Fix congestion_wait() sync/async vs read/write confusion
Subsystems (12)
PATCH 1/1 Fix congestion_wait() sync/async vs read/write confusion
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is 1faa16d22877f4839bd433547d770c676d1d964c.
Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
arch/x86/lib/usercopy_32.c +1 -1
@@ -751,7 +751,7 @@ survive:if (retval == -ENOMEM &&is_global_init (current)) {up_read (¤t->mm->mmap_sem); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto survive; }
drivers/block/pktcdvd.c +1 -1
@@ -2595,7 +2595,7 @@ static int pkt_make_request(struct request_queue *q, struct bio *bio)set_bdi_congested (&q->backing_dev_info,WRITE );do {spin_unlock (&pd->lock); -congestion_wait (WRITE ,HZ ); +congestion_wait (BLK_RW_ASYNC ,HZ );spin_lock (&pd->lock); }while (pd->bio_queue_size > pd->write_congestion_off); }
drivers/md/dm-crypt.c +1 -1
@@ -776,7 +776,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)* But don't wait if split was due to the io size restriction */ if (unlikely (out_of_pages)) -congestion_wait (WRITE ,HZ /100 ); +congestion_wait (BLK_RW_ASYNC ,HZ /100 );/* * With async crypto it is unsafe to share the crypto context
fs/fat/file.c +1 -1
@@ -134,7 +134,7 @@ static int fat_file_release(struct inode *inode, struct file *filp)if ((filp->f_mode &FMODE_WRITE ) &&MSDOS_SB (inode->i_sb)->options.flush) {fat_flush_inodes (inode->i_sb, inode,NULL ); -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }return 0 ; }
fs/reiserfs/journal.c +1 -1
@@ -997,7 +997,7 @@ static int reiserfs_async_progress_wait(struct super_block *s)DEFINE_WAIT (wait);struct reiserfs_journal *j =SB_JOURNAL (s);if (atomic_read (&j->j_async_throttle)) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );return 0 ; }
fs/xfs/linux-2.6/kmem.c +2 -2
@@ -53,7 +53,7 @@ kmem_alloc(size_t size, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }
@@ -130,7 +130,7 @@ kmem_zone_alloc(kmem_zone_t *zone, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }
fs/xfs/linux-2.6/xfs_buf.c +1 -1
@@ -412,7 +412,7 @@ _xfs_buf_lookup_pages(XFS_STATS_INC (xb_page_retries);xfsbufd_wakeup (0 , gfp_mask); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto retry; }
mm/backing-dev.c +3 -4
@@ -283,7 +283,6 @@ static wait_queue_head_t congestion_wqh[2] = { __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1]) }; -void clear_bdi_congested (struct backing_dev_info *bdi,int sync) {enum bdi_state bit;
@@ -308,18 +307,18 @@ EXPORT_SYMBOL(set_bdi_congested);/** * congestion_wait - wait for a backing_dev to become uncongested -* @rw: READ or WRITE +* @sync: SYNC or ASYNC IO * @timeout: timeout in jiffies * * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit * write congestion. If no backing_devs are congested then just wait for the * next write to be completed. */ -long congestion_wait (int rw,long timeout) +long congestion_wait (int sync,long timeout) {long ret;DEFINE_WAIT (wait); -wait_queue_head_t *wqh = &congestion_wqh[rw]; +wait_queue_head_t *wqh = &congestion_wqh[sync];prepare_to_wait (wqh, &wait,TASK_UNINTERRUPTIBLE ); ret =io_schedule_timeout (timeout);
mm/memcontrol.c +1 -1
@@ -1973,7 +1973,7 @@ try_to_free:if (!progress) { nr_retries--;/* maybe some writeback is necessary */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } }
mm/page-writeback.c +4 -4
@@ -575,7 +575,7 @@ static void balance_dirty_pages(struct address_space *mapping)if (pages_written >= write_chunk)break ;/* We've done our duty */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh &&
@@ -669,7 +669,7 @@ void throttle_vm_writeout(gfp_t gfp_mask)if (global_page_state (NR_UNSTABLE_NFS ) +global_page_state (NR_WRITEBACK ) <= dirty_thresh)break ; -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The caller might hold locks which can prevent IO completion
@@ -715,7 +715,7 @@ static void background_writeout(unsigned long _min_pages)if (wbc.nr_to_write >0 || wbc.pages_skipped >0 ) {/* Wrote less than expected */ if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ; }
@@ -787,7 +787,7 @@ static void wb_kupdate(unsigned long arg)writeback_inodes (&wbc);if (wbc.nr_to_write >0 ) {if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ;/* All the old data is written */ }
mm/page_alloc.c +2 -2
@@ -1666,7 +1666,7 @@ __alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order, preferred_zone, migratetype);if (!page && gfp_mask & __GFP_NOFAIL) -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (!page && (gfp_mask & __GFP_NOFAIL));return page;
@@ -1831,7 +1831,7 @@ rebalance: pages_reclaimed += did_some_progress;if (should_alloc_retry (gfp_mask, order, pages_reclaimed)) {/* Wait for some write requests to complete then retry */ -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto rebalance; }
mm/vmscan.c +4 -4
@@ -1104,7 +1104,7 @@ static unsigned long shrink_inactive_list(unsigned long max_scan,*/ if (nr_freed < nr_taken && !current_is_kswapd () && lumpy_reclaim) { -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The attempt at page out may have made some
@@ -1721,7 +1721,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist,/* Take a nap, wait for some writeback to complete */ if (sc->nr_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }/* top priority shrink_zones still had more to do? don't OOM, then */ if (!sc->all_unreclaimable &&scanning_global_lru (sc))
@@ -1960,7 +1960,7 @@ loop_again:* another pass across the zones. */ if (total_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * We do this so kswapd doesn't build up large priorities for
@@ -2233,7 +2233,7 @@ unsigned long shrink_all_memory(unsigned long nr_pages) goto out;if (sc.nr_scanned && prio <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } }
Signature · 1 line
Jens Axboe
Quoted 8 lines · Jens Axboe
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is
1faa16d22877f4839bd433547d770c676d1d964c.
