Re: [PATCH] blk-cgroup: Use cond_resched() when destroy blkgs

2021-01-27 Thread Baolin Wang
Hi Tejun, Hello, Baolin. On Tue, Jan 26, 2021 at 09:33:25PM +0800, Baolin Wang wrote: On !PREEMPT kernel, we can get below softlockup when doing stress testing with creating and destroying block cgroup repeatly. The reason is it may take a long time to acquire the queue's lock in the loop of

Re: [PATCH] blk-cgroup: Use cond_resched() when destroy blkgs

2021-01-27 Thread Tejun Heo
Hello, Baolin. On Tue, Jan 26, 2021 at 09:33:25PM +0800, Baolin Wang wrote: > On !PREEMPT kernel, we can get below softlockup when doing stress > testing with creating and destroying block cgroup repeatly. The > reason is it may take a long time to acquire the queue's lock in > the loop of

[PATCH] blk-cgroup: Use cond_resched() when destroy blkgs

2021-01-26 Thread Baolin Wang
On !PREEMPT kernel, we can get below softlockup when doing stress testing with creating and destroying block cgroup repeatly. The reason is it may take a long time to acquire the queue's lock in the loop of blkcg_destroy_blkgs(), thus we can use cond_resched() instead of cpu_relax() to avoid this