Re: [PATCH 1/2] HACK: ext3: mount fast even when recovering

[Eric Sandeen <sandeen@redhat.com>]

Andrew Morton wrote:
> On Tue, 14 Jul 2009 17:05:54 +0300
> Adrian Hunter <adrian.hunter@nokia.com> wrote:
> 
>> Speed up ext3 recovery mount time by not sync'ing the
>> block device.  Instead place all dirty buffers into the
>> I/O queue and add a write barrier.  This ensures that
>> no subsequent write will reach the disk before all the
>> recovery writes, but that we do not have to wait for the
>> I/O.
>>
>> Note that ext3 reads sectors the correct way: through the
>> buffer cache, so there is no risk of reading old metadata.
> 
> hm.  The change seems reasonable to me.  afaict it leaves no timing
> windows during which another crash could muck things up.
> 
> As long as those write barriers actually work.  Do they?  For all
> conceivable devices and IO schedulers?

Good point .... for many devices the barriers will fail, but by then I
think this code has already moved on, right?  (And some devices will
lie, but at that point, oh well).

You could do a test barrier IO at the start, and keep the old behavior
if it fails, perhaps?

(whoa, can barriers make something faster?  who woulda thunk it)

-Eric