Debugging Kernels that do not completely die (no Panic)
I am looking for a way to debug a kernel lockup condition. I am trying the new kernel 3.7.0 rc2. It is working for about 40 minutes and then it is freezing. The keyboard is frozen, the display is frozen and the sound card is continually playing the last second or so. What I would like to do is get a good trace of where this is happening. I have checked the /var/log/messages on the next boot, and the dmesg and do not see anything of note. Any ideas? I am a newbie with kernel debugging, although I have successfully built the kernel from source and set a few options in it.
- Raj Upadhyaya
The best advise is don't.
What was your last STABLE kernel? I say use that or recompile it with as few modifications as possible.
If there is some function that you need but isn't there, see if an app is available instead of messing with the kernel directly.
Modding the kernel is the option of last resort.
My last stable kernel is 3.6.2-4.fc17.i686 delived via yum from Fedora. There isn't any option that my workstation has that isn't in this kernel that I know of. I am a old programmer that used to do C programming a few decades ago, but haven't done it in the Linux / x86 platform. I thought it would be nice to give back to the community with any help I could assist with.
My thought was download the 3.7.0 rc2 from kernel.org, and compile it on my workstation and run it for awhile. I am also a user of Linux, so stability of the OS is important. I figure the sooner that a real kernel developer is informed of this lockup, the sooner it will be fixed. Since I found it in 3.7.0 rc1 and 3.7.0 rc2, it might not be obvious to many unless you run the system under load for awhile. I figured that linux-next: from github is way too experimental as it contains the checkins from all the developers but the mainline builds of Release Candidates would work as a test base.
|All times are GMT -5. The time now is 12:13 AM.|