Messages in this thread | | | Date | Tue, 12 Mar 2013 21:13:51 +0100 | From | Andi Kleen <> | Subject | Re: Regression with orderly_poweroff() |
| |
> Btw. There is another "strange" user, arch/x86/kernel/cpu/mcheck/mce.c. > It uses mce_trigger_work to call call_usermodehelper(UMH_NO_WAIT). > Why? UMH_NO_WAIT is already atomic. And the !work_pending() check is > confusing, schedule_work(schedule_work) checks it is not pending.
I think you're right, the additional step shouldn't be needed.
The MCE Handler uses the trick to do this in MCE context if the interrupts were enabled earlier.
-Andi
-- ak@linux.intel.com -- Speaking for myself only.
| |