You are not logged in.
Pages: 1
Hi.
I recently did a full system upgrade, including a kernel upgrade. Couple things broke, but one thing I can't seem to fix is conky. I keep getting this error when run conky
Conky: /etc/conky/conky.conf: 1: no such configuration: 'round'
Conky: can't open '/sys/bus/platform/devices/coretemp.1/temp1_input': No such file or directory
please check your device or remove this var from Conky
Segmentation fault
Reporting temperature on multiple cores was working fine before. The coretemp.0 directory exists, but that's it. If I remove references to coretemp from my conky file, then conky runs fine (but it would be nice to be able to see my cpu core temps again) I tried reinstalling conky and re-running sensors-detect, but it didn't work.
Anyone know how I can fix this?
Edit: I just remembered that I wrote this down after the kernel upgrade. Pacman said "kbd and rtc devices not world readable" or something like that. I don't know what this means, but maybe it's a clue. Pacman also told me to append "quiet" in the kernel command line, but I'm now sure where to do this.
Thanks,
Last edited by Pacopag (2011-09-23 14:49:09)
Offline
Same issue here. I think there's a bug in the coretemp module, I no longer see a coretemp.1 directory on my dual-core machine. Now I have the following under coretemp.0:
$ ls /sys/bus/platform/devices/coretemp.0/
driver name temp2_crit temp2_label temp3_crit_alarm temp3_max
hwmon power temp2_crit_alarm temp2_max temp3_input uevent
modalias subsystem temp2_input temp3_crit temp3_label
For the quick fix I changed my conkyrc to look at temp2_input and temp3_input, which appear to be the temps for each core. As a simple test, I did some stress testing to load the cores one at a time, and it seems that temp2 is representative of core 0, temp3 is core 1. However, this smells like a bug, so hopefully it's corrected upstream soon.
Offline
Great. Thanks for your reply. For some reason I have fewer tempX_inputs in coretemp.0 than I have cores. But anyway, I can monitor coretemp.0 for now, and like you say, hopefully it will take care of itself.
Last edited by Pacopag (2011-08-10 20:09:40)
Offline
You can compare the ouput of sensors and ls /sys/class/hwmon/*/* and update your conkyrc accordingly.
Now I can get the CPU temperature with ${hwmon 1 temp 2} and the 2 cores with ${hwmon 0 temp 2} and ${hwmon 0 temp 3} for my Core2Duo.
@Pacopag, you can find the kernel comman lind in your system bootloader (grub, syslinux). "queit" will decrease verbosity during boot.
Offline
Same here, breaking my temperature monitoring with vicious (awesome). I see no way of telling vicious to check temp2 instead of temp1 (which is missing), so I hope they get that bug fixed, soon
Offline
Forgive me for bumping an old post, but I fixed the problem and wanted to share with those whose conky broke too.
This is what I had in /etc/conky/conky.conf before the kernel upgrade and it worked fine (on a quad core cpu)
${platform coretemp.0 temp 1} C ${platform coretemp.1 temp 1} C ${platform coretemp.2 temp 1} C ${platform coretemp.3 temp 1} C
After kernel upgrade, this no longer worked.
Now I have
$alignr${platform coretemp.0 temp 1} C $alignr${platform coretemp.0 temp 2} C $alignr${platform coretemp.0 temp 3} C $alignr${platform coretemp.0 temp 4} C
and it seems to be working, as far as I can tell.
Cheers,
Offline
Pages: 1