Hello.
As I remember Tejun asked me to change the comment to this commit and
to clarify
things little bit that WRITE_SYNC is not related to integrity write at
all, the key is that
a caller waits for completion and block layer or block device should
prioritize the IO.
>From the original comment it is not evident, so I sent v2 of this
patch with the following
comment (I hope it clarifies things):
---
When data integrity operation happens (sync, fsync, fdatasync calls)
writeback control is set to WB_SYNC_ALL. In that case all write
requests are marked with WRITE_SYNC (WRITE | REQ_SYNC | REQ_NOIDLE)
indicating that caller is waiting for completion and block layer or
block device should prioritize the IO avoiding any possible delays.
But mpage writeback path ignores marking requests as WRITE_SYNC.
This patch fixes this.
---
If you are going to take this patch could you please take the v2 comment also?
--
Roman
On Tue, Dec 16, 2014 at 8:03 AM, <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
> From: Roman Pen <r.peniaev@xxxxxxxxx>
> Subject: fs/mpage.c: forgotten WRITE_SYNC in case of data integrity write
>
> In case of wbc->sync_mode == WB_SYNC_ALL we need to do data integrity
> write, thus mark request as WRITE_SYNC.
>
> akpm: afaict this change will cause the data integrity write bios to be
> placed onto the second queue in cfq_io_cq.cfqq[], which presumably results
> in special treatment. The documentation for REQ_SYNC is horrid.
>
> Signed-off-by: Roman Pen <r.peniaev@xxxxxxxxx>
> Reviewed-by: Jan Kara <jack@xxxxxxx>
> Cc: Jens Axboe <axboe@xxxxxxxxx>
> Cc: Tejun Heo <tj@xxxxxxxxxx>
> Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> ---
>
> fs/mpage.c | 23 +++++++++++++++--------
> 1 file changed, 15 insertions(+), 8 deletions(-)
>
> diff -puN fs/mpage.c~fs-mpagec-forgotten-write_sync-in-case-of-data-integrity-write fs/mpage.c
> --- a/fs/mpage.c~fs-mpagec-forgotten-write_sync-in-case-of-data-integrity-write
> +++ a/fs/mpage.c
> @@ -482,6 +482,7 @@ static int __mpage_writepage(struct page
> struct buffer_head map_bh;
> loff_t i_size = i_size_read(inode);
> int ret = 0;
> + int wr = (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE);
>
> if (page_has_buffers(page)) {
> struct buffer_head *head = page_buffers(page);
> @@ -590,7 +591,7 @@ page_is_mapped:
> * This page will go to BIO. Do we need to send this BIO off first?
> */
> if (bio && mpd->last_block_in_bio != blocks[0] - 1)
> - bio = mpage_bio_submit(WRITE, bio);
> + bio = mpage_bio_submit(wr, bio);
>
> alloc_new:
> if (bio == NULL) {
> @@ -614,7 +615,7 @@ alloc_new:
> */
> length = first_unmapped << blkbits;
> if (bio_add_page(bio, page, length, 0) < length) {
> - bio = mpage_bio_submit(WRITE, bio);
> + bio = mpage_bio_submit(wr, bio);
> goto alloc_new;
> }
>
> @@ -624,7 +625,7 @@ alloc_new:
> set_page_writeback(page);
> unlock_page(page);
> if (boundary || (first_unmapped != blocks_per_page)) {
> - bio = mpage_bio_submit(WRITE, bio);
> + bio = mpage_bio_submit(wr, bio);
> if (boundary_block) {
> write_boundary_block(boundary_bdev,
> boundary_block, 1 << blkbits);
> @@ -636,7 +637,7 @@ alloc_new:
>
> confused:
> if (bio)
> - bio = mpage_bio_submit(WRITE, bio);
> + bio = mpage_bio_submit(wr, bio);
>
> if (mpd->use_writepage) {
> ret = mapping->a_ops->writepage(page, wbc);
> @@ -692,8 +693,11 @@ mpage_writepages(struct address_space *m
> };
>
> ret = write_cache_pages(mapping, wbc, __mpage_writepage, &mpd);
> - if (mpd.bio)
> - mpage_bio_submit(WRITE, mpd.bio);
> + if (mpd.bio) {
> + int wr = (wbc->sync_mode == WB_SYNC_ALL ?
> + WRITE_SYNC : WRITE);
> + mpage_bio_submit(wr, mpd.bio);
> + }
> }
> blk_finish_plug(&plug);
> return ret;
> @@ -710,8 +714,11 @@ int mpage_writepage(struct page *page, g
> .use_writepage = 0,
> };
> int ret = __mpage_writepage(page, wbc, &mpd);
> - if (mpd.bio)
> - mpage_bio_submit(WRITE, mpd.bio);
> + if (mpd.bio) {
> + int wr = (wbc->sync_mode == WB_SYNC_ALL ?
> + WRITE_SYNC : WRITE);
> + mpage_bio_submit(wr, mpd.bio);
> + }
> return ret;
> }
> EXPORT_SYMBOL(mpage_writepage);
> _
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@xxxxxxxxx. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>