On Sat, 26 Feb 2022 09:31:37 -0800, Stephen Hemminger wrote:
> > Add parentheses to 'if' clause, otherwise will enlarged the
> > chance of error return.
> >
> > Fixes: 44c730b0e37971 ("sched: add PIE based congestion management")
> >
> > Signed-off-by: Weiguo Li <liw...@foxmail.com>
> > ---
> > lib/sched/rte_pie.c | 6 +++---
> > 1 file changed, 3 insertions(+), 3 deletions(-)
> >
> > diff --git a/lib/sched/rte_pie.c b/lib/sched/rte_pie.c
> > index cdb7bab697..51df403a25 100644
> > --- a/lib/sched/rte_pie.c
> > +++ b/lib/sched/rte_pie.c
> > @@ -18,10 +18,10 @@ rte_pie_rt_data_init(struct rte_pie *pie)
> > /* Allocate memory to use the PIE data structure */
> > pie = rte_malloc(NULL, sizeof(struct rte_pie), 0);
> >
> > - if (pie == NULL)
> > + if (pie == NULL) {
> > RTE_LOG(ERR, SCHED, "%s: Memory allocation fails\n",
> > __func__);
> > -
> > - return -1;
> > + return -1;
> > + }
> > }
> >
> > pie->active = 0;
>
> This will make the test in test_pie.c fail.
>
> The concept of passing NULL to the routine and expecting allocation
> is bad idea because the allocated structure is never initialized.
>
> Since rte_pie_rt_data_init(NULL) always returned -1.
> It would make more sense to take out the rte_malloc().
> And document it.
>
> P.s: the routing should return a negative rte_errno instead of -1
> as well.
>
Hi Stephen,
The 'rte_malloc' and null check is really misleading at the first sight...
Thanks for your suggestion!
-Weiguo
