Merge tag 'rfds-for-linus-2024-03-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

Pull x86 RFDS mitigation from Dave Hansen:
 "RFDS is a CPU vulnerability that may allow a malicious userspace to
  infer stale register values from kernel space. Kernel registers can
  have all kinds of secrets in them so the mitigation is basically to
  wait until the kernel is about to return to userspace and has user
  values in the registers. At that point there is little chance of
  kernel secrets ending up in the registers and the microarchitectural
  state can be cleared.

  This leverages some recent robustness fixes for the existing MDS
  vulnerability. Both MDS and RFDS use the VERW instruction for
  mitigation"

* tag 'rfds-for-linus-2024-03-11' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  KVM/x86: Export RFDS_NO and RFDS_CLEAR to guests
  x86/rfds: Mitigate Register File Data Sampling (RFDS)
  Documentation/hw-vuln: Add documentation for RFDS
  x86/mmio: Disable KVM mitigation when X86_FEATURE_CLEAR_CPU_BUF is set

6 files changed, 146 insertions, 9 deletions

diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index 8f7271d9f1d7..77e8c8c67950 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -2623,6 +2623,17 @@ config MITIGATION_GDS_FORCE
 
 	  If in doubt, say N.
 
+config MITIGATION_RFDS
+	bool "RFDS Mitigation"
+	depends on CPU_SUP_INTEL
+	default y
+	help
+	  Enable mitigation for Register File Data Sampling (RFDS) by default.
+	  RFDS is a hardware vulnerability which affects Intel Atom CPUs. It
+	  allows unprivileged speculative access to stale data previously
+	  stored in floating point, vector and integer registers.
+	  See also <file:Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst>
+
 endif
 
 config ARCH_HAS_ADD_PAGES
diff --git a/arch/x86/include/asm/cpufeatures.h b/arch/x86/include/asm/cpufeatures.h
index 0343caa016a9..f0337f7bcf16 100644
--- a/arch/x86/include/asm/cpufeatures.h
+++ b/arch/x86/include/asm/cpufeatures.h
@@ -506,4 +506,5 @@
 /* BUG word 2 */
 #define X86_BUG_SRSO			X86_BUG(1*32 + 0) /* AMD SRSO bug */
 #define X86_BUG_DIV0			X86_BUG(1*32 + 1) /* AMD DIV0 speculation bug */
+#define X86_BUG_RFDS			X86_BUG(1*32 + 2) /* CPU is vulnerable to Register File Data Sampling */
 #endif /* _ASM_X86_CPUFEATURES_H */
diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h
index 24c575cdd6b9..05956bd8bacf 100644
--- a/arch/x86/include/asm/msr-index.h
+++ b/arch/x86/include/asm/msr-index.h
@@ -176,6 +176,14 @@
 						 * CPU is not vulnerable to Gather
 						 * Data Sampling (GDS).
 						 */
+#define ARCH_CAP_RFDS_NO		BIT(27)	/*
+						 * Not susceptible to Register
+						 * File Data Sampling.
+						 */
+#define ARCH_CAP_RFDS_CLEAR		BIT(28)	/*
+						 * VERW clears CPU Register
+						 * File.
+						 */
 
 #define ARCH_CAP_XAPIC_DISABLE		BIT(21)	/*
 						 * IA32_XAPIC_DISABLE_STATUS MSR
diff --git a/arch/x86/kernel/cpu/bugs.c b/arch/x86/kernel/cpu/bugs.c
index 4dd00066c12a..e7ba936d798b 100644
--- a/arch/x86/kernel/cpu/bugs.c
+++ b/arch/x86/kernel/cpu/bugs.c
@@ -422,6 +422,13 @@ static void __init mmio_select_mitigation(void)
 	if (boot_cpu_has_bug(X86_BUG_MDS) || (boot_cpu_has_bug(X86_BUG_TAA) &&
 					      boot_cpu_has(X86_FEATURE_RTM)))
 		setup_force_cpu_cap(X86_FEATURE_CLEAR_CPU_BUF);
+
+	/*
+	 * X86_FEATURE_CLEAR_CPU_BUF could be enabled by other VERW based
+	 * mitigations, disable KVM-only mitigation in that case.
+	 */
+	if (boot_cpu_has(X86_FEATURE_CLEAR_CPU_BUF))
+		static_branch_disable(&mmio_stale_data_clear);
 	else
 		static_branch_enable(&mmio_stale_data_clear);
 