Quoted 4 lines
Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
Even if this does slow down some workloads, the bug is not in using the correct flag ;) So, I'd ack this one.
Jan Kara was able to reproduce the tiobench 2.6.30 regression, so I've cc'd him and kept the patch below.
-chris
arch/x86/lib/usercopy_32.c +1 -1
@@ -751,7 +751,7 @@ survive:if (retval == -ENOMEM &&is_global_init (current)) {up_read (¤t->mm->mmap_sem); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto survive; }
drivers/block/pktcdvd.c +1 -1
@@ -2595,7 +2595,7 @@ static int pkt_make_request(struct request_queue *q, struct bio *bio)set_bdi_congested (&q->backing_dev_info,WRITE );do {spin_unlock (&pd->lock); -congestion_wait (WRITE ,HZ ); +congestion_wait (BLK_RW_ASYNC ,HZ );spin_lock (&pd->lock); }while (pd->bio_queue_size > pd->write_congestion_off); }
drivers/md/dm-crypt.c +1 -1
@@ -776,7 +776,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)* But don't wait if split was due to the io size restriction */ if (unlikely (out_of_pages)) -congestion_wait (WRITE ,HZ /100 ); +congestion_wait (BLK_RW_ASYNC ,HZ /100 );/* * With async crypto it is unsafe to share the crypto context
fs/fat/file.c +1 -1
@@ -134,7 +134,7 @@ static int fat_file_release(struct inode *inode, struct file *filp)if ((filp->f_mode &FMODE_WRITE ) &&MSDOS_SB (inode->i_sb)->options.flush) {fat_flush_inodes (inode->i_sb, inode,NULL ); -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }return 0 ; }
fs/reiserfs/journal.c +1 -1
@@ -997,7 +997,7 @@ static int reiserfs_async_progress_wait(struct super_block *s)DEFINE_WAIT (wait);struct reiserfs_journal *j =SB_JOURNAL (s);if (atomic_read (&j->j_async_throttle)) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );return 0 ; }
fs/xfs/linux-2.6/kmem.c +2 -2
@@ -53,7 +53,7 @@ kmem_alloc(size_t size, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }
@@ -130,7 +130,7 @@ kmem_zone_alloc(kmem_zone_t *zone, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }
fs/xfs/linux-2.6/xfs_buf.c +1 -1
@@ -412,7 +412,7 @@ _xfs_buf_lookup_pages(XFS_STATS_INC (xb_page_retries);xfsbufd_wakeup (0 , gfp_mask); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto retry; }
mm/backing-dev.c +3 -4
@@ -283,7 +283,6 @@ static wait_queue_head_t congestion_wqh[2] = { __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1]) }; -void clear_bdi_congested (struct backing_dev_info *bdi,int sync) {enum bdi_state bit;
@@ -308,18 +307,18 @@ EXPORT_SYMBOL(set_bdi_congested);/** * congestion_wait - wait for a backing_dev to become uncongested -* @rw: READ or WRITE +* @sync: SYNC or ASYNC IO * @timeout: timeout in jiffies * * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit * write congestion. If no backing_devs are congested then just wait for the * next write to be completed. */ -long congestion_wait (int rw,long timeout) +long congestion_wait (int sync,long timeout) {long ret;DEFINE_WAIT (wait); -wait_queue_head_t *wqh = &congestion_wqh[rw]; +wait_queue_head_t *wqh = &congestion_wqh[sync];prepare_to_wait (wqh, &wait,TASK_UNINTERRUPTIBLE ); ret =io_schedule_timeout (timeout);
mm/memcontrol.c +1 -1
@@ -1973,7 +1973,7 @@ try_to_free:if (!progress) { nr_retries--;/* maybe some writeback is necessary */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } }
mm/page-writeback.c +4 -4
@@ -575,7 +575,7 @@ static void balance_dirty_pages(struct address_space *mapping)if (pages_written >= write_chunk)break ;/* We've done our duty */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh &&
@@ -669,7 +669,7 @@ void throttle_vm_writeout(gfp_t gfp_mask)if (global_page_state (NR_UNSTABLE_NFS ) +global_page_state (NR_WRITEBACK ) <= dirty_thresh)break ; -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The caller might hold locks which can prevent IO completion
@@ -715,7 +715,7 @@ static void background_writeout(unsigned long _min_pages)if (wbc.nr_to_write >0 || wbc.pages_skipped >0 ) {/* Wrote less than expected */ if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ; }
