On Mon, Apr 04, 2016 at 09:12:03AM +0200, Christoph Hellwig wrote: > On Mon, Apr 04, 2016 at 11:47:06AM +1000, Dave Chinner wrote: > > > Patch below fixes the issue, as well as the fact that > > > iomap_page_mkwrite_actor() needs to return the count of bytes > > > "written", not zero on success for iomap_write_segment() to do the > > > right thing on multi-segment writes. > > > > Hmm - this fix then breaks generic/029. Christoph, can you look into > > these issues? > > I've allocated some time for the series this week, I'll also add the > iomap based fiemap implementation from the gfs2 folks while I'm at it. Great! I've been running it here for the past week or so, and on 4k block size filesystems it hasn't caused any regressions. I've been running it with the straight-to-bio writeback paches for XFS as well, and it seems to be playing nicely with that, too. The above problems were seen independent of the witeback patch series, though... Cheers, Dave. -- Dave Chinner david@xxxxxxxxxxxxx _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
