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;
MIME-Version: 1.0
References: <20220820153506.607928-1-chenhuacai@loongson.cn> <CAK8P3a1Z2xR_Poiv4dTntpghWKES6Jk75jNiDy8q7_L8XfRyuQ@mail.gmail.com>
In-Reply-To: <CAK8P3a1Z2xR_Poiv4dTntpghWKES6Jk75jNiDy8q7_L8XfRyuQ@mail.gmail.com>
From:   Huacai Chen <chenhuacai@kernel.org>
Date:   Wed, 24 Aug 2022 20:49:55 +0800
Message-ID: <CAAhV-H7wLp2EiEL0zyLL0_+tZAqTwy+tfFcyE9WwJeDYTfQY-g@mail.gmail.com>
Subject: Re: [PATCH] LoongArch: Fix build warnings in VDSO
To:     Arnd Bergmann <arnd@kernel.org>
Cc:     Huacai Chen <chenhuacai@loongson.cn>, loongarch@lists.linux.dev,
        Xuefeng Li <lixuefeng@loongson.cn>,
        Guo Ren <guoren@kernel.org>, Xuerui Wang <kernel@xen0n.name>,
        Jiaxun Yang <jiaxun.yang@flygoat.com>,
        LKML <linux-kernel@vger.kernel.org>,
        kernel test robot <lkp@intel.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi, Arnd,

On Wed, Aug 24, 2022 at 6:10 PM Arnd Bergmann <arnd@kernel.org> wrote:
>
> On Sat, Aug 20, 2022 at 5:35 PM Huacai Chen <chenhuacai@loongson.cn> wrote:
> >
> > diff --git a/arch/loongarch/vdso/vgetcpu.c b/arch/loongarch/vdso/vgetcpu.c
> > index 43a0078e4418..e02e775f5360 100644
> > --- a/arch/loongarch/vdso/vgetcpu.c
> > +++ b/arch/loongarch/vdso/vgetcpu.c
> > @@ -24,6 +24,8 @@ static __always_inline const struct vdso_pcpu_data *get_pcpu_data(void)
> >         return (struct vdso_pcpu_data *)(get_vdso_base() - VDSO_DATA_SIZE);
> >  }
> >
> > +extern
> > +int __vdso_getcpu(unsigned int *cpu, unsigned int *node, struct getcpu_cache *unused);
> >  int __vdso_getcpu(unsigned int *cpu, unsigned int *node, struct getcpu_cache *unused)
> >  {
>
> I don't think that adding the declaration before the function is a
> good workaround here,
> that is likely to just trigger another warning about declarations in .c files.
Yes, this is not a good solution, I just use the same method as RISC-V. :)
But "another warning" is a checkpatch warning, not a build warning, so
lkp@intel.com will not send bug reports again and again. :)

>
> I would add the declarations into a header file instead. If the
> prototypes are the same
> across architectures, this can be a global header that also avoids the
> warnings elsewhere,
> otherwise just keep it local to this directory.
gettimeofday, clock_gettime and clock_getres can be declared across
architectures, but getcpu seems not suitable because not all
architectures have it.

Huacai

>
>       Arnd