[ofa-general] Re: [ewg] [PATCH] IB/ipoib: copy small SKBs in CM mode

[Roland Dreier]

 > +	if (wc->byte_len < SKB_TSHOLD) {
 > +		int dlen = wc->byte_len;
 > +
 > +		small_skb = dev_alloc_skb(dlen + 12);
 > +		if (small_skb) {
 > +			skb_reserve(small_skb, 12);
 > +			skb_copy_from_linear_data(skb, small_skb->data, dlen);
 > +			skb_put(small_skb, dlen);

Just noticed in the original patch: you need calls to
ib_dma_sync_single_for_cpu and ib_dma_sync_single_for_device around this
skb_copy_from_linear_data.

 > > Any reason why we wouldn't want this info in the patch changelog?

 > Not really. If you think it should be there, I'll add it to the
 > changelog along with an explanation to the question bellow.

Yes, definitely we want the performance info.  Imagine if you were
reading the patch in git history -- clearly this justification and
measurement would be very helpful in understanding why the patch was
added, and there's no reason to leave out the useful information you've
already written.

 - R.