Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.37 as permitted sender) client-ip=23.128.96.37;
MIME-Version: 1.0
References: <20230923053515.535607-1-irogers@google.com> <20230923053515.535607-2-irogers@google.com>
 <CAKwvOdmHg_43z_dTZrOLGubuBBvmHdPxSFjOWa3oWkbOp2qWWg@mail.gmail.com>
 <CAP-5=fV6c1tWAd2GjMwn4PQN=3BXNQGz=vbonHSjRjQ3fbEL+g@mail.gmail.com>
 <ZRWMWcNKvZMgiAMR@bergen.fjasle.eu> <CAK7LNASuOWtd-iv2C3HLWr1oq9gD8BiP4B=8wRUo81CSK=g5Yw@mail.gmail.com>
In-Reply-To: <CAK7LNASuOWtd-iv2C3HLWr1oq9gD8BiP4B=8wRUo81CSK=g5Yw@mail.gmail.com>
From:   Ian Rogers <irogers@google.com>
Date:   Thu, 5 Oct 2023 15:35:07 -0700
Message-ID: <CAP-5=fVz-Yqb_AnkJkoBAO8id7m2f0qaU_VnFVDoAXagwHfz6A@mail.gmail.com>
Subject: Re: [PATCH v1 01/18] gen_compile_commands: Allow the line prefix to
 still be cmd_
To:     Masahiro Yamada <masahiroy@kernel.org>
Cc:     Nicolas Schier <nicolas@fjasle.eu>,
        Nick Desaulniers <ndesaulniers@google.com>,
        Nathan Chancellor <nathan@kernel.org>,
        Tom Rix <trix@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Ingo Molnar <mingo@redhat.com>,
        Arnaldo Carvalho de Melo <acme@kernel.org>,
        Mark Rutland <mark.rutland@arm.com>,
        Alexander Shishkin <alexander.shishkin@linux.intel.com>,
        Jiri Olsa <jolsa@kernel.org>,
        Namhyung Kim <namhyung@kernel.org>,
        Adrian Hunter <adrian.hunter@intel.com>,
        Yicong Yang <yangyicong@hisilicon.com>,
        Jonathan Cameron <jonathan.cameron@huawei.com>,
        Yang Jihong <yangjihong1@huawei.com>,
        Kan Liang <kan.liang@linux.intel.com>,
        Ming Wang <wangming01@loongson.cn>,
        Huacai Chen <chenhuacai@kernel.org>,
        Sean Christopherson <seanjc@google.com>,
        K Prateek Nayak <kprateek.nayak@amd.com>,
        Yanteng Si <siyanteng@loongson.cn>,
        Yuan Can <yuancan@huawei.com>,
        Ravi Bangoria <ravi.bangoria@amd.com>,
        James Clark <james.clark@arm.com>,
        Paolo Bonzini <pbonzini@redhat.com>, llvm@lists.linux.dev,
        linux-kernel@vger.kernel.org, linux-perf-users@vger.kernel.org,
        bpf@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Tue, Oct 3, 2023 at 7:31=E2=80=AFAM Masahiro Yamada <masahiroy@kernel.or=
g> wrote:
>
> On Thu, Sep 28, 2023 at 11:26=E2=80=AFPM Nicolas Schier <nicolas@fjasle.e=
u> wrote:
> >
> > On Mon, 25 Sep 2023 09:06:11 -0700, Ian Rogers wrote:
> > > On Mon, Sep 25, 2023 at 8:49=E2=80=AFAM Nick Desaulniers
> > > <ndesaulniers@google.com> wrote:
> > > >
> > > > On Fri, Sep 22, 2023 at 10:35=E2=80=AFPM Ian Rogers <irogers@google=
.com> wrote:
> > > > >
> > > > > Builds in tools still use the cmd_ prefix in .cmd files, so don't
> > > > > require the saved part. Name the groups in the line pattern match=
 so
> > > >
> > > > Is that something that can be changed in the tools/ Makefiles?
> > > >
> > > > I'm fine with this change, just curious where the difference comes
> > > > from precisely.
> > > > Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
> > >
> > > I agree. The savedcmd_ change came from Masahiro in:
> > > https://lore.kernel.org/lkml/20221229091501.916296-1-masahiroy@kernel=
.org/
> > > I was reluctant to change the build logic in tools/ because of the
> > > potential to break things. Maybe Masahiro/Nicolas know of issues?
> >
> > I haven't seen any issues related to the introduction of savedcmd_; and
> > roughly searching through tools/ I cannot find a rule that matches the
> > pattern Masahiro described in commit 92215e7a801d ("kbuild: rename
> > cmd_$@ to savedcmd_$@ in *.cmd files", 2022-12-29).  For consistency,
> > I'd like to see the build rules in tools/ re-use the ones from scripts/
> > but as of now I don't see any necessity to introduce savedcmd in
> > tools/, yet.
> >
> > Kind regards,
> > Nicolas
>
>
> tools/build/Build.include mimics scripts/Kbuild.include
>
> That should be changed in the same way.

Thanks Masahiro,

I support that as a goal, I'm not sure I want to embrace it here. For
example, in tools/build/Build.include there is:
```
# Echo command
# Short version is used, if $(quiet) equals `quiet_', otherwise full one.
echo-cmd =3D $(if $($(quiet)cmd_$(1)),\
           echo '  $(call escsq,$($(quiet)cmd_$(1)))';)
```

This is relevant given the use of cmd_ and not savedcmd_. In
scripts/Kbuild.include there is no equivalent, nor is there in
scripts. So do I dig into the history of echo-cmd unfork that, and
then go to the next thing? I find a lot of things like
tools/selftests/bpf won't build for me. When I debug a failure is it
because of a change or a pre-existing issue? Given the scope of the
Build.include unification problem, and this 4 line change, I hope we
can move forward with this and keep unification as a separate problem
(I totally support solving that problem).

Thanks,
Ian

> --
> Best Regards
> Masahiro Yamada