Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#42720 - Sudden Poweroff- No powercyclet- with: 3.17-3.16.5-3.14 and Catalyst driver leaving no log/warnings

Attached to Project: Arch Linux
Opened by Edwin (chargen) - Sunday, 09 November 2014, 08:18 GMT
Last edited by Doug Newgard (Scimmia) - Monday, 17 November 2014, 23:46 GMT
Task Type Bug Report
Category Kernel
Status Closed
Assigned To No-one
Architecture x86_64
Severity Critical
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Description:

Experienced random hard reboots & sudden power down (like a silent /sbin/poweroff ipmi call) on a Dell XPS 710 - Core2 x86_64 - with kernels 3.17x 3.16x - crashes or failure don't appear in any log (others mention this phenomena as well). The HARD shutdown's poweroff occur randomly after several hours of uptime mostly when doing memory intensive operations or heavy browser usage. I used a AMD GPU 7x series and combined Catalyst driver-(havent tried free-pci, though) I've seen similar crash reports (but no 'poweroff' like a brownout/power supply failure). Reports suggest to use 3.14 kernel but the same reboot/ hard shutdown occur with all mentioned recent kernels. I know there have been backports in oct. to 3.14-> which also randomly crashes. This wasn't the case in september 2014. Several other crashreports suggest that hard-lockups/kerneloops/crashes occur with AMD drivers / hardware accell. issues with GPU intensive games/ Chrome/firefox hw acceleration. The main difference is that I'm having sudden powerdown and others are left with a garbled screen and the kernel in a crashed state.

At first I thought of faulty hardware but there wasn't any failure record in the BIOS. I didnt check the free drivers, but crashes were mentioned using the free nouveau drivers as well.

There is also a difference in the way it powers down: the DELL XPS 710 fan(s) behaviour: during a normal warm issued reboot the fans makes less noise. A powerdown crash makes all the fans blow much harder/louder, as if the system is overheated.

solution: a bit decadent. I bought an NVIDIA (730) a couple of days ago and the system has been up and stable since then. Linux wahnsee 3.17.0-1- #1 SMP PREEMPT Tue Oct 7 16:08:54 UTC 2014 x86_64 GNU/Linux- running stable for several days with Nvidia drivers. Haven't tried 3.14./3.16


Q: are there any radical changes done to powermanagement in regard to the drivers/kernel and AMD hardware, and/or catalyst driver kernel module accessing priviledged syscall to powerdown features? (in the kernel source: system/chassis dependend powerdown can be issued, especially for certain IPMI Dell boards.

Additional info:
* config and/or log files etc.
There are no traces of any crash or hw failure in logfiles. As a symptom: this seems to be the case for others as well.

Steps to reproduce:
Using radeon catalyst fglrx non-free driver with a 7x series GPU chipset with 3.14-3.17 kernels under heavy load.
This task depends upon

Closed by  Doug Newgard (Scimmia)
Monday, 17 November 2014, 23:46 GMT
Reason for closing:  No response
Comment by Doug Newgard (Scimmia) - Sunday, 09 November 2014, 16:23 GMT
I'm a bit confused here. It seems that the problem lies in the Catalyst driver (which are not in the repo, BTW), but then mention that others have crashes with nouveau, which is the free NVidia driver.
Comment by Doug Newgard (Scimmia) - Wednesday, 12 November 2014, 04:48 GMT
Ping? Any more clarification on where the issue is here? Still a bit confused over the mention of nouveau.

Loading...