[V2]fix ocfs2 aio/dio writing process hang

[V2]fix ocfs2 aio/dio writing process hang

[Junxiao Bi]

V2 changes:
- update the patch header of the first patch to make it more clear.


This patch list fixes an issue about ocfs2 aio/dio write process hang.
The call trace is like this:
@  cat /proc/15913/stack
@ [<ffffffffa06e1085>] ocfs2_aiodio_wait+0x85/0xc0 [ocfs2]
@ [<ffffffffa06e3e39>] ocfs2_file_aio_write+0x99/0xa0 [ocfs2]
@ [<ffffffff811ace97>] aio_rw_vect_retry+0x87/0x150
@ [<ffffffff811aebd1>] aio_run_iocb+0x71/0x170
@ [<ffffffff811af59b>] io_submit_one+0x1ab/0x280
@ [<ffffffff811af77c>] do_io_submit+0x10c/0x1c0
@ [<ffffffff811af840>] sys_io_submit+0x10/0x20
@ [<ffffffff81509b42>] system_call_fastpath+0x16/0x1b
@ [<ffffffffffffffff>] 0xffffffffffffffff

[V2 PATCH 1/2] aio: make kiocb->private NUll in init_sync_kiocb()

[Junxiao Bi]

Ocfs2 uses kiocb.*private as a flag of unsigned long size. In
commit a11f7e6 ocfs2: serialize unaligned aio, the unaligned
io flag is involved in it to serialize the unaligned aio. As
*private is not initialized in init_sync_kiocb() of do_sync_write(),
this unaligned io flag may be unexpectly set in an aligned dio.
And this will cause OCFS2_I(inode)->ip_unaligned_aio decreased
to -1 in ocfs2_dio_end_io(), thus the following unaligned dio
will hang forever at ocfs2_aiodio_wait() in ocfs2_file_aio_write().

Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>
---
 include/linux/aio.h |    1 +
 1 file changed, 1 insertion(+)

diff --git a/include/linux/aio.h b/include/linux/aio.h
index 2314ad8..b1a520e 100644
--- a/include/linux/aio.h
+++ b/include/linux/aio.h
@@ -140,6 +140,7 @@ struct kiocb {
 		(x)->ki_dtor = NULL;			\
 		(x)->ki_obj.tsk = tsk;			\
 		(x)->ki_user_data = 0;                  \
+		(x)->private = NULL;			\
 	} while (0)
 
 #define AIO_RING_MAGIC			0xa10a10a1
-- 
1.7.9.5

Re: [V2 PATCH 1/2] aio: make kiocb->private NUll in init_sync_kiocb()

[Greg KH]

On Tue, Jun 05, 2012 at 03:26:49PM +0800, Junxiao Bi wrote:
> Ocfs2 uses kiocb.*private as a flag of unsigned long size. In
> commit a11f7e6 ocfs2: serialize unaligned aio, the unaligned
> io flag is involved in it to serialize the unaligned aio. As
> *private is not initialized in init_sync_kiocb() of do_sync_write(),
> this unaligned io flag may be unexpectly set in an aligned dio.
> And this will cause OCFS2_I(inode)->ip_unaligned_aio decreased
> to -1 in ocfs2_dio_end_io(), thus the following unaligned dio
> will hang forever at ocfs2_aiodio_wait() in ocfs2_file_aio_write().
> 
> Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>
> ---
>  include/linux/aio.h |    1 +
>  1 file changed, 1 insertion(+)

<formletter>

This is not the correct way to submit patches for inclusion in the
stable kernel tree.  Please read Documentation/stable_kernel_rules.txt
for how to do this properly.

</formletter>

Re: [V2 PATCH 1/2] aio: make kiocb->private NUll in init_sync_kiocb()

[Jeff Moyer]

Junxiao Bi <junxiao.bi@oracle.com> writes:

> Ocfs2 uses kiocb.*private as a flag of unsigned long size. In
> commit a11f7e6 ocfs2: serialize unaligned aio, the unaligned
> io flag is involved in it to serialize the unaligned aio. As
> *private is not initialized in init_sync_kiocb() of do_sync_write(),
> this unaligned io flag may be unexpectly set in an aligned dio.
> And this will cause OCFS2_I(inode)->ip_unaligned_aio decreased
> to -1 in ocfs2_dio_end_io(), thus the following unaligned dio
> will hang forever at ocfs2_aiodio_wait() in ocfs2_file_aio_write().
>
> Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>

Acked-by: Jeff Moyer <jmoyer@redhat.com>

[V2 PATCH 2/2] ocfs2: clear unaligned io flag when dio fails

[Junxiao Bi]

The unaligned io flag is set in the kiocb when an unaligned
dio is issued, it should be cleared even when the dio fails,
or it may affect the following io which are using the same
kiocb.

Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>
---
 fs/ocfs2/file.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 061591a..98513c8 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -2422,8 +2422,10 @@ out_dio:
 		unaligned_dio = 0;
 	}
 
-	if (unaligned_dio)
+	if (unaligned_dio) {
+		ocfs2_iocb_clear_unaligned_aio(iocb);
 		atomic_dec(&OCFS2_I(inode)->ip_unaligned_aio);
+	}
 
 out:
 	if (rw_level != -1)
-- 
1.7.9.5

Re: [V2 PATCH 2/2] ocfs2: clear unaligned io flag when dio fails

[Greg KH]

On Tue, Jun 05, 2012 at 03:26:50PM +0800, Junxiao Bi wrote:
> The unaligned io flag is set in the kiocb when an unaligned
> dio is issued, it should be cleared even when the dio fails,
> or it may affect the following io which are using the same
> kiocb.
> 
> Signed-off-by: Junxiao Bi <junxiao.bi@oracle.com>
> ---
>  fs/ocfs2/file.c |    4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 

<formletter>

This is not the correct way to submit patches for inclusion in the
stable kernel tree.  Please read Documentation/stable_kernel_rules.txt
for how to do this properly.

</formletter>