Re: 2.6.0-test6-mm4 - oops in __aio_run_iocbs()

[Daniel McNeil]

On Thu, 2003-10-09 at 05:59, Suparna Bhattacharya wrote:
> On Thu, Oct 09, 2003 at 04:46:24PM +0530, Suparna Bhattacharya wrote:
> > On Wed, Oct 08, 2003 at 04:18:15PM -0700, Daniel McNeil wrote:
> > > I'm been testing AIO on test6-mm4 using a ext3 file system and
> > > copying a 88MB file to an already existing preallocated file of 88MB.
> > > I been using my aiocp program to copy the file using i/o sizes of
> > > 1k to 512k and outstanding aio requests of between 1 and 64 using
> > > O_DIRECT, O_SYNC and O_DIRECT & O_SYNC.  Everything works as long
> > > as the file is pre-allocated.  When copying the file to a new file
> > > (O_CREAT|O_DIRECT), I get the following oops:
> > 
> > What are the i/o sizes and block sizes for which you get the oops ?
> > Is this only for large i/o sizes ?
> > 
> > __aio_run_iocbs should have been called only for buffered i/o, 
> > so this sounds like an O_DIRECT fallback to buffered i/o.
> > Possibly after already submitting some blocks direct to BIO,
> > the i/o completion path for which ends up calling aio_complete
> > releasing the iocb. That could explain the use-after-free situation
> > you see.
> > 
> > But, O_DIRECT write should fallback to buffered i/o only if it 
> > encounters holes in the middle of the file, not for simple appends 
> > as in your case. Need to figure out how this could have happened ...
> 
> Took a quick look at aiocp.c - wondering if its possible that
> some of the later read requests complete earlier and trigger
> a write to higher offset first. Resulting in the file being
> extended with holes in between - holes which get overwritten
> at a later point as the earlier read requests complete.

That is an interesting idea.  I can change my aiocp program to
do a printf if it hits this situation.

> 
> Though I don't yet see how a situation could arise in the 
> single threaded case where part of the request gets submitted 
> direct to BIO and the rest falls back to buffered-io ... Need 
> to think about it a bit more. 
> Are your writes all block aligned ?

All i/o is aligned to the size of the block size (128k in this
test).

Daniel

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/