mm/cgroup/reclaim: Fix dirty pages throttling on cgroup v1
Subsystems (4)
PATCH 1/1 mm/cgroup/reclaim: Fix dirty pages throttling on cgroup v1
balance_dirty_pages doesn't do the required dirty throttling on cgroupv1. See commit 9badce000e2c ("cgroup, writeback: don't enable cgroup writeback on traditional hierarchies"). Instead, the kernel depends on writeback throttling in shrink_folio_list to achieve the same goal. With large memory systems, the flusher may not be able to writeback quickly enough such that we will start finding pages in the shrink_folio_list already in writeback. Hence for cgroupv1 let's do a reclaim throttle after waking up the flusher.
The below test which used to fail on a 256GB system completes till the the file system is full with this change.
root@lp2:/sys/fs/cgroup/memory# mkdir test root@lp2:/sys/fs/cgroup/memory# cd test/ root@lp2:/sys/fs/cgroup/memory/test# echo 120M > memory.limit_in_bytes root@lp2:/sys/fs/cgroup/memory/test# echo $$ > tasks root@lp2:/sys/fs/cgroup/memory/test# dd if=/dev/zero of=/home/kvaneesh/test bs=1M Killed
--- mm/vmscan.c | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-)
mm/vmscan.c +13 -1
@@ -2514,8 +2514,20 @@ static unsigned long shrink_inactive_list(unsigned long nr_to_scan,* the flushers simply cannot keep up with the allocation * rate. Nudge the flusher threads in case they are asleep. */ -if (stat.nr_unqueued_dirty == nr_taken) +if (stat.nr_unqueued_dirty == nr_taken) {wakeup_flusher_threads (WB_REASON_VMSCAN ); +/* +* For cgroupv1 dirty throttling is achieved by waking up +* the kernel flusher here and later waiting on folios +* which are in writeback to finish (see shrink_folio_list()). +* +* Flusher may not be able to issue writeback quickly +* enough for cgroupv1 writeback throttling to work +* on a large system. +*/ +if (!writeback_throttling_sane (sc)) +reclaim_throttle (pgdat,VMSCAN_THROTTLE_WRITEBACK ); + } sc->nr.dirty += stat.nr_dirty; sc->nr.congested += stat.nr_congested;
Signature · 1 line
2.38.1
Quoted 18 lines · Aneesh Kumar K.V
balance_dirty_pagesdoesn't do the required dirty throttling on cgroupv1. See commit9badce000e2c("cgroup, writeback: don't enable cgroup writeback on traditional hierarchies"). Instead, the kernel depends on writeback throttling inshrink_folio_listto achieve the same goal. With large memory systems, the flusher may not be able to writeback quickly enough such that we will start finding pages in theshrink_folio_listalready in writeback. Hence for cgroupv1 let's do a reclaim throttle after waking up the flusher.The below test which used to fail on a 256GB system completes till the the file system is full with this change.
root@lp2:/sys/fs/cgroup/memory# mkdir test root@lp2:/sys/fs/cgroup/memory# cd test/ root@lp2:/sys/fs/cgroup/memory/test# echo 120M > memory.limit_in_bytes root@lp2:/sys/fs/cgroup/memory/test# echo $$ > tasks root@lp2:/sys/fs/cgroup/memory/test# dd if=/dev/zero of=/home/kvaneesh/test bs=1M Killed
Quoted 3 lines
Cc <[email protected]>Suggested-by Johannes Weiner <[email protected]>Signed-off-by Aneesh Kumar K.V <[email protected]>
Thanks Aneesh