On 08/24/2017 04:33 PM, Dietmar Eggemann wrote:
Hi Joonwoo,
On 08/25/2017 12:06 AM, Joonwoo Park wrote:
On 08/24/2017 10:55 AM, Dietmar Eggemann wrote:
Hi Leo,
On 08/23/2017 10:00 AM, Leo Yan wrote:
Hi Leo, Can you please try this patch and let me know if it helps? There seems indeed missing change in google's tree I'm on since it's reverted. The tree Dietmar gave never had this change. You will still need ERRATUM 858921 to avoid panic(curr_run_sum < 0).
Hi Dietmar, do you have https://android-review.googlesource.com/#/c/kernel/common/+/426442/? If not, the bug that 426442 aims to fix can mask the problem so that it will be really hard to hit the cr_avg < 0.
Ah OK. No, my tree doesn't have change 426442 'sched: WALT: fix broken cumulative runnable average accounting'.
e107985-lin:/opt/git/msm_build/kernel:__hikey-linaro/android-hikey-linaro-4.4$ glg2 kernel/sched/walt.c d368c6faa19b - sched: walt: fix window misalignment when HZ=300 (2017-07-12 Joonwoo Park) e487a24793bb - sched/walt: kill {min,max}_capacity (2016-12-08 Juri Lelli) cf43809d7aa0 - sched/walt: Drop arch-specific timer access (2016-09-20 Chris Redpath) 1c81f45725d2 - sched/walt: include missing header for arm_timer_read_counter() (2016-09-09 Amit Pundir) 23ed57dbcc14 - arch_timer: add error handling when the MPM global timer is cleared (2016-08-11 Chris Redpath) 519c62750eb6 - sched/walt: Accounting for number of irqs pending on each core (2016-08-11 Srinath Sridharan) efb86bd08a2e - sched: Introduce Window Assisted Load Tracking (WALT) (2016-08-11 Srivatsa Vaddagiri)
So you're saying that I will need change 426442 and then the patch below so that this BUG_ON((s64)rq->cumulative_runnable_avg < 0) doesn't fire?
Hi Dietmar,
Yes. That's what I'm hoping :)
The log Leo provided showed that BUG_ON(cr_avg < 0) triggered while doing rt task migration. The earlier patch I provided fixes rt/stop/deadline classes migration fixup correctness. So I think the patch should fix Leo's case.
Thanks, Joonwoo