]> git.ipfire.org Git - thirdparty/linux.git/commit
KVM: SVM: Check EFER.SVME and CPL on #GP intercept of SVM instructions
authorYosry Ahmed <yosry@kernel.org>
Mon, 16 Mar 2026 20:27:28 +0000 (20:27 +0000)
committerSean Christopherson <seanjc@google.com>
Fri, 3 Apr 2026 23:08:03 +0000 (16:08 -0700)
commit783cf7d01fb8788f37735c0a6c3955024189287c
tree7f653f065740a093f0d4603fd97538a9d7d61cf7
parentd2fbeb61e1451eba09eb3249aaf1f01d4c5c1f8b
KVM: SVM: Check EFER.SVME and CPL on #GP intercept of SVM instructions

When KVM intercepts #GP on an SVM instruction from L2, it checks the
legality of RAX, and injects a #GP if RAX is illegal, or otherwise
synthesizes a #VMEXIT to L1. However, checking EFER.SVME and CPL takes
precedence over both the RAX check and the intercept. Call
nested_svm_check_permissions() first to cover both.

Note that if #GP is intercepted on SVM instruction in L1, the intercept
handlers of VMRUN/VMLOAD/VMSAVE already perform these checks.

Note #2, if KVM does not intercept #GP, the check for EFER.SVME is not
done in the correct order, because KVM handles it by intercepting the
instructions when EFER.SVME=0 and injecting #UD.  However, a #GP
injected by hardware would happen before the instruction intercept,
leading to #GP taking precedence over #UD from the guest's perspective.
Opportunistically add a FIXME for this.

Fixes: 82a11e9c6fa2 ("KVM: SVM: Add emulation support for #GP triggered by SVM instructions")
Signed-off-by: Yosry Ahmed <yosry@kernel.org>
Link: https://patch.msgid.link/20260316202732.3164936-6-yosry@kernel.org
Signed-off-by: Sean Christopherson <seanjc@google.com>
arch/x86/kvm/svm/svm.c