Messages in this thread | | | Date | Sat, 31 May 1997 19:23:04 -0700 (PDT) | From | "B. James Phillippe" <> | Subject | [2.1.42] Weird CPU usage? |
| |
Greetings,
I'm using 2.1.42 on a typical x86-based dialup workstation. The only problem I've seen with it so far is a weird CPU utilization phenomenon that I cannot pinpoint. I use a Cyrix 6x86 with a huge fan; when I do CPU-intensive operations, the HLT instruction is not being used so the voltage on the board changes and the noise of the fan does also. I've had this machine for a long time and am well accustomed to the sounds of it; when I do compiles, I can hear the fan speed up a little, for instance. I know this sounds far out there; just keep reading.. What I'm noticing, is that I'll be working away on my machine while it's otherwise idle, and I'll hear the familiar speed change of the fan, lasting about 5 to 15 seconds. During this time, response on the machine is terrible, with key strokes appearing in chunks with delays of up to 2 seconds. Network IO is affected similarly. If I move the mouse around in X, it moves in hops with 2 second delays in tracking. Then, it suddenly returns to normal. I've done all the <META>-<SCROLL-LOCK> stuff to see what's going on, as well as ps -aux's and cat /proc/meminfo's. Nothing looks weird at all. However, uptime and xload both agree with the CPU spinning on something; uptime goes up to 2 and xload shows the spikes on the graph. I can't for life of me figure out what's getting spun because I'm not doing it. I have a creeping suspicion that it could be related to my BSD CCP compression module(??). I've only noticed this when I'm in X, but I'm almost always in X when I'm sitting there. Anyway, please trust me that this is not a silly coincidence; it only does this on 2.1.42.
Any ideas?
thanks, -bp -- B. James Phillippe Seattle Software Labs, Inc Network Administrator Phone: (206) 521-8346 NIC Handle: BJP4 Fax: (206) 521-8340 http://w3.terran.org/~bryan http://www.sealabs.com
| |