1 files changed, 71 insertions, 30 deletions

diff --git a/arch/x86/xen/enlighten_pv.c b/arch/x86/xen/enlighten_pv.c
index 9b1a58dda935..f82857e48815 100644
--- a/arch/x86/xen/enlighten_pv.c
+++ b/arch/x86/xen/enlighten_pv.c
@@ -108,11 +108,21 @@ struct tls_descs {
  */
 static DEFINE_PER_CPU(struct tls_descs, shadow_tls_desc);
 
+static __read_mostly bool xen_msr_safe = IS_ENABLED(CONFIG_XEN_PV_MSR_SAFE);
+
+static int __init parse_xen_msr_safe(char *str)
+{
+	if (str)
+		return strtobool(str, &xen_msr_safe);
+	return -EINVAL;
+}
+early_param("xen_msr_safe", parse_xen_msr_safe);
+
 static void __init xen_pv_init_platform(void)
 {
 	/* PV guests can't operate virtio devices without grants. */
 	if (IS_ENABLED(CONFIG_XEN_VIRTIO))
-		virtio_set_mem_acc_cb(virtio_require_restricted_mem_acc);
+		virtio_set_mem_acc_cb(xen_virtio_restricted_mem_acc);
 
 	populate_extra_pte(fix_to_virt(FIX_PARAVIRT_BOOTMAP));
 
@@ -917,14 +927,18 @@ static void xen_write_cr4(unsigned long cr4)
 	native_write_cr4(cr4);
 }
 
-static u64 xen_read_msr_safe(unsigned int msr, int *err)
+static u64 xen_do_read_msr(unsigned int msr, int *err)
 {
-	u64 val;
+	u64 val = 0;	/* Avoid uninitialized value for safe variant. */
 
 	if (pmu_msr_read(msr, &val, err))
 		return val;
 
-	val = native_read_msr_safe(msr, err);
+	if (err)
+		val = native_read_msr_safe(msr, err);
+	else
+		val = native_read_msr(msr);
+
 	switch (msr) {
 	case MSR_IA32_APICBASE:
 		val &= ~X2APIC_ENABLE;
@@ -933,23 +947,39 @@ static u64 xen_read_msr_safe(unsigned int msr, int *err)
 	return val;
 }
 
-static int xen_write_msr_safe(unsigned int msr, unsigned low, unsigned high)
+static void set_seg(unsigned int which, unsigned int low, unsigned int high,
+		    int *err)
 {
-	int ret;
-	unsigned int which;
-	u64 base;
+	u64 base = ((u64)high << 32) | low;
 
-	ret = 0;
+	if (HYPERVISOR_set_segment_base(which, base) == 0)
+		return;
 
+	if (err)
+		*err = -EIO;
+	else
+		WARN(1, "Xen set_segment_base(%u, %llx) failed\n", which, base);
+}
+
+/*
+ * Support write_msr_safe() and write_msr() semantics.
+ * With err == NULL write_msr() semantics are selected.
+ * Supplying an err pointer requires err to be pre-initialized with 0.
+ */
+static void xen_do_write_msr(unsigned int msr, unsigned int low,
+			     unsigned int high, int *err)
+{
 	switch (msr) {
-	case MSR_FS_BASE:		which = SEGBASE_FS; goto set;
-	case MSR_KERNEL_GS_BASE:	which = SEGBASE_GS_USER; goto set;
-	case MSR_GS_BASE:		which = SEGBASE_GS_KERNEL; goto set;
-
-	set:
-		base = ((u64)high << 32) | low;
-		if (HYPERVISOR_set_segment_base(which, base) != 0)
-			ret = -EIO;
+	case MSR_FS_BASE:
+		set_seg(SEGBASE_FS, low, high, err);
+		break;
+
+	case MSR_KERNEL_GS_BASE:
+		set_seg(SEGBASE_GS_USER, low, high, err);
+		break;
+
+	case MSR_GS_BASE:
+		set_seg(SEGBASE_GS_KERNEL, low, high, err);
 		break;
 
 	case MSR_STAR:
@@ -965,31 +995,42 @@ static int xen_write_msr_safe(unsigned int msr, unsigned low, unsigned high)
 		break;
 
 	default:
-		if (!pmu_msr_write(msr, low, high, &ret))
-			ret = native_write_msr_safe(msr, low, high);
+		if (!pmu_msr_write(msr, low, high, err)) {
+			if (err)
+				*err = native_write_msr_safe(msr, low, high);
+			else
+				native_write_msr(msr, low, high);
+		}
 	}
+}
 
-	return ret;
+static u64 xen_read_msr_safe(unsigned int msr, int *err)
+{
+	return xen_do_read_msr(msr, err);
+}
+
+static int xen_write_msr_safe(unsigned int msr, unsigned int low,
+			      unsigned int high)
+{
+	int err = 0;
+
+	xen_do_write_msr(msr, low, high, &err);
+
+	return err;
 }
 
 static u64 xen_read_msr(unsigned int msr)
 {
-	/*
-	 * This will silently swallow a #GP from RDMSR.  It may be worth
-	 * changing that.
-	 */
 	int err;
 
-	return xen_read_msr_safe(msr, &err);
+	return xen_do_read_msr(msr, xen_msr_safe ? &err : NULL);
 }
 
 static void xen_write_msr(unsigned int msr, unsigned low, unsigned high)
 {
-	/*
-	 * This will silently swallow a #GP from WRMSR.  It may be worth
-	 * changing that.
-	 */
-	xen_write_msr_safe(msr, low, high);
+	int err;
+
+	xen_do_write_msr(msr, low, high, xen_msr_safe ? &err : NULL);
 }
 
 /* This is called once we have the cpu_possible_mask */