Re: [tip:perf/core] perf/x86: Fix USER/KERNEL tagging of samples

From: Linus Torvalds <torvalds@linux-foundation.org>
To: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: mingo@kernel.org, hpa@zytor.com, eranian@google.com,
	linux-kernel@vger.kernel.org, fweisbec@gmail.com,
	akpm@linux-foundation.org, tglx@linutronix.de,
	linux-tip-commits@vger.kernel.org
Subject: Re: [tip:perf/core] perf/x86: Fix USER/KERNEL tagging of samples
Date: Fri, 6 Jul 2012 11:34:35 -0700	[thread overview]
Message-ID: <CA+55aFw+vE4zMe1bJohLuoMDaFpqGCz6sBNRE8gkV7JOvtro0g@mail.gmail.com> (raw)
In-Reply-To: <CA+55aFzqoz7dcUCQfSrXxX2T8=2wgWdb1H2vaBhk7V-r0uaogw@mail.gmail.com>

On Fri, Jul 6, 2012 at 11:16 AM, Linus Torvalds
<torvalds@linux-foundation.org> wrote:
>
> If some code then changes the values in the pt_regs, it is *that* code
> that needs to think twice about what it does. Where is that code?

>From a quick grep it looks like it is __intel_pmu_pebs_event() that does this.

THAT is where you would possibly have a huge honking big comment about
how you have to fake the CS register contents because the PEBS
information is incomplete. But make it clear that it is a total hack.

Also, somebody should check. Is the PEBS information *actually* the
instruction pointer (address within the code segment), or is it the
"linear address" (segment base + rip)? I hope it is the latter,
because in the absense of CS, the segment-based address is very
unclear indeed.

And if it *is* the linear address, then at that point you could do

   regs->cs = kernel_ip(ip) ? __KERNEL_CS : __USER_CS;
   regs->eflags &= ~X86_EFLAGS_VM;

and document this as a "we fake the CS and vm86 mode, using the known
zero-based code segments". At that point it would be technically
correct.

But any code that does "kernel_ip(regs->ip)" is just terminally
confused and can never be sane.

                  Linus