Hello Eli,
> > In this case, how many tx drop packets from ifconfig output? Should we
> > see ifconfig tx drop packets + tx successfully transmit packets close
> > to netperf packets?
> That's right.
I am looking at ipoib_cm_handle_tx_wc(), there is no tx drop packets
increased in this situat
> In this case, how many tx drop packets from ifconfig output? Should we
> see ifconfig tx drop packets + tx successfully transmit packets close
> to netperf packets?
That's right.
>
> Any TCP STREAM test results to share here?
TCP won't demonstrate the problem since it uses Nagle's algorithm t
On Tue, 2008-05-27 at 13:00 -0700, Roland Dreier wrote:
> I changed this to a build bug, to avoid waiting until runtime to notice
> this problem:
>
> + /*
> +* When copying small received packets, we only copy from the
> +* linear data part of the SKB, so we rely on this cond
thanks, applied for 2.6.27...
> +/*
> + * we rely on this condition when copying small skbs and we
> + * pass ownership of the first fragment only.
> + */
> +if (SKB_TSHOLD > IPOIB_CM_HEAD_SIZE) {
> +printk("%s: SKB_TSHOLD(%d) must not be larger then %d\n",
>