Messages in this thread |  | | Date | Wed, 19 Sep 2001 21:40:08 -0400 (EDT) | From | Ignacio Vazquez-Abrams <> | Subject | Re: [PATCH] Preemption Latency Measurement Tool |
| |
I'm posting this in the hope that someone will find it useful.
Worst 20 latency times of 4647 measured in this period. usec cause mask start line/file address end line/file 35735 spin_lock 1 358/memory.c c012c74a 379/memory.c 14003 spin_lock 1 341/vmscan.c c013977b 362/vmscan.c 13701 spin_lock 1 1376/sched.c c0118ef9 697/sched.c 12533 BKL 7 2689/buffer.c c0149acc 697/sched.c 12459 spin_lock 7 547/sched.c c0116ca3 697/sched.c 10479 spin_lock 1 703/vmscan.c c013a213 685/vmscan.c 10443 spin_lock 1 669/inode.c c015e8dd 696/inode.c 9163 reacqBKL 1 1375/sched.c c0118edb 1381/sched.c 8727 BKL 9 164/inode.c c0171885 188/inode.c 8066 usb-uhci d 76/softirq.c c0120cf2 119/softirq.c 8041 BKL 9 130/attr.c c015fbd4 143/attr.c 7982 spin_lock 8 547/sched.c c0116ca3 204/namei.c 6920 reacqBKL 5 1375/sched.c c0118edb 697/sched.c 6720 spin_lock 0 1376/sched.c c0118ef9 1380/sched.c 6567 spin_lock 0 661/vmscan.c c013a0f2 764/vmscan.c 6365 spin_lock 0 703/vmscan.c c013a213 764/vmscan.c 5680 BKL 1 452/exit.c c011f6c6 697/sched.c 5646 spin_lock 0 227/inode.c c015df35 696/inode.c 5428 spin_lock 1 710/inode.c c015ea08 696/inode.c 4933 spin_lock 0 547/sched.c c0116ca3 1380/sched.c
This was taken after compiling several packages and running dbench a few times.
I also ran it previously while running 'while true ; do python -c range\(3e7\) ; done', and the first line had minimum latencies of 10000, going as high as about 96000 when swapping.
This is on an Athlon 1050/100 Asus A7V (KT133) with PC133 CL2 SDRAM running 2.4.9-ac12-preempt1.
-- Ignacio Vazquez-Abrams <ignacio@openservices.net>
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
|  |