On Mon, Jan 21, 2019 at 12:42:56PM -0500, Brian Foster wrote: > > It protects against the case where we migrate a COW fork mapping to the > > data fork, which is not protected by the page lock. But I guess the > > check warrants a comment and an assert. > > > > Yeah, probably. It's not really clear to me what that means. > > > > 2. If so, then it also seems that the whole "eof:" thing in > > > xfs_map_blocks() should never happen for data forks. If that's the case, > > > the use of the eof label there seems gratuitous. > > > > Let me try with asserts enabled. > > > > Ok, thanks. So, the nimaps == 0 case hits even for the data fork when running xfs/442 on a 1k block size file system. That test has generally been a fun source of bugs in the always_cow series.
