be released, allowing the filesystem to be unmounted while I/O is still in flight. When I/O completion occurs, core data structures may have been freed, causing completion to run into invalid memory accesses and likely to panic. This problem is reproduced on XFS via directory readahead. A filesystem is mounted, a directory is opened/closed and the filesystem immediately unmounted. The open/close cycle triggers a directory readahead that if delayed long enough, runs buffer I/O completion after the unmount has completed. To address this problem, add a mechanism to track all in-flight, asynchronous buffers using per-cpu counters in the buftarg. The buffer is accounted on the first I/O submission after the current reference is acquired and unaccounted once the buffer is returned to the LRU or freed. Update xfs_wait_buftarg() to wait on all in-flight I/O before walking the LRU list. Once in-flight I/O has completed and the workqueue has drained, all new buffers should have been released onto the LRU. Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-20 01:15:28 +00:00
btp->bt_shrinker.count_objects = xfs_buftarg_shrink_count;
btp->bt_shrinker.scan_objects = xfs_buftarg_shrink_scan;
btp->bt_shrinker.seeks = DEFAULT_SEEKS;
btp->bt_shrinker.flags = SHRINKER_NUMA_AWARE;
if (register_shrinker(&btp->bt_shrinker))
goto error_pcpu;
return btp;
error_pcpu:
percpu_counter_destroy(&btp->bt_io_count);
error_lru:
list_lru_destroy(&btp->bt_lru);
error_free:
kmem_free(btp);
return NULL;
}
/*
* Cancel a delayed write list.
*
* Remove each buffer from the list, clear the delwri queue flag and drop the
* associated buffer reference.
*/
void
xfs_buf_delwri_cancel(
struct list_head *list)
{
struct xfs_buf *bp;
while (!list_empty(list)) {
bp = list_first_entry(list, struct xfs_buf, b_list);
xfs_buf_lock(bp);
bp->b_flags &= ~_XBF_DELWRI_Q;
list_del_init(&bp->b_list);
xfs_buf_relse(bp);
}
}
/*
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
* Add a buffer to the delayed write list.
*
* This queues a buffer for writeout if it hasn't already been. Note that
* neither this routine nor the buffer list submission functions perform
* any internal synchronization. It is expected that the lists are thread-local
* to the callers.
*
* Returns true if we queued up the buffer, or false if it already had
* been on the buffer list.
*/
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
bool
xfs_buf_delwri_queue(
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
struct xfs_buf *bp,
struct list_head *list)
{
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
ASSERT(xfs_buf_islocked(bp));
ASSERT(!(bp->b_flags & XBF_READ));
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
/*
* If the buffer is already marked delwri it already is queued up
* by someone else for imediate writeout. Just ignore it in that
* case.
*/
if (bp->b_flags & _XBF_DELWRI_Q) {
trace_xfs_buf_delwri_queued(bp, _RET_IP_);
return false;
}
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
trace_xfs_buf_delwri_queue(bp, _RET_IP_);
/*
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
* If a buffer gets written out synchronously or marked stale while it
* is on a delwri list we lazily remove it. To do this, the other party
* clears the _XBF_DELWRI_Q flag but otherwise leaves the buffer alone.
* It remains referenced and on the list. In a rare corner case it
* might get readded to a delwri list after the synchronous writeout, in
* which case we need just need to re-add the flag here.
*/
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
bp->b_flags |= _XBF_DELWRI_Q;
if (list_empty(&bp->b_list)) {
atomic_inc(&bp->b_hold);
list_add_tail(&bp->b_list, list);
}
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
return true;
}
/*
* Compare function is more complex than it needs to be because
* the return value is only 32 bits and we are doing comparisons
* on 64 bit values
*/
static int
xfs_buf_cmp(
void *priv,
struct list_head *a,
struct list_head *b)
{
struct xfs_buf *ap = container_of(a, struct xfs_buf, b_list);
struct xfs_buf *bp = container_of(b, struct xfs_buf, b_list);
xfs_daddr_t diff;
diff = ap->b_maps[0].bm_bn - bp->b_maps[0].bm_bn;
if (diff < 0)
return -1;
if (diff > 0)
return 1;
return 0;
}
/*
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
* Submit buffers for write. If wait_list is specified, the buffers are
* submitted using sync I/O and placed on the wait list such that the caller can
* iowait each buffer. Otherwise async I/O is used and the buffers are released
* at I/O completion time. In either case, buffers remain locked until I/O
* completes and the buffer is released from the queue.