@@ -787,7 +787,7 @@ static void wb_kupdate(unsigned long arg)writeback_inodes (&wbc);if (wbc.nr_to_write >0 ) {if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ;/* All the old data is written */ }
mm/page_alloc.c +2 -2
@@ -1666,7 +1666,7 @@ __alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order, preferred_zone, migratetype);if (!page && gfp_mask & __GFP_NOFAIL) -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (!page && (gfp_mask & __GFP_NOFAIL));return page;
@@ -1831,7 +1831,7 @@ rebalance: pages_reclaimed += did_some_progress;if (should_alloc_retry (gfp_mask, order, pages_reclaimed)) {/* Wait for some write requests to complete then retry */ -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto rebalance; }
mm/vmscan.c +4 -4
@@ -1104,7 +1104,7 @@ static unsigned long shrink_inactive_list(unsigned long max_scan,*/ if (nr_freed < nr_taken && !current_is_kswapd () && lumpy_reclaim) { -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The attempt at page out may have made some
@@ -1721,7 +1721,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist,/* Take a nap, wait for some writeback to complete */ if (sc->nr_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }/* top priority shrink_zones still had more to do? don't OOM, then */ if (!sc->all_unreclaimable &&scanning_global_lru (sc))
@@ -1960,7 +1960,7 @@ loop_again:* another pass across the zones. */ if (total_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * We do this so kswapd doesn't build up large priorities for
@@ -2233,7 +2233,7 @@ unsigned long shrink_all_memory(unsigned long nr_pages) goto out;if (sc.nr_scanned && prio <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } }
Signature · 1 line
Jens Axboe
Quoted 8 lines · Jens Axboe
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is
1faa16d22877f4839bd433547d770c676d1d964c.
Quoted 4 lines
Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
Thanks, but I'm not sure which one do you mean in particular.
http://bugzilla.kernel.org/show_bug.cgi?id=13408 looks like it might be somewhat related.
The complete list is at http://lkml.org/lkml/2009/7/6/383 .
Best,
Rafael
Quoted 245 lines · Jens Axboe
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is
1faa16d22877f4839bd433547d770c676d1d964c.Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
arch/x86/lib/usercopy_32.c +1
-1@@ -751,7 +751,7 @@ survive:if (retval == -ENOMEM &&is_global_init (current)) {up_read (¤t->mm->mmap_sem); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto survive; }drivers/block/pktcdvd.c +1
-1@@ -2595,7 +2595,7 @@ static int pkt_make_request(struct request_queue *q, struct bio *bio)set_bdi_congested (&q->backing_dev_info,WRITE );do {spin_unlock (&pd->lock); -congestion_wait (WRITE ,HZ ); +congestion_wait (BLK_RW_ASYNC ,HZ );spin_lock (&pd->lock); }while (pd->bio_queue_size > pd->write_congestion_off); }drivers/md/dm-crypt.c +1
-1@@ -776,7 +776,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)* But don't wait if split was due to the io size restriction */ if (unlikely (out_of_pages)) -congestion_wait (WRITE ,HZ /100 ); +congestion_wait (BLK_RW_ASYNC ,HZ /100 );/* * With async crypto it is unsafe to share the crypto context fs/fat/file.c +1
-1@@ -134,7 +134,7 @@ static int fat_file_release(struct inode *inode, struct file *filp)if ((filp->f_mode &FMODE_WRITE ) &&MSDOS_SB (inode->i_sb)->options.flush) {fat_flush_inodes (inode->i_sb, inode,NULL ); -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }return 0 ; }fs/reiserfs/journal.c +1
-1@@ -997,7 +997,7 @@ static int reiserfs_async_progress_wait(struct super_block *s)DEFINE_WAIT (wait);struct reiserfs_journal *j =SB_JOURNAL (s);if (atomic_read (&j->j_async_throttle)) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );return 0 ; }fs/xfs/linux-2.6/kmem.c +2
