[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAM9d7cgFAc7WyZY4n82YLU=5yAifrVqoCtHh2iE-aLKx=uC4=w@mail.gmail.com>
Date: Fri, 30 Aug 2019 12:56:02 +0900
From: Namhyung Kim <namhyung@...nel.org>
To: Tejun Heo <tj@...nel.org>
Cc: Ingo Molnar <mingo@...nel.org>,
Peter Zijlstra <a.p.zijlstra@...llo.nl>,
Arnaldo Carvalho de Melo <acme@...nel.org>,
LKML <linux-kernel@...r.kernel.org>,
Jiri Olsa <jolsa@...hat.com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
Stephane Eranian <eranian@...gle.com>,
Li Zefan <lizefan@...wei.com>,
Johannes Weiner <hannes@...xchg.org>,
Adrian Hunter <adrian.hunter@...el.com>
Subject: Re: [PATCH 1/9] perf/core: Add PERF_RECORD_CGROUP event
Hi Tejun,
On Wed, Aug 28, 2019 at 11:48 PM Tejun Heo <tj@...nel.org> wrote:
>
> Hello, Namhyung.
>
> On Wed, Aug 28, 2019 at 04:31:22PM +0900, Namhyung Kim wrote:
> > + * struct {
> > + * struct perf_event_header header;
> > + * u64 ino;
> > + * u64 path_len;
> > + * char path[];
>
> ino and path aren't great identifers for cgroups because both can get
> recycled pretty quickly. Can you please take a look at
> KERNFS_ROOT_SUPPORT_EXPORTOP? That's the fhandle that cgroup uses,
> currently the standard ino+gen which isn't ideal but good enough.
> Another benefit is that the path can also be resolved efficiently from
> userspace given the handle.
Thanks for the info, I'll take a look at it.
Namhyung
Powered by blists - more mailing lists