As the documentation for netif_trans_update() says, netdev_start_xmit()
already updates the last-tx time after every good xmit. So don't
duplicate that effort.

One odd case is that qeth_flush_buffers() also gets called from our
TX completion handler, to flush out any partially filled buffer when
we switch the queue to non-packing mode. But as the TX completion
handler will _always_ wake the txq, we don't have to worry about
the TX watchdog there.

Signed-off-by: Julian Wiedmann <j...@linux.ibm.com>
---
 drivers/s390/net/qeth_core_main.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/s390/net/qeth_core_main.c 
b/drivers/s390/net/qeth_core_main.c
index ac592d610c5e..9e495df742cb 100644
--- a/drivers/s390/net/qeth_core_main.c
+++ b/drivers/s390/net/qeth_core_main.c
@@ -3371,7 +3371,6 @@ static void qeth_flush_buffers(struct qeth_qdio_out_q 
*queue, int index,
        }
 
        QETH_TXQ_STAT_ADD(queue, bufs, count);
-       netif_trans_update(queue->card->dev);
        qdio_flags = QDIO_FLAG_SYNC_OUTPUT;
        if (atomic_read(&queue->set_pci_flags_count))
                qdio_flags |= QDIO_FLAG_PCI_OUT;
-- 
2.16.4

Reply via email to