Re: [PATCH 01/15] writeback: introduce .tagged_writepages for the WB_SYNC_NONE sync stage

[Wu Fengguang <fengguang.wu@intel.com>]

On Wed, Jun 08, 2011 at 07:02:34AM +0800, Andrew Morton wrote:
> On Wed, 08 Jun 2011 05:32:37 +0800
> Wu Fengguang <fengguang.wu@intel.com> wrote:
> 
> > sync(2) is performed in two stages: the WB_SYNC_NONE sync and the
> > WB_SYNC_ALL sync. Identify the first stage with .tagged_writepages and
> > do livelock prevention for it, too.
> > 
> > Note that writeback_inodes_sb() is called by not only sync(), they are
> > treated the same because the other callers also need livelock prevention.
> > 
> > Impact:  It changes the order in which pages/inodes are synced to disk.
> > Now in the WB_SYNC_NONE stage, it won't proceed to write the next inode
> > until finished with the current inode.
> 
> What problem is this patch actually fixing?  It sounds like there's
> some livelock scenario in the WB_SYNC_NONE phase.  otoh the final
> paragraph implies that the WB_SYNC_NONE phase is failing to write some
> pages under some situations.

Problem is: the WB_SYNC_NONE phase has no livelock prevention _at all_.

Jan's commit f446daaea9 ("mm: implement writeback livelock avoidance
using page tagging") is a partial fix in that it only fixed the
WB_SYNC_ALL phase livelock.

Although ext4 is tested to no longer livelock with commit f446daaea9,
it may due to some "redirty_tail() after pages_skipped" effect which
is by no means a guarantee for _all_ the file systems.

> Suggest that the changelog be fleshed out to cover all of this.

OK, I'll add the above two paragraphs to the changelog.

Thanks,
Fengguang