RE: ftrace function_graph causes system crash

From: Bean Huo (beanhuo)
Date: Thu Sep 22 2016 - 04:37:36 EST


> -----Original Message-----
> From: Steven Rostedt [mailto:rostedt@xxxxxxxxxxx]
> Sent: Mittwoch, 21. September 2016 20:17
> To: Jisheng Zhang <jszhang@xxxxxxxxxxx>
> Cc: Bean Huo (beanhuo) <beanhuo@xxxxxxxxxx>; Zoltan Szubbocsev
> (zszubbocsev) <zszubbocsev@xxxxxxxxxx>; catalin.marinas@xxxxxxx;
> will.deacon@xxxxxxx; rfi@xxxxxxxxxxxxxxxxxxxxxx; linux-
> kernel@xxxxxxxxxxxxxxx; mingo@xxxxxxxxxx; linux-arm-
> kernel@xxxxxxxxxxxxxxxxxxx
> Subject: Re: ftrace function_graph causes system crash
>
> On Wed, 21 Sep 2016 17:13:07 +0800
> Jisheng Zhang <jszhang@xxxxxxxxxxx> wrote:
>
> > I'm not sure whether the commit d6df3576e6b4
> ("clocksource/drivers/arm_global_timer
> > : Prevent ftrace recursion") can fix this issue.
> >
> > this commit is merged since v4.3, I noticed your kernel version is v4.0
>
> BTW, yes, that would be the fix.
>
> -- Steve
>
> >
> > Thanks,
> > Jisheng
> >
> > > Do you know now how to deeply debug and trace which line is wrong
> through Ftrace?
> > >

Hi, Steven and Jisheng
Thanks to both warm-hearted guys. I merged d6df3576e6b4 patch into my kernel 4.0.
Then it is true, no cash appears again.

I have one more question that current ftrace can trace DMA latency, include mapping and unmapping?
Means I want to know when one BIO request be completed. Just like blktrace.
But blktrace can not tell me the function calling sequence.

--Bean