nb: obviously, netscape crashed the day before (1376 div 60 = 22), and
was since then consuming nearly all of the CPU% ... maybe the scheduling
algorithm could be adapted to give lower priority to processes
with high cpu-utilization and make a task-switch occur less frequently.
(the machine is a server with about 100 accounts, so I am
not alone)
/herp
herp@wildsau # ps u
USER PID %CPU %MEM SIZE RSS TTY STAT START TIME COMMAND
herp 5428 0.0 1.3 1148 656 p3 S 01:11 0:00 -bash
herp 5482 0.0 0.4 884 232 p3 R 01:21 0:00 ps u
herp 26138 96.2 4.0 8104 1904 ? R 01:32 1375:13 (netscape)
herp@wildsau # strace -p 26138
--- SIGSEGV (Segmentation fault) ---
getpid() = 26138
kill(26138, SIGUNUSED) = 0
sigreturn() = ? (mask now ~[KILL SEGV STOP])
--- SIGSEGV (Segmentation fault) ---
getpid() = 26138
kill(26138, SIGUNUSED) = 0
sigreturn() = ? (mask now ~[KILL SEGV STOP])
--- SIGSEGV (Segmentation fault) ---
getpid() = 26138
kill(26138, SIGUNUSED) = 0
sigreturn() = ? (mask now ~[KILL SEGV STOP])
--- SIGSEGV (Segmentation fault) ---
getpid() = 26138
kill(26138, SIGUNUSED) = 0
... etc