| Message ID | 20260319163524.45554-1-tgamblin@baylibre.com |
|---|---|
| State | New |
| Headers | show |
| Series | qemuboot-x86: add qemux86 kvm workaround | expand |
diff --git a/meta/conf/machine/include/x86/qemuboot-x86.inc b/meta/conf/machine/include/x86/qemuboot-x86.inc index 82ef4a0b73..abdbc12fa1 100644 --- a/meta/conf/machine/include/x86/qemuboot-x86.inc +++ b/meta/conf/machine/include/x86/qemuboot-x86.inc @@ -2,7 +2,7 @@ IMAGE_CLASSES += "qemuboot" QB_SMP ?= "-smp 4" QB_CPU:x86 ?= "-cpu IvyBridge -machine q35,i8042=off" -QB_CPU_KVM:x86 ?= "-cpu IvyBridge -machine q35,i8042=off" +QB_CPU_KVM:x86 ?= "-cpu IvyBridge,kvmclock=off -machine q35,i8042=off" QB_CPU:x86-64 ?= "-cpu Skylake-Client -machine q35,i8042=off" QB_CPU_KVM:x86-64 ?= "-cpu Skylake-Client -machine q35,i8042=off"
[YOCTO #16074] As per the bug listed, investigation of an intermittent failure with KVM-enabled qemux86 VMs has shown that there is a possible issue with a spectre v2 retpoline mitigation when running on hosts with certain Intel CPUs, e.g. the Intel Ultra 7 265. We have confirmed that disabling the KVM paravirtualized clock prevents the failure; so for now, avoid it by specifying '-cpu IvyBridge,kvmclock=off' instead of simply '-cpu IvyBridge' in qemuboot-x86 for the KVM case. Note that this issue does not affect qemux86-64. Signed-off-by: Trevor Gamblin <tgamblin@baylibre.com> --- meta/conf/machine/include/x86/qemuboot-x86.inc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)