- Monday, November 3, 2014

openafs@conference.openafs.org

Monday, November 3, 2014< ^ >

Room Configuration

Room Occupants

GMT+0
[03:41:04] nwf leaves the room
[03:41:27] nwf joins the room
[03:49:21] nwf leaves the room
[03:49:35] nwf joins the room
[04:00:18] nwf leaves the room
[04:00:47] nwf joins the room
[04:47:43] Jeffrey Altman leaves the room
[04:52:22] Jeffrey Altman joins the room
[07:59:31] Simon Wilkinson joins the room
[09:41:03] Simon Wilkinson leaves the room
[12:27:12] Simon Wilkinson joins the room
[13:20:33] Simon Wilkinson leaves the room
[14:28:24] mvita joins the room
[14:52:51] meffie joins the room
[14:55:17] <meffie> is the sol11 x86 builder still failing the git clean step? it should have failed only once.
[15:15:26] <Jeffrey Altman> look at the buildbot history
[15:16:23] <meffie> ok
[15:22:42] <meffie> yes, i spent much of friday evening trying to figure out why builds 5257 to 5260 failed. for some reason that i dont understand, the changes to acinclude.m4 broke how configure generates config.status (with ksh93, not bash).  using git bisect, i found just fixing the whitespace in acinclude.m4 (gerrit 11566) fixes configure. it is very ood. the git clean error happened because i reset the build directory, thinking there was something left over.
[15:27:43] <Jeffrey Altman> there were five builds that failed.  the four from this morning succeeded.  http://buildbot.openafs.org:8010/tgrid?length=40
[15:28:44] deason joins the room
[15:30:46] <meffie> yes, the first 4 were from the ksh93 error, the fifth was because git clean failed when i reset the build directory.
[15:32:29] <meffie> the 4 from this morning are on the 1.6.x branch, so dont have the whitespace changes in acinclude.m4 to trigger the ksh93 bug.
[15:38:06] Simon Wilkinson joins the room
[16:04:30] meffie leaves the room
[16:15:17] meffie joins the room
[16:16:59] kaduk joins the room
[16:19:03] <kaduk> Simon Wilkinson: did you see my gss-openssh mail?  "Corrupted MAC on input" is not exactly something that sounds easy to diagnose...
[16:21:27] <Simon Wilkinson> I haven't had a chance to think about it yet. My quick question would be - is the server using a different KEX than the client?
[16:23:05] <kaduk> It's possible; I think the KEX from my original testing is not the one that shows up in my recent testing.
It looked like it was the third call to mac_compute() which was returning garbage
[16:27:48] meffie leaves the room
[17:46:27] meffie joins the room
[19:16:16] nwf leaves the room
[19:16:38] nwf joins the room
[19:27:06] nwf leaves the room
[19:27:22] nwf joins the room
[20:10:21] meffie leaves the room
[20:55:14] meffie joins the room
[20:55:50] meffie leaves the room
[20:55:50] meffie joins the room
[21:29:32] meffie leaves the room
[22:27:15] ballbery leaves the room
[22:37:41] ballbery joins the room
[22:47:19] kaduk leaves the room
[22:51:06] ballbery leaves the room
[22:51:08] ballbery joins the room
[23:54:28] kaduk joins the room