please try this patch on top of my v5. It should fix the issue. I tested
on a RHEL 9 host and a fc43 SEV-SNP host, 1 or multiple vcpus, two
different guests and all looks good.
From 26db44eba8c727160dd9e97c9d5582a0ddc5884d Mon Sep 17 00:00:00 2001
From: Ani Sinha <[email protected]>
Date: Thu, 19 Feb 2026 13:12:02 +0530
Subject: [PATCH] Call cpu_synchronize_all_post_init for coco as well
Fixes issue reported by Cedric
Signed-off-by: Ani Sinha <[email protected]>
---
accel/kvm/kvm-all.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/accel/kvm/kvm-all.c b/accel/kvm/kvm-all.c
index d7ea60f582..0610bf8434 100644
--- a/accel/kvm/kvm-all.c
+++ b/accel/kvm/kvm-all.c
@@ -2816,7 +2816,7 @@ static int kvm_reset_vmfd(MachineState *ms)
}
s->vmfd = ret;
-
+ kvm_state->guest_state_protected = false;
kvm_setup_dirty_ring(s);
/* rebind memory to new vm fd */
@@ -2872,6 +2872,9 @@ static int kvm_reset_vmfd(MachineState *ms)
* kvm fd has changed. Commit the irq routes to KVM once more.
*/
kvm_irqchip_commit_routes(s);
+ if (ms->cgs) {
+ cpu_synchronize_all_post_init();
+ }
trace_kvm_reset_vmfd();
return ret;
}
Tested (OS reboot, system_reset) with a RHEL10 guest (2 vCPUs) on
a RHEL9 and a RHEL10 host, using a SATA device and active NIC VFs.
All Looks good.
There is still one message :
qemu-system-x86_64: info: virtual machine state has been rebuilt with new
guest file handle.
which seems normal.
Thanks,
C.
