[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20210920185736.40285450@canb.auug.org.au>
Date: Mon, 20 Sep 2021 18:57:36 +1000
From: Stephen Rothwell <sfr@...b.auug.org.au>
To: Yafang Shao <laoar.shao@...il.com>
Cc: Thomas Gleixner <tglx@...utronix.de>, Ingo Molnar <mingo@...e.hu>,
"H. Peter Anvin" <hpa@...or.com>,
Peter Zijlstra <peterz@...radead.org>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
Linux Next Mailing List <linux-next@...r.kernel.org>
Subject: Re: linux-next: build warning after merge of the tip tree
Hi all,
On Mon, 20 Sep 2021 18:30:16 +1000 Stephen Rothwell <sfr@...b.auug.org.au> wrote:
>
> On Mon, 20 Sep 2021 14:55:29 +0800 Yafang Shao <laoar.shao@...il.com> wrote:
> >
> > On Mon, Sep 20, 2021 at 9:33 AM Stephen Rothwell <sfr@...b.auug.org.au> wrote:
> > >
> > > After merging the tip tree, today's linux-next build (powerpc_ppc64
> > > defconfig) produced this warning:
> > >
> > > kernel/sched/debug.c: In function 'print_cfs_group_stats':
> > > kernel/sched/debug.c:460:41: warning: unused variable 'stats' [-Wunused-variable]
> > > 460 | struct sched_statistics *stats = __schedstats_from_se(se);
> > > | ^~~~~
> > >
> > > Caused by commit
> > >
> > > cb3e971c435d ("sched: Make struct sched_statistics independent of fair sched class")
> > >
> > > # CONFIG_SCHEDSTATS is not set
> > >
> >
> > Thanks for the report.
> >
> > We have discussed this issue before[1].
> > This warning happens when CONFIG_SCHEDSTATS is not set and
> > schedstat_enabled() is 0, so the whole scope should be not compiled.
> > It seems that we don't need to fix this warning.
> >
> > [1]. https://lore.kernel.org/lkml/20210911082505.115758-1-laoar.shao@gmail.com/
>
> Clearly it will be compiled if CONFIG_SCHEDSTATS is not set as that is
> exactly what this build has ... even sections of code guarded by "if
> (0)" are compiled, they may just not produce any output in the binary.
>
> Also, I do not have W=1 for this build.
>
> If you turned schedstat_val() into a static inline function, then this
> warning would go away. That also means that argument types and return
> values will be better checked.
I take it back, that won't work :-(
> So, please fix this.
Still it needs to be fixed to keep unnecessary warnings out of our builds.
--
Cheers,
Stephen Rothwell
Content of type "application/pgp-signature" skipped
Powered by blists - more mailing lists