2018-06-06 05:42:14 +03:00
// SPDX-License-Identifier: GPL-2.0
2005-04-17 02:20:36 +04:00
/*
2005-11-02 06:59:41 +03:00
* Copyright ( c ) 2000 - 2003 Silicon Graphics , Inc .
* All Rights Reserved .
2005-04-17 02:20:36 +04:00
*/
# include "xfs.h"
# include "xfs_fs.h"
2019-06-29 05:25:35 +03:00
# include "xfs_shared.h"
2013-08-12 14:49:26 +04:00
# include "xfs_format.h"
2013-10-23 03:50:10 +04:00
# include "xfs_log_format.h"
# include "xfs_trans_resv.h"
2005-04-17 02:20:36 +04:00
# include "xfs_mount.h"
# include "xfs_inode.h"
2013-10-23 03:51:50 +04:00
# include "xfs_quota.h"
2013-10-23 03:50:10 +04:00
# include "xfs_trans.h"
2005-04-17 02:20:36 +04:00
# include "xfs_buf_item.h"
# include "xfs_trans_priv.h"
# include "xfs_qm.h"
2013-10-23 03:50:10 +04:00
# include "xfs_log.h"
2005-04-17 02:20:36 +04:00
2010-06-23 12:11:15 +04:00
static inline struct xfs_dq_logitem * DQUOT_ITEM ( struct xfs_log_item * lip )
{
return container_of ( lip , struct xfs_dq_logitem , qli_item ) ;
}
2005-04-17 02:20:36 +04:00
/*
* returns the number of iovecs needed to log the given dquot item .
*/
2013-08-12 14:50:04 +04:00
STATIC void
2005-04-17 02:20:36 +04:00
xfs_qm_dquot_logitem_size (
2013-08-12 14:50:04 +04:00
struct xfs_log_item * lip ,
int * nvecs ,
int * nbytes )
2005-04-17 02:20:36 +04:00
{
2013-08-12 14:50:04 +04:00
* nvecs + = 2 ;
* nbytes + = sizeof ( struct xfs_dq_logformat ) +
sizeof ( struct xfs_disk_dquot ) ;
2005-04-17 02:20:36 +04:00
}
/*
* fills in the vector of log iovecs for the given dquot log item .
*/
STATIC void
xfs_qm_dquot_logitem_format (
2010-06-23 12:11:15 +04:00
struct xfs_log_item * lip ,
2013-12-13 04:34:02 +04:00
struct xfs_log_vec * lv )
2005-04-17 02:20:36 +04:00
{
2020-07-14 20:37:22 +03:00
struct xfs_disk_dquot ddq ;
2010-06-23 12:11:15 +04:00
struct xfs_dq_logitem * qlip = DQUOT_ITEM ( lip ) ;
2013-12-13 04:34:02 +04:00
struct xfs_log_iovec * vecp = NULL ;
2013-12-13 04:34:07 +04:00
struct xfs_dq_logformat * qlf ;
2005-04-17 02:20:36 +04:00
2013-12-13 04:34:07 +04:00
qlf = xlog_prepare_iovec ( lv , & vecp , XLOG_REG_TYPE_QFORMAT ) ;
qlf - > qlf_type = XFS_LI_DQUOT ;
qlf - > qlf_size = 2 ;
2020-07-14 20:37:30 +03:00
qlf - > qlf_id = qlip - > qli_dquot - > q_id ;
2013-12-13 04:34:07 +04:00
qlf - > qlf_blkno = qlip - > qli_dquot - > q_blkno ;
qlf - > qlf_len = 1 ;
qlf - > qlf_boffset = qlip - > qli_dquot - > q_bufoffset ;
xlog_finish_iovec ( lv , vecp , sizeof ( struct xfs_dq_logformat ) ) ;
2013-12-13 04:34:02 +04:00
2020-07-14 20:37:22 +03:00
xfs_dquot_to_disk ( & ddq , qlip - > qli_dquot ) ;
xlog_copy_iovec ( lv , & vecp , XLOG_REG_TYPE_DQUOT , & ddq ,
2013-12-13 04:00:43 +04:00
sizeof ( struct xfs_disk_dquot ) ) ;
2005-04-17 02:20:36 +04:00
}
/*
* Increment the pin count of the given dquot .
