public inbox for linux-xfs@vger.kernel.org
 help / color / mirror / Atom feed
From: Mark Tinguely <tinguely@sgi.com>
To: Dave Chinner <david@fromorbit.com>
Cc: "Michael L. Semon" <mlsemon35@gmail.com>, xfs@oss.sgi.com
Subject: Re: [PATCH 44/49] xfs: Reduce allocations during CIL insertion
Date: Thu, 25 Jul 2013 10:02:40 -0500	[thread overview]
Message-ID: <51F13E10.1010805@sgi.com> (raw)
In-Reply-To: <20130725002108.GA11222@dastard>

On 07/24/13 19:21, Dave Chinner wrote:
> On Wed, Jul 24, 2013 at 08:28:42AM -0500, Mark Tinguely wrote:
>> If you could please redo the test and get the stack traces with
>> /proc/sysrq-trigger and if you kernel works with crash, a core dump.
>> For the stack trace, I mostly want to know if it has several
>> "xlog_grant_head_wait" entries in it, because ...
>>
>> ...I seemed to have triggered a couple log space reservation hangs
>> with fsstress one XFS partition and a mega-copy on another
>> partition, but will have to graft the new XFS tree onto a Linux 3.10
>> kernel to get crash (and one of my sata controllers) to work again.
>
> They are unrelated to this patchset.
>
> Somewhere in the code there
> is a mismatch between what we reserve as the base requirement for an
> actual log write and what the CIL actually steals, and that is, most
> likely, what is leading to log hangs.
>
> This is demonstratable in the fact that generic/070 on 512 byte
> block size filesystems regularly hits a transaction reservation
> exhausted assert failure on transaction commit of the periodic log
> dummy transaction on my test rigs.
>
> Cheers,
>
> Dave.

In testing patch 44, I did not trip over any cil stealing asserts before 
the hang. I think the cil steal assert is a different and a legitimate 
complaint. When I tripped over the ASSERT in with the v3 inode enabled, 
the writeid only reserves space for the sb but there were occasions of 
root btree and attribute fork entry that were also logged.

patch 43 runs for hours without incident. Previous to this series, I ran 
the same tests with parent pointer testing with much higher log 
reservations for day or two and never got a hang.

I tested patch 44 with copy like tests and both times it hung both times 
- not a convincing number of tests. A quick look, I see an empty AIL, 
empty CIL, the CTX is using 0 bytes, doesn't look like there are any cil 
pushes going nor any older ctx, the ctx has an empty ticket reservation. 
The log tail is  0xd000014d7 and reserve/grant is 0xe00204d04. The next 
reservation is for a rename transaction that uses just over the log 
space left. There has to be a log space leak. I will go back patch 43 on 
one machine and patch 44 on another and make sure it is patch 44 is 
causing the problem.

--Mark.

_______________________________________________
xfs mailing list
xfs@oss.sgi.com
http://oss.sgi.com/mailman/listinfo/xfs

  reply	other threads:[~2013-07-25 15:02 UTC|newest]

