On Tue, Apr 9, 2013 at 8:29 PM, Sedat Dilek wrote: > On Tue, Apr 9, 2013 at 8:26 PM, Sedat Dilek wrote: >> On Tue, Apr 9, 2013 at 6:51 PM, Viresh Kumar wrote: >>> On 9 April 2013 21:38, Sedat Dilek wrote: >>>> With x=3 the system gets in an unuseable state. >>>> >>>> root# echo 0 > /sys/devices/system/cpu/cpu3/online >>>> >>>> I could not write my reply and had to do a hard/cold reboot. >>>> The dmesg log I saw looked similiar to my digicam-shot. >>> >>> Few things i need from you. First is output of cpufreq-info. Then >>> all the steps you did to reproduce above? Removed any other cpus? >>> >>> I am not able to find next-20130326 tag in my repo, only have 23 and 28. >>> Can you debug it a bit to find exact line of code causing this issue using >>> objdump? >>> >>> HINT: Documentation/BUG-HUNTING.. >>> >>> Give me line numbers of both of these functions: __cpufreq_governor() and >>> __cpufreq_remove_dev(). >> >> I have recompiled next-20130326 and the REGRESSION is still reproducible. >> >> Attached are my dmesg, kernel-config, tarball of my drivers/cpufreq >> build-dir, objdump of cpufreq_governor.o and the list of my current >> amd64-toolchain. >> >> Hope this helps you! >> >> Regards, >> - Sedat - > > $ cd linux-next/ > > $ objdump --disassemble-all drivers/cpufreq/cpufreq.o > > /tmp/cpufreq_o--disassemble-all.txt > > ...attached. > > - Sedat - Hmm, I remembered Thorsten Glaser told be to pass also "-Mintel" parameter ("-D" shortform for "--disassemble-all"): $ objdump -D -Mintel drivers/cpufreq/cpufreq.o > /tmp/cpufreq_o-D-Mintel.txt File attached. Hope this helps you. - Sedat -