<html><body>
<p><tt>Hello Eli,</tt><br>
<br>
<tt>> > In this case, how many tx drop packets from ifconfig output? Should we<br>
> > see ifconfig tx drop packets + tx successfully transmit packets close<br>
> > to netperf packets? <br>
> That's right.<br>
</tt><br>
<tt>I am looking at ipoib_cm_handle_tx_wc(), there is no tx drop packets increased in this situation, so tx transmit packets should be around netperf send packets.</tt><br>
<br>
<tt>void ipoib_cm_handle_tx_wc(struct net_device *dev, struct ib_wc *wc)</tt><br>
<tt>{</tt><br>
<tt> ...</tt><br>
<br>
<tt> tx_req = &tx->tx_ring[wr_id];</tt><br>
<br>
<tt> ib_dma_unmap_single(priv->ca, tx_req->mapping[0], tx_req->skb->len, DMA_TO_DEVICE);</tt><br>
<br>
<tt> /* FIXME: is this right? Shouldn't we only increment on success? */</tt><br>
<tt> ++dev->stats.tx_packets;</tt><br>
<tt> dev->stats.tx_bytes += tx_req->skb->len;</tt><br>
<tt>...</tt><br>
<tt>}</tt><br>
<tt> <br>
> > Any TCP STREAM test results to share here?<br>
> TCP won't demonstrate the problem since it uses Nagle's algorithm to<br>
> aggregate data into full sized packets.<br>
</tt><br>
<tt>So when hitting this RNR retry, the error status return was flush err, so the packets were silently dropped instead of "failed cm send event" and clear the interface up flag? </tt><br>
<br>
<tt>Please correct me if wrong.</tt><br>
<br>
<tt>thanks</tt><br>
<tt>Shirley</tt></body></html>