| commit | 1d727bac5ba47d4ea407a72ede506d0a452755ae | |
| author | Matthew Dillon <dillon@dragonflybsd.org> | |
| Tue, 23 Sep 2008 17:43:41 +0000 (23 17:43 +0000) | ||
| committer | Matthew Dillon <dillon@dragonflybsd.org> | |
| Tue, 23 Sep 2008 17:43:41 +0000 (23 17:43 +0000) | ||
| tree | 81fa4a73bc4666bbb3bf141967b3ae83903ee88a | treesnapshot (tar.gz zip) |
| parent | 553d42f9f3cbeceeac9185e347a7ea7eed5b7f21 | commitdiff |
Make two more changes to the ata request queue sorting code.
* Do not try to sort write requests. There's no point, the hard drive
has a write cache and a much better understanding of the media topology
and will sort the requests for us.
* Move the freeze point every 16 requests instead of every 128 to further
reduce the amount of reordering we allow to prevent older requests from
being delayed indefinitely by newer requests.
Note: the original ata code didn't code any limits at all, causing insanely
long latencies during heavy disk I/O for no good reason.
Reported-by: Hasso Tepper <hasso@estpak.ee>
* Do not try to sort write requests. There's no point, the hard drive
has a write cache and a much better understanding of the media topology
and will sort the requests for us.
* Move the freeze point every 16 requests instead of every 128 to further
reduce the amount of reordering we allow to prevent older requests from
being delayed indefinitely by newer requests.
Note: the original ata code didn't code any limits at all, causing insanely
long latencies during heavy disk I/O for no good reason.
Reported-by: Hasso Tepper <hasso@estpak.ee>
| sys/dev/disk/nata/ata-queue.c | diffblobblamehistory |