Messages in this thread | | | Date | Wed, 09 Sep 2015 22:06:10 -0400 | From | Waiman Long <> | Subject | Re: [4.2, Regression] Queued spinlocks cause major XFS performance regression |
| |
On 09/04/2015 11:54 AM, Peter Zijlstra wrote: > On Fri, Sep 04, 2015 at 05:30:35PM +0200, Peter Zijlstra wrote: >> Should I place the virt_spin_lock() thing under CONFIG_PARAVIRT (maybe >> even _SPINLOCKS) such that only paravirt enabled kernels when ran on a >> hypervisor that does not support paravirt patching (HyperV, VMware, >> etc..) revert to the test-and-set? > Ah, CONFIG_HYPERVISOR_GUEST seems fitting, that's a prerequisite for all > the PARAVIRT options too. >
Yes, I also used that in my queued unfair lock patch.
Cheers, Longman
| |