*/
STATIC void
xfs_qm_dquot_logitem_pin (
2010-06-23 12:11:15 +04:00
struct xfs_log_item * lip )
2005-04-17 02:20:36 +04:00
{
2010-06-23 12:11:15 +04:00
struct xfs_dquot * dqp = DQUOT_ITEM ( lip ) - > qli_dquot ;
2005-04-17 02:20:36 +04:00
ASSERT ( XFS_DQ_IS_LOCKED ( dqp ) ) ;
2008-10-30 09:05:18 +03:00
atomic_inc ( & dqp - > q_pincount ) ;
2005-04-17 02:20:36 +04:00
}
/*
* Decrement the pin count of the given dquot , and wake up
* anyone in xfs_dqwait_unpin ( ) if the count goes to 0. The
2008-10-30 09:05:04 +03:00
* dquot must have been previously pinned with a call to
* xfs_qm_dquot_logitem_pin ( ) .
2005-04-17 02:20:36 +04:00
*/
STATIC void
xfs_qm_dquot_logitem_unpin (
2010-06-23 12:11:15 +04:00
struct xfs_log_item * lip ,
2010-06-23 12:11:15 +04:00
int remove )
2005-04-17 02:20:36 +04:00
{
2010-06-23 12:11:15 +04:00
struct xfs_dquot * dqp = DQUOT_ITEM ( lip ) - > qli_dquot ;
2005-04-17 02:20:36 +04:00
2008-10-30 09:05:04 +03:00
ASSERT ( atomic_read ( & dqp - > q_pincount ) > 0 ) ;
if ( atomic_dec_and_test ( & dqp - > q_pincount ) )
wake_up ( & dqp - > q_pinwait ) ;
2005-04-17 02:20:36 +04:00
}
/*
* This is called to wait for the given dquot to be unpinned .
* Most of these pin / unpin routines are plagiarized from inode code .
*/
void
xfs_qm_dqunpin_wait (
2010-06-23 12:11:15 +04:00
struct xfs_dquot * dqp )
2005-04-17 02:20:36 +04:00
{
ASSERT ( XFS_DQ_IS_LOCKED ( dqp ) ) ;
2008-10-30 09:05:04 +03:00
if ( atomic_read ( & dqp - > q_pincount ) = = 0 )
2005-04-17 02:20:36 +04:00
return ;
/*
* Give the log a push so we don ' t wait here too long .
*/
2010-01-19 12:56:46 +03:00
xfs_log_force ( dqp - > q_mount , 0 ) ;
2008-10-30 09:05:04 +03:00
wait_event ( dqp - > q_pinwait , ( atomic_read ( & dqp - > q_pincount ) = = 0 ) ) ;
2005-04-17 02:20:36 +04:00
}
STATIC uint
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 09:58:39 +04:00
xfs_qm_dquot_logitem_push (
struct xfs_log_item * lip ,
2018-03-08 01:59:39 +03:00
struct list_head * buffer_list )
__releases ( & lip - > li_ailp - > ail_lock )
__acquires ( & lip - > li_ailp - > ail_lock )
2005-04-17 02:20:36 +04:00
{
2010-06-23 12:11:15 +04:00
struct xfs_dquot * dqp = DQUOT_ITEM ( lip ) - > qli_dquot ;
2017-11-28 19:54:10 +03:00
struct xfs_buf * bp = lip - > li_buf ;
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 09:58:39 +04:00
uint rval = XFS_ITEM_SUCCESS ;
int error ;
2005-04-17 02:20:36 +04:00
2008-10-30 09:05:04 +03:00
if ( atomic_read ( & dqp - > q_pincount ) > 0 )
2010-02-02 02:13:42 +03:00
return XFS_ITEM_PINNED ;
2005-04-17 02:20:36 +04:00
2011-12-07 01:58:14 +04:00
if ( ! xfs_dqlock_nowait ( dqp ) )
2010-02-02 02:13:42 +03:00
return XFS_ITEM_LOCKED ;
2005-04-17 02:20:36 +04:00
2012-04-23 09:58:37 +04:00
/*
* Re - check the pincount now that we stabilized the value by
* taking the quota lock .
*/
if ( atomic_read ( & dqp - > q_pincount ) > 0 ) {
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 09:58:39 +04:00
rval = XFS_ITEM_PINNED ;
goto out_unlock ;
2012-04-23 09:58:37 +04:00
}
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 09:58:39 +04:00
/*
* Someone else is already flushing the dquot . Nothing we can do
* here but wait for the flush to finish and remove the item from
* the AIL .