Thread overview: 158+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-07-19  6:24 [PATCH 00/49] current patch queue for 3.12 Dave Chinner
2013-07-19  6:24 ` [PATCH 01/49] xfs: separate out log format definitions Dave Chinner
2013-07-23 14:00   ` Brian Foster
2013-07-19  6:24 ` [PATCH 02/49] xfs: split out inode log item format definition Dave Chinner
2013-07-23 14:00   ` Brian Foster
2013-07-30 16:20   ` Mark Tinguely
2013-08-01  8:50     ` Christoph Hellwig
2013-07-19  6:24 ` [PATCH 03/49] xfs: split out buf log item format definitions Dave Chinner
2013-07-23 14:01   ` Brian Foster
2013-07-19  6:24 ` [PATCH 04/49] xfs: split out inode log item format definition Dave Chinner
2013-07-23 14:01   ` Brian Foster
2013-07-19  6:24 ` [PATCH 05/49] xfs: separate dquot on disk format definitions out of xfs_quota.h Dave Chinner
2013-07-24 12:09   ` Brian Foster
2013-08-01  8:51   ` Christoph Hellwig
2013-08-02  1:44     ` Dave Chinner
2013-07-19  6:24 ` [PATCH 06/49] xfs: separate icreate log format definitions from xfs_icreate_item.h Dave Chinner
2013-07-24 12:09   ` Brian Foster
2013-07-19  6:24 ` [PATCH 07/49] xfs: split out on-disk transaction definitions Dave Chinner
2013-07-24 12:09   ` Brian Foster
2013-07-19  6:24 ` [PATCH 08/49] xfs: introduce xfs_rtalloc_defs.h Dave Chinner
2013-07-24 12:09   ` Brian Foster
2013-07-19  6:24 ` [PATCH 09/49] xfs: introduce xfs_quota_defs.h Dave Chinner
2013-07-25 12:54   ` Brian Foster
2013-07-19  6:24 ` [PATCH 10/49] xfs: sync minor header differences needed by userspace Dave Chinner
2013-07-25 12:54   ` Brian Foster
2013-07-19  6:24 ` [PATCH 11/49] xfs: split out transaction reservation code Dave Chinner
2013-07-25 12:55   ` Brian Foster
2013-07-19  6:24 ` [PATCH 12/49] xfs: move inode fork definitions to a new header file Dave Chinner
2013-07-25 20:40   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 13/49] xfs: move unrealted definitions out of xfs_inode.h Dave Chinner
2013-07-25 19:24   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 14/49] xfs: introduce xfs_inode_buf.c for inode buffer operations Dave Chinner
2013-07-25 19:17   ` Mark Tinguely
2013-07-26  0:24     ` Dave Chinner
2013-07-26 12:41   ` Brian Foster
2013-07-19  6:24 ` [PATCH 15/49] xfs: move getdents code into it's own file Dave Chinner
2013-07-26 13:00   ` Brian Foster
2013-07-19  6:24 ` [PATCH 16/49] xfs: reshuffle dir2 definitions around for userspace Dave Chinner
2013-07-26 13:18   ` Brian Foster
2013-07-19  6:24 ` [PATCH 17/49] xfs: split out attribute listing code into separate file Dave Chinner
2013-07-27 20:23   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 18/49] xfs: split out attribute fork truncation " Dave Chinner
2013-07-27 19:25   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 19/49] xfs: split out the remote symlink handling Dave Chinner
2013-07-27 19:48   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 20/49] xfs: introduce xfs_sb.c for sharing with libxfs Dave Chinner
2013-07-27 17:54   ` Mark Tinguely
2013-07-28  1:08     ` Dave Chinner
2013-07-19  6:24 ` [PATCH 21/49] xfs: create xfs_bmap_util.[ch] Dave Chinner
2013-07-27 17:57   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 22/49] xfs: minor cleanups Dave Chinner
2013-07-27 18:00   ` Mark Tinguely
2013-07-28  1:07     ` Dave Chinner
2013-07-19  6:24 ` [PATCH 23/49] xfs: fix issues that cause userspace warnings Dave Chinner
2013-07-27 18:02   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 24/49] xfs: kill xfs_vnodeops.[ch] Dave Chinner
2013-07-26 19:18   ` Mark Tinguely
2013-07-27  1:55     ` Dave Chinner
2013-07-27 18:58       ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 25/49] xfs: consolidate xfs_rename.c Dave Chinner
2013-07-26 19:33   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 26/49] xfs: consolidate xfs_utils.c Dave Chinner
2013-07-26 20:16   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 27/49] xfs: consolidate extent swap code Dave Chinner
2013-07-26 21:16   ` Mark Tinguely
2013-07-19  6:24 ` [PATCH 28/49] xfs: don't special case shared superblock mounts Dave Chinner
2013-07-26 15:32   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 29/49] xfs: kill __KERNEL__ check for debug code in allocation code Dave Chinner
2013-07-26 15:07   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 30/49] xfs: remove __KERNEL__ from debug code Dave Chinner
2013-07-26 15:03   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 31/49] xfs: remove __KERNEL__ check from xfs_dir2_leaf.c Dave Chinner
2013-07-26 14:16   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 32/49] xfs: xfs_filestreams.h doesn't need __KERNEL__ Dave Chinner
2013-07-26 14:10   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 33/49] xfs: move kernel specific type definitions to xfs.h Dave Chinner