*/
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
static int
xfs_buf_delwri_submit_buffers(
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
struct list_head *buffer_list,
struct list_head *wait_list)
{
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
struct xfs_buf *bp, *n;
int pinned = 0;
struct blk_plug plug;
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
list_sort(NULL, buffer_list, xfs_buf_cmp);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
blk_start_plug(&plug);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
list_for_each_entry_safe(bp, n, buffer_list, b_list) {
if (!wait_list) {
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
if (xfs_buf_ispinned(bp)) {
pinned++;
continue;
}
if (!xfs_buf_trylock(bp))
continue;
} else {
xfs_buf_lock(bp);
}
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
/*
* Someone else might have written the buffer synchronously or
* marked it stale in the meantime. In that case only the
* _XBF_DELWRI_Q flag got cleared, and we have to drop the
* reference and remove it from the list here.
*/
if (!(bp->b_flags & _XBF_DELWRI_Q)) {
list_del_init(&bp->b_list);
xfs_buf_relse(bp);
continue;
}
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
trace_xfs_buf_delwri_split(bp, _RET_IP_);
/*
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
* If we have a wait list, each buffer (and associated delwri
* queue reference) transfers to it and is submitted
* synchronously. Otherwise, drop the buffer from the delwri
* queue and submit async.
*/
bp->b_flags &= ~(_XBF_DELWRI_Q | XBF_WRITE_FAIL);
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
bp->b_flags |= XBF_WRITE;
if (wait_list) {
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
bp->b_flags &= ~XBF_ASYNC;
list_move_tail(&bp->b_list, wait_list);
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
} else {
bp->b_flags |= XBF_ASYNC;
list_del_init(&bp->b_list);
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
}
__xfs_buf_submit(bp, false);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
}
blk_finish_plug(&plug);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
return pinned;
}
/*
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
* Write out a buffer list asynchronously.
*
* This will take the @buffer_list, write all non-locked and non-pinned buffers
* out and not wait for I/O completion on any of the buffers. This interface
* is only safely useable for callers that can track I/O completion by higher
* level means, e.g. AIL pushing as the @buffer_list is consumed in this
* function.
xfs: clear ail delwri queued bufs on unmount of shutdown fs In the typical unmount case, the AIL is forced out by the unmount sequence before the xfsaild task is stopped. Since AIL items are removed on writeback completion, this means that the AIL ->ail_buf_list delwri queue has been drained. This is not always true in the shutdown case, however. It's possible for buffers to sit on a delwri queue for a period of time across submission attempts if said items are locked or have been relogged and pinned since first added to the queue. If the attempt to log such an item results in a log I/O error, the error processing can shutdown the fs, remove the item from the AIL, stale the buffer (dropping the LRU reference) and clear its delwri queue state. The latter bit means the buffer will be released from a delwri queue on the next submission attempt, but this might never occur if the filesystem has shutdown and the AIL is empty. This means that such buffers are held indefinitely by the AIL delwri queue across destruction of the AIL. Aside from being a memory leak, these buffers can also hold references to in-core perag structures. The latter problem manifests as a generic/475 failure, reproducing the following asserts at unmount time: XFS: Assertion failed: atomic_read(&pag->pag_ref) == 0, file: fs/xfs/xfs_mount.c, line: 151 XFS: Assertion failed: atomic_read(&pag->pag_ref) == 0, file: fs/xfs/xfs_mount.c, line: 132 To prevent this problem, clear the AIL delwri queue as a final step before xfsaild() exit. The !empty state should never occur in the normal case, so add an assert to catch unexpected problems going forward. [dgc: add comment explaining need for xfs_buf_delwri_cancel() after calling xfs_buf_delwri_submit_nowait().] Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Dave Chinner <david@fromorbit.com>
2018-10-18 06:21:49 +00:00
*
* Note: this function will skip buffers it would block on, and in doing so
* leaves them on @buffer_list so they can be retried on a later pass. As such,
* it is up to the caller to ensure that the buffer list is fully submitted or
* cancelled appropriately when they are finished with the list. Failure to
* cancel or resubmit the list until it is empty will result in leaked buffers
* at unmount time.
