netmap: set IFCAP_NETMAP in if_capabilities

[vmaffione]

Summary

On netmap_attach(), we need to set the IFCAP_NETMAP flag in ifp->if_capabilities.
Also remove the redundant code in cxgbe(4) that does the same.

Test Plan

Integration tests and unit tests ran successfully.

[np]

The cxgbe part of this diff reverts r309725. The commit message for r309725 indicates the changes were made to catch up with r307394, which was the a netmap update. Can you confirm that various releases of netmap have had different behavior regarding if_capabilities? If netmap has finally settled on handling the caps this way then I have no objection to one last catch-up in the driver.

[vmaffione]

The truth is that netmap not setting the IFCAP_NETMAP flag in ifp->if_capabilities after r307394 is a mistake that we did not notice so far.
It got introduced by refactoring, 3 years ago (see https://github.com/luigirizzo/netmap/commit/5d0796f93a1107eb14422c7b8ea416f7fd750a2e).
Sorry for that, it was unintended.
I just happened to notice that and fix it.

[np]

In D17987#389775, @v.maffione_gmail.com wrote:

The truth is that netmap not setting the IFCAP_NETMAP flag in ifp->if_capabilities after r307394 is a mistake that we did not notice so far.
It got introduced by refactoring, 3 years ago (see https://github.com/luigirizzo/netmap/commit/5d0796f93a1107eb14422c7b8ea416f7fd750a2e).
Sorry for that, it was unintended.
I just happened to notice that and fix it.

Ok, in that case can you revert r309725 all by itself as a separate commit? I think that will leave a better trail in svn history.

[vmaffione]

Sounds good, that's a better approach. I'll do it as soon as @gnn or @hrs approve.

[gnn]

I approve the plan laid out in the comments with np@