[00:22:31] Locking suggestions from kib have resulted in a patch that, using supported interfaces, allows parallel buildworld to succeed. I still need to clean it up, though -- something for tomorrow. [00:56:24] --- Russ has left: Disconnected [01:47:09] --- Simon Wilkinson has become available [02:31:38] --- reuteras has become available [02:52:07] --- Simon Wilkinson has left [03:21:19] --- Simon Wilkinson has become available [04:45:28] --- meffie has become available [04:54:50] --- Simon Wilkinson has left [05:23:22] --- reuteras has left [06:05:19] --- Simon Wilkinson has become available [06:33:00] --- abo has left [06:56:09] --- abo has become available [07:09:36] --- edgester has become available [07:10:03] yay! buildbot is working correctly again [07:10:20] I'm still testing my script to re-inject missed changes, though. [07:11:22] --- edgester has left [07:13:32] We seem to have had multiple builds of a load of recent changes ? [07:23:49] --- deason has become available [07:52:52] --- jaltman has left: Disconnected [07:58:01] --- Simon Wilkinson has left [08:08:29] --- jaltman has become available [08:47:14] --- edgester has become available [08:47:55] that's correct. my re-injection script seems to be re-adding them. I've disabled the re-injection until I can fix it. [08:48:37] the good news is that everything has been attempted to be built at least once ;) [08:49:14] --- edgester has left [08:50:00] Hm. I don't remember getting buildbot mail for my one-liner that Derrick insta-merged. [08:52:22] (3186) [08:58:01] I'm assuming jason's injection thing wasn't looking at merged commits [10:22:27] --- Simon Wilkinson has become available [10:25:42] --- edgester has become available [10:28:12] my change re-injection script only looks at unmerged and un-verified commits [10:28:44] er, rather commits that are both unmerged and unverified [10:29:48] my current problem is that I have to exclude commits where buildbot has already commented [10:30:17] --- edgester has left [11:08:08] --- rra has become available [12:01:09] hmm, a little odd thing now that we're using osi_Assert in lots of places... it tends to say "Fatal Rx Error" if it fails; I assume we want to change that, since it's being used in a lot of non-rx places? [12:02:02] i changed it once. i bet at some point i lost the change [12:09:30] --- meffie has left [12:09:41] Derrick, that built cleanly on your 8.X VM? [12:09:55] --- meffie has become available [12:10:16] (I sanatized/conditionalized the patch a bit after the long run on my test box.) [12:11:13] it build on my 8.(something) vm which is no longer pure [12:11:30] i should have tried it on a trueblood machine, not the one which has een mongrelized testing axe drivers [12:11:32] Cool. [12:12:11] I would be more concerned about build errors on 7.X, but I don't think we know anyone running that. Runtime errors, now ... [12:23:59] --- deason has left [12:39:23] This would be a whole lot easier if the majority of our source tree didn't end up including rx_packet.h [12:50:41] builds clean on a nonmongrelized VM too btw [12:52:05] I should really go through and make the vnode locking wrappers (thread vs. no thread) consistently named throughout the various files. [13:49:05] --- Simon Wilkinson has left [14:17:30] --- jaltman has left: Disconnected [14:29:55] --- edgester has become available [15:43:43] --- meffie has left [16:12:32] --- Simon Wilkinson has become available [17:22:51] --- Simon Wilkinson has left [17:37:25] --- rra has left: Disconnected [18:02:38] --- Russ has become available [19:02:23] --- jaltman has become available [19:43:48] yay, buildbot is caught up [19:44:30] my change injection script is working well so far [19:44:36] * edgester crosses fingers [19:46:39] --- edgester has left [21:31:57] Hm, I suppose I should worry about this memory leak at some point. [21:33:59] --- jaltman has left: Replaced by new connection [21:34:00] --- jaltman has become available [21:35:40] Also, with this assert stuff: In file included from /usr/ports/net/openafs-devel/work/openafs/include/ubik.h:60, from admin_tools.c:33: /usr/ports/net/openafs-devel/work/openafs/include/afs/afs_assert.h:15:1: warning: "assert" redefined In file included from admin_tools.c:20: /usr/ports/net/openafs-devel/work/openafs/include/afs/debug.h:104:1: warning: this is the location of the previous definition [21:44:40] What does error_table.y want timeb.h for? [21:48:17] --- jaltman has left: Replaced by new connection [21:48:19] --- jaltman has become available [22:09:16] --- jaltman has left: Replaced by new connection [22:09:17] --- jaltman has become available [22:10:39] --- jaltman has left: Disconnected [22:10:49] --- jaltman has become available [22:11:37] --- jaltman has left: Disconnected