*/
int
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
xfs_buf_delwri_submit_nowait(
struct list_head *buffer_list)
{
return xfs_buf_delwri_submit_buffers(buffer_list, NULL);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
}
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
/*
* Write out a buffer list synchronously.
*
* This will take the @buffer_list, write all buffers out and wait for I/O
* completion on all of the buffers. @buffer_list is consumed by the function,
* so callers must have some other way of tracking buffers if they require such
* functionality.
*/
int
xfs_buf_delwri_submit(
struct list_head *buffer_list)
{
LIST_HEAD (wait_list);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
int error = 0, error2;
struct xfs_buf *bp;
xfs_buf_delwri_submit_buffers(buffer_list, &wait_list);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
/* Wait for IO to complete. */
while (!list_empty(&wait_list)) {
bp = list_first_entry(&wait_list, struct xfs_buf, b_list);
list_del_init(&bp->b_list);
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
/*
* Wait on the locked buffer, check for errors and unlock and
* release the delwri queue reference.
*/
error2 = xfs_buf_iowait(bp);
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
xfs_buf_relse(bp);
if (!error)
error = error2;
}
xfs: on-stack delayed write buffer lists Queue delwri buffers on a local on-stack list instead of a per-buftarg one, and write back the buffers per-process instead of by waking up xfsbufd. This is now easily doable given that we have very few places left that write delwri buffers: - log recovery: Only done at mount time, and already forcing out the buffers synchronously using xfs_flush_buftarg - quotacheck: Same story. - dquot reclaim: Writes out dirty dquots on the LRU under memory pressure. We might want to look into doing more of this via xfsaild, but it's already more optimal than the synchronous inode reclaim that writes each buffer synchronously. - xfsaild: This is the main beneficiary of the change. By keeping a local list of buffers to write we reduce latency of writing out buffers, and more importably we can remove all the delwri list promotions which were hitting the buffer cache hard under sustained metadata loads. The implementation is very straight forward - xfs_buf_delwri_queue now gets a new list_head pointer that it adds the delwri buffers to, and all callers need to eventually submit the list using xfs_buf_delwi_submit or xfs_buf_delwi_submit_nowait. Buffers that already are on a delwri list are skipped in xfs_buf_delwri_queue, assuming they already are on another delwri list. The biggest change to pass down the buffer list was done to the AIL pushing. Now that we operate on buffers the trylock, push and pushbuf log item methods are merged into a single push routine, which tries to lock the item, and if possible add the buffer that needs writeback to the buffer list. This leads to much simpler code than the previous split but requires the individual IOP_PUSH instances to unlock and reacquire the AIL around calls to blocking routines. Given that xfsailds now also handle writing out buffers, the conditions for log forcing and the sleep times needed some small changes. The most important one is that we consider an AIL busy as long we still have buffers to push, and the other one is that we do increment the pushed LSN for buffers that are under flushing at this moment, but still count them towards the stuck items for restart purposes. Without this we could hammer on stuck items without ever forcing the log and not make progress under heavy random delete workloads on fast flash storage devices. [ Dave Chinner: - rebase on previous patches. - improved comments for XBF_DELWRI_Q handling - fix XBF_ASYNC handling in queue submission (test 106 failure) - rename delwri submit function buffer list parameters for clarity - xfs_efd_item_push() should return XFS_ITEM_PINNED ] Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Dave Chinner <dchinner@redhat.com> Reviewed-by: Mark Tinguely <tinguely@sgi.com> Signed-off-by: Ben Myers <bpm@sgi.com>
