On Fri, Jul 8, 2016 at 9:19 AM, Pravin B Shelar <[email protected]> wrote:
> diff --git a/datapath/linux/compat/gso.h b/datapath/linux/compat/gso.h
> index 0f2b09a..5441973 100644
> --- a/datapath/linux/compat/gso.h
> +++ b/datapath/linux/compat/gso.h
> @@ -23,6 +23,11 @@ struct ovs_gso_cb {
> #ifndef HAVE_INNER_NETWORK_HEADER
> unsigned int inner_network_header;
> #endif
> +#ifndef HAVE_NDO_FILL_METADATA_DST
> + /* Keep original tunnel info during userspace action execution. */
> + struct metadata_dst *tmp;
Can you give this a little bit more descriptive name? I always find
'tmp' a little too vague.
> +#define SKB_SETUP_FILL_METADATA_DST(skb) { \
> + struct metadata_dst *new_md_dst; \
> + struct metadata_dst *md_dst; \
> + int md_size; \
> + \
> + SKB_RESTORE_FILL_METADATA_DST(skb); \
> + new_md_dst = kmalloc(sizeof(struct metadata_dst) + 256, GFP_ATOMIC); \
As I mentioned in the other patch, I think it would be good to zero
out this newly allocated memory to avoid any possible subtle problems.
But otherwise I am happy with this:
Acked-by: Jesse Gross <[email protected]>
I think that's everything on this series at this point! Thanks for all
of the revisions. (I don't feel the need to take another look at any
of the changes that I just noted in either of these two patches.)
_______________________________________________
dev mailing list
[email protected]
http://openvswitch.org/mailman/listinfo/dev
