Messages in this thread | | | Date | Mon, 14 Sep 2015 17:07:16 +0800 | From | "Wangnan (F)" <> | Subject | Re: [GIT PULL 00/15] perf/core improvements and fixes |
| |
On 2015/9/14 16:41, Ingo Molnar wrote: > * Arnaldo Carvalho de Melo <acme@redhat.com> wrote: > >> Em Tue, Sep 08, 2015 at 04:21:47PM +0200, Ingo Molnar escreveu: >>> * Arnaldo Carvalho de Melo <acme@redhat.com> wrote: >>> >>>> Em Fri, Sep 04, 2015 at 10:06:28PM -0300, Arnaldo Carvalho de Melo escreveu: >>>>> Hi Ingo, >>>>> >>>>> Please consider pulling, this is on top of the previous pull request, >>>>> perf-core-for-mingo. >>>> Ingo, please do not pull this 'perf-core-for-mingo-2' tag, there were >>>> some misunderstandings about the acks for "Move sw clock metrics >>>> printout to stat-shadow" and Jiri and Andi are working that out. >>>> >>>> I'll remove those patches and get a new perf-core-for-mingo-2 >>>> tag in place, before continuing today's batch, which possibly will be >>>> available as 'perf-core-for-mingo-3' >>>> >>>> What is in 'perf-core-for-mingo" should be Ok. >>> Ok! >> Thanks! I have already removed that problematic changeset and resigned >> the 'perf-core-for-mingo-2' tag, same contents modulo that cset. > Hm, so I pulled it (commit 1765d9b26f84), but with an old perf.data I'm getting > this crash: > > triton:~/tip/tools/perf> perf report > perf: Segmentation fault > -------- backtrace -------- > perf[0x52bc0b] > /lib/x86_64-linux-gnu/libc.so.6(+0x352f0)[0x7f51a583c2f0] > perf[0x42ce95] > perf[0x4bc6c3] > perf[0x4bcfa1] > perf[0x4bf939] > perf(perf_session__process_events+0x390)[0x4be430] > perf(cmd_report+0x1070)[0x42e2e0] > perf[0x478e03] > perf(main+0x60a)[0x41f1ba] > /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf0)[0x7f51a5827a40] > perf(_start+0x29)[0x41f2d9] > [0x0] > > I also re-tested qand it still crashes. > > Bisected it to: > > e1e499aba570 perf tools: Add processor socket info to hist_entry and addr_location > > Running on Ubuntu, 1 socket box, 12 CPUs.
Hi Ingo,
It seems you met a bug we are discussing these days. Please have a look at the following discussions:
http://lkml.kernel.org/r/1441630315-189525-1-git-send-email-wangnan0@huawei.com
http://lkml.kernel.org/r/1441828225-667-1-git-send-email-acme@kernel.org
Thank you.
> I went back to perf/core 8f3e5684d3fb and it doesn't crash anymore - so I unpulled > your tree for now. (Will send you the perf.data privately.) > > 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/
| |