2012-04-23 05:58:39 +00:00
return error;
}
xfs: push buffer of flush locked dquot to avoid quotacheck deadlock Reclaim during quotacheck can lead to deadlocks on the dquot flush lock: - Quotacheck populates a local delwri queue with the physical dquot buffers. - Quotacheck performs the xfs_qm_dqusage_adjust() bulkstat and dirties all of the dquots. - Reclaim kicks in and attempts to flush a dquot whose buffer is already queud on the quotacheck queue. The flush succeeds but queueing to the reclaim delwri queue fails as the backing buffer is already queued. The flush unlock is now deferred to I/O completion of the buffer from the quotacheck queue. - The dqadjust bulkstat continues and dirties the recently flushed dquot once again. - Quotacheck proceeds to the xfs_qm_flush_one() walk which requires the flush lock to update the backing buffers with the in-core recalculated values. It deadlocks on the redirtied dquot as the flush lock was already acquired by reclaim, but the buffer resides on the local delwri queue which isn't submitted until the end of quotacheck. This is reproduced by running quotacheck on a filesystem with a couple million inodes in low memory (512MB-1GB) situations. This is a regression as of commit 43ff2122e6 ("xfs: on-stack delayed write buffer lists"), which removed a trylock and buffer I/O submission from the quotacheck dquot flush sequence. Quotacheck first resets and collects the physical dquot buffers in a delwri queue. Then, it traverses the filesystem inodes via bulkstat, updates the in-core dquots, flushes the corrected dquots to the backing buffers and finally submits the delwri queue for I/O. Since the backing buffers are queued across the entire quotacheck operation, dquot reclaim cannot possibly complete a dquot flush before quotacheck completes. Therefore, quotacheck must submit the buffer for I/O in order to cycle the flush lock and flush the dirty in-core dquot to the buffer. Add a delwri queue buffer push mechanism to submit an individual buffer for I/O without losing the delwri queue status and use it from quotacheck to avoid the deadlock. This restores quotacheck behavior to as before the regression was introduced. Reported-by: Martin Svec <martin.svec@zoner.cz> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-06-15 04:21:45 +00:00
/*
* Push a single buffer on a delwri queue.
*
* The purpose of this function is to submit a single buffer of a delwri queue
* and return with the buffer still on the original queue. The waiting delwri
* buffer submission infrastructure guarantees transfer of the delwri queue
* buffer reference to a temporary wait list. We reuse this infrastructure to
* transfer the buffer back to the original queue.
*
* Note the buffer transitions from the queued state, to the submitted and wait
* listed state and back to the queued state during this call. The buffer
* locking and queue management logic between _delwri_pushbuf() and
* _delwri_queue() guarantee that the buffer cannot be queued to another list
* before returning.
*/
int
xfs_buf_delwri_pushbuf(
struct xfs_buf *bp,
struct list_head *buffer_list)
{
LIST_HEAD (submit_list);
int error;
ASSERT(bp->b_flags & _XBF_DELWRI_Q);
trace_xfs_buf_delwri_pushbuf(bp, _RET_IP_);
/*
* Isolate the buffer to a new local list so we can submit it for I/O
* independently from the rest of the original list.
