Suspend/Resume Should Be Faster With Large Servers On Linux 3.18
With the x86/cpu changes there is a change by Lan Tianyu of Intel to speed up the suspend/resume process by avoiding a 10mms sleep for CPU offlining during the S3 state. This is a fix to a timing related issue and in the tested Intel hardware led to a sleep time from 100ms to less than 5ms. For large servers the suspend time can be reduced by much greater times (in one reported case by 2.3 seconds).
Ingo's pull request for x86/cpu had just this change for Linux 3.18. "This tree includes a single commit that speeds up x86 suspend/resume by replacing a naive 100msec sleep basedpolling loop with proper completion notification. This gives some real suspend/resume benefit on servers with larger core counts."
Add A Comment