*/
2008-08-13 10:41:43 +04:00
if ( ! xfs_dqflock_nowait ( dqp ) ) {
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 09:58:39 +04:00
rval = XFS_ITEM_FLUSHING ;
goto out_unlock ;
}
2018-03-08 01:59:39 +03:00
spin_unlock ( & lip - > li_ailp - > ail_lock ) ;
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 09:58:39 +04:00
error = xfs_qm_dqflush ( dqp , & bp ) ;
2018-05-05 01:30:20 +03:00
if ( ! error ) {
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 09:58:39 +04:00
if ( ! xfs_buf_delwri_queue ( bp , buffer_list ) )
rval = XFS_ITEM_FLUSHING ;
xfs_buf_relse ( bp ) ;
2020-03-27 18:29:45 +03:00
} else if ( error = = - EAGAIN )
rval = XFS_ITEM_LOCKED ;
2005-04-17 02:20:36 +04:00
2018-03-08 01:59:39 +03:00
spin_lock ( & lip - > li_ailp - > ail_lock ) ;
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 09:58:39 +04:00
out_unlock :
xfs_dqunlock ( dqp ) ;
return rval ;
2005-04-17 02:20:36 +04:00
}
STATIC void
2019-06-29 05:27:32 +03:00
xfs_qm_dquot_logitem_release (
2010-06-23 12:11:15 +04:00
struct xfs_log_item * lip )
2005-04-17 02:20:36 +04:00
{
2010-06-23 12:11:15 +04:00
struct xfs_dquot * dqp = DQUOT_ITEM ( lip ) - > qli_dquot ;
2005-04-17 02:20:36 +04:00
ASSERT ( XFS_DQ_IS_LOCKED ( dqp ) ) ;
/*
* dquots are never ' held ' from getting unlocked at the end of
* a transaction . Their locking and unlocking is hidden inside the
* transaction layer , within trans_commit . Hence , no LI_HOLD flag
* for the logitem .
*/
xfs_dqunlock ( dqp ) ;
}
2019-06-29 05:27:32 +03:00
STATIC void
xfs_qm_dquot_logitem_committing (
struct xfs_log_item * lip ,
2021-06-18 18:21:52 +03:00
xfs_csn_t seq )
2019-06-29 05:27:32 +03:00
{
return xfs_qm_dquot_logitem_release ( lip ) ;
}
2011-10-28 13:54:24 +04:00
static const struct xfs_item_ops xfs_dquot_item_ops = {
2010-06-23 12:11:15 +04:00
. iop_size = xfs_qm_dquot_logitem_size ,
. iop_format = xfs_qm_dquot_logitem_format ,
. iop_pin = xfs_qm_dquot_logitem_pin ,
. iop_unpin = xfs_qm_dquot_logitem_unpin ,
2019-06-29 05:27:32 +03:00
. iop_release = xfs_qm_dquot_logitem_release ,
. iop_committing = xfs_qm_dquot_logitem_committing ,
2010-06-23 12:11:15 +04:00
. iop_push = xfs_qm_dquot_logitem_push ,
2005-04-17 02:20:36 +04:00
} ;
/*
* Initialize the dquot log item for a newly allocated dquot .
* The dquot isn ' t locked at this point , but it isn ' t on any of the lists
* either , so we don ' t care .
*/
void
xfs_qm_dquot_logitem_init (
2010-06-23 12:11:15 +04:00
struct xfs_dquot * dqp )
2005-04-17 02:20:36 +04:00
{
2010-06-23 12:11:15 +04:00
struct xfs_dq_logitem * lp = & dqp - > q_logitem ;
2005-04-17 02:20:36 +04:00
2010-03-23 02:10:00 +03:00
xfs_log_item_init ( dqp - > q_mount , & lp - > qli_item , XFS_LI_DQUOT ,
& xfs_dquot_item_ops ) ;
2005-04-17 02:20:36 +04:00
lp - > qli_dquot = dqp ;
}
/*------------------ QUOTAOFF LOG ITEMS -------------------*/
2010-06-23 12:11:15 +04:00
static inline struct xfs_qoff_logitem * QOFF_ITEM ( struct xfs_log_item * lip )
{
return container_of ( lip , struct xfs_qoff_logitem , qql_item ) ;
}
2005-04-17 02:20:36 +04:00
/*
* This returns the number of iovecs needed to log the given quotaoff item .
* We only need 1 iovec for an quotaoff item . It just logs the
* quotaoff_log_format structure .
