linux-kernel - Re: [PATCH] perf metricgroup: Fix segmentation fault for metrics with no pmu event

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 3 Feb 2021 08:46:52 +0800
From:   "Jin, Yao" <yao.jin@...ux.intel.com>
To:     John Garry <john.garry@...wei.com>, acme@...nel.org,
        jolsa@...nel.org, peterz@...radead.org, mingo@...hat.com,
        alexander.shishkin@...ux.intel.com
Cc:     Linux-kernel@...r.kernel.org, ak@...ux.intel.com,
        kan.liang@...el.com, yao.jin@...el.com
Subject: Re: [PATCH] perf metricgroup: Fix segmentation fault for metrics with
 no pmu event

Hi John,

On 2/2/2021 4:15 PM, John Garry wrote:
> On 02/02/2021 02:24, Jin Yao wrote:
>> Hit a segmentation fault for DRAM_BW_Use on SKL/KBL.
>>
>>    # perf stat -M DRAM_BW_Use -a -- sleep 1
>>
>>    Segmentation fault (core dumped)
>>
>>    (gdb) backtrace
>>    #0  __strcmp_sse2_unaligned () at ../sysdeps/x86_64/multiarch/strcmp-sse2-unaligned.S:31
>>    #1  0x0000555c9facd9ce in find_evsel_group (evlist_used=0x555ca16d09c0, 
>> metric_events=0x555ca16e9160, has_constraint=false, metric_no_merge=false,
>>        pctx=0x555ca16c8ff0, perf_evlist=0x555ca16c5340) at util/metricgroup.c:281
>>    #2  metricgroup__setup_events (metric_events_list=0x555ca0044378 <stat_config+280>, 
>> perf_evlist=0x555ca16c5340, metric_no_merge=<optimized out>,
>>        groups=0x7ffc599193f0) at util/metricgroup.c:323
>>    #3  parse_groups (perf_evlist=perf_evlist@...ry=0x555ca16c5340, str=str@...ry=0x7ffc599205f8 
>> "DRAM_BW_Use", metric_no_group=metric_no_group@...ry=false,
>>        metric_no_merge=metric_no_merge@...ry=false, fake_pmu=fake_pmu@...ry=0x0, 
>> metric_events=metric_events@...ry=0x555ca0044378 <stat_config+280>,
>>        map=0x555ca004e780 <pmu_events_map+768>) at util/metricgroup.c:1235
>>    #4  0x0000555c9face096 in metricgroup__parse_groups (opt=<optimized out>, str=0x7ffc599205f8 
>> "DRAM_BW_Use", metric_no_group=<optimized out>,
>>        metric_no_merge=<optimized out>, metric_events=0x555ca0044378 <stat_config+280>) at 
>> util/metricgroup.c:1253
>>    #5  0x0000555c9fbd084e in get_value (p=p@...ry=0x7ffc599196b0, opt=0x555ca0043f60 
>> <stat_options+4032>, flags=flags@...ry=1) at parse-options.c:251
>>    #6  0x0000555c9fbd1aa2 in parse_short_opt (options=<optimized out>, p=<optimized out>) at 
>> parse-options.c:351
>>    #7  parse_options_step (usagestr=0x7ffc59919830, options=0x555ca0042fa0 <stat_options>, 
>> ctx=0x7ffc599196b0) at parse-options.c:539
>>    #8  parse_options_subcommand (argc=argc@...ry=7, argv=argv@...ry=0x7ffc5991e6c0, 
>> options=options@...ry=0x555ca0042fa0 <stat_options>,
>>        subcommands=subcommands@...ry=0x7ffc59919840, usagestr=usagestr@...ry=0x7ffc59919830, 
>> flags=flags@...ry=2) at parse-options.c:654
>>    #9  0x0000555c9fa15154 in cmd_stat (argc=7, argv=0x7ffc5991e6c0) at builtin-stat.c:2136
>>    #10 0x0000555c9fa8bafd in run_builtin (p=0x555ca004df20 <commands+288>, argc=7, 
>> argv=0x7ffc5991e6c0) at perf.c:312
>>    #11 0x0000555c9f9f413a in handle_internal_command (argv=0x7ffc5991e6c0, argc=7) at perf.c:364
>>    #12 run_argv (argcp=<synthetic pointer>, argv=<synthetic pointer>) at perf.c:408
>>    #13 main (argc=7, argv=0x7ffc5991e6c0) at perf.c:538
>>
>> DRAM_BW_Use uses an event 'duration_time' but it doesn't have pmu, so
>> ev->leader->pmu_name is NULL for this case. See following code piece:
>>
>> !strcmp(ev->leader->pmu_name,
>>     metric_events[i]->leader->pmu_name)
>>
>> It causes the segmentation fault. Now check the pmu_name before strcmp.
>>
>> Fixes: c2337d67199a("perf metricgroup: Fix metrics using aliases covering multiple PMUs")
>> Signed-off-by: Jin Yao <yao.jin@...ux.intel.com>
> 
> This should be fixed in v5.11-rc6 - please check it.
> 
> 9c880c24cb0d perf metricgroup: Fix for metrics containing duration_time
> 
> Thanks,
> John

Oh, I see, your patch is in Arnaldo's perf/urgent branch. Sorry about that, I just tested Arnaldo's 
perf/core branch. :(

Thanks
Jin Yao