Messages in this thread | | | Date | Thu, 11 Nov 2010 09:58:25 +0100 | From | Ingo Molnar <> | Subject | Re: [PATCH 03/20] x86, mce: Add persistent MCE event |
| |
* Borislav Petkov <bp@amd64.org> wrote:
> On Wed, Nov 10, 2010 at 05:21:57PM -0500, Ingo Molnar wrote: > > > > * Ben Gamari <bgamari@gmail.com> wrote: > > > > > On Thu, 4 Nov 2010 16:36:39 +0100, Borislav Petkov <bp@amd64.org> wrote: > > > > From: Borislav Petkov <borislav.petkov@amd.com> > > > > > > > > Add the necessary glue to enable the mce_record tracepoint on boot > > > > turning it into a persistent event. This exports the MCE buffer > > > > read-only to a userspace daemon which will hook into it through debugfs > > > > when booting is finished. > > > > > > While MCE technically is a diagnostics service, do we want to start adding > > > dependencies on debugfs in long-running, common-place daemons? I was under the > > > impression we were to avoid using debugfs for anything other than values for > > > one-off debugging. > > > > See this current discussion thread on lkml: > > > > [RFC][PATCH] perf: sysfs type id > > > > Events are being added to sysfs as we want to avoid the debugfs dependency. > > Yep, event enumeration will land in /sysfs eventually but in this case debugfs is > used to export the buffers with the events recorded so far to the daemon which > mmaps them. And I remember also being mentioned that debugfs will thus be always > compiled in because of that.
The event ring-buffer can be mmap()-ed off the fd that sys_perf_event_open() gives. This is what all of tools/perf/ does - it uses debugfs only for even enumeration (which will move to sysfs).
Thanks,
Ingo
| |