[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <MN2PR21MB1518440CF0E4C5A335EBD8E3F78F0@MN2PR21MB1518.namprd21.prod.outlook.com>
Date: Fri, 29 May 2020 21:44:55 +0000
From: Steve MacLean <Steve.MacLean@...rosoft.com>
To: Steve MacLean <Steve.MacLean@...rosoft.com>,
Ian Rogers <irogers@...gle.com>
CC: Steve MacLean <steve.maclean@...ux.microsoft.com>,
Peter Zijlstra <peterz@...radead.org>,
Ingo Molnar <mingo@...hat.com>,
Arnaldo Carvalho de Melo <acme@...nel.org>,
Mark Rutland <mark.rutland@....com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
Jiri Olsa <jolsa@...hat.com>,
Namhyung Kim <namhyung@...nel.org>,
Stephane Eranian <eranian@...gle.com>,
LKML <linux-kernel@...r.kernel.org>
Subject: RE: [EXTERNAL] Re: [PATCH v4] perf inject --jit: Remove //anon mmap
events
>/tmp/perf/perf record -k 1 -e cycles:u -o /tmp/perf.data java
>-agentpath:/tmp/perf/libperf-jvmti.so -XX:+PreserveFramePointer
>-XX:InitialCodeCacheSize=20M -XX:ReservedCodeCacheSize=1G -jar
>dacapo-9.12-bach.jar jython
Its also possible the `InitialCodeCacheSize=20M` argument is masking the problem. Something like 4K would make the problem much easier to see.
I see the problem every time .NET grows the cache across a 64K page boundary. 20M may mean you are allocating huge pages or something which might mask the problem. We may be allocating code pages 64K at a time.
Powered by blists - more mailing lists