Hi,
I got a problem report with fio where larger block size random reads
where markedly slower with buffered IO than with O_DIRECT, and the
initial thought was that perhaps this was some fio oddity. The reporter
eventually discovered that turning off the fadvise hint made it work
fine. So I took a look, and it seems we never do readahead for
FMODE_RANDOM even if the request size is larger than 1 page. That seems
like a bug, if an application is doing eg 16kb random reads, you want to
readahead the 12kb remaining data. On devices where smaller transfer
sizes are slower than larger ones, this can make a large difference.
This patch makes us readahead even for FMODE_RANDOM, iff we'll be
reading more pages in that single read. I ran a quick test here, and it
appears to fix the problem (no difference with fadvise POSIX_FADV_RANDOM
being passed in or not).
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
diff --git a/mm/readahead.c b/mm/readahead.c
index 337b20e..d4b201c 100644
--- a/mm/readahead.c
+++ b/mm/readahead.c
@@ -501,8 +501,11 @@ void page_cache_sync_readahead(struct address_space *mapping,
if (!ra->ra_pages)
return;
- /* be dumb */
- if (filp->f_mode & FMODE_RANDOM) {
+ /*
+ * Be dumb for files marked as randomly accessed, but do readahead
+ * inside the original request (req_size > 1).
+ */
+ if ((filp->f_mode & FMODE_RANDOM) && req_size == 1) {
force_page_cache_readahead(mapping, filp, offset, req_size);
return;
}
--
Jens Axboe
--
