Bruce Evans
fbbeeb6cd6
Uninlined CURSIG() and unpolluted <sys/signalvar.h>. CURSIG() had become
...
very bloated, first with 128-bit sigset_t's, then with locking in the
SMP case, then with locking in all cases. The space bloat was probably
also time bloat, partly because the fast path through CURSIG() was
pessimized by the sigset_t changes. This change speeds up lmbench's
pipe-based latency benchmark by 4% on a Celeron. <sys/signalvar.h>
had become very polluted to support the bloat.
2000-09-17 14:28:33 +00:00
..
2000-09-16 18:28:24 +00:00
2000-09-16 18:55:05 +00:00
2000-09-16 21:04:49 +00:00
2000-09-14 21:38:44 +00:00
2000-05-03 07:31:38 +00:00
2000-09-09 14:35:35 +00:00
2000-09-16 20:45:11 +00:00
2000-09-16 18:57:18 +00:00
2000-07-04 16:35:15 +00:00
2000-07-04 11:25:35 +00:00
2000-09-17 13:26:25 +00:00
2000-09-16 12:06:58 +00:00
2000-09-02 19:17:34 +00:00
2000-09-12 17:10:39 +00:00
2000-09-13 18:33:25 +00:00
2000-09-16 19:04:24 +00:00
2000-09-13 18:33:25 +00:00
2000-09-07 01:33:02 +00:00
2000-09-17 14:28:33 +00:00
2000-09-11 19:34:04 +00:00
2000-09-11 05:45:06 +00:00
2000-09-13 03:20:35 +00:00
2000-08-29 14:45:49 +00:00
2000-09-16 14:17:15 +00:00
2000-05-24 21:16:56 +00:00
2000-05-24 21:16:56 +00:00
2000-09-07 01:33:02 +00:00
2000-09-17 13:50:12 +00:00
2000-09-15 17:21:35 +00:00
2000-05-01 20:32:07 +00:00
2000-08-28 08:24:46 +00:00
2000-05-26 02:09:24 +00:00
2000-07-04 11:25:35 +00:00
2000-02-13 03:32:07 +00:00
2000-09-07 01:33:02 +00:00
2000-09-05 22:29:36 +00:00
2000-09-07 01:33:02 +00:00
2000-08-29 14:45:49 +00:00
2000-07-04 11:25:35 +00:00
2000-09-15 05:47:56 +00:00
2000-09-13 18:33:25 +00:00
2000-09-17 13:26:25 +00:00
2000-05-01 20:32:07 +00:00
2000-09-14 20:15:16 +00:00
2000-09-17 14:28:33 +00:00
2000-06-27 19:09:55 +00:00
2000-09-17 07:26:42 +00:00
2000-09-15 22:00:23 +00:00
2000-08-24 18:53:09 +00:00