[00:14:16] --- jaltman/FrogsLeap has left: Replaced by new connection [00:14:17] --- jaltman/FrogsLeap has become available [00:18:06] --- jaltman has left: Disconnected [00:19:36] --- jaltman has become available [04:57:17] --- mdionne has become available [09:15:04] --- Russ has become available [10:17:54] kaduk_mitedu: so, the symptoms are that some rx replies seem to get ignored [10:56:37] --- pod has left [10:57:43] --- jaltman/FrogsLeap has left: Disconnected [10:57:56] --- jaltman/FrogsLeap has become available [11:03:04] it doesn't usually happen in my local cell, it's mostly first stalling on cells across the WAN [11:05:12] --- jaltman/FrogsLeap has left: Disconnected [11:07:16] --- jaltman has left: Disconnected [11:57:31] jakllsch: They get ignored, but later ones do (so it is not a deadlock)? (Also, 'kaduk' and 'kaduk_mitedu' are equally likely to grab my attention if that matters to you) [12:02:08] no, it's a deadlock [12:08:19] Ah. When this happened in FreeBSD, it was because getevent() was dropping the glock when it allocated memory. [12:22:18] --- jaltman/FrogsLeap has become available [13:14:48] --- deason has become available [13:30:28] --- jaltman/FrogsLeap has left: Disconnected [13:41:39] --- jaltman/FrogsLeap has become available [14:08:56] --- jaltman/FrogsLeap has left: Disconnected [14:20:07] --- jaltman/FrogsLeap has become available [15:04:14] --- jaltman/FrogsLeap has left: Disconnected [15:12:48] --- jaltman/FrogsLeap has become available [17:27:29] --- mdionne has left [21:19:55] --- jaltman has become available [23:53:52] --- Russ has left: Disconnected