*/
2013-08-12 14:50:04 +04:00
STATIC void
2010-06-23 12:11:15 +04:00
xfs_qm_qoff_logitem_size (
2013-08-12 14:50:04 +04:00
struct xfs_log_item * lip ,
int * nvecs ,
int * nbytes )
2005-04-17 02:20:36 +04:00
{
2013-08-12 14:50:04 +04:00
* nvecs + = 1 ;
* nbytes + = sizeof ( struct xfs_qoff_logitem ) ;
2005-04-17 02:20:36 +04:00
}
STATIC void
2010-06-23 12:11:15 +04:00
xfs_qm_qoff_logitem_format (
struct xfs_log_item * lip ,
2013-12-13 04:34:02 +04:00
struct xfs_log_vec * lv )
2005-04-17 02:20:36 +04:00
{
2010-06-23 12:11:15 +04:00
struct xfs_qoff_logitem * qflip = QOFF_ITEM ( lip ) ;
2013-12-13 04:34:02 +04:00
struct xfs_log_iovec * vecp = NULL ;
2013-12-13 04:34:08 +04:00
struct xfs_qoff_logformat * qlf ;
2005-04-17 02:20:36 +04:00
2013-12-13 04:34:08 +04:00
qlf = xlog_prepare_iovec ( lv , & vecp , XLOG_REG_TYPE_QUOTAOFF ) ;
qlf - > qf_type = XFS_LI_QUOTAOFF ;
qlf - > qf_size = 1 ;
qlf - > qf_flags = qflip - > qql_flags ;
xlog_finish_iovec ( lv , vecp , sizeof ( struct xfs_qoff_logitem ) ) ;
2005-04-17 02:20:36 +04:00
}
/*
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 09:58:39 +04:00
* There isn ' t much you can do to push a quotaoff item . It is simply
* stuck waiting for the log to be flushed to disk .
2005-04-17 02:20:36 +04:00
*/
STATIC uint
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 09:58:39 +04:00
xfs_qm_qoff_logitem_push (
struct xfs_log_item * lip ,
struct list_head * buffer_list )
2005-04-17 02:20:36 +04:00
{
return XFS_ITEM_LOCKED ;
}
STATIC xfs_lsn_t
xfs_qm_qoffend_logitem_committed (
2010-06-23 12:11:15 +04:00
struct xfs_log_item * lip ,
xfs_lsn_t lsn )
2005-04-17 02:20:36 +04:00
{
2010-06-23 12:11:15 +04:00
struct xfs_qoff_logitem * qfe = QOFF_ITEM ( lip ) ;
struct xfs_qoff_logitem * qfs = qfe - > qql_start_lip ;
2005-04-17 02:20:36 +04:00
2020-03-17 00:26:09 +03:00
xfs_qm_qoff_logitem_relse ( qfs ) ;
2010-06-23 12:11:15 +04:00
xfs: allocate log vector buffers outside CIL context lock
One of the problems we currently have with delayed logging is that
under serious memory pressure we can deadlock memory reclaim. THis
occurs when memory reclaim (such as run by kswapd) is reclaiming XFS
inodes and issues a log force to unpin inodes that are dirty in the
CIL.
The CIL is pushed, but this will only occur once it gets the CIL
context lock to ensure that all committing transactions are complete
and no new transactions start being committed to the CIL while the
push switches to a new context.
The deadlock occurs when the CIL context lock is held by a
committing process that is doing memory allocation for log vector
buffers, and that allocation is then blocked on memory reclaim
making progress. Memory reclaim, however, is blocked waiting for
a log force to make progress, and so we effectively deadlock at this
point.
To solve this problem, we have to move the CIL log vector buffer
allocation outside of the context lock so that memory reclaim can
always make progress when it needs to force the log. The problem
with doing this is that a CIL push can take place while we are
determining if we need to allocate a new log vector buffer for
an item and hence the current log vector may go away without
warning. That means we canot rely on the existing log vector being
present when we finally grab the context lock and so we must have a
replacement buffer ready to go at all times.
To ensure this, introduce a "shadow log vector" buffer that is
always guaranteed to be present when we gain the CIL context lock
and format the item. This shadow buffer may or may not be used
during the formatting, but if the log item does not have an existing
log vector buffer or that buffer is too small for the new
modifications, we swap it for the new shadow buffer and format
the modifications into that new log vector buffer.
The result of this is that for any object we modify more than once
in a given CIL checkpoint, we double the memory required
to track dirty regions in the log. For single modifications then
we consume the shadow log vectorwe allocate on commit, and that gets
consumed by the checkpoint. However, if we make multiple
modifications, then the second transaction commit will allocate a
shadow log vector and hence we will end up with double the memory
usage as only one of the log vectors is consumed by the CIL
checkpoint. The remaining shadow vector will be freed when th elog
item is freed.