*/
xfs_buf_lock(bp);
list_move(&bp->b_list, &submit_list);
xfs_buf_unlock(bp);
/*
* Delwri submission clears the DELWRI_Q buffer flag and returns with
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
* the buffer on the wait list with the original reference. Rather than
xfs: push buffer of flush locked dquot to avoid quotacheck deadlock Reclaim during quotacheck can lead to deadlocks on the dquot flush lock: - Quotacheck populates a local delwri queue with the physical dquot buffers. - Quotacheck performs the xfs_qm_dqusage_adjust() bulkstat and dirties all of the dquots. - Reclaim kicks in and attempts to flush a dquot whose buffer is already queud on the quotacheck queue. The flush succeeds but queueing to the reclaim delwri queue fails as the backing buffer is already queued. The flush unlock is now deferred to I/O completion of the buffer from the quotacheck queue. - The dqadjust bulkstat continues and dirties the recently flushed dquot once again. - Quotacheck proceeds to the xfs_qm_flush_one() walk which requires the flush lock to update the backing buffers with the in-core recalculated values. It deadlocks on the redirtied dquot as the flush lock was already acquired by reclaim, but the buffer resides on the local delwri queue which isn't submitted until the end of quotacheck. This is reproduced by running quotacheck on a filesystem with a couple million inodes in low memory (512MB-1GB) situations. This is a regression as of commit 43ff2122e6 ("xfs: on-stack delayed write buffer lists"), which removed a trylock and buffer I/O submission from the quotacheck dquot flush sequence. Quotacheck first resets and collects the physical dquot buffers in a delwri queue. Then, it traverses the filesystem inodes via bulkstat, updates the in-core dquots, flushes the corrected dquots to the backing buffers and finally submits the delwri queue for I/O. Since the backing buffers are queued across the entire quotacheck operation, dquot reclaim cannot possibly complete a dquot flush before quotacheck completes. Therefore, quotacheck must submit the buffer for I/O in order to cycle the flush lock and flush the dirty in-core dquot to the buffer. Add a delwri queue buffer push mechanism to submit an individual buffer for I/O without losing the delwri queue status and use it from quotacheck to avoid the deadlock. This restores quotacheck behavior to as before the regression was introduced. Reported-by: Martin Svec <martin.svec@zoner.cz> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-06-15 04:21:45 +00:00
* bounce the buffer from a local wait list back to the original list
* after I/O completion, reuse the original list as the wait list.
*/
xfs_buf_delwri_submit_buffers(&submit_list, buffer_list);
/*
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
* The buffer is now locked, under I/O and wait listed on the original
* delwri queue. Wait for I/O completion, restore the DELWRI_Q flag and
* return with the buffer unlocked and on the original queue.
xfs: push buffer of flush locked dquot to avoid quotacheck deadlock Reclaim during quotacheck can lead to deadlocks on the dquot flush lock: - Quotacheck populates a local delwri queue with the physical dquot buffers. - Quotacheck performs the xfs_qm_dqusage_adjust() bulkstat and dirties all of the dquots. - Reclaim kicks in and attempts to flush a dquot whose buffer is already queud on the quotacheck queue. The flush succeeds but queueing to the reclaim delwri queue fails as the backing buffer is already queued. The flush unlock is now deferred to I/O completion of the buffer from the quotacheck queue. - The dqadjust bulkstat continues and dirties the recently flushed dquot once