Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation

To:	Christoph Hellwig <hch@xxxxxxxxxxxxx>
Subject:	Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation
From:	Dave Chinner <david@xxxxxxxxxxxxx>
Date:	Wed, 22 Dec 2010 08:42:40 +1100
Cc:	xfs@xxxxxxxxxxx
In-reply-to:	<20101221151511.GA26127@xxxxxxxxxxxxx>
References:	<1292916570-25015-1-git-send-email-david@xxxxxxxxxxxxx> <1292916570-25015-7-git-send-email-david@xxxxxxxxxxxxx> <20101221151511.GA26127@xxxxxxxxxxxxx>
User-agent:	Mutt/1.5.20 (2009-06-14)

On Tue, Dec 21, 2010 at 10:15:11AM -0500, Christoph Hellwig wrote:
> This patch causes tesr 014 to take ~ 870 seconds instead of 6, thus
> beeing almost 150 times slower on mt 32-bit test VM, so I'll have to NAK
> it for now.

It's not the speculative preallocation changes - ithey are just
exposing some other regression. That is, using MOUNT_OPTIONS="-o
allocsize=4k" gives the previous behaviour, while allocsize=512m
gives the same behaviour as the dynamic preallocation.

The dynamic behaviour is resulting in megabyte sized IOs being
issued for random 512 byte writes (which is wrong), so I'm tending
towards it being a regression caused by the reecent writeback path
changes. I'll dig deeper today.

Cheers,

Dave.
-- 
Dave Chinner
david@xxxxxxxxxxxxx

<Prev in Thread]	Current Thread	[Next in Thread>
[PATCH 02/34] xfs: use KM_NOFS for allocations during attribute list operations, (continued) [PATCH 02/34] xfs: use KM_NOFS for allocations during attribute list operations, Dave Chinner Re: [PATCH 02/34] xfs: use KM_NOFS for allocations during attribute list operations, Christoph Hellwig [PATCH 03/34] lib: percpu counter add unless less than functionality, Dave Chinner Re: [PATCH 03/34] lib: percpu counter add unless less than functionality, Alex Elder Re: [PATCH 03/34] lib: percpu counter add unless less than functionality, Dave Chinner [PATCH 01/34] xfs: provide a inode iolock lockdep class, Dave Chinner Re: [PATCH 01/34] xfs: provide a inode iolock lockdep class, Christoph Hellwig [PATCH 05/34] xfs: demultiplex xfs_icsb_modify_counters(), Dave Chinner [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Dave Chinner Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Christoph Hellwig Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Dave Chinner <= Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Dave Chinner Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Alex Elder Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Christoph Hellwig Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Alex Elder Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Alex Elder [PATCH 11/34] xfs: convert xfsbud shrinker to a per-buftarg shrinker., Dave Chinner [PATCH 08/34] xfs: rcu free inodes, Dave Chinner [PATCH 10/34] xfs: convert pag_ici_lock to a spin lock, Dave Chinner [PATCH 12/34] xfs: add a lru to the XFS buffer cache, Dave Chinner [PATCH 09/34] xfs: convert inode cache lookups to use RCU locking, Dave Chinner

Previous by Date:	Re: Issue on kernel 2.6.35.9 vanilla, Dave Chinner
Next by Date:	Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Dave Chinner
Previous by Thread:	Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Christoph Hellwig
Next by Thread:	Re: [PATCH 06/34] xfs: dynamic speculative EOF preallocation, Dave Chinner
Indexes:	[Date] [Thread] [Top] [All Lists]