-2@@ -53,7 +53,7 @@ kmem_alloc(size_t size, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }@@ -130,7 +130,7 @@ kmem_zone_alloc(kmem_zone_t *zone, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }fs/xfs/linux-2.6/xfs_buf.c +1
-1@@ -412,7 +412,7 @@ _xfs_buf_lookup_pages(XFS_STATS_INC (xb_page_retries);xfsbufd_wakeup (0 , gfp_mask); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto retry; }mm/backing-dev.c +3
-4@@ -283,7 +283,6 @@ static wait_queue_head_t congestion_wqh[2] = { __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1]) }; -void clear_bdi_congested (struct backing_dev_info *bdi,int sync) {enum bdi_state bit;@@ -308,18 +307,18 @@ EXPORT_SYMBOL(set_bdi_congested);/** * congestion_wait - wait for a backing_dev to become uncongested -* @rw: READ or WRITE +* @sync: SYNC or ASYNC IO * @timeout: timeout in jiffies * * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit * write congestion. If no backing_devs are congested then just wait for the * next write to be completed. */ -long congestion_wait (int rw,long timeout) +long congestion_wait (int sync,long timeout) {long ret;DEFINE_WAIT (wait); -wait_queue_head_t *wqh = &congestion_wqh[rw]; +wait_queue_head_t *wqh = &congestion_wqh[sync];prepare_to_wait (wqh, &wait,TASK_UNINTERRUPTIBLE ); ret =io_schedule_timeout (timeout);mm/memcontrol.c +1
-1@@ -1973,7 +1973,7 @@ try_to_free:if (!progress) { nr_retries--;/* maybe some writeback is necessary */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } }mm/page-writeback.c +4
-4@@ -575,7 +575,7 @@ static void balance_dirty_pages(struct address_space *mapping)if (pages_written >= write_chunk)break ;/* We've done our duty */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh &&@@ -669,7 +669,7 @@ void throttle_vm_writeout(gfp_t gfp_mask)if (global_page_state (NR_UNSTABLE_NFS ) +global_page_state (NR_WRITEBACK ) <= dirty_thresh)break ; -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The caller might hold locks which can prevent IO completion @@ -715,7 +715,7 @@ static void background_writeout(unsigned long _min_pages)if (wbc.nr_to_write >0 || wbc.pages_skipped >0 ) {/* Wrote less than expected */ if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ; }@@ -787,7 +787,7 @@ static void wb_kupdate(unsigned long arg)writeback_inodes (&wbc);if (wbc.nr_to_write >0 ) {if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ;/* All the old data is written */ }mm/page_alloc.c +2
-2@@ -1666,7 +1666,7 @@ __alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order, preferred_zone, migratetype);if (!page && gfp_mask & __GFP_NOFAIL) -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (!page && (gfp_mask & __GFP_NOFAIL));return page;@@ -1831,7 +1831,7 @@ rebalance: pages_reclaimed += did_some_progress;if (should_alloc_retry (gfp_mask, order, pages_reclaimed)) {/* Wait for some write requests to complete then retry */ -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto rebalance; }mm/vmscan.c +2
-2@@ -1104,7 +1104,7 @@ static unsigned long shrink_inactive_list(unsigned long max_scan,*/ if (nr_freed < nr_taken && !current_is_kswapd () && lumpy_reclaim) { -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The attempt at page out may have made some @@ -1721,7 +1721,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist,/* Take a nap, wait for some writeback to complete */ if (sc->nr_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }/* top priority shrink_zones still had more to do? don't OOM, then */ if (!sc->all_unreclaimable &&scanning_global_lru (sc))
mm/vmscan.c loop_again: · line 1963
Quoted 5 lines
mm/vmscan.c +1
-1@@ -1960,7 +1960,7 @@ loop_again:* another pass across the zones. */ if (total_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );
Oh, great...
This particular change will affect _all_ users of set_bdi_congested(WRITE)/clear_bdi_congested(WRITE). If you're going to do this, then you had better be prepared to change them all. There's one in fs/nfs/write.c...