again. - Quotacheck proceeds to the xfs_qm_flush_one() walk which requires the flush lock to update the backing buffers with the in-core recalculated values. It deadlocks on the redirtied dquot as the flush lock was already acquired by reclaim, but the buffer resides on the local delwri queue which isn't submitted until the end of quotacheck. This is reproduced by running quotacheck on a filesystem with a couple million inodes in low memory (512MB-1GB) situations. This is a regression as of commit 43ff2122e6 ("xfs: on-stack delayed write buffer lists"), which removed a trylock and buffer I/O submission from the quotacheck dquot flush sequence. Quotacheck first resets and collects the physical dquot buffers in a delwri queue. Then, it traverses the filesystem inodes via bulkstat, updates the in-core dquots, flushes the corrected dquots to the backing buffers and finally submits the delwri queue for I/O. Since the backing buffers are queued across the entire quotacheck operation, dquot reclaim cannot possibly complete a dquot flush before quotacheck completes. Therefore, quotacheck must submit the buffer for I/O in order to cycle the flush lock and flush the dirty in-core dquot to the buffer. Add a delwri queue buffer push mechanism to submit an individual buffer for I/O without losing the delwri queue status and use it from quotacheck to avoid the deadlock. This restores quotacheck behavior to as before the regression was introduced. Reported-by: Martin Svec <martin.svec@zoner.cz> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-06-15 04:21:45 +00:00
*/
xfs: use sync buffer I/O for sync delwri queue submission If a delwri queue occurs of a buffer that sits on a delwri queue wait list, the queue sets _XBF_DELWRI_Q without changing the state of ->b_list. This occurs, for example, if another thread beats the current delwri waiter thread to the buffer lock after I/O completion. Once the waiter acquires the lock, it removes the buffer from the wait list and leaves a buffer with _XBF_DELWRI_Q set but not populated on a list. This results in a lost buffer submission and in turn can result in assert failures due to _XBF_DELWRI_Q being set on buffer reclaim or filesystem lockups if the buffer happens to cover an item in the AIL. This problem has been reproduced by repeated iterations of xfs/305 on high CPU count (28xcpu) systems with limited memory (~1GB). Dirty dquot reclaim races with an xfsaild push of a separate dquot backed by the same buffer such that the buffer sits on the reclaim wait list at the time xfsaild attempts to queue it. Since the latter dquot has been flush locked but the underlying buffer not submitted for I/O, the dquot pins the AIL and causes the filesystem to livelock. This race is essentially made possible by the buffer lock cycle involved with waiting on a synchronous delwri queue submission. Close the race by using synchronous buffer I/O for respective delwri queue submission. This means the buffer remains locked across the I/O and so is inaccessible from other contexts while in the intermediate wait list state. The sync buffer I/O wait mechanism is factored into a helper such that sync delwri buffer submission and serialization are batched operations. Designed-by: Dave Chinner <dchinner@redhat.com> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2018-07-12 05:26:34 +00:00
error = xfs_buf_iowait(bp);