2013-07-26 13:51   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 34/49] xfs: make struct xfs_perag kernel only Dave Chinner
2013-07-26 13:32   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 35/49] xfs: Introduce a new structure to hold transaction reservation items Dave Chinner
2013-07-22 13:05   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 36/49] xfs: Introduce tr_fsyncts to m_reservation Dave Chinner
2013-07-22 13:22   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 37/49] xfs: Make writeid transaction use tr_writeid Dave Chinner
2013-07-22 13:23   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 38/49] xfs: refactor xfs_trans_reserve() interface Dave Chinner
2013-07-22 13:27   ` Mark Tinguely
2013-07-22 23:37     ` Dave Chinner
2013-07-19  6:25 ` [PATCH 39/49] xfs: Get rid of all XFS_XXX_LOG_RES() macro Dave Chinner
2013-07-22 13:31   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 40/49] xfs: Refactor xfs_ticket_alloc() to extract a new helper Dave Chinner
2013-07-22 13:49   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 41/49] xfs: Add xfs_log_rlimit.c Dave Chinner
2013-07-23 15:15   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 42/49] xfs: Validate log space at mount time Dave Chinner
2013-07-22 13:55   ` Mark Tinguely
2013-07-25  4:11   ` Dave Chinner
2013-07-19  6:25 ` [PATCH 43/49] xfs: return log item size in IOP_SIZE Dave Chinner
2013-07-23 18:22   ` Mark Tinguely
2013-08-01  8:18   ` Christoph Hellwig
2013-07-19  6:25 ` [PATCH 44/49] xfs: Reduce allocations during CIL insertion Dave Chinner
2013-07-23 21:15   ` Mark Tinguely
2013-07-23 21:44     ` Michael L. Semon
2013-07-24 13:28       ` Mark Tinguely
2013-07-24 19:20         ` Michael L. Semon
2013-07-25  0:21         ` Dave Chinner
2013-07-25 15:02           ` Mark Tinguely [this message]
2013-07-26  0:32             ` Dave Chinner
2013-07-26 20:46         ` Michael L. Semon
2013-07-26 21:06           ` Mark Tinguely
2013-07-26 22:19             ` Michael L. Semon
2013-07-27  1:58             ` Dave Chinner
2013-07-27 18:32               ` Mark Tinguely
2013-07-28  1:12                 ` Dave Chinner
2013-07-29 14:15                   ` Mark Tinguely
2013-07-30  0:30                     ` Dave Chinner
2013-07-30 13:31                       ` Mark Tinguely
2013-07-30 22:19                         ` Dave Chinner
2013-07-19  6:25 ` [PATCH 45/49] xfs: avoid CIL allocation during insert Dave Chinner
2013-07-29 18:13   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 46/49] xfs: Combine CIL insert and prepare passes Dave Chinner
2013-07-23 21:21   ` Mark Tinguely
2013-07-25  0:23     ` Dave Chinner
2013-07-29 21:07   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 47/49] xfs: split the CIL lock Dave Chinner
2013-07-29 22:24   ` Mark Tinguely
2013-07-19  6:25 ` [PATCH 48/49] xfs: Add read-only support for dirent filetype field Dave Chinner
2013-07-30 19:10   ` Mark Tinguely
2013-08-12  0:59     ` ***** SUSPECTED SPAM ***** " Dave Chinner
2013-08-12 13:25       ` Mark Tinguely
2013-08-13  0:50         ` Dave Chinner
2013-08-13 15:42           ` Mark Tinguely
2013-08-13 15:57             ` Christoph Hellwig
2013-08-14  7:50             ` Dave Chinner
2013-08-14 18:47               ` Geoffrey Wehrman
2013-08-15  4:22                 ` Dave Chinner
2013-08-15 17:39                   ` Geoffrey Wehrman
2013-08-15  5:59                 ` Ric Wheeler
2013-08-15 18:04                   ` Geoffrey Wehrman
2013-08-15 21:41                     ` Ric Wheeler
2013-08-16 14:08                       ` Geoffrey Wehrman
2013-08-19  5:28                         ` Dave Chinner
2013-08-19 18:48                           ` Geoffrey Wehrman
2013-08-20  2:23                             ` Dave Chinner
2013-08-20 14:30                               ` Geoffrey Wehrman
2013-08-20 18:27                                 ` Ric Wheeler
2013-08-20 19:47                         ` Eric Sandeen
2013-08-15 16:50                 ` Eric Sandeen
2013-08-15 18:32                   ` Geoffrey Wehrman
2013-08-15 18:41                     ` Eric Sandeen
2013-08-15 19:06                       ` Geoffrey Wehrman
2013-08-15 18:54                   ` Ben Myers
2013-08-15 22:40   ` Ben Myers
2013-07-19  6:25 ` [PATCH 49/49] xfs: Add write " Dave Chinner
2013-07-21  6:23 ` [PATCH 00/49] current patch queue for 3.12 Michael L. Semon
2013-07-22 23:43   ` Dave Chinner
2013-07-23  1:00     ` Michael L. Semon
2013-08-01 21:21 ` Ben Myers

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=51F13E10.1010805@sgi.com \
    --to=tinguely@sgi.com \
    --cc=david@fromorbit.com \
    --cc=mlsemon35@gmail.com \
    --cc=xfs@oss.sgi.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox