On Monday 25 June 2007 12:03, Hugo Vanwoerkom wrote: > Prismatic Plasma wrote: > > On Monday 25 June 2007 05:31, Andrew Sackville-West wrote: > >> On Mon, Jun 25, 2007 at 03:19:22AM +0000, Prismatic Plasma wrote: > >>> I've had this problem for a while, and I can't seem to solve or even > >>> fully diagnose what's wrong. The 2 most common symptoms are: > >>> 1) random segmentation faults during compiling. It's most apparent (and > >>> annoying) during long compiles. There's sometimes an assembler message > >>> instead, complaining about unknown variables or junk at end of line. In > >>> the error message it's often apparent that something got corrupted by > >>> one character. The file is usually a header and it isn't corrupted on > >>> disk. > >> > >> memory. > >> > >>> 2) occasionally the system goes wild and thrashing, sucking up nearly > >>> all cpu. Sometimes it causes the machine to lock up, but usually I can > >>> kill the process that triggered the problem and everything settles down > >>> after 10 seconds or so. If I restart the process, sometimes > >>> everything's okay, sometimes it goes wild again. > >> > >> maybe memory. > >> > >>> Is this a familiar problem? Is it the mobo or ram? Or is it a software > >>> issue messing up virtual memory? My bios is updated, and I've tried > >>> several kernels in the 2.6 line, including a couple of custom compiles. > >>> The problem exists even in single-user mode, so it doesn't have > >>> anything to do with the windowing system. I'm currently using kernel > >>> 2.6.21, etch amd-64, ext3 fs, and sata disk (WD Cavalier, I think). > >> > >> definitely memory. > >> > >> A > > > > Do you mean bad ram? Or is it a timing issue from bios that need > > twiddling? The machine's been flakey like this since I got it. > > I had just that problem but with my current mobo (EP-8VTAI). When I just > had it: random segmentation faults in places where that should never > happen (compiles) and hard stops. > > It turned out that I had to raise the DIMM voltage by 0.4 volts. That is > 2.5 volts by default and can be raised on this board by steps of 0.2 > volts. I raised it first 0.2 volts and the problem lessened but did not > go away. Then raised it by 0.4 volts and the board is like a rock. > > I would try to raise the DIMM voltage on that board and see what happenes. > > Hugo
Didn't help. The DIMM voltage specs are 1.9-2.0 V. Mobo lets me adjust from 1.8-2.0 V. Setting higher than 1.90 V causes crashes more often. Lower doesn't help either. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
