Bug 202009

Summary: powernow-k8: fid trans failed, fid 0xc, curr 0xe
Product: [Fedora] Fedora Reporter: Alan Cox <alan>
Component: kernel-xenAssignee: Xen Maintainance List <xen-maint>
Status: CLOSED CURRENTRELEASE QA Contact: Brian Brock <bbrock>
Severity: high Docs Contact:
Priority: high    
Version: rawhideCC: bstein, carenas, clalance, davej, jr-redhatbugs2, wtogami
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: kernel-xen-2.6.18-1.2767.fc6 Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2006-11-07 14:36:30 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On: 191566    
Bug Blocks:    

Description Alan Cox 2006-08-10 11:23:15 UTC
+++ This bug was initially created as a clone of Bug #191566 +++

Description of problem:

Seeing a fair number of:

powernow-k8: fid trans failed, fid 0xc, curr 0xe
powernow-k8: transition frequency failed

# cat /proc/cpuinfo
processor       : 0
vendor_id       : AuthenticAMD
cpu family      : 15
model           : 43
model name      : AMD Athlon(tm)64 X2 Dual Core Processor  4200+
stepping        : 1
cpu MHz         : 2200.000
cache size      : 512 KB
physical id     : 0
siblings        : 1
core id         : 0
cpu cores       : 0
fpu             : yes
fpu_exception   : yes
cpuid level     : 1
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov
pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt lm 3dnowext
3dnow pni lahf_lm cmp_legacy
bogomips        : 5527.18
TLB size        : 1024 4K pages
clflush size    : 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp

(cpu #1 deleted)


Version-Release number of selected component (if applicable):
2.6.16-1.2111_FC5xen0

Comment 1 Alan Cox 2006-08-10 11:32:20 UTC
Seeing the same with FC6test2 and the Xen kernel - spews powernow errors every
second or so, with high urgency level that spew all over the console. Not sure
if this is a Xen related bug (no powernow and it shouldnt be enabled) or a
kernel one.


Comment 2 Chris Lalancette 2006-08-10 14:25:40 UTC
This seems to be Xen specific.  When the Xen dom0 boots up, the powernow-k8_init
function is called, which goes around to each CPU to query whether it supports
frequency and voltage transition.  All of this is read-only, so all of the calls
to access the various feature flags of the CPU succeed, and the code thinks that
the processor is capable of the transitions.  However, later on, when it
actually tries to do the transition, the Hypervisor prevents the write to the
MSR register, which fails the frequency change, which causes the messages to be
continually output.  This can be further seen by looking at the messages coming
out of both the Hypervisor and the dom0 kernel:

(XEN) (file=traps.c, line=1385) Domain attempted WRMSR 00000000c0010041 from
00.powernow-k8: fid trans failed, fid 0x10, curr 0x12
powernow-k8: transition frequency failed

Comment 3 Jordan Russell 2006-08-14 00:33:16 UTC
Patch here (haven't tested myself):

http://groups.google.com/group/xen-devel-archive/browse_thread/thread/dd1794f10ecf69aa/23a55071d651ee28

Comment 4 Alan Cox 2006-08-31 09:32:27 UTC
Still present in the -200 kernel from rawhide yesterday

Comment 5 Alan Cox 2006-10-02 22:54:34 UTC
Still present in FC6Pre - spewed all over the console, message level regardless
and makes Xen unusuable on the afflicted systems.


Comment 6 Dave Jones 2006-11-07 04:48:52 UTC
I was under the impression we were disabling CPU_FREQ for Xen for now until some
sanity gets beaten into it on the Xen side.


Comment 7 Stephen Tweedie 2006-11-07 14:36:30 UTC
It was disabled on Oct. 11.