From: Ido Segev <id...@amazon.com>
As the refill called as part of ena_start(), we end up the refill
progress with stuck buffers at the caller core cache.
Calling to flush the cache results with invalidate this cache and free
those stuck buffers.
Fixes: 1173fca25af9 ("ena: add polling-mode driver")
Cc: sta...@dpdk.org
Signed-off-by: Ido Segev <id...@amazon.com>
Reviewed-by: Michal Krawczyk <m...@semihalf.com>
Reviewed-by: Igor Chauskin <igo...@amazon.com>
---
drivers/net/ena/ena_ethdev.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/net/ena/ena_ethdev.c b/drivers/net/ena/ena_ethdev.c
index 20ff3653c6..516e244295 100644
--- a/drivers/net/ena/ena_ethdev.c
+++ b/drivers/net/ena/ena_ethdev.c
@@ -1246,6 +1246,10 @@ static int ena_queue_start(struct ena_ring *ring)
PMD_INIT_LOG(ERR, "Failed to populate rx ring !");
return ENA_COM_FAULT;
}
+ /* Flush per-core RX buffers pools cache as they can be used on other
+ * cores as well.
+ */
+ rte_mempool_cache_flush(NULL, ring->mb_pool);
return 0;
}
--
2.25.1
