> > -   if (src_addr)
> > -           memcpy(&req->src_addr, src_addr, ip_addr_size(src_addr));
> > +   if (src_addr->sa_family != dst_addr->sa_family)
> > +           return -EINVAL;
>
>The old code seemed to allow for the possibility of src_addr being
>NULL.  There only seems to be one in-tree caller of rdma_resolve_ip,
>which does pass in a src_addr, but the API definition in the header says
>
> * @src_addr: An optional source address to use in the resolution.  If a
>
>so it seems passing in NULL would be allowed?

I forgot about the header file comments.  (I only though about rdma_resolve_addr
supporting a NULL src_addr.)  I took the check out when I noticed that src_addr
was always coming in non-NULL.  The address itself doesn't need to be set beyond
the address family, so it's still optional in that regard.  (A pretty weak
argument, but it's all I have. :)

I can go either way here - either update the comments in the header file, or
allow passing in NULL.  Do you or does anyone else have a preference?

- Sean

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to