Re: [libvirt] [PATCH 09/11] qemu: Move capability checks for IOMMU features

All current IOMMU features are specific to Intel IOMMU, so understandably we check for the corresponding capabilities inside the Intel-specific switch() branch; however, we want to make sure SMMUv3 IOMMU users get an error if they try to enable any of those features in their guest, and performing the capability checks unconditionally is both the easiest way to achieve that, as well as the one least likely to result in us inadvertently letting users enable some new Intel-specific IOMMU feature for ARM guests later on. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com&gt; --- src/qemu/qemu_domain.c | 63 +++++++++++++++++++++++------------------- 1 file changed, 35 insertions(+), 28 deletions(-) @@ -6158,6 +6130,41 @@ qemuDomainDeviceDefValidateIOMMU(const virDomainIOMMUDef *iommu, return -1; } + /* These capability checks ensure we're not trying to use features + * of Intel IOMMU that the QEMU binary does not support, but they + * also make sure we report an error when trying to use features + * that are not implemented by SMMUv3,

so they must be here instead + * of inside the switch() statement above */

+ + if (iommu->intremap != VIR_TRISTATE_SWITCH_ABSENT && + !virQEMUCapsGet(qemuCaps, QEMU_CAPS_INTEL_IOMMU_INTREMAP)) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("iommu: interrupt remapping is not supported " + "with this QEMU binary")); + return -1; + }