[Top] [All Lists]

Re: [PATCH 0/4] writeback: tracing and wbc->nr_to_write fixes

To: Richard Kennedy <richard@xxxxxxxxxxxxxxx>
Subject: Re: [PATCH 0/4] writeback: tracing and wbc->nr_to_write fixes
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Wed, 21 Apr 2010 09:29:04 +1000
Cc: linux-fsdevel@xxxxxxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, xfs@xxxxxxxxxxx
In-reply-to: <4BCD97C8.6090406@xxxxxxxxxxxxxxx>
References: <1271731314-5893-1-git-send-email-david@xxxxxxxxxxxxx> <4BCD97C8.6090406@xxxxxxxxxxxxxxx>
User-agent: Mutt/1.5.20 (2009-06-14)
On Tue, Apr 20, 2010 at 01:02:16PM +0100, Richard Kennedy wrote:
> On 20/04/10 03:41, Dave Chinner wrote:
> > This series contains the initial writeback tracing patches from
> > Jens, as well as the extensions I added to provide visibility into
> > writeback control structures as the are used by the writeback code.
> > The visibility given is sufficient to understand what is happening
> > in the writeback path - what path is writing data, what path is
> > blocking on congestion, etc, and to determine the differences in
> > behaviour for different sync modes and calling contexts. This
> > tracing really needs to be integrated into mainline so that anyone
> > can improve the tracing as they use it to track down problems
> > in our convoluted writeback paths.
> > 
> > The remaining patches are fixes to problems that the new tracing
> > highlighted.
> Hi Dave,
> Thanks for adding tracing to this, it will be really useful.
> The fix to write_cache_pages looks really interesting, I'm going to test
> it on my machine. Maybe it should be a separate patch to get more
> visibility?

I don't see a big need to separate the series at this point. Once
there's been a review and testing we can decide how to push them
into mainline. IMO, the tracing is just as important as the bug

> Ext4 also multiplies nr_to_write, so will that need fixing too?

No idea. I don't claim to understand ext4's convoluted delayed
allocation path and all it's constraints, so I guess you'd need to
ask the ext4 developers about that one. After all, with the tracing
they'd be able to see if there is a problem. ;)


Dave Chinner

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