[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <96c37bf7-7dc9-5895-be76-cc52a032709f@intel.com>
Date: Wed, 30 Sep 2020 06:51:35 -0700
From: Dave Hansen <dave.hansen@...el.com>
To: Stephane Eranian <eranian@...gle.com>,
"Liang, Kan" <kan.liang@...ux.intel.com>
Cc: Peter Zijlstra <peterz@...radead.org>,
Ingo Molnar <mingo@...hat.com>,
Arnaldo Carvalho de Melo <acme@...nel.org>,
LKML <linux-kernel@...r.kernel.org>,
Mark Rutland <mark.rutland@....com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
Jiri Olsa <jolsa@...hat.com>, Andi Kleen <ak@...ux.intel.com>,
kirill.shutemov@...ux.intel.com,
Michael Ellerman <mpe@...erman.id.au>,
benh@...nel.crashing.org, Paul Mackerras <paulus@...ba.org>
Subject: Re: [PATCH V8 1/4] perf/core: Add PERF_SAMPLE_DATA_PAGE_SIZE
On 9/30/20 12:15 AM, Stephane Eranian wrote:
>> + /*
>> + * Software page-table walkers must disable IRQs,
>> + * which prevents any tear down of the page tables.
>> + */
>> + local_irq_save(flags);
>> +
>> + size = __perf_get_page_size(current->active_mm, addr);
>> +
> When I tested on my kernel, it panicked because I suspect
> current->active_mm could be NULL. Adding a check for NULL avoided the
> problem. But I suspect this is not the correct solution.
Did you happen to capture the oops? I can _imagine_ scenarios where
current->active_mm could be NULL, I just can't find any obvious ones in
the code.
Powered by blists - more mailing lists