xfs: push buffer of flush locked dquot to avoid quotacheck deadlock Reclaim during quotacheck can lead to deadlocks on the dquot flush lock: - Quotacheck populates a local delwri queue with the physical dquot buffers. - Quotacheck performs the xfs_qm_dqusage_adjust() bulkstat and dirties all of the dquots. - Reclaim kicks in and attempts to flush a dquot whose buffer is already queud on the quotacheck queue. The flush succeeds but queueing to the reclaim delwri queue fails as the backing buffer is already queued. The flush unlock is now deferred to I/O completion of the buffer from the quotacheck queue. - The dqadjust bulkstat continues and dirties the recently flushed dquot once again. - Quotacheck proceeds to the xfs_qm_flush_one() walk which requires the flush lock to update the backing buffers with the in-core recalculated values. It deadlocks on the redirtied dquot as the flush lock was already acquired by reclaim, but the buffer resides on the local delwri queue which isn't submitted until the end of quotacheck. This is reproduced by running quotacheck on a filesystem with a couple million inodes in low memory (512MB-1GB) situations. This is a regression as of commit 43ff2122e6 ("xfs: on-stack delayed write buffer lists"), which removed a trylock and buffer I/O submission from the quotacheck dquot flush sequence. Quotacheck first resets and collects the physical dquot buffers in a delwri queue. Then, it traverses the filesystem inodes via bulkstat, updates the in-core dquots, flushes the corrected dquots to the backing buffers and finally submits the delwri queue for I/O. Since the backing buffers are queued across the entire quotacheck operation, dquot reclaim cannot possibly complete a dquot flush before quotacheck completes. Therefore, quotacheck must submit the buffer for I/O in order to cycle the flush lock and flush the dirty in-core dquot to the buffer. Add a delwri queue buffer push mechanism to submit an individual buffer for I/O without losing the delwri queue status and use it from quotacheck to avoid the deadlock. This restores quotacheck behavior to as before the regression was introduced. Reported-by: Martin Svec <martin.svec@zoner.cz> Signed-off-by: Brian Foster <bfoster@redhat.com> Reviewed-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
2017-06-15 04:21:45 +00:00
bp->b_flags |= _XBF_DELWRI_Q;
xfs_buf_unlock(bp);
return error;
}
int __init
xfs_buf_init(void)
{
xfs_buf_zone = kmem_zone_init_flags(sizeof(xfs_buf_t), "xfs_buf",
KM_ZONE_HWALIGN, NULL);
if (!xfs_buf_zone)
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
goto out;
return 0;
xfs: event tracing support Convert the old xfs tracing support that could only be used with the out of tree kdb and xfsidbg patches to use the generic event tracer. To use it make sure CONFIG_EVENT_TRACING is enabled and then enable all xfs trace channels by: echo 1 > /sys/kernel/debug/tracing/events/xfs/enable or alternatively enable single events by just doing the same in one event subdirectory, e.g. echo 1 > /sys/kernel/debug/tracing/events/xfs/xfs_ihold/enable or set more complex filters, etc. In Documentation/trace/events.txt all this is desctribed in more detail. To reads the events do a cat /sys/kernel/debug/tracing/trace Compared to the last posting this patch converts the tracing mostly to the one tracepoint per callsite model that other users of the new tracing facility also employ. This allows a very fine-grained control of the tracing, a cleaner output of the traces and also enables the perf tool to use each tracepoint as a virtual performance counter, allowing us to e.g. count how often certain workloads git various spots in XFS. Take a look at http://lwn.net/Articles/346470/ for some examples. Also the btree tracing isn't included at all yet, as it will require additional core tracing features not in mainline yet, I plan to deliver it later. And the really nice thing about this patch is that it actually removes many lines of code while