linux-kernel - Re: [PATCH RESEND v4] perf/core: Fix reentry problem in perf_output_read

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:   Mon, 5 Sep 2022 09:33:43 +0800
From:   Yang Jihong <yangjihong1@...wei.com>
To:     Steven Rostedt <rostedt@...dmis.org>
CC:     Ingo Molnar <mingo@...hat.com>,
        Arnaldo Carvalho de Melo <acme@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        Namhyung Kim <namhyung@...nel.org>,
        linux-kernel <linux-kernel@...r.kernel.org>,
        linux-perf-users <linux-perf-users@...r.kernel.org>,
        Yang Jihong <yangjihong1@...wei.com>,
        Jiri Olsa <jolsa@...nel.org>
Subject: Re: [PATCH RESEND v4] perf/core: Fix reentry problem in
 perf_output_read_group

Hello,

On 2022/9/5 9:25, Yang Jihong wrote:
> On Fri, Sep 02, 2022 at 04:29:18PM +0800, Yang Jihong wrote:
>  > perf_output_read_group may respond to IPI request of other cores and
>  > invoke __perf_install_in_context function. As a result, hwc 
> configuration is modified.
>  > causing inconsistency and unexpected consequences.
>  >
>  > Interrupts are not disabled when perf_output_read_group reads PMU 
> counter.
>  > In this case, IPI request may be received from other cores.
>  > As a result, PMU configuration is modified and an error occurs when
>  > reading PMU counter:
>  >
>  >                    CPU0                                         CPU1
>  > __se_sys_perf_event_open
>  > perf_install_in_context
>  > perf_output_read_group smp_call_function_single
>  >   for_each_sibling_event(sub, leader) { generic_exec_single
>  >     if ((sub != event) && remote_function
>  >         (sub->state == PERF_EVENT_STATE_ACTIVE))                    |
>  > <enter IPI handler: __perf_install_in_context>   <----RAISE IPI-----+
>  > __perf_install_in_context
>  >   ctx_resched
>  >     event_sched_out
>  >       armpmu_del
>  >         ...
>  >         hwc->idx = -1; // event->hwc.idx is set to -1 ...
>  > <exit IPI>
>  >             sub->pmu->read(sub);
>  >               armpmu_read
>  >                 armv8pmu_read_counter
>  >                   armv8pmu_read_hw_counter
>  >                     int idx = event->hw.idx; // idx = -1
>  >                     u64 val = armv8pmu_read_evcntr(idx);
>  >                       u32 counter = ARMV8_IDX_TO_COUNTER(idx); // 
> invalid counter = 30
>  >                       read_pmevcntrn(counter) // undefined instruction
>  >
>  > Signed-off-by: Yang Jihong <yangjihong1@...wei.com>
>  > ---
> 
> Right; thanks for reminding me. I had hoped you'd pick up and test the 
> extra assertion some, but I'll just do that on top.
> .
OK, if you want me to do assertion testing on any scenarios,
please send me the patch.

Thanks,
Yang