Trond
Quoted 12 lines
mm/vmscan.c +1
-1/* * We do this so kswapd doesn't build up large priorities for @@ -2233,7 +2233,7 @@ unsigned long shrink_all_memory(unsigned long nr_pages) goto out;if (sc.nr_scanned && prio <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } }
Quoted 254 lines · Trond Myklebust
Quoted 245 lines · Jens Axboe
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is
1faa16d22877f4839bd433547d770c676d1d964c.Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
arch/x86/lib/usercopy_32.c +1
-1@@ -751,7 +751,7 @@ survive:if (retval == -ENOMEM &&is_global_init (current)) {up_read (¤t->mm->mmap_sem); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto survive; }drivers/block/pktcdvd.c +1
-1@@ -2595,7 +2595,7 @@ static int pkt_make_request(struct request_queue *q, struct bio *bio)set_bdi_congested (&q->backing_dev_info,WRITE );do {spin_unlock (&pd->lock); -congestion_wait (WRITE ,HZ ); +congestion_wait (BLK_RW_ASYNC ,HZ );spin_lock (&pd->lock); }while (pd->bio_queue_size > pd->write_congestion_off); }drivers/md/dm-crypt.c +1
-1@@ -776,7 +776,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)* But don't wait if split was due to the io size restriction */ if (unlikely (out_of_pages)) -congestion_wait (WRITE ,HZ /100 ); +congestion_wait (BLK_RW_ASYNC ,HZ /100 );/* * With async crypto it is unsafe to share the crypto context fs/fat/file.c +1
-1@@ -134,7 +134,7 @@ static int fat_file_release(struct inode *inode, struct file *filp)if ((filp->f_mode &FMODE_WRITE ) &&MSDOS_SB (inode->i_sb)->options.flush) {fat_flush_inodes (inode->i_sb, inode,NULL ); -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }return 0 ; }fs/reiserfs/journal.c +1
-1@@ -997,7 +997,7 @@ static int reiserfs_async_progress_wait(struct super_block *s)DEFINE_WAIT (wait);struct reiserfs_journal *j =SB_JOURNAL (s);if (atomic_read (&j->j_async_throttle)) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );return 0 ; }fs/xfs/linux-2.6/kmem.c +2
-2@@ -53,7 +53,7 @@ kmem_alloc(size_t size, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }@@ -130,7 +130,7 @@ kmem_zone_alloc(kmem_zone_t *zone, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }fs/xfs/linux-2.6/xfs_buf.c +1
-1@@ -412,7 +412,7 @@ _xfs_buf_lookup_pages(XFS_STATS_INC (xb_page_retries);xfsbufd_wakeup (0 , gfp_mask); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto retry; }mm/backing-dev.c +3
-4@@ -283,7 +283,6 @@ static wait_queue_head_t congestion_wqh[2] = { __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1]) }; -void clear_bdi_congested (struct backing_dev_info *bdi,int sync) {enum bdi_state bit;@@ -308,18 +307,18 @@ EXPORT_SYMBOL(set_bdi_congested);/** * congestion_wait - wait for a backing_dev to become uncongested -* @rw: READ or WRITE +* @sync: SYNC or ASYNC IO * @timeout: timeout in jiffies * * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit * write congestion. If no backing_devs are congested then just wait for the * next write to be completed. */ -long congestion_wait (int rw,long timeout) +long congestion_wait (int sync,long timeout) {long ret;DEFINE_WAIT (wait); -wait_queue_head_t *wqh = &congestion_wqh[rw]; +wait_queue_head_t *wqh = &congestion_wqh[sync];prepare_to_wait (wqh, &wait,TASK_UNINTERRUPTIBLE ); ret =io_schedule_timeout (timeout);mm/memcontrol.c +1
-1@@ -1973,7 +1973,7 @@ try_to_free:if (!progress) { nr_retries--;/* maybe some writeback is necessary */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } }mm/page-writeback.c +4
-4@@ -575,7 +575,7 @@ static void balance_dirty_pages(struct address_space *mapping)if (pages_written >= write_chunk)break ;/* We've done our duty */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh &&@@ -669,7 +669,7 @@ void throttle_vm_writeout(gfp_t gfp_mask)if (global_page_state (NR_UNSTABLE_NFS ) +global_page_state (NR_WRITEBACK ) <= dirty_thresh)break ; -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The caller might hold locks which can prevent IO completion @@ -715,7 +715,7 @@ static void background_writeout(unsigned long _min_pages)if (wbc.nr_to_write >0 || wbc.pages_skipped >0 ) {/* Wrote less than expected */ if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ; }@@ -787,7 +787,7 @@ static void wb_kupdate(unsigned long arg)writeback_inodes (&wbc);if (wbc.nr_to_write >0 ) {if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ;/* All the old data is written */ }mm/page_alloc.c +2
-2@@ -1666,7 +1666,7 @@ __alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order, preferred_zone, migratetype);if (!page && gfp_mask & __GFP_NOFAIL) -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (!page && (gfp_mask & __GFP_NOFAIL));return page;@@ -1831,7 +1831,7 @@ rebalance: pages_reclaimed += did_some_progress;if (should_alloc_retry (gfp_mask, order, pages_reclaimed)) {/* Wait for some write requests to complete then retry */ -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto rebalance; }mm/vmscan.c +2
-2@@ -1104,7 +1104,7 @@ static unsigned long shrink_inactive_list(unsigned long max_scan,*/ if (nr_freed < nr_taken && !current_is_kswapd () && lumpy_reclaim) { -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The attempt at page out may have made some @@ -1721,7 +1721,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist,/* Take a nap, wait for some writeback to complete */ if (sc->nr_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }/* top priority shrink_zones still had more to do? don't OOM, then */ if (!sc->all_unreclaimable &&scanning_global_lru (sc))mm/vmscan.c loop_again: · line 1963
Quoted 5 lines
mm/vmscan.c +1
-1@@ -1960,7 +1960,7 @@ loop_again:* another pass across the zones. */ if (total_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );Oh, great...
