Messages in this thread Patch in this message |  | | | Date | Sat, 18 Sep 2010 12:11:58 -0700 | | From | Greg KH <> | | Subject | [020/129] ext4: move aio completion after unwritten extent conversion |
| |
2.6.35-stable review patch. If anyone has any objections, please let us know.
------------------ From: Jiaying Zhang <jiayingz@google.com>
commit 5b3ff237bef43b9e7fb7d1eb858e29b73fd664f9 upstream.
This patch is to be applied upon Christoph's "direct-io: move aio_complete into ->end_io" patch. It adds iocb and result fields to struct ext4_io_end_t, so that we can call aio_complete from ext4_end_io_nolock() after the extent conversion has finished.
I have verified with Christoph's aio-dio test that used to fail after a few runs on an original kernel but now succeeds on the patched kernel.
See http://thread.gmane.org/gmane.comp.file-systems.ext4/19659 for details.
Signed-off-by: Jiaying Zhang <jiayingz@google.com> Signed-off-by: "Theodore Ts'o" <tytso@mit.edu> Cc: Chuck Ebbert <cebbert@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
--- fs/ext4/ext4.h | 4 +++- fs/ext4/inode.c | 17 ++++++++++++----- 2 files changed, 15 insertions(+), 6 deletions(-) --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -167,13 +167,15 @@ struct mpage_da_data { }; #define EXT4_IO_UNWRITTEN 0x1 typedef struct ext4_io_end { - struct list_head list; /* per-file finished AIO list */ + struct list_head list; /* per-file finished IO list */ struct inode *inode; /* file being written to */ unsigned int flag; /* unwritten or not */ struct page *page; /* page struct for buffer write */ loff_t offset; /* offset in the file */ ssize_t size; /* size of the extent */ struct work_struct work; /* data work queue */ + struct kiocb *iocb; /* iocb struct for AIO */ + int result; /* error value for AIO */ } ext4_io_end_t; /* --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -3668,6 +3668,8 @@ static int ext4_end_io_nolock(ext4_io_en return ret; } + if (io->iocb) + aio_complete(io->iocb, io->result, 0); /* clear the DIO AIO unwritten flag */ io->flag = 0; return ret; @@ -3767,6 +3769,8 @@ static ext4_io_end_t *ext4_init_io_end ( io->offset = 0; io->size = 0; io->page = NULL; + io->iocb = NULL; + io->result = 0; INIT_WORK(&io->work, ext4_end_io_work); INIT_LIST_HEAD(&io->list); } @@ -3796,12 +3800,18 @@ static void ext4_end_io_dio(struct kiocb if (io_end->flag != EXT4_IO_UNWRITTEN){ ext4_free_io_end(io_end); iocb->private = NULL; - goto out; +out: + if (is_async) + aio_complete(iocb, ret, 0); + return; } io_end->offset = offset; io_end->size = size; - io_end->flag = EXT4_IO_UNWRITTEN; + if (is_async) { + io_end->iocb = iocb; + io_end->result = ret; + } wq = EXT4_SB(io_end->inode->i_sb)->dio_unwritten_wq; /* queue the work to convert unwritten extents to written */ @@ -3813,9 +3823,6 @@ static void ext4_end_io_dio(struct kiocb list_add_tail(&io_end->list, &ei->i_completed_io_list); spin_unlock_irqrestore(&ei->i_completed_io_lock, flags); iocb->private = NULL; -out: - if (is_async) - aio_complete(iocb, ret, 0); } static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate)
|  |