<div dir="ltr">On Thu, May 11, 2017 at 3:30 AM, Peter Carlsson <span dir="ltr"><<a href="mailto:maillist.peter@home.se" target="_blank">maillist.peter@home.se</a>></span> wrote:<br><div class="gmail_extra"><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hello!<br>
<br>
My system has recently frozen about 15 minutes after boot and I see the<br>
messages below in syslog at that time. I need to do a full power reboot,<br>
and then it usually works without problem. The temperatures reported by<br>
sensors at that time seems reasonable to me (< 40 degree C).<br>
<br>
Google results suggest different solutions, but nothing unambiguous that<br>
seems to work for everyone. It could be a hardware or a software issue.<br>
<br>
The system is a combined 0.28 BE/FE with Debian Jessie with a first (?)<br>
generation Intel Core i5-661 CPU with integrated Intel HD Graphics (not<br>
used) and a PCIe NVIDIA GT215 GeForce GT 240 GPU. I have not done changes<br>
or upgrades recently except security upgrades.<br>
<br>
Since my system horsepower-wise is more than enough I hope I could keep<br>
all hardware and just replace the GPU if that is the problem. I could<br>
either buy a new PCIe GPU or use the integrated GPU if the is enough for<br>
good quality HDTV recordings.<br>
<br>
I would appreciate any help to either identify/solve the problem or to<br>
recommend a good replacement GPU. Back when I bought the PCIe GPU it was<br>
quite new with VDPAU support in MythTV and it was important to choose the<br>
correct GPU. Today, maybe it has become easier?<br></blockquote></div><br></div><div class="gmail_extra">I have had this happen to me twice - once with a fanless GT220 that ran for years and once with a fan'ed GT240 that was just plagued with issues.<br><br></div><div class="gmail_extra">The GT220 was actually overheating. I had used nvidia-settings commands to run it full throttle with no power management (for some reason???), and it finally just puked on me after years of service. The 240 I replaced it with was apparently some special model I did not investigate properly and it had 1/2 the RAM installed that nVidia had suggested to OEMs, so it was falling off the bus when there were not enough internal resources for it to to VDPAU C media properly. I returned it and replaced ith with a GT 640 (I know - overkill, but it had the newer VDPAU D capabilities, and I am not budget constrained).<br><br></div><div class="gmail_extra">Back to the fanless card - it actually ran for another 6 months without issue once I let it have power management back and it could throttle the GPU speed down. Then it just died all-together one day. I do not disable throttling any longer and really cannot remember why I ever did - it probably had something carried over from an ION-1 config or something...? That was over 6 years ago though.<br><br></div><div class="gmail_extra">Point being, it could be motherboard, GPU, RAM on the GPU..... Hard to isolate intermittent hardware failures unfortunately.<br></div><div class="gmail_extra"><br></div><div class="gmail_extra">-Greg<br></div></div>