On Tue, 22 Feb 2022 11:02:24 +0000
Reshma Pattan <reshma.pat...@intel.com> wrote:
> Check lcore id value is not the maximum core supported.
> Using lcore id without this check might cause
> out of bound access inside the rte_eal_wait_lcore.
>
> Coverity issue: 375841
> Fixes: b2854d5317e8 ("app/pdump: support multi-core capture")
> Cc: vipin.vargh...@intel.com
> Cc: sta...@dpdk.org
>
> Signed-off-by: Reshma Pattan <reshma.pat...@intel.com>
> ---
> v3: add new function to get next core id and validate it.
> ---
> app/pdump/main.c | 13 +++++++++++--
> 1 file changed, 11 insertions(+), 2 deletions(-)
>
> diff --git a/app/pdump/main.c b/app/pdump/main.c
> index 04a38e8911..e4e62811c9 100644
> --- a/app/pdump/main.c
> +++ b/app/pdump/main.c
> @@ -900,6 +900,15 @@ dump_packets_core(void *arg)
> return 0;
> }
>
> +static inline void
> +get_next_core(uint32_t *lcore_id)
> +{
> + *lcore_id = rte_get_next_lcore(*lcore_id, 1, 0);
> + if (*lcore_id == RTE_MAX_LCORE)
> + rte_exit(EXIT_FAILURE,
> + "Max core limit %u reached for packet capture",
> *lcore_id);
> +}
> +
This looks good, can I make a some suggestions.
Since function either exits or returns a good value, I would prefer
that it returned the lcore id. Avoiding call by reference if possible.
Also, the lcore is currently typed unsigned int in rte_lcore.h
therefore use that type?
Inline is certainly unnecessary here, not critical path and compiler
is likely to do it anyway.
Also, DPDK uses lcore for most places (rather than core) so use that name.
Result is:
static unsigned int
get_next_lcore(unsigned int lcore)
{
lcore = rte_get_next_lcore(lcore, 1, 0);
if (lcore >= RTE_MAX_LCORE)
"Max core limit %u reached for packet capture", lcore);
return lcore;
}
> static inline void
> dump_packets(void)
> {
> @@ -930,12 +939,12 @@ dump_packets(void)
> return;
> }
>
> - lcore_id = rte_get_next_lcore(lcore_id, 1, 0);
> + get_next_core(&lcore_id);
>
> for (i = 0; i < num_tuples; i++) {
> rte_eal_remote_launch(dump_packets_core,
> &pdump_t[i], lcore_id);
> - lcore_id = rte_get_next_lcore(lcore_id, 1, 0);
> + get_next_core(&lcore_id);
>
> if (rte_eal_wait_lcore(lcore_id) < 0)
> rte_exit(EXIT_FAILURE, "failed to wait\n");
