On Thu, 2022-04-28 at 19:22 +0200, Paolo Bonzini wrote: > On 4/28/22 17:32, Sean Christopherson wrote: > > On Tue, Apr 26, 2022, Maxim Levitsky wrote: > > > I can reproduce this in a VM, by running and CTRL+C'in my ipi_stress test, > > > > Can you post your ipi_stress test? I'm curious to see if I can repro, and also > > very curious as to what might be unique about your test. I haven't been able to > > repro the syzbot test, nor have I been able to repro by killing VMs/tests. > > Did you test with CONFIG_PREEMPT=y? yes, I test with CONFIG_PREEMPT but I only enabled it a day ago, I think I had seen this warning before, but could bit, I'll check if that fails without CONFIG_PREEMPT as well. What I recently changed, is that I enabled lockdep and related settings on all my machines and VMs, I enabled CONFIG_PREEMPT, and I also switched to tdp_mmu on all systems. Bugs are biting but it is better this way, especially to weed out the last bugs of my nested avic code :) Best regards, Maxim Levitsky > > (BTW, the fact that it reproduces under 5.17 is a mixed blessing, > because it means that we can analyze/stare at a simpler codebase). > > Paolo >
