[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <YwfTJ3k1zgTcejEn@kernel.org>
Date: Thu, 25 Aug 2022 16:53:11 -0300
From: Arnaldo Carvalho de Melo <acme@...nel.org>
To: Rob Herring <robh@...nel.org>
Cc: Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
Ingo Molnar <mingo@...hat.com>,
Catalin Marinas <catalin.marinas@....com>,
Peter Zijlstra <peterz@...radead.org>,
Mark Rutland <mark.rutland@....com>,
Will Deacon <will@...nel.org>, Jiri Olsa <jolsa@...nel.org>,
Namhyung Kim <namhyung@...nel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
linux-arm-kernel <linux-arm-kernel@...ts.infradead.org>,
linux-perf-users <linux-perf-users@...r.kernel.org>
Subject: Re: [PATCH RFC v1 2/3] perf: Add perf_event_attr::config3
Em Thu, Aug 25, 2022 at 02:43:42PM -0500, Rob Herring escreveu:
> On Thu, Aug 25, 2022 at 2:31 PM Arnaldo Carvalho de Melo <acme@...nel.org> wrote:
> > Em Thu, Aug 25, 2022 at 01:08:01PM -0500, Rob Herring escreveu:
> > > Arm SPEv1.2 adds another 64-bits of event filtering control. As the
> > > existing perf_event_attr::configN fields are all used up for SPE PMU, an
> > > additional field is needed. Add a new 'config3' field.
> > Try not to have tools/ and kernel code in the same patch, else you'll
> > burden kernel developers into testing tools/, which so far has been
> > refrained.
> I knew that, but assumed the header was special...
So, we have tools/perf/check-headers.sh for that purpose, so that tools
developers can have the opportunity to see that the kernel ABI changed,
think about it, check what is needed to best support the new kernel ABI,
etc.
I really love when kernel developers take the time and help with
supporting new features by working on the tools/ bits, that would be
perfect!
But, as we saw just a few days ago, when changes were made in the kernel
and in tools/ in tandem, tools/ got broken just before -rc2:
"perf tools: Fix compile error for x86"
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=cfd2b5c1106fa20254d9f24970232cdf24860005>
So it seems better to first do the kernel work, then do the user part,
to avoid burdening kernel developers with tools/ work.
- Arnaldo
Powered by blists - more mailing lists