Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
Date:   Thu, 14 Apr 2022 19:51:46 +0200
From:   Nicolas Schier <nicolas@fjasle.eu>
To:     Masahiro Yamada <masahiroy@kernel.org>
Cc:     Randy Dunlap <rdunlap@infradead.org>,
        Nick Desaulniers <ndesaulniers@google.com>,
        Yann Droneaud <ydroneaud@opteya.com>,
        Michal Marek <michal.lkml@markovi.net>,
        Linux Kbuild mailing list <linux-kbuild@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        Linus Torvalds <torvalds@linux-foundation.org>
Subject: Re: [PATCHv1] kbuild: support W=e to make build abort in case of
 warning
Message-ID: <YlhfMqQ+ypf4OdkN@fjasle.eu>
References: <1422803720-14723-1-git-send-email-ydroneaud@opteya.com>
 <20220408084607.106468-1-ydroneaud@opteya.com>
 <CAK7LNAQZLt_OecOogOQiSu5snW+sffsMoFgVcjPTx_idj_=_tQ@mail.gmail.com>
 <CAKwvOd=yNnKsHJo0QWvoTuFF9p-y=cTftTD+7FY-wJ_f23zFTQ@mail.gmail.com>
 <81585705-6ed8-12e5-1355-332a6a5d2b17@infradead.org>
 <CAK7LNAS6ap9dR=kzRgQgt+d7FBBbVrwEqGU9g_pFD+nzMUt+gQ@mail.gmail.com>
 <YlgdF9qmJyYGHKXZ@bergen.fjasle.eu>
 <CAK7LNARVKL4vCRHz5=_sW_oBpkpqhDu5Hp1hfG+Hnwqg-61HVA@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <CAK7LNARVKL4vCRHz5=_sW_oBpkpqhDu5Hp1hfG+Hnwqg-61HVA@mail.gmail.com>
Precedence: bulk

On Fri, Apr 15, 2022 at 12:15:40AM +0900 Masahiro Yamada wrote:
> On Thu, Apr 14, 2022 at 10:19 PM Nicolas Schier <nicolas@fjasle.eu> wrote:
> >
> > P? l?. 09. april 2022 kl. 10.47 +0000 skrev Masahiro Yamada:
> > > On Sat, Apr 9, 2022 at 5:36 AM Randy Dunlap <rdunlap@infradead.org> Wrote:
> > > >
> > > >
> > > >
> > > > On 4/8/22 13:29, Nick Desaulniers wrote:
> > > > > On Fri, Apr 8, 2022 at 4:06 AM Masahiro Yamada <masahiroy@kernel.org> wrote:
> > > > >>
> > > > >> On Fri, Apr 8, 2022 at 5:46 PM Yann Droneaud <ydroneaud@opteya.com> wrote:
> > > > >>>
> > > > >>> When developing new code/feature, CONFIG_WERROR is most
> > > > >>> often turned off, especially for people using make W=12 to
> > > > >>> get more warnings.
> > > > >>>
> > > > >>> In such case, turning on -Werror temporarily would require
> > > > >>> switching on CONFIG_WERROR in the configuration, building,
> > > > >>> then switching off CONFIG_WERROR.
> > > > >>>
> > > > >>> For this use case, this patch introduces a new 'e' modifier
> > > > >>> to W= as a short hand for KCFLAGS+=-Werror" so that -Werror
> > > > >>> got added to the kernel (built-in) and modules' CFLAGS.
> > > > >>>
> > > > >>> Signed-off-by: Yann Droneaud <ydroneaud@opteya.com>
> > > > >>> ---
> > > > >>>  Makefile                   |  1 +
> > > > >>>  scripts/Makefile.extrawarn | 13 +++++++++++--
> > > > >>>  2 files changed, 12 insertions(+), 2 deletions(-)
> > > > >>>
> > > > >>> Changes since v0[0]:
> > > > >>>
> > > > >>>  - rebase on top of commit 64a91907c896 ("kbuild: refactor scripts/Makefile.extrawarn")
> > > > >>>  - document use case after commit 3fe617ccafd6 ("Enable '-Werror' by default for all kernel builds")
> > > > >>>
> > > > >>> [0] https://lore.kernel.org/all/1422803720-14723-1-git-send-email-ydroneaud@opteya.com/
> > > > >>
> > > > >>
> > > > >> I remembered the previous submission, I liked it, but I had lost it.
> > > > >>
> > > > >> It seems already 7 years ago, (before I became the Kbuild maintainer).
> > > > >> Thanks for coming back to this.
> > > > >>
> > > > >>
> > > > >> I like this, but I will wait some time for review comments.
> > > > >
> > > > > Dunno, this seems pretty simple:
> > > > >
> > > > > $ ./scripts/config -e WERROR
> > > > > $ make ... W=12
> > > >
> > > > Yeah, that's about what I was thinking too..
> > >
> > >
> > >
> > > But, you cannot change the .config
> > > when you build external modules.
> > >
> > > "make W=e" might be useful for people who strive to
> > > keep their downstream modules warning-free.
> > >
> > >
> > > W=e is the same pattern.
> > > I do not see much downside.
> >
> > If I set CONFIG_WERROR=y on the make command line, I could have the
> > same result, don't I?
> >
> >   make CONFIG_WERROR=1 ...
> >
> > no matter if in-tree or for external kernel modules.
> 
> Yes.
> 
> If you can change the kernel configuration,
> you can enable CONFIG_WERROR.
> 
> To build external modules against the read-only
> /lib/modules/$(uname -r)/build/,
> it is not so feasible to change the .config file, though.

hm, I wanted to point out something different.  When I build an external module
against a read-only kbuild-tree, I _can_ change kconfig values by specifying
them on the make command line, just as I can add 'W=e':

  make -C /lib/modules/$(uname -r)/build M=~+ CONFIG_SAMPLE_KOBJECT=m CONFIG_WERROR=y

Thus, I suspect, that this is the very same as if I'd run

  make -C /lib/modules/$(uname -r)/build M=~+ CONFIG_SAMPLE_KOBJECT=m W=e

So, "W=e" is actually just a shortcut for "CONFIG_WERROR=1".  Or have I missed
something?

Kind regards,
Nicolas