> On Sat, May 02, 2020 at 04:24:19PM +0300, Dan Carpenter wrote:
> > Hello Wu Hao,
> >
> > The patch 724142f8c42a: "fpga: dfl: fme: add performance reporting
> > support" from Apr 27, 2020, leads to the following static checker
> > warning:
> >
> > drivers/fpga/dfl-fme-perf.c:948 fme_perf_offline_cpu()
> > error: potential NULL dereference 'priv'.
> >
> > drivers/fpga/dfl-fme-perf.c
> > 941 static int fme_perf_offline_cpu(unsigned int cpu, struct hlist_node
> *node)
> > 942 {
> > 943 struct fme_perf_priv *priv;
> > 944 int target;
> > 945
> > 946 priv = hlist_entry_safe(node, struct fme_perf_priv, node);
> > ^^^^^
> > It doesn't make sense to use the _safe() version if we're not going to
> > check for NULL.
>
> Thanks for catching this. Will fix it.
Yes, it's not necessary to use _safe version here.
Thanks Dan for reporting this, and thanks Yilun for covering this when
I was on leave.
Hao
>
> Regards,
> Yilun.
>
> >
> > 947
> > 948 if (cpu != priv->cpu)
> > 949 return 0;
> > 950
> > 951 target = cpumask_any_but(cpu_online_mask, cpu);
> > 952 if (target >= nr_cpu_ids)
> > 953 return 0;
> > 954
> > 955 priv->cpu = target;
> > 956 return 0;
> > 957 }
> >
> > regards,
> > dan carpenter