This can probably be optimised in future - access to the shadow log
vector is serialised by the object lock (as opposited to the active
log vector, which is controlled by the CIL context lock) and so we
can probably free shadow log vector from some objects when the log
item is marked clean on removal from the AIL.
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
2016-07-22 02:52:35 +03:00
kmem_free ( lip - > li_lv_shadow ) ;
2008-05-19 10:31:57 +04:00
kmem_free ( qfe ) ;
2005-04-17 02:20:36 +04:00
return ( xfs_lsn_t ) - 1 ;
}
2020-03-17 00:26:09 +03:00
STATIC void
xfs_qm_qoff_logitem_release (
struct xfs_log_item * lip )
{
struct xfs_qoff_logitem * qoff = QOFF_ITEM ( lip ) ;
if ( test_bit ( XFS_LI_ABORTED , & lip - > li_flags ) ) {
if ( qoff - > qql_start_lip )
xfs_qm_qoff_logitem_relse ( qoff - > qql_start_lip ) ;
xfs_qm_qoff_logitem_relse ( qoff ) ;
}
}
2011-10-28 13:54:24 +04:00
static const struct xfs_item_ops xfs_qm_qoffend_logitem_ops = {
2010-06-23 12:11:15 +04:00
. iop_size = xfs_qm_qoff_logitem_size ,
. iop_format = xfs_qm_qoff_logitem_format ,
. iop_committed = xfs_qm_qoffend_logitem_committed ,
. iop_push = xfs_qm_qoff_logitem_push ,
2020-03-17 00:26:09 +03:00
. iop_release = xfs_qm_qoff_logitem_release ,
2005-04-17 02:20:36 +04:00
} ;
2011-10-28 13:54:24 +04:00
static const struct xfs_item_ops xfs_qm_qoff_logitem_ops = {
2010-06-23 12:11:15 +04:00
. iop_size = xfs_qm_qoff_logitem_size ,
. iop_format = xfs_qm_qoff_logitem_format ,
. iop_push = xfs_qm_qoff_logitem_push ,
2020-03-17 00:26:09 +03:00
. iop_release = xfs_qm_qoff_logitem_release ,
2005-04-17 02:20:36 +04:00
} ;
2020-03-17 00:26:09 +03:00
/*
* Delete the quotaoff intent from the AIL and free it . On success ,
* this should only be called for the start item . It can be used for
* either on shutdown or abort .
*/
void
xfs_qm_qoff_logitem_relse (
struct xfs_qoff_logitem * qoff )
{
struct xfs_log_item * lip = & qoff - > qql_item ;
ASSERT ( test_bit ( XFS_LI_IN_AIL , & lip - > li_flags ) | |
test_bit ( XFS_LI_ABORTED , & lip - > li_flags ) | |
XFS_FORCED_SHUTDOWN ( lip - > li_mountp ) ) ;
2020-05-06 23:27:04 +03:00
xfs_trans_ail_delete ( lip , 0 ) ;
2020-03-17 00:26:09 +03:00
kmem_free ( lip - > li_lv_shadow ) ;
kmem_free ( qoff ) ;
}
2005-04-17 02:20:36 +04:00
/*
* Allocate and initialize an quotaoff item of the correct quota type ( s ) .
*/
2010-06-23 12:11:15 +04:00
struct xfs_qoff_logitem *
2005-04-17 02:20:36 +04:00
xfs_qm_qoff_logitem_init (
2010-06-23 12:11:15 +04:00
struct xfs_mount * mp ,
struct xfs_qoff_logitem * start ,
uint flags )
2005-04-17 02:20:36 +04:00
{
2010-06-23 12:11:15 +04:00
struct xfs_qoff_logitem * qf ;
2005-04-17 02:20:36 +04:00
2019-08-26 22:06:22 +03:00
qf = kmem_zalloc ( sizeof ( struct xfs_qoff_logitem ) , 0 ) ;
2005-04-17 02:20:36 +04:00
2010-03-23 02:10:00 +03:00
xfs_log_item_init ( mp , & qf - > qql_item , XFS_LI_QUOTAOFF , start ?
& xfs_qm_qoffend_logitem_ops : & xfs_qm_qoff_logitem_ops ) ;
2005-04-17 02:20:36 +04:00
qf - > qql_item . li_mountp = mp ;
qf - > qql_start_lip = start ;
2013-12-13 04:34:08 +04:00
qf - > qql_flags = flags ;
2010-06-23 12:11:15 +04:00
return qf ;
2005-04-17 02:20:36 +04:00
}