[PATCH] app: fix mbuf_autotest in case of defined RTE_LIBRTE_MBUF_DEBUG

[Olivier Matz]

Hello,

On Wed, Mar 15, 2023 at 11:29:50AM +0100, David Marchand wrote:
> Hello,
> 
> On Mon, Mar 13, 2023 at 5:57 PM Pavel Ivashchenko
> <pivashchenko at nfware.com> wrote:
>
> app: fix mbuf_autotest in case of defined RTE_LIBRTE_MBUF_DEBUG

I suggest this title instead:

test/mbuf: fix mbuf autotest when mbuf debug is enabled

> >
> > How to reproduce:
> >
> > 1. Define RTE_LIBRTE_MBUF_DEBUG
> > 2. MALLOC_PERTURB_=178 DPDK_TEST=mbuf_autotest gdb --args obj-x86_64-linux-gnu/app/test/dpdk-test --file-prefix=mbuf_autotest
> >
> >    PANIC in rte_mbuf_sanity_check():
> >    bad pkt_len
> >
> >    ...
> >    #6  0x00007ffff7d3d4cc in rte_mbuf_sanity_check (m=m at entry=0x17f8c3400, is_header=is_header at entry=1) at ../lib/mbuf/rte_mbuf.c:384
> >    #7  0x0000555555653d57 in rte_pktmbuf_free (m=0x17f8c3400) at ../lib/mbuf/rte_mbuf.h:1385
> >    #8  0x000055555565c7a6 in test_nb_segs_and_next_reset () at ../app/test/test_mbuf.c:2752
> >    #9  test_mbuf () at ../app/test/test_mbuf.c:2967
> >    ...
> >
> >    (gdb) frame 6
> >    #6  0x00007ffff7d3d4cc in rte_mbuf_sanity_check (m=m at entry=0x17f8c3400, is_header=is_header at entry=1) at ../lib/mbuf/rte_mbuf.c:384
> >    384                  rte_panic("%s\n", reason);
> >    (gdb) p/d m->pkt_len
> >    $4 = 1500
> >
> > Fixes: efc6f9104c80 ("mbuf: fix reset on mbuf free")
> > Cc: stable at dpdk.org
> >
> > Signed-off-by: Pavel Ivashchenko <pivashchenko at nfware.com>
> 
> LGTM, thanks.
> Just a small comment.
> 
> 
> > ---
> >  app/test/test_mbuf.c | 1 +
> >  1 file changed, 1 insertion(+)
> >
> > diff --git a/app/test/test_mbuf.c b/app/test/test_mbuf.c
> > index 6cbb03b0af..d471a23805 100644
> > --- a/app/test/test_mbuf.c
> > +++ b/app/test/test_mbuf.c
> > @@ -2744,6 +2744,7 @@ test_nb_segs_and_next_reset(void)
> >
> >         /* split m0 chain in two, between m1 and m2 */
> >         m0->nb_segs = 2;
> > +       m0->pkt_len -= 500;
> 
> m0->pkt_len -= m2->data_len seems more readable and robust to me.
> 
> Opinions?

Even if the 500 is hardcoded right above, this looks indeed better.

Or this seems fine too:
m0->pkt_len = m0->data_len + m1->data_len;

Thanks,
Olivier

> 
> 
> >         m1->next = NULL;
> >         m2->nb_segs = 1;
> >
> 
> 
> -- 
> David Marchand
>