Am 16.03.2016 um 19:54 hat Max Reitz geschrieben: > This series appears to reinvent itself with every revision. This time, > its main implication is that BBs are no longer automatically treated as > monitor-owned, and that a BB's name is tightly tied to the monitor > reference (it is considered equivalent to that reference).
Thanks, applied to the block branch. > v4: > - blk_hide_on_behalf_of_hmp_drive_del() does a bit more than what > monitor_remove_blk() does; namely, it invokes bdrv_make_anon() on the > BDS. We should probably continue to do so. Probably not, removing the node-name of a BDS just because a BB disappears certainly feels like a bug. Even more so now that node-names are auto-generated and the general assumption is that all BDSes have one. But that's a preexisting problem, so we can fix it on top. Kevin