adding this nice functionality: fs/xfs/Makefile | 8 fs/xfs/linux-2.6/xfs_acl.c | 1 fs/xfs/linux-2.6/xfs_aops.c | 52 - fs/xfs/linux-2.6/xfs_aops.h | 2 fs/xfs/linux-2.6/xfs_buf.c | 117 +-- fs/xfs/linux-2.6/xfs_buf.h | 33 fs/xfs/linux-2.6/xfs_fs_subr.c | 3 fs/xfs/linux-2.6/xfs_ioctl.c | 1 fs/xfs/linux-2.6/xfs_ioctl32.c | 1 fs/xfs/linux-2.6/xfs_iops.c | 1 fs/xfs/linux-2.6/xfs_linux.h | 1 fs/xfs/linux-2.6/xfs_lrw.c | 87 -- fs/xfs/linux-2.6/xfs_lrw.h | 45 - fs/xfs/linux-2.6/xfs_super.c | 104 --- fs/xfs/linux-2.6/xfs_super.h | 7 fs/xfs/linux-2.6/xfs_sync.c | 1 fs/xfs/linux-2.6/xfs_trace.c | 75 ++ fs/xfs/linux-2.6/xfs_trace.h | 1369 +++++++++++++++++++++++++++++++++++++++++ fs/xfs/linux-2.6/xfs_vnode.h | 4 fs/xfs/quota/xfs_dquot.c | 110 --- fs/xfs/quota/xfs_dquot.h | 21 fs/xfs/quota/xfs_qm.c | 40 - fs/xfs/quota/xfs_qm_syscalls.c | 4 fs/xfs/support/ktrace.c | 323 --------- fs/xfs/support/ktrace.h | 85 -- fs/xfs/xfs.h | 16 fs/xfs/xfs_ag.h | 14 fs/xfs/xfs_alloc.c | 230 +----- fs/xfs/xfs_alloc.h | 27 fs/xfs/xfs_alloc_btree.c | 1 fs/xfs/xfs_attr.c | 107 --- fs/xfs/xfs_attr.h | 10 fs/xfs/xfs_attr_leaf.c | 14 fs/xfs/xfs_attr_sf.h | 40 - fs/xfs/xfs_bmap.c | 507 +++------------ fs/xfs/xfs_bmap.h | 49 - fs/xfs/xfs_bmap_btree.c | 6 fs/xfs/xfs_btree.c | 5 fs/xfs/xfs_btree_trace.h | 17 fs/xfs/xfs_buf_item.c | 87 -- fs/xfs/xfs_buf_item.h | 20 fs/xfs/xfs_da_btree.c | 3 fs/xfs/xfs_da_btree.h | 7 fs/xfs/xfs_dfrag.c | 2 fs/xfs/xfs_dir2.c | 8 fs/xfs/xfs_dir2_block.c | 20 fs/xfs/xfs_dir2_leaf.c | 21 fs/xfs/xfs_dir2_node.c | 27 fs/xfs/xfs_dir2_sf.c | 26 fs/xfs/xfs_dir2_trace.c | 216 ------ fs/xfs/xfs_dir2_trace.h | 72 -- fs/xfs/xfs_filestream.c | 8 fs/xfs/xfs_fsops.c | 2 fs/xfs/xfs_iget.c | 111 --- fs/xfs/xfs_inode.c | 67 -- fs/xfs/xfs_inode.h | 76 -- fs/xfs/xfs_inode_item.c | 5 fs/xfs/xfs_iomap.c | 85 -- fs/xfs/xfs_iomap.h | 8 fs/xfs/xfs_log.c | 181 +---- fs/xfs/xfs_log_priv.h | 20 fs/xfs/xfs_log_recover.c | 1 fs/xfs/xfs_mount.c | 2 fs/xfs/xfs_quota.h | 8 fs/xfs/xfs_rename.c | 1 fs/xfs/xfs_rtalloc.c | 1 fs/xfs/xfs_rw.c | 3 fs/xfs/xfs_trans.h | 47 + fs/xfs/xfs_trans_buf.c | 62 - fs/xfs/xfs_vnodeops.c | 8 70 files changed, 2151 insertions(+), 2592 deletions(-) Signed-off-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Alex Elder <aelder@sgi.com>
2009-12-14 23:14:59 +00:00
out:
return -ENOMEM;
}
void
xfs_buf_terminate(void)
{
kmem_zone_destroy(xfs_buf_zone);
}
void xfs_buf_set_ref(struct xfs_buf *bp, int lru_ref)
{
/*
* Set the lru reference count to 0 based on the error injection tag.
* This allows userspace to disrupt buffer caching for debug/testing
* purposes.
*/
if (XFS_TEST_ERROR(false, bp->b_target->bt_mount,
XFS_ERRTAG_BUF_LRU_REF))
lru_ref = 0;
atomic_set(&bp->b_lru_ref, lru_ref);
}
/*
* Verify an on-disk magic value against the magic value specified in the
* verifier structure. The verifier magic is in disk byte order so the caller is
* expected to pass the value directly from disk.
*/
bool
xfs_verify_magic(
struct xfs_buf *bp,
__be32 dmagic)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
int idx;
idx = xfs_sb_version_hascrc(&mp->m_sb);
if (unlikely(WARN_ON(!bp->b_ops || !bp->b_ops->magic[idx])))
return false;
return dmagic == bp->b_ops->magic[idx];
}
/*
* Verify an on-disk magic value against the magic value specified in the
* verifier structure. The verifier magic is in disk byte order so the caller is
* expected to pass the value directly from disk.
*/
bool
xfs_verify_magic16(
struct xfs_buf *bp,
__be16 dmagic)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
int idx;
idx = xfs_sb_version_hascrc(&mp->m_sb);
if (unlikely(WARN_ON(!bp->b_ops || !bp->b_ops->magic16[idx])))
return false;
return dmagic == bp->b_ops->magic16[idx];
}