distributor: fix crash in flush

Occasionally, the distributor single-packet-at-a-time mode will
segfault because it inadvertently calls some burst mode code when
flushing packets.

This patch ensures that only the v20 (single mode) codepath is used, and
returns without falling through to the burst mode code.

Fixes: 775003ad2f96 ("distributor: add new burst-capable library")

Signed-off-by: David Hunt <david.hunt@intel.com>
Tested-by: Yong Liu <yong.liu@intel.com>
Acked-by: Bruce Richardson <bruce.richardson@intel.com>
This commit is contained in:
David Hunt 2017-04-14 09:59:53 +01:00 committed by Thomas Monjalon
parent 284c908cc5
commit 917f49cfd0

View File

@ -555,7 +555,7 @@ total_outstanding(const struct rte_distributor *d)
int
rte_distributor_flush_v1705(struct rte_distributor *d)
{
const unsigned int flushed = total_outstanding(d);
unsigned int flushed;
unsigned int wkr;
if (d->alg_type == RTE_DIST_ALG_SINGLE) {
@ -563,6 +563,8 @@ rte_distributor_flush_v1705(struct rte_distributor *d)
return rte_distributor_flush_v20(d->d_v20);
}
flushed = total_outstanding(d);
while (total_outstanding(d) > 0)
rte_distributor_process(d, NULL, 0);
@ -590,6 +592,7 @@ rte_distributor_clear_returns_v1705(struct rte_distributor *d)
if (d->alg_type == RTE_DIST_ALG_SINGLE) {
/* Call the old API */
rte_distributor_clear_returns_v20(d->d_v20);
return;
}
/* throw away returns, so workers can exit */