Eric Lee / linux-smarc-t335x-v3.2

Download as
Browse Code ยป

Commit 55c022bbddb2c056b5dff1bd1b1758d31b6d64c9

Authored by Shaohua Li
Committed by Jens Axboe
1 parent 719c0c5906
Exists in master and in 4 other branches v3.2_AM335xPSP_04.06.00.11, v3.2_NEWT335xPSP_04.06.00.11, v3.2_SBC_SMARTMEN, v3.2_SMARCT335xPSP_04.06.00.11

block: avoid building too big plug list 

When I test fio script with big I/O depth, I found the total throughput drops
compared to some relative small I/O depth. The reason is the thread accumulates
big requests in its plug list and causes some delays (surely this depends
on CPU speed).
I thought we'd better have a threshold for requests. When a threshold reaches,
this means there is no request merge and queue lock contention isn't severe
when pushing per-task requests to queue, so the main advantages of blk plug
don't exist. We can force a plug list flush in this case.
With this, my test throughput actually increases and almost equals to small
I/O depth. Another side effect is irq off time decreases in blk_flush_plug_list()
for big I/O depth.
The BLK_MAX_REQUEST_COUNT is choosen arbitarily, but 16 is efficiently to
reduce lock contention to me. But I'm open here, 32 is ok in my test too.

Signed-off-by: Shaohua Li <shaohua.li@intel.com>
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>

Showing 2 changed files with 8 additions and 0 deletions Side-by-side Diff

... ... @@ -1302,7 +1302,10 @@
1302 1302 plug->should_sort = 1;
1303 1303 }
1304 1304 list_add_tail(&req->queuelist, &plug->list);
  1305 + plug->count++;
1305 1306 drive_stat_acct(req, 1);
  1307 + if (plug->count >= BLK_MAX_REQUEST_COUNT)
  1308 + blk_flush_plug_list(plug, false);
1306 1309 } else {
1307 1310 spin_lock_irq(q->queue_lock);
1308 1311 add_acct_request(q, req, where);
... ... @@ -2626,6 +2629,7 @@
2626 2629 INIT_LIST_HEAD(&plug->list);
2627 2630 INIT_LIST_HEAD(&plug->cb_list);
2628 2631 plug->should_sort = 0;
  2632 + plug->count = 0;
2629 2633  
2630 2634 /*
2631 2635 * If this is a nested plug, don't actually assign it. It will be
... ... @@ -2709,6 +2713,7 @@
2709 2713 return;
2710 2714  
2711 2715 list_splice_init(&plug->list, &list);
  2716 + plug->count = 0;
2712 2717  
2713 2718 if (plug->should_sort) {
2714 2719 list_sort(NULL, &list, plug_rq_cmp);
include/linux/blkdev.h
Diff comments   View file @ 55c022b
... ... @@ -862,7 +862,10 @@
862 862 struct list_head list;
863 863 struct list_head cb_list;
864 864 unsigned int should_sort;
  865 + unsigned int count;
865 866 };
  867 +#define BLK_MAX_REQUEST_COUNT 16
  868 +
866 869 struct blk_plug_cb {
867 870 struct list_head list;
868 871 void (*callback)(struct blk_plug_cb *);