Messages in this thread Patch in this message | | | Date | Thu, 23 Oct 2008 18:15:03 +0200 | From | Miquel van Smoorenburg <> | Subject | [PATCH] do_mpage_readpage: don't submit lots of small bios on boundary |
| |
(not really sure who to put in the Cc list here)
While tracing I/O patterns with blktrace (a great tool) a few weeks ago I identified a minor issue in fs/mpage.c
[PATCH] do_mpage_readpage: don't submit lots of small bios on boundary
As the comment above mpage_readpages() says, a fs's get_block function will set BH_Boundary when it maps a block just before a block for which extra I/O is required.
Since get_block() can map a range of pages, for all these pages the BH_Boundary flag will be set. But we only need to push what I/O we have accumulated at the last block of this range.
This makes do_mpage_readpage() send out the largest possible bio instead of a bunch of page-sized ones in the BH_Boundary case.
Signed-off-by: Miquel van Smoorenburg <mikevs@xs4all.net>
diff -ruN linux-2.6.26.5.orig/fs/mpage.c linux-2.6.26.5/fs/mpage.c --- linux-2.6.26.5.orig/fs/mpage.c 2008-09-08 19:40:20.000000000 +0200 +++ linux-2.6.26.5/fs/mpage.c 2008-10-23 00:32:11.000000000 +0200 @@ -307,7 +307,10 @@ goto alloc_new; } - if (buffer_boundary(map_bh) || (first_hole != blocks_per_page)) + relative_block = block_in_file - *first_logical_block; + nblocks = map_bh->b_size >> blkbits; + if ((buffer_boundary(map_bh) && relative_block == nblocks) || + (first_hole != blocks_per_page)) bio = mpage_bio_submit(READ, bio); else *last_block_in_bio = blocks[blocks_per_page - 1];
| |