[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 12 Dec 2006 16:18:32 +0300
From: Dmitriy Monakhov <dmonakhov@...ru>
To: Andrew Morton <akpm@...l.org>
Cc: Dmitriy Monakhov <dmonakhov@...ru>,
Dmitriy Monakhov <dmonakhov@...nvz.org>,
linux-kernel@...r.kernel.org,
Linux Memory Management <linux-mm@...ck.org>,
<devel@...nvz.org>, xfs@....sgi.com
Subject: Re: [PATCH] incorrect error handling inside generic_file_direct_write
Andrew Morton <akpm@...l.org> writes:
> On Tue, 12 Dec 2006 15:20:52 +0300
> Dmitriy Monakhov <dmonakhov@...ru> wrote:
>
>> > XFS (at least) can call generic_file_direct_write() with i_mutex not held.
>> > And vmtruncate() expects i_mutex to be held.
>> >
>> > I guess a suitable solution would be to push this problem back up to the
>> > callers: let them decide whether to run vmtruncate() and if so, to ensure
>> > that i_mutex is held.
>> >
>> > The existence of generic_file_aio_write_nolock() makes that rather messy
>> > though.
>> This means we may call generic_file_aio_write_nolock() without i_mutex, right?
>> but call trace is :
>> generic_file_aio_write_nolock()
>> ->generic_file_buffered_write() /* i_mutex not held here */
>> but according to filemaps locking rules: mm/filemap.c:77
>> ..
>> * ->i_mutex (generic_file_buffered_write)
>> * ->mmap_sem (fault_in_pages_readable->do_page_fault)
>> ..
>> I'm confused a litle bit, where is the truth?
>
> xfs_write() calls generic_file_direct_write() without taking i_mutex for
> O_DIRECT writes.
Yes, but my quastion is about __generic_file_aio_write_nolock().
As i understand _nolock sufix means that i_mutex was already locked
by caller, am i right ?
If yes, than __generic_file_aio_write_nolock() is beter place for vmtrancate()
acclivity after generic_file_direct_write() has fail.
Signed-off-by: Dmitriy Monakhov <dmonakhov@...nvz.org>
-------
View attachment "diff-generic-direct-io-write-fix" of type "text/plain" (588 bytes)
Powered by blists - more mailing lists