Quoted 4 lines
This particular change will affect _all_ users of set_bdi_congested(WRITE)/clear_bdi_congested(WRITE). If you're going to do this, then you had better be prepared to change them all. There's one in
fs/nfs/write.c...
More specifically, you need to audit and fix:
git grep '\(set\|clear\)_bdi_congested *(.*, *\(READ\|WRITE\) *)'
drivers/block/pktcdvd.c: clear_bdi_congested(&pd->disk->queue->ba drivers/block/pktcdvd.c: set_bdi_congested(&q->backing_dev_info, fs/fuse/dev.c: clear_bdi_congested(&fc->bdi, READ); fs/fuse/dev.c: clear_bdi_congested(&fc->bdi, WRITE); fs/fuse/dev.c: set_bdi_congested(&fc->bdi, READ); fs/fuse/dev.c: set_bdi_congested(&fc->bdi, WRITE); fs/nfs/write.c: set_bdi_congested(&nfss->backing_dev_info, WRITE fs/nfs/write.c: clear_bdi_congested(&nfss->backing_dev_info, WRITE);
Quoted 263 lines · Trond Myklebust
Quoted 254 lines · Trond Myklebust
Quoted 245 lines · Jens Axboe
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is
1faa16d22877f4839bd433547d770c676d1d964c.Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
arch/x86/lib/usercopy_32.c +1
-1@@ -751,7 +751,7 @@ survive:if (retval == -ENOMEM &&is_global_init (current)) {up_read (¤t->mm->mmap_sem); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto survive; }drivers/block/pktcdvd.c +1
-1@@ -2595,7 +2595,7 @@ static int pkt_make_request(struct request_queue *q, struct bio *bio)set_bdi_congested (&q->backing_dev_info,WRITE );do {spin_unlock (&pd->lock); -congestion_wait (WRITE ,HZ ); +congestion_wait (BLK_RW_ASYNC ,HZ );spin_lock (&pd->lock); }while (pd->bio_queue_size > pd->write_congestion_off); }drivers/md/dm-crypt.c +1
-1@@ -776,7 +776,7 @@ static void kcryptd_crypt_write_convert(struct dm_crypt_io *io)* But don't wait if split was due to the io size restriction */ if (unlikely (out_of_pages)) -congestion_wait (WRITE ,HZ /100 ); +congestion_wait (BLK_RW_ASYNC ,HZ /100 );/* * With async crypto it is unsafe to share the crypto context fs/fat/file.c +1
-1@@ -134,7 +134,7 @@ static int fat_file_release(struct inode *inode, struct file *filp)if ((filp->f_mode &FMODE_WRITE ) &&MSDOS_SB (inode->i_sb)->options.flush) {fat_flush_inodes (inode->i_sb, inode,NULL ); -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }return 0 ; }fs/reiserfs/journal.c +1
-1@@ -997,7 +997,7 @@ static int reiserfs_async_progress_wait(struct super_block *s)DEFINE_WAIT (wait);struct reiserfs_journal *j =SB_JOURNAL (s);if (atomic_read (&j->j_async_throttle)) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );return 0 ; }fs/xfs/linux-2.6/kmem.c +2
-2@@ -53,7 +53,7 @@ kmem_alloc(size_t size, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }@@ -130,7 +130,7 @@ kmem_zone_alloc(kmem_zone_t *zone, unsigned int __nocast flags)printk (KERN_ERR "XFS: possible memory allocation " "deadlock in %s (mode:0x%x)\n" , __func__, lflags); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (1 ); }fs/xfs/linux-2.6/xfs_buf.c +1
-1@@ -412,7 +412,7 @@ _xfs_buf_lookup_pages(XFS_STATS_INC (xb_page_retries);xfsbufd_wakeup (0 , gfp_mask); -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto retry; }mm/backing-dev.c +3
-4@@ -283,7 +283,6 @@ static wait_queue_head_t congestion_wqh[2] = { __WAIT_QUEUE_HEAD_INITIALIZER(congestion_wqh[1]) }; -void clear_bdi_congested (struct backing_dev_info *bdi,int sync) {enum bdi_state bit;@@ -308,18 +307,18 @@ EXPORT_SYMBOL(set_bdi_congested);/** * congestion_wait - wait for a backing_dev to become uncongested -* @rw: READ or WRITE +* @sync: SYNC or ASYNC IO * @timeout: timeout in jiffies * * Waits for up to @timeout jiffies for a backing_dev (any backing_dev) to exit * write congestion. If no backing_devs are congested then just wait for the * next write to be completed. */ -long congestion_wait (int rw,long timeout) +long congestion_wait (int sync,long timeout) {long ret;DEFINE_WAIT (wait); -wait_queue_head_t *wqh = &congestion_wqh[rw]; +wait_queue_head_t *wqh = &congestion_wqh[sync];prepare_to_wait (wqh, &wait,TASK_UNINTERRUPTIBLE ); ret =io_schedule_timeout (timeout);mm/memcontrol.c +1
-1@@ -1973,7 +1973,7 @@ try_to_free:if (!progress) { nr_retries--;/* maybe some writeback is necessary */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } }mm/page-writeback.c +4
