[PATCH] [273/275] KVM: Correct ordering of ldt reload wrt fs/gs reload

[Andi Kleen]

2.6.35-longterm review patch.  If anyone has any objections, please let me know.

------------------
From: Avi Kivity <avi@xxxxxxxxxx>

commit 0a77fe4c188e25917799f2356d4aa5e6d80c39a2 upstream.

If fs or gs refer to the ldt, they must be reloaded after the ldt.  Reorder
the code to that effect.

Userspace code that uses the ldt with kvm is nonexistent, so this doesn't fix
a user-visible bug.

Signed-off-by: Avi Kivity <avi@xxxxxxxxxx>
Signed-off-by: Marcelo Tosatti <mtosatti@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxx>
Signed-off-by: Andi Kleen <ak@xxxxxxxxxxxxxxx>

---
 arch/x86/kvm/svm.c |    2 +-
 arch/x86/kvm/vmx.c |    4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

Index: linux-2.6.35.y/arch/x86/kvm/svm.c
===================================================================
--- linux-2.6.35.y.orig/arch/x86/kvm/svm.c	2011-03-29 23:53:08.796312600 -0700
+++ linux-2.6.35.y/arch/x86/kvm/svm.c	2011-03-29 23:53:15.697136026 -0700
@@ -3185,6 +3185,7 @@
 	vcpu->arch.regs[VCPU_REGS_RIP] = svm->vmcb->save.rip;
 
 	load_host_msrs(vcpu);
+	kvm_load_ldt(ldt_selector);
 	loadsegment(fs, fs_selector);
 #ifdef CONFIG_X86_64
 	load_gs_index(gs_selector);
@@ -3192,7 +3193,6 @@
 #else
 	loadsegment(gs, gs_selector);
 #endif
-	kvm_load_ldt(ldt_selector);
 
 	reload_tss(vcpu);
 
Index: linux-2.6.35.y/arch/x86/kvm/vmx.c
===================================================================
--- linux-2.6.35.y.orig/arch/x86/kvm/vmx.c	2011-03-29 23:53:08.868310757 -0700
+++ linux-2.6.35.y/arch/x86/kvm/vmx.c	2011-03-29 23:53:32.094716453 -0700
@@ -789,8 +789,6 @@
 
 	++vmx->vcpu.stat.host_state_reload;
 	vmx->host_state.loaded = 0;
-	if (vmx->host_state.fs_reload_needed)
-		loadsegment(fs, vmx->host_state.fs_sel);
 	if (vmx->host_state.gs_ldt_reload_needed) {
 		kvm_load_ldt(vmx->host_state.ldt_sel);
 #ifdef CONFIG_X86_64
@@ -800,6 +798,8 @@
 		loadsegment(gs, vmx->host_state.gs_sel);
 #endif
 	}
+	if (vmx->host_state.fs_reload_needed)
+		loadsegment(fs, vmx->host_state.fs_sel);
 	reload_tss();
 #ifdef CONFIG_X86_64
 	if (is_long_mode(&vmx->vcpu)) {
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/