Red Hat Bugzilla – Bug 485582
RV350 hardlocks with DRI + kernel-184.108.40.206-170.2.5.fc10.i686
Last modified: 2009-05-17 15:25:42 EDT
Created attachment 331938 [details]
Description of problem:
Sometime between kernel-PAE-220.127.116.11-68.fc10.i686 and kernel-PAE-18.104.22.168-170.2.5.fc10.i686 DRI has caused my RV350 to hardlock any time any accelerated 3D graphics are displayed with KMS enabled
Version-Release number of selected component (if applicable):
Almost always with kernel-PAE-22.214.171.124-170.2.5.fc10.i686, never with kernel-PAE-126.96.36.199-68.fc10.i686
Steps to Reproduce:
1. Boot kernel-PAE-188.8.131.52-170.2.5.fc10.i686
2. Run a DRI application, glxgears isn't always enough to trigger this, supertuxkart will 100% of the time
Screen goes black "Input out of range" and system hardlocks
I have attached the output of lspci -vvv; an Xorg.0.log and the output of dmesg.
FWIW, this also occurs with the non-PAE kernels.
Created attachment 331939 [details]
Created attachment 331940 [details]
Can you download some kernels from koji and test them to try and pinpoint when this started? Or just test whatever kernels are currently installed?
Ok, my results:
184.108.40.206-68.fc10.i686.PAE - Good
220.127.116.11-117.fc10.i686.PAE - "Input signal out of range"
18.104.22.168-159.fc10.i686.PAE - System Freezes, but no out of range
22.214.171.124-170.2.5.fc10.i686.PAE - "Input signal out of range"
126.96.36.199-170.2.24.fc10.i686.PAE - "Input signal out of range"
It seems that something introduced in 188.8.131.52 has broken DRI.
Just thought I'd report that this issue seems to have gone away now (kernel-PAE-184.108.40.206-170.2.35.fc10.i686 or possibly earlier).