[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <0b71d4f9-f707-3d39-c358-7c06c5689a9d@linux.intel.com>
Date: Thu, 3 Jun 2021 06:52:19 -0700
From: Andi Kleen <ak@...ux.intel.com>
To: Josh Poimboeuf <jpoimboe@...hat.com>,
Peter Zijlstra <peterz@...radead.org>
Cc: syzbot <syzbot+84fe685c02cd112a2ac3@...kaller.appspotmail.com>,
bp@...en8.de, hpa@...or.com, inglorion@...gle.com,
linux-kernel@...r.kernel.org, mingo@...hat.com,
syzkaller-bugs@...glegroups.com, tglx@...utronix.de,
x86@...nel.org, Andy Lutomirski <luto@...nel.org>,
Steven Rostedt <rostedt@...dmis.org>
Subject: Re: [syzbot] KASAN: stack-out-of-bounds Read in profile_pc
> True, ftrace does have function profiling (function_profile_enabled).
>
> Steve, is there a way to enable that on the kernel cmdline?
That's not really comparable. function profiling has a lot more
overhead. Also there is various code which has ftrace instrumentation
disabled.
I don't think why you want to kill the old profiler. It's rarely used,
but when you need it usually works. It's always good to have simple fall
backs. And it's not that it's a lot of difficult code.
-Andi
Powered by blists - more mailing lists