Bug 43056
Summary: | Divide by 0 error in task_utime resulting in a kernel loop | ||
---|---|---|---|
Product: | Process Management | Reporter: | m_joshi (m_joshi) |
Component: | Scheduler | Assignee: | Ingo Molnar (mingo) |
Status: | RESOLVED WILL_NOT_FIX | ||
Severity: | normal | CC: | m_joshi |
Priority: | P1 | ||
Hardware: | All | ||
OS: | Linux | ||
Kernel Version: | 2.6.24.7 | Subsystem: | |
Regression: | No | Bisected commit-id: |
Description
m_joshi@hotmail.com
2012-04-06 11:48:06 UTC
Subsequently any other process wanting to take the lock for this process (whose lock has been acquired by this process which has created the exception) is looping for the spin lock This is happening as the device had been up and running for 497 datys +2 hrs + and we have application whose instances are binded to a core each and the cores are dedicated for that processing (nothing else runs on these cores except some non-maskable interrupts). So the utime values for these processes have overflown there lower 32 bits (it is a 64 bit kernel). Then are other processes/scripts which monitor these applications by reading their stat files (or calling pidof) |