Hi! I try to compile an vanilla kernel 2.6.32.2 on centos 5.4 and i have this error :
CC mm/page-writeback.o /home/root_data/linux-2.6.32.2/mm/page-writeback.c: In function ‘test_clear_page_writeback’: /home/root_data/linux-2.6.32.2/mm/page-writeback.c:1280: internal compiler error: Segmentation fault Please submit a full bug report, with preprocessed source if appropriate. See URL:http://bugzilla.redhat.com/bugzilla for instructions. The bug is not reproducible, so it is likely a hardware or OS problem. make[2]: *** [mm/page-writeback.o] Error 1 make[1]: *** [mm] Error 2 make: *** [sub-make] Error 2
Have anyone any idea about this ?
The hardware is in via vb8001 with this cpu [root@sevcenco linux-2.6.32.2]# cat /proc/cpuinfo processor : 0 vendor_id : CentaurHauls cpu family : 6 model : 15 model name : VIA Nano processor L2200@1600MHz stepping : 2 cpu MHz : 1615.851 cache size : 1024 KB fpu : yes fpu_exception : yes cpuid level : 10 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat clflush acpi mmx fxsr sse sse2 ss tm syscall nx fxsr_opt rdtscp lm up pni monitor vmx est tm2 ssse3 cx16 xtpr lahf_lm bogomips : 3231.70 clflush size : 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management:
The config can be found at http://sevcenco.homelinux.org/adrian/kernel_nano/
Thanks, Adrian
On Wed, Jan 6, 2010 at 5:15 PM, Adrian Sevcenco Adrian.Sevcenco@cern.ch wrote:
Hi! I try to compile an vanilla kernel 2.6.32.2 on centos 5.4 and i have this error :
Out of curiosity, why are you rebuilding the kernel? Is there a driver you need which isn't supplied by the elrepo repository folks?
Jim Perrin wrote:
On Wed, Jan 6, 2010 at 5:15 PM, Adrian Sevcenco Adrian.Sevcenco@cern.ch wrote:
Hi! I try to compile an vanilla kernel 2.6.32.2 on centos 5.4 and i have this error :
Out of curiosity, why are you rebuilding the kernel? Is there a driver you need which isn't supplied by the elrepo repository folks?
Well, the thing is that i lack power scaling (i understood that the module should have name something like overhaul .. or something) Given that this is an samba home server with very little load is a pity that it stays all the time at maximum freq.. elrepo guys are fantastic and from them i have the vt1211 driver for the sensors .. but i dint see that they would have also scaling drivers in their repo. Thanks, Adrian
On 01/07/2010 10:12 AM, Adrian Sevcenco wrote:
Jim Perrin wrote:
On Wed, Jan 6, 2010 at 5:15 PM, Adrian SevcencoAdrian.Sevcenco@cern.ch wrote:
Hi! I try to compile an vanilla kernel 2.6.32.2 on centos 5.4 and i have this error :
Out of curiosity, why are you rebuilding the kernel? Is there a driver you need which isn't supplied by the elrepo repository folks?
Well, the thing is that i lack power scaling (i understood that the module should have name something like overhaul .. or something) Given that this is an samba home server with very little load is a pity that it stays all the time at maximum freq.. elrepo guys are fantastic and from them i have the vt1211 driver for the sensors .. but i dint see that they would have also scaling drivers in their repo. Thanks, Adrian
Depends what hardware you have. ELRepo has a fixed powernow-k8 driver for AMD processors that allows correct scaling on multicore processors. See here:
http://blog.toracat.org/2009/08/go-green-with-newer-amd-processors/ http://elrepo.org/tiki/kmod-powernow-k8
Other than that, generally I've found most scaling issues to be hardware/bios related, rather than an issue within CentOS itself so I'd suggest checking you have the latest bios for your motherboard and that all power saving options within the bios are correctly set.
Hope that helps.
On 01/07/2010 11:36 AM, Ned Slider wrote:
Depends what hardware you have. ELRepo has a fixed powernow-k8 driver for AMD processors that allows correct scaling on multicore processors. See here:
OP has a VIA nano, the K8 stuff isnt going to work there.
I wonder if its possible to manually set clockrate.
- KB
Karanbir Singh schrieb:
On 01/07/2010 11:36 AM, Ned Slider wrote:
Depends what hardware you have. ELRepo has a fixed powernow-k8 driver for AMD processors that allows correct scaling on multicore processors. See here:
OP has a VIA nano, the K8 stuff isnt going to work there.
I wonder if its possible to manually set clockrate.
- KB
the OP's posting had the hint
The bug is not reproducible, so it is likely a hardware or OS problem.
To rule out a hardware problem it would be good to try the compilation with e.g. Fedora 12 (run from LiveCD or USB stick).
I did not look at the config, but I had good results with the 2.6.31 kernel compiled on CentOS 5.4 with the Fedora 12 config (but that was on x86_64 so YMMV).
HTH,
Kay
Kay Diederichs wrote:
Karanbir Singh schrieb:
On 01/07/2010 11:36 AM, Ned Slider wrote:
Depends what hardware you have. ELRepo has a fixed powernow-k8 driver for AMD processors that allows correct scaling on multicore processors. See here:
OP has a VIA nano, the K8 stuff isnt going to work there.
I wonder if its possible to manually set clockrate.
- KB
the OP's posting had the hint
The bug is not reproducible, so it is likely a hardware or OS problem.
To rule out a hardware problem it would be good to try the compilation with e.g. Fedora 12 (run from LiveCD or USB stick).
k thx!
I did not look at the config, but I had good results with the 2.6.31 kernel compiled on CentOS 5.4 with the Fedora 12 config (but that was on x86_64 so YMMV).
this is also on x86_64 .. well, nevermind, i just tried a compile with original config-2.6.18-164.9.1.el5 and it worked .. that means that i was over zealous in configuration :D ;)
Thanks for advices, Adrian
Adrian Sevcenco wrote:
Kay Diederichs wrote:
I did not look at the config, but I had good results with the 2.6.31 kernel compiled on CentOS 5.4 with the Fedora 12 config (but that was on x86_64 so YMMV).
this is also on x86_64 .. well, nevermind, i just tried a compile with original config-2.6.18-164.9.1.el5 and it worked .. that means that i was over zealous in configuration :D ;)
So once again after many tries i am stuck! it seems that on x86_64 i cannot enable longhaul cpufreq driver!! in 2.6.32.3 X86_LONGHAUL depends on CPU_FREQ [=y] && X86_32 [=n] && ACPI_PROCESSOR [=y] !! i am wondering why is there that X86_32 ..... the thing it seems that is defined at arch/x86/kernel/cpu/cpufreq/Kconfig:209 i am trying the get in touch with cpufreq guys, but in mean time have anyone any idea about this? what is the reason of X86_32 requirement?
Thanks, Adrian
On Thu, 2010-01-07 at 11:46 +0000, Karanbir Singh wrote:
On 01/07/2010 11:36 AM, Ned Slider wrote:
Depends what hardware you have. ELRepo has a fixed powernow-k8 driver for AMD processors that allows correct scaling on multicore processors. See here:
OP has a VIA nano, the K8 stuff isnt going to work there.
I wonder if its possible to manually set clockrate.
- KB
--- Yes you can do that if scaling works in /etc/sysconfig/apmd.../cpuspeed/
John
-
Adrian Sevcenco -
Jim Perrin -
JohnS -
Karanbir Singh -
Kay Diederichs -
Ned Slider