<br><br><div class="gmail_quote">On 1 February 2010 01:37, steve <span dir="ltr"><<a href="mailto:networks1@cox.net">networks1@cox.net</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<div class="im">>Just for the record, a memory corruption problem is not necessarily a<br>
memory problem. Memory corruption<br>
>could be caused by software, for example, copying memory into an array that<br>
exceeds the array's size and<br>
>stomping on a pointer in some data structure that follows the array. Memory<br>
corruption is really hard to<br>
>debug because the offending event can occur long before the crash.<br>
<br>
</div>Good point. Most of the things I've found about kernel panic say it's<br>
almost always a hardware problem. I hope that's the case and it's not what<br>
you're describing.<br>
<br></blockquote><div>Hardware or driver/kernel code related.<br><br>What version of the nVidia drivers are you using? I get a kernel panic when starting X with nVidia drivers newer than 190.18, I think related to my Xorg.conf (although as it's not a huge issue I've not investigated further.) It might be worth looking at though?<br>
<br><br></div><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
OTOH I just finished running 10 passes of memtest86+ and it found no<br>
problems :-( so if that's believable it's not a hardware problem in the<br>
memory sticks. I still can't find any way to test the video hardware in<br>
linux.<br>
<br></blockquote><div>Yeah, that sounds fairly okay, I've had memtest take 15-20 hours to find issues with marginal memory modules before, but that was causing a crash every few days, not in a repeatable fashion like you're describing.<br>
</div><blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
About a week ago I was having problems with choppy audio because I did not<br>
have the video playback profile set to use VDPAU. Interestingly, when I<br>
played back under that scenario, the audio was hosed but I did not have this<br>
lockup problem--i.e. a recording would play all the way through. So<br>
something must be amiss in the hardware or software associated with VDPAU.<br>
<div><div></div><div class="h5"><br></div></div></blockquote><div>Do you have any issues running things like Compiz, or any graphically intensive benchmarks?<br><br>Ian<br><br></div></div>