Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Zhangjin Wu <falcon@tinylab.org>
To:     w@1wt.eu
Cc:     arnd@arndb.de, falcon@tinylab.org, linux-kernel@vger.kernel.org,
        linux-kselftest@vger.kernel.org, thomas@t-8ch.de
Subject: Re: [PATCH v3 12/12] selftests/nolibc: speed up some targets with multiple jobs
Date:   Sat, 29 Jul 2023 17:11:56 +0800
Message-Id: <20230729091156.8761-1-falcon@tinylab.org>
In-Reply-To: <20230729083128.GM956@1wt.eu>
References: <20230729083128.GM956@1wt.eu>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Feedback-ID: bizesmtp:tinylab.org:qybglogicsvrgz:qybglogicsvrgz5a-1
Precedence: bulk

> On Sat, Jul 29, 2023 at 08:44:32AM +0200, Thomas Wei?schuh wrote:
> > On 2023-07-28 04:35:01+0800, Zhangjin Wu wrote:
> > > The sysroot install and kernel build targets are time cost, let's use
> > > -j<N> to parallelize them with multiple jobs.
> > > 
> > > Signed-off-by: Zhangjin Wu <falcon@tinylab.org>
> > > ---
> > >  tools/testing/selftests/nolibc/Makefile | 4 ++--
> > >  1 file changed, 2 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/tools/testing/selftests/nolibc/Makefile b/tools/testing/selftests/nolibc/Makefile
> > > index ad2538ec5eb0..1b45c22f9a94 100644
> > > --- a/tools/testing/selftests/nolibc/Makefile
> > > +++ b/tools/testing/selftests/nolibc/Makefile
> > > @@ -171,7 +171,7 @@ PHONY = sysroot/$(ARCH)/include
> > >  sysroot/$(ARCH)/include:
> > >  	$(Q)rm -rf sysroot/$(ARCH) sysroot/sysroot
> > >  	$(QUIET_MKDIR)mkdir -p sysroot
> > > -	$(Q)$(MAKE) -C ../../../include/nolibc ARCH=$(ARCH) OUTPUT=$(CURDIR)/sysroot/ headers_standalone
> > > +	$(Q)$(MAKE) -j$$(nproc) -C ../../../include/nolibc ARCH=$(ARCH) OUTPUT=$(CURDIR)/sysroot/ headers_standalone
> > 
> > This should already work when the users specify -j on the make command
> > line themselves.
> > I'm not a fan of force-enabling it here.
> 
> Indeed, we must not do that, because some users might for instance
> prefer to build multiple archs in parallel and benefit from a better
> parallelism and now they'd end up with too many processes.

Ok, let users do what they want.

Zhangjin

> 
> Willy