direct-io: Submit bio after boundary buffer is added to it
Jan Kara
reviewerPATCH 1/1 direct-io: Submit bio after boundary buffer is added to it
Currently, dio_send_cur_page() submits bio before current page and cached sdio->cur_page is added to the bio if sdio->boundary is set. This is actually wrong because sdio->boundary means the current buffer is the last one before metadata needs to be read. So we should rather submit the bio after the current page is added to it.
--- fs/direct-io.c | 28 +++++++++++----------------- 1 files changed, 11 insertions(+), 17 deletions(-)
Andrew, can you add this patch the previous direct IO boundary handling fix?
fs/direct-io.c +11 -17
@@ -672,12 +672,6 @@ static inline int dio_send_cur_page(struct dio *dio, struct dio_submit *sdio,if (sdio->final_block_in_bio != sdio->cur_page_block || cur_offset != bio_next_offset)dio_bio_submit (dio, sdio); -/* -* Submit now if the underlying fs is about to perform a -* metadata read -*/ -else if (sdio->boundary) -dio_bio_submit (dio, sdio); }if (sdio->bio ==NULL ) {
@@ -737,16 +731,6 @@ submit_page_section(struct dio *dio, struct dio_submit *sdio, struct page *page, sdio->cur_page_block + (sdio->cur_page_len >> sdio->blkbits) == blocknr) { sdio->cur_page_len += len; - -/* -* If sdio->boundary then we want to schedule the IO now to -* avoid metadata seeks. -*/ -if (sdio->boundary) { - ret =dio_send_cur_page (dio, sdio, map_bh); -page_cache_release (sdio->cur_page); - sdio->cur_page =NULL ; - } goto out; }
@@ -758,7 +742,7 @@ submit_page_section(struct dio *dio, struct dio_submit *sdio, struct page *page,page_cache_release (sdio->cur_page); sdio->cur_page =NULL ;if (ret) - goto out; +return ret; }page_cache_get (page);/* It is in dio */
@@ -768,6 +752,16 @@ submit_page_section(struct dio *dio, struct dio_submit *sdio, struct page *page, sdio->cur_page_block = blocknr; sdio->cur_page_fs_offset = sdio->block_in_file << sdio->blkbits; out: +/* +* If sdio->boundary then we want to schedule the IO now to +* avoid metadata seeks. +*/ +if (sdio->boundary) { + ret =dio_send_cur_page (dio, sdio, map_bh); +dio_bio_submit (dio, sdio); +page_cache_release (sdio->cur_page); + sdio->cur_page =NULL ; + }return ret; }
Signature · 1 line
1.7.1
Quoted 6 lines · On Tue, 9 Apr 2013 17:42:49 +0200 Jan Kara <[email protected]> wrote:
Currently,
dio_send_cur_page()submits bio before current page and cachedsdio->cur_pageis added to the bio ifsdio->boundaryis set. This is actually wrong becausesdio->boundarymeans the current buffer is the last one before metadata needs to be read. So we should rather submit the bio after the current page is added to it.
Quoted 7 lines
Reported-by Kazuya Mio <[email protected]>Signed-off-by Jan Kara <[email protected]>--- fs/direct-io.c | 28 +++++++++++----------------- 1 files changed, 11 insertions(+), 17 deletions(-)Andrew, can you add this patch the previous direct IO boundary handling fix?
What is the user-visible impact of these two patches?
Jan Kara
reviewerQuoted 14 lines · Andrew Morton
Quoted 12 lines · On Tue, 9 Apr 2013 17:42:49 +0200 Jan Kara <[email protected]> wrote:
Currently,
dio_send_cur_page()submits bio before current page and cachedsdio->cur_pageis added to the bio ifsdio->boundaryis set. This is actually wrong becausesdio->boundarymeans the current buffer is the last one before metadata needs to be read. So we should rather submit the bio after the current page is added to it.Reported-by Kazuya Mio <[email protected]>Signed-off-by Jan Kara <[email protected]>--- fs/direct-io.c | 28 +++++++++++----------------- 1 files changed, 11 insertions(+), 17 deletions(-)
Quoted 3 lines
Quoted 1 line
Andrew, can you add this patch the previous direct IO boundary handling fix?
What is the user-visible impact of these two patches?
I've measured about 10% throughput improvement of direct IO reads on ext3
with SATA harddrive (from 90 MB/s to 100 MB/s) with these patches applied. With ramdisk, the improvement was about 3-fold (from 350 MB/s to 1.2 GB/s). For other filesystems (such as ext4), the improvements won't be as visible because the frequency of BH_Boundary flag being set is much smaller...
Honza
Signature · 2 lines
Jan Kara <[email protected]> SUSE Labs, CR