-4@@ -575,7 +575,7 @@ static void balance_dirty_pages(struct address_space *mapping)if (pages_written >= write_chunk)break ;/* We've done our duty */ -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); } if (bdi_nr_reclaimable + bdi_nr_writeback < bdi_thresh &&@@ -669,7 +669,7 @@ void throttle_vm_writeout(gfp_t gfp_mask)if (global_page_state (NR_UNSTABLE_NFS ) +global_page_state (NR_WRITEBACK ) <= dirty_thresh)break ; -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The caller might hold locks which can prevent IO completion @@ -715,7 +715,7 @@ static void background_writeout(unsigned long _min_pages)if (wbc.nr_to_write >0 || wbc.pages_skipped >0 ) {/* Wrote less than expected */ if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ; }@@ -787,7 +787,7 @@ static void wb_kupdate(unsigned long arg)writeback_inodes (&wbc);if (wbc.nr_to_write >0 ) {if (wbc.encountered_congestion || wbc.more_io) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );else break ;/* All the old data is written */ }mm/page_alloc.c +2
-2@@ -1666,7 +1666,7 @@ __alloc_pages_high_priority(gfp_t gfp_mask, unsigned int order, preferred_zone, migratetype);if (!page && gfp_mask & __GFP_NOFAIL) -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); }while (!page && (gfp_mask & __GFP_NOFAIL));return page;@@ -1831,7 +1831,7 @@ rebalance: pages_reclaimed += did_some_progress;if (should_alloc_retry (gfp_mask, order, pages_reclaimed)) {/* Wait for some write requests to complete then retry */ -congestion_wait (WRITE ,HZ /50 ); +congestion_wait (BLK_RW_ASYNC ,HZ /50 ); goto rebalance; }mm/vmscan.c +2
-2@@ -1104,7 +1104,7 @@ static unsigned long shrink_inactive_list(unsigned long max_scan,*/ if (nr_freed < nr_taken && !current_is_kswapd () && lumpy_reclaim) { -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );/* * The attempt at page out may have made some @@ -1721,7 +1721,7 @@ static unsigned long do_try_to_free_pages(struct zonelist *zonelist,/* Take a nap, wait for some writeback to complete */ if (sc->nr_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 ); }/* top priority shrink_zones still had more to do? don't OOM, then */ if (!sc->all_unreclaimable &&scanning_global_lru (sc))mm/vmscan.c loop_again: · line 1963
Quoted 5 lines
mm/vmscan.c +1
-1@@ -1960,7 +1960,7 @@ loop_again:* another pass across the zones. */ if (total_scanned && priority <DEF_PRIORITY -2 ) -congestion_wait (WRITE ,HZ /10 ); +congestion_wait (BLK_RW_ASYNC ,HZ /10 );Oh, great...
Quoted 4 lines
This particular change will affect _all_ users of set_bdi_congested(WRITE)/clear_bdi_congested(WRITE). If you're going to do this, then you had better be prepared to change them all. There's one in
fs/nfs/write.c...More specifically, you need to audit and fix:
git grep '\(set\|clear\)_bdi_congested *(.*, *\(READ\|WRITE\) *)'
Quoted 8 lines
drivers/block/pktcdvd.c: clear_bdi_congested(&pd->disk->queue->ba drivers/block/pktcdvd.c: set_bdi_congested(&q->backing_dev_info, fs/fuse/dev.c: clear_bdi_congested(&fc->bdi, READ); fs/fuse/dev.c: clear_bdi_congested(&fc->bdi, WRITE); fs/fuse/dev.c: set_bdi_congested(&fc->bdi, READ); fs/fuse/dev.c: set_bdi_congested(&fc->bdi, WRITE); fs/nfs/write.c: set_bdi_congested(&nfss->backing_dev_info, WRITE fs/nfs/write.c: clear_bdi_congested(&nfss->backing_dev_info, WRITE);
Thanks, not sure why those were missed. I'll go over everything again, thanks for checking!
Signature · 1 line
Jens Axboe
Quoted 8 lines
Quoted 6 lines
Oh, great...
This particular change will affect _all_ users of set_bdi_congested(WRITE)/clear_bdi_congested(WRITE). If you're going to do this, then you had better be prepared to change them all. There's one in
fs/nfs/write.c...More specifically, you need to audit and fix:
Please stop quoting 250+ lines of patch for a couple of lines reply. Seriously that's just extremly rude and annoying.
Jan Kara
maintainerQuoted 14 lines · Chris Mason
Quoted 8 lines · Jens Axboe
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is
1faa16d22877f4839bd433547d770c676d1d964c.Quoted 4 lines
Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
Quoted 5 lines
Even if this does slow down some workloads, the bug is not in using the correct flag ;) So, I'd ack this one.