@@ -474,6 +481,57 @@ static int __init mmio_stale_data_parse_cmdline(char *str)
 early_param("mmio_stale_data", mmio_stale_data_parse_cmdline);
 
 #undef pr_fmt
+#define pr_fmt(fmt)	"Register File Data Sampling: " fmt
+
+enum rfds_mitigations {
+	RFDS_MITIGATION_OFF,
+	RFDS_MITIGATION_VERW,
+	RFDS_MITIGATION_UCODE_NEEDED,
+};
+
+/* Default mitigation for Register File Data Sampling */
+static enum rfds_mitigations rfds_mitigation __ro_after_init =
+	IS_ENABLED(CONFIG_MITIGATION_RFDS) ? RFDS_MITIGATION_VERW : RFDS_MITIGATION_OFF;
+
+static const char * const rfds_strings[] = {
+	[RFDS_MITIGATION_OFF]			= "Vulnerable",
+	[RFDS_MITIGATION_VERW]			= "Mitigation: Clear Register File",
+	[RFDS_MITIGATION_UCODE_NEEDED]		= "Vulnerable: No microcode",
+};
+
+static void __init rfds_select_mitigation(void)
+{
+	if (!boot_cpu_has_bug(X86_BUG_RFDS) || cpu_mitigations_off()) {
+		rfds_mitigation = RFDS_MITIGATION_OFF;
+		return;
+	}
+	if (rfds_mitigation == RFDS_MITIGATION_OFF)
+		return;
+
+	if (x86_read_arch_cap_msr() & ARCH_CAP_RFDS_CLEAR)
+		setup_force_cpu_cap(X86_FEATURE_CLEAR_CPU_BUF);
+	else
+		rfds_mitigation = RFDS_MITIGATION_UCODE_NEEDED;
+}
+
+static __init int rfds_parse_cmdline(char *str)
+{
+	if (!str)
+		return -EINVAL;
+
+	if (!boot_cpu_has_bug(X86_BUG_RFDS))
+		return 0;
+
+	if (!strcmp(str, "off"))
+		rfds_mitigation = RFDS_MITIGATION_OFF;
+	else if (!strcmp(str, "on"))
+		rfds_mitigation = RFDS_MITIGATION_VERW;
+
+	return 0;
+}
+early_param("reg_file_data_sampling", rfds_parse_cmdline);
+
+#undef pr_fmt
 #define pr_fmt(fmt)     "" fmt
 
 static void __init md_clear_update_mitigation(void)
@@ -498,11 +556,19 @@ static void __init md_clear_update_mitigation(void)
 		taa_mitigation = TAA_MITIGATION_VERW;
 		taa_select_mitigation();
 	}
-	if (mmio_mitigation == MMIO_MITIGATION_OFF &&
-	    boot_cpu_has_bug(X86_BUG_MMIO_STALE_DATA)) {
+	/*
+	 * MMIO_MITIGATION_OFF is not checked here so that mmio_stale_data_clear
+	 * gets updated correctly as per X86_FEATURE_CLEAR_CPU_BUF state.
+	 */
+	if (boot_cpu_has_bug(X86_BUG_MMIO_STALE_DATA)) {
 		mmio_mitigation = MMIO_MITIGATION_VERW;
 		mmio_select_mitigation();
 	}
+	if (rfds_mitigation == RFDS_MITIGATION_OFF &&
+	    boot_cpu_has_bug(X86_BUG_RFDS)) {
+		rfds_mitigation = RFDS_MITIGATION_VERW;
+		rfds_select_mitigation();
+	}
 out:
 	if (boot_cpu_has_bug(X86_BUG_MDS))
 		pr_info("MDS: %s\n", mds_strings[mds_mitigation]);
@@ -512,6 +578,8 @@ out:
 		pr_info("MMIO Stale Data: %s\n", mmio_strings[mmio_mitigation]);
 	else if (boot_cpu_has_bug(X86_BUG_MMIO_UNKNOWN))
 		pr_info("MMIO Stale Data: Unknown: No mitigations\n");
+	if (boot_cpu_has_bug(X86_BUG_RFDS))
+		pr_info("Register File Data Sampling: %s\n", rfds_strings[rfds_mitigation]);
 }
 
 static void __init md_clear_select_mitigation(void)
@@ -519,11 +587,12 @@ static void __init md_clear_select_mitigation(void)
 	mds_select_mitigation();
 	taa_select_mitigation();
 	mmio_select_mitigation();
+	rfds_select_mitigation();
 
