Page MenuHomeFreeBSD

Add M_FLOWID to M_COPYFLAGS
ClosedPublic

Authored by bryanv on Oct 8 2014, 5:02 AM.
Tags
None
Referenced Files
F103839080: D914.diff
Sat, Nov 30, 2:42 AM
Unknown Object (File)
Sep 18 2024, 12:57 AM
Unknown Object (File)
Sep 17 2024, 4:17 PM
Unknown Object (File)
Sep 17 2024, 8:00 AM
Unknown Object (File)
Sep 9 2024, 2:53 AM
Unknown Object (File)
Sep 7 2024, 12:00 PM
Unknown Object (File)
Sep 1 2024, 10:33 AM
Unknown Object (File)
Aug 13 2024, 10:02 PM
Subscribers

Details

Reviewers
adrian
Group Reviewers
network
Summary

I'm sorting through changes made in our work tree, and came across this commit
that seems just to be an oversight. Shouldn't M_FLOWID be included in M_COPYFLAGS
so that places like m_{move,dup}_pkthdr() carry over the flag in situations like
when we have to prepend a new mbuf pkthdr? The actual flowid value gets copied
over as a part of copying the pkthdr, but the corresponding flag is not set.

Or does somebody have some historical context why M_FLOWID was omitted?

Diff Detail

Repository
rS FreeBSD src repository - subversion
Lint
No Lint Coverage
Unit
No Test Coverage

Event Timeline

bryanv retitled this revision from to Add M_FLOWID to M_COPYFLAGS.
bryanv updated this object.
bryanv edited the test plan for this revision. (Show Details)
bryanv added a reviewer: network.
adrian added a reviewer: adrian.
adrian added a subscriber: adrian.

Looks good to me. Good catch!

This revision is now accepted and ready to land.Oct 8 2014, 5:20 AM

What if the packet was encapsulated? Does the flag get copied over to the inner packet? If it does, then that's a bug because the flowid needs to be recalculated.

In D914#9, @rpaulo wrote:

What if the packet was encapsulated? Does the flag get copied over to the inner packet? If it does, then that's a bug because the flowid needs to be recalculated.

I think that should be up to the caller to recalculate handle that. The RSS values were being propagated over so at least we are consistent now. Our handling of flow information with regards to encapsulation isn't the greatest, and something that I plan to work on to make vxlan perform better.