15/04/2021 14:04, Kevin Traynor:
> On 15/04/2021 01:52, Min Hu (Connor) wrote:
> > + if (iter == NULL) {
> > + RTE_ETHDEV_LOG(ERR, "Failed to iterator init for NULL\n");
>
> "Failed to init iterator for NULL iterator\n"
The word "Failed" looks weird in these checks.
What about "Cannot"?
Example: "Cannot init NULL iterator"
> > + if (devargs_str == NULL) {
> > + RTE_ETHDEV_LOG(ERR, "Failed to iterate matching NULL\n");
>
> "Failed to init iterator for NULL devargs\n"
"Cannot init iterator for NULL devargs"
> > + if (owner == NULL) {
> > + RTE_ETHDEV_LOG(ERR, "Failed to get ethdev port %u owner by
> > NULL\n",
> > + port_id);
> > + return -EINVAL;
> > + }
>
> This fn uses both %u and %"PRIu16" for port_id
I don't see the benefit of PRIu16.
> > + if (str == NULL) {
> > + RTE_ETHDEV_LOG(ERR, "Failed to convert link to NULL\n");
>
> "Failed to convert link to NULL string\n"
"Cannot convert link in NULL string"
> > + if (eth_link == NULL) {
> > + RTE_ETHDEV_LOG(ERR, "Failed to convert NULL to string\n");
>
> "Failed to convert NULL link to string\n"
"Cannot convert NULL link"
