| commit | c61f09ed855b5009f816242ce281fd01586d4646 | |
| author | Michael S. Tsirkin <mst@redhat.com> | |
| Mon, 23 Nov 2015 10:48:52 +0000 (23 12:48 +0200) | ||
| committer | Michael S. Tsirkin <mst@redhat.com> | |
| Wed, 25 Nov 2015 11:42:38 +0000 (25 13:42 +0200) | ||
| tree | 4885044e0e45d871d21fec0321a75f8bcd1d15b1 | treesnapshot (tar.gz zip) |
| parent | d39c87d70763f2755d1d7a719817b06f0281fb01 | commitdiff |
vhost-user: clarify start and enable
It seems that we currently have some duplication between
started and enabled states.
The actual reason is that enable is not documented correctly:
what it does is connecting ring to the backend.
This is important for MQ, because a Linux guest expects TX
packets to be completed even if it disables some queues
temporarily.
Cc: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Cc: Victor Kaplansky <victork@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
It seems that we currently have some duplication between
started and enabled states.
The actual reason is that enable is not documented correctly:
what it does is connecting ring to the backend.
This is important for MQ, because a Linux guest expects TX
packets to be completed even if it disables some queues
temporarily.
Cc: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Cc: Victor Kaplansky <victork@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
| docs/specs/vhost-user.txt | diffblobblamehistory |