Date: Wed, 18 Nov 2015 09:20:33 +0100
From: Ingo Molnar <mingo@kernel.org>
To: "Wangnan (F)" <wangnan0@huawei.com>
Cc: Jiri Olsa <jolsa@kernel.org>, Arnaldo Carvalho de Melo <acme@kernel.org>,
        David Ahern <dsahern@gmail.com>, Peter Zijlstra <peterz@infradead.org>,
        Milian Wolff <milian.wolff@kdab.com>, linux-kernel@vger.kernel.org,
        pi3orama <pi3orama@163.com>, lizefan 00213767 <lizefan@huawei.com>
Subject: Re: [BUG REPORT] perf tools: x86_64: Broken calllchain when sampling
 taken at 'callq' instruction
Message-ID: <20151118082033.GA24726@gmail.com>
References: <564C26C4.2040603@huawei.com>
 <564C3011.8090002@huawei.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <564C3011.8090002@huawei.com>
User-Agent: Mutt/1.5.23 (2014-03-12)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1577
Lines: 58


* Wangnan (F) <wangnan0@huawei.com> wrote:

> On 2015/11/18 15:20, Wangnan (F) wrote:
> >Hi all,
> >
> >When analysising Jiri's patchset [1] I found a dwarf unwind problem.
> >On x86 platform, when sample is at a 'callq' instruction, dwarf based
> >stack unwind always fail.
> >
> >I compile a small C source file with debug information, turn off
> >frame pointer and disable optimization:
> >
> >$ gcc -g -O0 -fomit-frame-pointer ./test_dwarf_unwind.c -o
> >./test_dwarf_unwind
> 
> For whom want to test it: here is the test code I used.
> 
> #include <stdio.h>
> #include <unistd.h>
> #include <sys/time.h>
> 
> static volatile int x = 0;
> 
> int funcc(void)
> {
>     struct timeval tv1, tv2;
>     unsigned long us1, us2;
> 
>     gettimeofday(&tv1, NULL);
> 
>     us1 = tv1.tv_sec * 1000000 + tv1.tv_usec;
> 
>     while(1) {
>         x = x + 100;
>         gettimeofday(&tv2, NULL);
>         us2 = tv2.tv_sec * 1000000 + tv2.tv_usec;
>         if (us2 - us1 >= 3000000)
>             break;
>     }
>     return x;
> }
> int funcb(void) { return funcc();}
> int funca(void) { return funcb();}
> int main() { funca(); return 0;}

What CPU model is this, and what event was used - PEBS perhaps? This might be some 
sort of PMU sampling bug/quirk/misfeature - or perhaps a kernel side fixup that 
went bad?

Thanks,

	Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/