Backport the fix for deadlock in CFS-bandwidth timer locking
Low-latency profile of StarlingX is affected by a deadlock in
CFS scheduler. spin_lock is used in IRQ handler there instead of
spin_lock_irqsave. This leads to an attempt to lock the same
spinlock twice and inevitable system freeze. Backporting c0ad4aa4d8
commit from upstream kernel to cure the issue.
Change-Id: I5416c0e0886f42d2bcec8e3e5da063e6af6916f8
Closes-bug: 1832854
Signed-off-by: Alex Kozyrev <email address hidden>
Reviewed: https:/ /review. opendev. org/673028 /git.openstack. org/cgit/ starlingx/ integ/commit/ ?id=56a91fec13c d018f998a0bdf0b eed753efd94df5
Committed: https:/
Submitter: Zuul
Branch: master
commit 56a91fec13cd018 f998a0bdf0beed7 53efd94df5
Author: Alex Kozyrev <email address hidden>
Date: Fri Jul 26 13:58:17 2019 -0400
Backport the fix for deadlock in CFS-bandwidth timer locking
Low-latency profile of StarlingX is affected by a deadlock in lock_irqsave. This leads to an attempt to lock the same
CFS scheduler. spin_lock is used in IRQ handler there instead of
spin_
spinlock twice and inevitable system freeze. Backporting c0ad4aa4d8
commit from upstream kernel to cure the issue.
Change-Id: I5416c0e0886f42 d2bcec8e3e5da06 3e6af6916f8
Closes-bug: 1832854
Signed-off-by: Alex Kozyrev <email address hidden>