kvm_flush_coalesced_mmio_buffer explicitly checks for this to be non-null. Since kvm_init_vcpu sets this as an offset from the mapped cpu->kvm_run it should be reset to NULL after kvm_run is unmapped.
Signed-off-by: Sid Manning <sidn...@quicinc.com> --- accel/kvm/kvm-all.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c index c65b790433..b67964e552 100644 --- a/accel/kvm/kvm-all.c +++ b/accel/kvm/kvm-all.c @@ -512,6 +512,8 @@ static int do_kvm_destroy_vcpu(CPUState *cpu) goto err; } + s->coalesced_mmio_ring = NULL; + if (cpu->kvm_dirty_gfns) { ret = munmap(cpu->kvm_dirty_gfns, s->kvm_dirty_ring_bytes); if (ret < 0) { -- 2.34.1