 	/*
-	 * As MDS, TAA and MMIO Stale Data mitigations are inter-related, update
-	 * and print their mitigation after MDS, TAA and MMIO Stale Data
-	 * mitigation selection is done.
+	 * As these mitigations are inter-related and rely on VERW instruction
+	 * to clear the microarchitural buffers, update and print their status
+	 * after mitigation selection is done for each of these vulnerabilities.
 	 */
 	md_clear_update_mitigation();
 }
@@ -2613,6 +2682,11 @@ static ssize_t mmio_stale_data_show_state(char *buf)
 			  sched_smt_active() ? "vulnerable" : "disabled");
 }
 
+static ssize_t rfds_show_state(char *buf)
+{
+	return sysfs_emit(buf, "%s\n", rfds_strings[rfds_mitigation]);
+}
+
 static char *stibp_state(void)
 {
 	if (spectre_v2_in_eibrs_mode(spectre_v2_enabled) &&
@@ -2772,6 +2846,9 @@ static ssize_t cpu_show_common(struct device *dev, struct device_attribute *attr
 	case X86_BUG_GDS:
 		return gds_show_state(buf);
 
+	case X86_BUG_RFDS:
+		return rfds_show_state(buf);
+
 	default:
 		break;
 	}
@@ -2846,6 +2923,11 @@ ssize_t cpu_show_gds(struct device *dev, struct device_attribute *attr, char *bu
 {
 	return cpu_show_common(dev, attr, buf, X86_BUG_GDS);
 }
+
+ssize_t cpu_show_reg_file_data_sampling(struct device *dev, struct device_attribute *attr, char *buf)
+{
+	return cpu_show_common(dev, attr, buf, X86_BUG_RFDS);
+}
 #endif
 
 void __warn_thunk(void)
diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c
index 39792077f65b..ba8cf5e9ce56 100644
--- a/arch/x86/kernel/cpu/common.c
+++ b/arch/x86/kernel/cpu/common.c
@@ -1224,6 +1224,8 @@ static const __initconst struct x86_cpu_id cpu_vuln_whitelist[] = {
 #define SRSO		BIT(5)
 /* CPU is affected by GDS */
 #define GDS		BIT(6)
+/* CPU is affected by Register File Data Sampling */
+#define RFDS		BIT(7)
 
 static const struct x86_cpu_id cpu_vuln_blacklist[] __initconst = {
 	VULNBL_INTEL_STEPPINGS(IVYBRIDGE,	X86_STEPPING_ANY,		SRBDS),
@@ -1251,9 +1253,18 @@ static const struct x86_cpu_id cpu_vuln_blacklist[] __initconst = {
 	VULNBL_INTEL_STEPPINGS(TIGERLAKE,	X86_STEPPING_ANY,		GDS),
 	VULNBL_INTEL_STEPPINGS(LAKEFIELD,	X86_STEPPING_ANY,		MMIO | MMIO_SBDS | RETBLEED),
 	VULNBL_INTEL_STEPPINGS(ROCKETLAKE,	X86_STEPPING_ANY,		MMIO | RETBLEED | GDS),
-	VULNBL_INTEL_STEPPINGS(ATOM_TREMONT,	X86_STEPPING_ANY,		MMIO | MMIO_SBDS),
-	VULNBL_INTEL_STEPPINGS(ATOM_TREMONT_D,	X86_STEPPING_ANY,		MMIO),
-	VULNBL_INTEL_STEPPINGS(ATOM_TREMONT_L,	X86_STEPPING_ANY,		MMIO | MMIO_SBDS),
+	VULNBL_INTEL_STEPPINGS(ALDERLAKE,	X86_STEPPING_ANY,		RFDS),
+	VULNBL_INTEL_STEPPINGS(ALDERLAKE_L,	X86_STEPPING_ANY,		RFDS),
+	VULNBL_INTEL_STEPPINGS(RAPTORLAKE,	X86_STEPPING_ANY,		RFDS),
+	VULNBL_INTEL_STEPPINGS(RAPTORLAKE_P,	X86_STEPPING_ANY,		RFDS),
+	VULNBL_INTEL_STEPPINGS(RAPTORLAKE_S,	X86_STEPPING_ANY,		RFDS),
+	VULNBL_INTEL_STEPPINGS(ATOM_GRACEMONT,	X86_STEPPING_ANY,		RFDS),
+	VULNBL_INTEL_STEPPINGS(ATOM_TREMONT,	X86_STEPPING_ANY,		MMIO | MMIO_SBDS | RFDS),
+	VULNBL_INTEL_STEPPINGS(ATOM_TREMONT_D,	X86_STEPPING_ANY,		MMIO | RFDS),
+	VULNBL_INTEL_STEPPINGS(ATOM_TREMONT_L,	X86_STEPPING_ANY,		MMIO | MMIO_SBDS | RFDS),
+	VULNBL_INTEL_STEPPINGS(ATOM_GOLDMONT,	X86_STEPPING_ANY,		RFDS),
+	VULNBL_INTEL_STEPPINGS(ATOM_GOLDMONT_D,	X86_STEPPING_ANY,		RFDS),
+	VULNBL_INTEL_STEPPINGS(ATOM_GOLDMONT_PLUS, X86_STEPPING_ANY,		RFDS),
 
