Most recent kernel where this bug did not occur: unknown Distribution: Gentoo Hardware Environment: amd64 Problem Description: I'm using the k8-powernow for powersaving on my X2-4200+ with cpufreq and cpudyn. When I start up, the k8-powernow statements are as follows: powernow-k8: Found 2 AMD Athlon 64 / Opteron processors (version 1.50.4) powernow-k8: 0 : fid 0xe (2200 MHz), vid 0x8 (1350 mV) powernow-k8: 1 : fid 0x2 (1000 MHz), vid 0x12 (1100 mV) This is confirmed by scaling_available_frequencies which echos 2200000 1000000 However, scaling_min_freq is 1452000 and that's what the powersave module pushes down to. if I force it to 1000000, my machine locks up. The Gentoo devs told me to post it here after replicating it on the 2.6.16-rc1 Specs: Motherboard: gigabyte GA-K8NF-9 BIOS F10 (latest) Processor: Athlon 64 4200+
Link to origiinal gentoo bug report : http://bugs.gentoo.org/show_bug.cgi?id=117879
Your system has a very weird PN! table. I'd expect to see several P-states going between 2.2 GHz and 1.6 GHz, then a 1.0 GHz P-state. It's also weird that the driver is reporting a 1.0 GHz minimum P-state, but cpufreq is not reporting that minimum frequency. Please make sure you have the latest BIOS. Also, could you turn on cpufreq debugging and recompile your kernel? I'd like to see if the driver is providing any additional messages.
I will be going on a 2 month sabbatical starting February 2nd, 2006 and not returning until mid-April. I will not be reachable by email or phone while on sabbatical. My co-workers will be covering for me on technical issues, but I encourage everyone to close as any outstanding technical issue before I leave. -Mark Langsdorf Linux Validation Tools and Support AMD, Inc. <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN"> <HTML> <HEAD> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1"> <META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7232.82"> <TITLE>Out of Office AutoReply: [Bug 5934] System locks if cpufreq set to lowest level.</TITLE> </HEAD> <BODY> <!-- Converted from text/plain format --> <P><FONT SIZE=2>I will be going on a 2 month sabbatical starting February 2nd, 2006 and not returning until mid-April. I will not be reachable by email or phone while on sabbatical. My co-workers will be covering for me on technical issues, but I encourage everyone to close as any outstanding technical issue before I leave.<BR> <BR> -Mark Langsdorf<BR> Linux Validation Tools and Support<BR> AMD, Inc.<BR> </FONT> </P> </BODY> </HTML>
Created attachment 7109 [details] My dmesg with full debug on. Here's the dmesg with full debugging turned on. Seems, there's either a bad table or a hard to read table.
I was funning F10 until now, the dmesg attached is with the latest F11 bios, just released.
That BIOS is bad, but that still doesn't explain why cpufreq is sending the wrong transition frequency to the driver. Can you hook up a serial console and see what is transmitted before the hardlock?
Please reopen this bug: - if it is still present in kernel 2.6.19-rc5 and - you can provide the requested data.
I will be at at a conference from November 5th until November 11th. I wil have intermittent access to email until November 13th. In my absence, refer Red Hat related Linux issues to Bhavana Nagendra at bhavana.nagendra@amd.com, and Novel related Linux issues to Joachim Deguara at joachim.deguara@amd.com. -Mark Langsdorf New Technology Products AMD, Inc. <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN"> <HTML> <HEAD> <META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1"> <META NAME="Generator" CONTENT="MS Exchange Server version 6.5.7651.14"> <TITLE>Out of Office AutoReply: [Bug 5934] System locks if cpufreq set to lowest level.</TITLE> </HEAD> <BODY> <!-- Converted from text/plain format --> <P><FONT SIZE=2>I will be at at a conference from November 5th until November 11th. I wil have intermittent access to email until November 13th.<BR> <BR> In my absence, refer Red Hat related Linux issues to Bhavana Nagendra at bhavana.nagendra@amd.com, and Novel related Linux issues to Joachim Deguara at joachim.deguara@amd.com.<BR> <BR> -Mark Langsdorf<BR> New Technology Products<BR> AMD, Inc.<BR> </FONT> </P> </BODY> </HTML>