Hi Frederic, FYI, On Tue, 12 Nov 2019 at 22:38, Naresh Kamboju <naresh.kamboju@xxxxxxxxxx> wrote: > > INFO: rcu_sched self-detected stall on CPU detected on Linux next > 5.4.0-rc7-next-20191112 running on x86_64 device and continuously > popped up kernel. > > Steps to reproduce: > 1. boot x86_64 with Linux next 5.4.0-rc7-next-20191112 > 2. Run "perf test -v" > > x86_64 device: > Running perf test cases > perf test -v > ... > test child forked, pid 418 > mmap size 528384B > [ 74.040659] rcu: INFO: rcu_sched self-detected stall on CPU > [ 74.046275] rcu: 0-....: (26000 ticks this GP) > idle=25a/1/0x4000000000000002 softirq=7882/7882 fqs=6495 > [ 74.151951] RIP: 0010:irq_work_sync+0x51/0x60 Thanks for providing fix patch. [PATCH] irq_work: Fix IRQ_WORK_BUZY bit clearing Applied this patch and tested perf test on x86_64 and reported issue is gone. ref: https://lore.kernel.org/lkml/20191113171201.14032-1-frederic@xxxxxxxxxx/T/#u - Naresh
