Re: [PATCH] sched/vtime: Fix an unitialized variable warning

[Peter Zijlstra <peterz@infradead.org>]

On Fri, Mar 27, 2020 at 10:43:34PM +0100, Borislav Petkov wrote:
> Hi dude,
> 
> right before I was going to send the trivial, shut-up-gcc variant, I
> thought that maybe we should do this instead.
> 
> Thoughts?
> 
> ---
> Fix:
> 
>   kernel/sched/cputime.c: In function ‘kcpustat_field’:
>   kernel/sched/cputime.c:1007:6: warning: ‘val’ may be used \
> 	  uninitialized in this function [-Wmaybe-uninitialized]
>    1007 |  u64 val;
>         | ^~~
> 
> because gcc can't see that val is used only when err is 0.
> 
> Signed-off-by: Borislav Petkov <bp@suse.de>
> ---
>  kernel/sched/cputime.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
> index dac9104d126f..ff9435dee1df 100644
> --- a/kernel/sched/cputime.c
> +++ b/kernel/sched/cputime.c
> @@ -1003,12 +1003,12 @@ u64 kcpustat_field(struct kernel_cpustat *kcpustat,
>  		   enum cpu_usage_stat usage, int cpu)
>  {
>  	u64 *cpustat = kcpustat->cpustat;
> +	u64 val = cpustat[usage];
>  	struct rq *rq;
> -	u64 val;
>  	int err;
>  
>  	if (!vtime_accounting_enabled_cpu(cpu))
> -		return cpustat[usage];
> +		return val;
>  
>  	rq = cpu_rq(cpu);


Hurph.. this might result in an unconditional load (and extra
cache-miss) for the vtime_accounting_enabled_cpu() case.

I suspesct the =0 this would be better. Stupid stupid compiler!