Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-05-06 Thread Steven Rostedt
On Wed, 29 Apr 2020 15:12:24 -0400 Joel Fernandes wrote: > > Mine ;-) > > > > The "kthread_should_stop()" is the normal way to end kernel threads. Using > > get/put is not the orthodox way. > > Agreed :-) thanks. And while I'm working on getting a new "urgent" branch out, I found this code i

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-29 Thread Joel Fernandes
On Wed, Apr 29, 2020 at 12:31:41PM -0400, Steven Rostedt wrote: > On Wed, 29 Apr 2020 19:54:16 +0800 > Xiao Yang wrote: > > > On 2020/4/28 22:45, Steven Rostedt wrote: > > > diff --git a/kernel/trace/preemptirq_delay_test.c > > > b/kernel/trace/preemptirq_delay_test.c > > >> index 1c28ca20e30

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-29 Thread Steven Rostedt
On Wed, 29 Apr 2020 19:54:16 +0800 Xiao Yang wrote: > On 2020/4/28 22:45, Steven Rostedt wrote: > > diff --git a/kernel/trace/preemptirq_delay_test.c > > b/kernel/trace/preemptirq_delay_test.c > >> index 1c28ca20e30b..6d9131ae7e8c 100644 > >> --- a/kernel/trace/preemptirq_delay_test.c > >>

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-29 Thread Xiao Yang
On 2020/4/28 22:45, Steven Rostedt wrote: diff --git a/kernel/trace/preemptirq_delay_test.c b/kernel/trace/preemptirq_delay_test.c index 1c28ca20e30b..6d9131ae7e8c 100644 --- a/kernel/trace/preemptirq_delay_test.c +++ b/kernel/trace/preemptirq_delay_test.c @@ -113,15 +113,27 @@ static int p

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-29 Thread Xiao Yang
On 2020/4/28 22:15, Joel Fernandes wrote: I am wondering if it is because in your test, the kthread exits too quickly. We have these comments in kthread_stop(): * If threadfn() may call do_exit() itself, the caller must ensure * task_struct can't go away. Does the below diff on top of the pr

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-29 Thread Xiao Yang
Hi Joel, Steven First of all, the previous dmesg may not be reliable and I often got the following dmesg by inserting and remving preemptirq_delay_test in loops: (btw: I think these dmesgs are caused by the same bug). --- [ 178.484520] [

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-28 Thread Steven Rostedt
On Tue, 28 Apr 2020 10:44:09 -0400 Steven Rostedt wrote: > On Tue, 28 Apr 2020 18:19:05 +0800 > Xiao Yang wrote: > > > Hi Joel, > > > > Thanks for your quick fix. > > > > Unfortunately, it fixes my original panic but introduces other > > issues(two wanings and one panic) on my arm64 vm, as be

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-28 Thread Steven Rostedt
On Tue, 28 Apr 2020 18:19:05 +0800 Xiao Yang wrote: > Hi Joel, > > Thanks for your quick fix. > > Unfortunately, it fixes my original panic but introduces other > issues(two wanings and one panic) on my arm64 vm, as below: > Does the following on top of Joel's patch fix it? -- Steve diff --g

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-28 Thread Joel Fernandes
On Tue, Apr 28, 2020 at 06:19:05PM +0800, Xiao Yang wrote: > Hi Joel, > > Thanks for your quick fix. > > Unfortunately, it fixes my original panic but introduces other > issues(two wanings and one panic) on my arm64 vm, as below: > -

Re: [PATCH] kernel/trace: Stop and wait for kthread on preempt irq module unload

2020-04-28 Thread Xiao Yang
Hi Joel, Thanks for your quick fix. Unfortunately, it fixes my original panic but introduces other issues(two wanings and one panic) on my arm64 vm, as below: [ 3465.434942] [ cut here ] [ 3465.435481] re