Oren On 3/13/19 8:05 PM, Oren Twaig wrote: > Hi Suravee, > > Turns out, the _same_ bug was already discussed in the past by > yourself, Paolo and Radim (both now 'cc'-ed) > > Please read it here: > https://patchwork.kernel.org/patch/8292231/ > > > After reading that thread, I have couple of questions: > > First, > > You wrote : "I have tried NOT setting the IRR, and only > kick_vcpu(). And things seem to work fine. Therefore, I think your > analysis is likely to be correct." > > AFAIU, it means that the below patch is wrong just as Paolo > suggested in his original answer and you did fixed it back than, but the > code is now back ? Thanks for the recap. I was chasing a bug and totally forgot about this discussion. I just found the reason why irr_pending was not set to 1 in my test, which then causes the target vcpu to never get wake up and scheduled to process the IRR bit that was set by AVIC as part of sending IPI. I will send out a patch to revert and clean up this mess. Thanks for catching this. And sorry for confusion. Thanks, Suravee
