On Fri, Apr 07, 2017 at 10:42:20AM -0400, Ian D. Leroux wrote: > On Wed, Apr 5, 2017, at 08:01, Patrick Welche wrote: > > It seems something recentish broke FFS, and WAPBL is blameless... > > I won't be much use to you, but could you file a PR or keep current- > users@ advised of any progress on this? I'd rather not update my main > machine until this is resolved ...
I built a 30 March 2017 10:00 UTC kernel on morden.netbsd.org (so the src tree would be OK). Using that kernel on the affected machine, a cvs checkout appears to be fine, and 4 simultaneous untarrings of pkgsrc.tar.xz produced identical results. Only one strange observation during all that: scp'ing a 24Mb kernel from 2 Apr -current laptop to the machine running the what looks like a good 30 March kernel resulted in a 9.1Mb kernel. 2nd scp worked. So, there are still many possibilities, and it might just be my box... (Running the /usr/tests after generating the bad cvs checkouts earlier, the tests didn't fail in a different way to other setups.) Cheers, Patrick