 	VULNBL_AMD(0x15, RETBLEED),
 	VULNBL_AMD(0x16, RETBLEED),
@@ -1287,6 +1298,24 @@ static bool arch_cap_mmio_immune(u64 ia32_cap)
 		ia32_cap & ARCH_CAP_SBDR_SSDP_NO);
 }
 
+static bool __init vulnerable_to_rfds(u64 ia32_cap)
+{
+	/* The "immunity" bit trumps everything else: */
+	if (ia32_cap & ARCH_CAP_RFDS_NO)
+		return false;
+
+	/*
+	 * VMMs set ARCH_CAP_RFDS_CLEAR for processors not in the blacklist to
+	 * indicate that mitigation is needed because guest is running on a
+	 * vulnerable hardware or may migrate to such hardware:
+	 */
+	if (ia32_cap & ARCH_CAP_RFDS_CLEAR)
+		return true;
+
+	/* Only consult the blacklist when there is no enumeration: */
+	return cpu_matches(cpu_vuln_blacklist, RFDS);
+}
+
 static void __init cpu_set_bug_bits(struct cpuinfo_x86 *c)
 {
 	u64 ia32_cap = x86_read_arch_cap_msr();
@@ -1403,6 +1432,9 @@ static void __init cpu_set_bug_bits(struct cpuinfo_x86 *c)
 	    boot_cpu_has(X86_FEATURE_AVX))
 		setup_force_cpu_bug(X86_BUG_GDS);
 
+	if (vulnerable_to_rfds(ia32_cap))
+		setup_force_cpu_bug(X86_BUG_RFDS);
+
 	if (cpu_matches(cpu_vuln_whitelist, NO_MELTDOWN))
 		return;
 
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index e02cc710f56d..ffe580169c93 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -1623,7 +1623,8 @@ static bool kvm_is_immutable_feature_msr(u32 msr)
 	 ARCH_CAP_SKIP_VMENTRY_L1DFLUSH | ARCH_CAP_SSB_NO | ARCH_CAP_MDS_NO | \
 	 ARCH_CAP_PSCHANGE_MC_NO | ARCH_CAP_TSX_CTRL_MSR | ARCH_CAP_TAA_NO | \
 	 ARCH_CAP_SBDR_SSDP_NO | ARCH_CAP_FBSDP_NO | ARCH_CAP_PSDP_NO | \
-	 ARCH_CAP_FB_CLEAR | ARCH_CAP_RRSBA | ARCH_CAP_PBRSB_NO | ARCH_CAP_GDS_NO)
+	 ARCH_CAP_FB_CLEAR | ARCH_CAP_RRSBA | ARCH_CAP_PBRSB_NO | ARCH_CAP_GDS_NO | \
+	 ARCH_CAP_RFDS_NO | ARCH_CAP_RFDS_CLEAR)
 
 static u64 kvm_get_arch_capabilities(void)
 {
@@ -1655,6 +1656,8 @@ static u64 kvm_get_arch_capabilities(void)
 		data |= ARCH_CAP_SSB_NO;
 	if (!boot_cpu_has_bug(X86_BUG_MDS))
 		data |= ARCH_CAP_MDS_NO;
+	if (!boot_cpu_has_bug(X86_BUG_RFDS))
+		data |= ARCH_CAP_RFDS_NO;
 
 	if (!boot_cpu_has(X86_FEATURE_RTM)) {
 		/*