Re: [PATCH] jbd_commit_transaction() races with journal_try_to_drop_buffers() causing DIO failures

[Badari Pulavarty <pbadari@us.ibm.com>]

On Mon, 2008-05-05 at 19:06 +0200, Jan Kara wrote:
> On Thu 01-05-08 08:16:21, Badari Pulavarty wrote:
> > Hi Andrew & Jan,
> > 
> > I was able to reproduce the customer problem involving DIO
> > (invalidate_inode_pages2) problem by writing simple testcase
> > to keep writing to a file using buffered writes and DIO writes
> > forever in a loop. I see DIO writes fail with -EIO.
> > 
> > After a long debug, found 2 cases how this could happen.
> > These are race conditions with journal_try_to_free_buffers()
> > and journal_commit_transaction().
> > 
> > 1) journal_submit_data_buffers() tries to get bh_state lock. If
> > try lock fails, it drops the j_list_lock and sleeps for
> > bh_state lock, while holding a reference on the buffer.
> > In the meanwhile, journal_try_to_free_buffers() can clean up the
> > journal head could call try_to_free_buffers(). try_to_free_buffers()
> > would fail due to the reference held by journal_submit_data_buffers()
> > - which in turn causes failues for DIO (invalidate_inode_pages2()).
> > 
> > 2) When the buffer is on t_locked_list waiting for IO to finish,
> > we hold a reference and give up the cpu, if we can't get
> > bh_state lock. This causes try_to_free_buffers() to fail.
> > 
> > Fix is to drop the reference on the buffer if we can't get
> > bh_state lock, give up the cpu and re-try the whole operation -
> > instead of waiting for the vh_state lock.
> > 
> > Does this look like a resonable fix ?
>   As Mingming pointed out there are few other places where we could hold
> the bh reference. Note also that we accumulate references to buffers in the
> wbuf[] list and we need that for submit_bh() which consumes one bh
> reference. Generally, it seems to me as a too fragile and impractical
> rule "nobody can hold bh reference when not holding page lock" which is
> basically what it comes down to if you really want to be sure that
> journal_try_to_free_buffers() succeeds. And also note that in principle
> there are other places which hold references to buffers without holding the
> page lock - for example writepage() in ordered mode (although this one is
> in practice hardly triggerable). So how we could fix at least the races
> with commit code is to implement launder_page() callback for ext3/4 which
> would wait for the previous transaction commit in case the page has buffers
> that are part of that commit (I don't want this logic in
> journal_try_to_free_buffers() as that is called also on memory-reclaim
> path, but journal_launder_page() is fine with me). This would be correct
> but could considerably slow down O_DIRECT writes in cases they're mixed
> with buffered writes so I'm not sure if this is acceptable.

Yes. I have been discussing all of these with Mingming. I agree with you
that it looks silly to update all the places to not hold a ref on the
buffer while waiting for spinlocks. 

Adding a launder_page() seems to be the right approach. I wouldn't worry
about making O_DIRECT slower. Currently, its failing with -EIO in this
case anyway. 

Unfortunately, this happens at a customer site and I can't effort to
give them a complete re-write type patch - I need to get this into 
older distro release :( Thats why I am trying to fix the cases.
I am finding more and more of these ..

Thanks,
Badari