[Top] [All Lists]

Re: [PATCH 09/10] xfs: on-stack delayed write buffer lists

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject: Re: [PATCH 09/10] xfs: on-stack delayed write buffer lists
From: Mark Tinguely <tinguely@xxxxxxx>
Date: Fri, 20 Apr 2012 13:19:46 -0500
Cc: xfs@xxxxxxxxxxx
In-reply-to: <20120327164646.975031281@xxxxxxxxxxxxxxxxxxxxxx>
References: <20120327164400.967415009@xxxxxxxxxxxxxxxxxxxxxx> <20120327164646.975031281@xxxxxxxxxxxxxxxxxxxxxx>
User-agent: Mozilla/5.0 (X11; FreeBSD amd64; rv:9.0) Gecko/20120122 Thunderbird/9.0
On 03/27/12 11:44, Christoph Hellwig wrote:
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 handles 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.

Signed-off-by: Christoph Hellwig<hch@xxxxxx>

Test 106 runs to completion with patch 06.

Patch 07 and 08 do not compile without patch 09.

Starting with patch 09, I get the following hang on every test 106:

ID: 27992  TASK: ffff8808310d00c0  CPU: 2   COMMAND: "mount"
 #0 [ffff880834237938] __schedule at ffffffff81417200
 #1 [ffff880834237a80] schedule at ffffffff81417574
 #2 [ffff880834237a90] schedule_timeout at ffffffff81415805
 #3 [ffff880834237b30] wait_for_common at ffffffff81416a67
 #4 [ffff880834237bc0] wait_for_completion at ffffffff81416bd8
 #5 [ffff880834237bd0] xfs_buf_iowait at ffffffffa04fc5a5 [xfs]
 #6 [ffff880834237c00] xfs_buf_delwri_submit at ffffffffa04fe4b9 [xfs]
 #7 [ffff880834237c40] xfs_qm_quotacheck at ffffffffa055cb2d [xfs]
 #8 [ffff880834237cc0] xfs_qm_mount_quotas at ffffffffa055cdf0 [xfs]
 #9 [ffff880834237cf0] xfs_mountfs at ffffffffa054c041 [xfs]
#10 [ffff880834237d40] xfs_fs_fill_super at ffffffffa050ca80 [xfs]
#11 [ffff880834237d70] mount_bdev at ffffffff81150c5c
#12 [ffff880834237de0] xfs_fs_mount at ffffffffa050ac00 [xfs]
#13 [ffff880834237df0] mount_fs at ffffffff811505f8
#14 [ffff880834237e40] vfs_kern_mount at ffffffff8116c070
#15 [ffff880834237e80] do_kern_mount at ffffffff8116c16e
#16 [ffff880834237ec0] do_mount at ffffffff8116d6f0
#17 [ffff880834237f20] sys_mount at ffffffff8116d7f3
#18 [ffff880834237f80] system_call_fastpath at ffffffff814203b9

The workers seem to be idle. For example the xfsaild:

PID: 27676  TASK: ffff880832880240  CPU: 3   COMMAND: "xfsaild/sda7"
 #0 [ffff880832933cb0] __schedule at ffffffff81417200
 #1 [ffff880832933df8] schedule at ffffffff81417574
 #2 [ffff880832933e08] schedule_timeout at ffffffff81415805
 #3 [ffff880832933ea8] xfsaild at ffffffffa0555935 [xfs]
 #4 [ffff880832933ee8] kthread at ffffffff8105dd6e
 #5 [ffff880832933f48] kernel_thread_helper at ffffffff814216a4

The hang is on the third quotacheck.

Should be easy to duplicate this.

--Mark Tinguely.

<Prev in Thread] Current Thread [Next in Thread>