Jan Kara was able to reproduce the tiobench 2.6.30 regression, so I've cc'd him and kept the patch below.
Thanks for the patch Chris. I've remeasured tiobench with the 2.6.30 +
the fix but it didn't help (which is not too surprising as what I observe is most likely CFQ related as there's no regression with NOOP scheduler).
Just to recall:
2.6.29 (CFQ) Avg StdDev
8 38.01 40.26 39.69 -> 39.32 0.955092 16 40.09 38.18 40.05 -> 39.44 0.891104
2.6.30-rc8 (CFQ) 8 36.67 36.81 38.20 -> 37.23 0.69062 16 37.45 36.47 37.46 -> 37.13 0.464351
2.6.30-rc8+fix (CFQ) 8 37.56 37.38 37.98 -> 37.64 0.251396 16 38.11 36.71 37.18 -> 37.33 0.581741
So with the fix there's no statistically significant difference and we
are still below 2.6.29 results. I'm now going to retest with the WRITE_SYNC changes reverted.
Honza
--
Jan Kara <[email protected]>
SUSE Labs, CR
Quoted 32 lines · Jan Kara
Quoted 14 lines · Chris Mason
Quoted 8 lines · Jens Axboe
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is
1faa16d22877f4839bd433547d770c676d1d964c.Quoted 4 lines
Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
Quoted 5 lines
Even if this does slow down some workloads, the bug is not in using the correct flag ;) So, I'd ack this one.
Jan Kara was able to reproduce the tiobench 2.6.30 regression, so I've cc'd him and kept the patch below.
Thanks for the patch Chris. I've remeasured tiobench with the 2.6.30 +
the fix but it didn't help (which is not too surprising as what I observe is most likely CFQ related as there's no regression with NOOP scheduler).
Just to recall:
2.6.29 (CFQ) Avg StdDev8 38.01 40.26 39.69 -> 39.32 0.955092 16 40.09 38.18 40.05 -> 39.44 0.8911042.6.30-rc8 (CFQ) 8 36.67 36.81 38.20 -> 37.23 0.69062 16 37.45 36.47 37.46 -> 37.13 0.464351
Quoted 7 lines
2.6.30-rc8+fix (CFQ) 8 37.56 37.38 37.98 -> 37.64 0.251396 16 38.11 36.71 37.18 -> 37.33 0.581741So with the fix there's no statistically significant difference and we
are still below 2.6.29 results. I'm now going to retest with theWRITE_SYNCchanges reverted.
Well, its good the patch didn't make things worse ;) I didn't have the highest hopes that it would resolve the regression, but thanks for testing!
-chris
Jan Kara
maintainerQuoted 37 lines · Chris Mason
Quoted 36 lines · Jan Kara
Quoted 14 lines · Chris Mason
Quoted 8 lines · Jens Axboe
Hi,
This one isn't great, we currently have broken congestion wait logic in the kernel. 2.6.30 is impacted as well, so this patch should go to stable too once it's in -git. I'll let this one simmer until tomorrow, then ask Linus to pull it. The offending commit breaking this is
1faa16d22877f4839bd433547d770c676d1d964c.Quoted 4 lines
Meanwhile, it could potentially cause buffered writeout slowdowns in the kernel. Perhaps the 2.6.30 regression in that area is caused by this? Would be interesting if the submitter could test. I can't find the list, CC'ing Rafael.
Quoted 5 lines
Even if this does slow down some workloads, the bug is not in using the correct flag ;) So, I'd ack this one.
Jan Kara was able to reproduce the tiobench 2.6.30 regression, so I've cc'd him and kept the patch below.
Thanks for the patch Chris. I've remeasured tiobench with the 2.6.30 +
the fix but it didn't help (which is not too surprising as what I observe is most likely CFQ related as there's no regression with NOOP scheduler).
Just to recall:
2.6.29 (CFQ) Avg StdDev8 38.01 40.26 39.69 -> 39.32 0.955092 16 40.09 38.18 40.05 -> 39.44 0.8911042.6.30-rc8 (CFQ) 8 36.67 36.81 38.20 -> 37.23 0.69062 16 37.45 36.47 37.46 -> 37.13 0.4643512.6.30-rc8+fix (CFQ) 8 37.56 37.38 37.98 -> 37.64 0.251396 16 38.11 36.71 37.18 -> 37.33 0.581741
Quoted 7 lines
Quoted 3 lines
So with the fix there's no statistically significant difference and we
are still below 2.6.29 results. I'm now going to retest with theWRITE_SYNCchanges reverted.Well, its good the patch didn't make things worse ;) I didn't have the highest hopes that it would resolve the regression, but thanks for testing!
I've now tried to revert everything which looked WRITE_SYNC related but
it didn't help either. Now, I'm trying to basically bisect CFQ changes and I'll see whether it goes somewhere...
Honza
Signature · 2 lines
Jan Kara <[email protected]> SUSE Labs, CR