----- Original Message -----
From: Felipe Contreras <felipe.contreras@xxxxxxxxx>
>>On Fri, Jun 6, 2014 at 4:58 AM, <zhdxzx@xxxxxxxx> wrote:
>> Alternatively can we try wait_iff_congested(zone, BLK_RW_ASYNC, HZ/10) ?
>>
> I see the same problem with that code.
>
The comments around the congestion_wait,
[1]
*
* Once a zone is flagged ZONE_WRITEBACK, kswapd will count the number
* of pages under pages flagged for immediate reclaim and stall if any
* are encountered in the nr_immediate check below.
*/
if (nr_writeback && nr_writeback == nr_taken)
zone_set_flag(zone, ZONE_WRITEBACK);
[2]
/*
* If dirty pages are scanned that are not queued for IO, it
* implies that flushers are not keeping up. In this case, flag
* the zone ZONE_TAIL_LRU_DIRTY and kswapd will start writing
* pages from reclaim context. It will forcibly stall in the
* next check.
*/
if (nr_unqueued_dirty == nr_taken)
zone_set_flag(zone, ZONE_TAIL_LRU_DIRTY);
The "force stall" in [2] conflicts with "start writing pages" in [2], and
conflicts with "nr_immediate check below" in [1] as well, IIUC.
Would you please try again based only on comment [1](based on v3.15-rc8)?
thanks
Hillf
--- a/mm/vmscan.c Sat Jun 7 18:38:08 2014
+++ b/mm/vmscan.c Sat Jun 7 20:08:36 2014
@@ -1566,7 +1566,7 @@ shrink_inactive_list(unsigned long nr_to
* implies that pages are cycling through the LRU faster than
* they are written so also forcibly stall.
*/
- if (nr_unqueued_dirty == nr_taken || nr_immediate)
+ if (nr_immediate)
congestion_wait(BLK_RW_ASYNC, HZ/10);
}
--?韬{.n?壏?檩jg??a旃??迻)钋?韰骅w+h櫒璀?ｙ/i?⒏?婃⒎?湼Щ 烚m?? ?)钋?痂燐^櫒?觥櫒ザ?v夗﨣趓O璁?f?i?⒏?