Skip to contentSkip to conversation

Enable CET userspace support

PATCH New Yang Weijiang 1 participant 5 messages Jul 20, 2023 KVM

lore
Share & tools
lore
#0 lore raw

v2 cover Enable CET userspace support

CET virtualization requires VMM userspace setup for CET feature bits enumeration, this series enables all related settings.

Guest CET user and kernel mode SHSTK/IBT are both supported.

Changes in v2:
1. Added supervisor CET states support.
2. Consolidated all CET states in a struct of vmstate.

v1 link:
https://lore.kernel.org/all/[email protected]/

Yang Weijiang (4):
target/i386: Enable XSAVES support for CET states target/i386: Add CET MSRs access interface target/i386: Add CET states to vmstate target/i386: Advertise CET related flags in feature words

 target/i386/cpu.c     | 54 +++++++++++++++++++++++++++++----------
 target/i386/cpu.h     | 41 ++++++++++++++++++++++++++++++
 target/i386/kvm/kvm.c | 59 +++++++++++++++++++++++++++++++++++++++++++
 target/i386/machine.c | 28 ++++++++++++++++++++
 4 files changed, 169 insertions(+), 13 deletions(-)

base-commit: a342ce9dfeed8088c426e5d51d4a7e47f3764b84

Signature · 1 line
2.27.0
#1 lore raw

PATCH v2 1/4 target/i386: Enable XSAVES support for CET states

Add CET_U/S bits in xstate area and report support in xstate feature mask.
MSR_XSS[bit 11] corresponds to CET user mode states. MSR_XSS[bit 12] corresponds to CET supervisor mode states.

CET Shadow Stack(SHSTK) and Indirect Branch Tracking(IBT) features are enumerated via CPUID.(EAX=07H,ECX=0H):ECX[7] and EDX[20] respectively, two featues share the same state bits in XSS,so if either of the features is enabled, set CET_U and CET_S bits together.

Opportunistically fix the array format issue.

---
 target/i386/cpu.c | 45 ++++++++++++++++++++++++++++++++++++---------
 target/i386/cpu.h | 23 +++++++++++++++++++++++
 2 files changed, 59 insertions(+), 9 deletions(-)
target/i386/cpu.c +36 -9
index f083ff4335..ea11b589e3 100644
@@ -944,8 +944,8 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
         .feat_names = {
             NULL, NULL, NULL, NULL,
             NULL, NULL, NULL, NULL,
-            NULL, NULL, NULL, NULL,
-            NULL, NULL, NULL, NULL,
+            NULL, NULL, NULL, "cet-u",
+            "cet-s", NULL, NULL, NULL,
             NULL, NULL, NULL, NULL,
             NULL, NULL, NULL, NULL,
             NULL, NULL, NULL, NULL,
@@ -1421,7 +1421,8 @@ static const X86RegisterInfo32 x86_reg_info_32[CPU_NB_REGS32] = {
 #undef REGISTER
 
 /* CPUID feature bits available in XSS */
-#define CPUID_XSTATE_XSS_MASK    (XSTATE_ARCH_LBR_MASK)
+#define CPUID_XSTATE_XSS_MASK    (XSTATE_ARCH_LBR_MASK | XSTATE_CET_U_MASK | \
+                                  XSTATE_CET_S_MASK)
 
 ExtSaveArea x86_ext_save_areas[XSAVE_STATE_AREA_COUNT] = {
     [XSTATE_FP_BIT] = {
@@ -1439,7 +1440,7 @@ ExtSaveArea x86_ext_save_areas[XSAVE_STATE_AREA_COUNT] = {
             .size = sizeof(XSaveAVX) },
     [XSTATE_BNDREGS_BIT] =
           { .feature = FEAT_7_0_EBX, .bits = CPUID_7_0_EBX_MPX,
-            .size = sizeof(XSaveBNDREG)  },
+            .size = sizeof(XSaveBNDREG) },
     [XSTATE_BNDCSR_BIT] =
           { .feature = FEAT_7_0_EBX, .bits = CPUID_7_0_EBX_MPX,
             .size = sizeof(XSaveBNDCSR)  },
@@ -1459,14 +1460,24 @@ ExtSaveArea x86_ext_save_areas[XSAVE_STATE_AREA_COUNT] = {
             .feature = FEAT_7_0_EDX, .bits = CPUID_7_0_EDX_ARCH_LBR,
             .offset = 0 /*supervisor mode component, offset = 0 */,
             .size = sizeof(XSavesArchLBR) },
+    [XSTATE_CET_U_BIT] = {
+        .feature = FEAT_7_0_ECX, .bits = CPUID_7_0_ECX_CET_SHSTK,
+        /*
+         * The features enabled in XSS MSR always use compacted format
+         * to store the data, in this case .offset == 0.
+         */
+        .offset = 0,
+        .size = sizeof(XSavesCETU) },
+    [XSTATE_CET_S_BIT] = {
+        .feature = FEAT_7_0_ECX, .bits = CPUID_7_0_ECX_CET_SHSTK,
+        .offset = 0,
+        .size = sizeof(XSavesCETS) },
     [XSTATE_XTILE_CFG_BIT] = {
         .feature = FEAT_7_0_EDX, .bits = CPUID_7_0_EDX_AMX_TILE,
-        .size = sizeof(XSaveXTILECFG),
-    },
+        .size = sizeof(XSaveXTILECFG) },
     [XSTATE_XTILE_DATA_BIT] = {
         .feature = FEAT_7_0_EDX, .bits = CPUID_7_0_EDX_AMX_TILE,
-        .size = sizeof(XSaveXTILEDATA)
-    },
+        .size = sizeof(XSaveXTILEDATA) }
 };
 
 uint32_t xsave_area_size(uint64_t mask, bool compacted)
@@ -6259,9 +6270,25 @@ static void x86_cpu_enable_xsave_components(X86CPU *cpu)
         if (env->features[esa->feature] & esa->bits) {
             mask |= (1ULL << i);
         }
+
+        /*
+         * Both CET SHSTK and IBT feature depend on XSAVES support, and two
+         * features can be enabled independently, so if either of the two
+         * features is enabled, we set the XSAVES support bits to make the
+         * enabled feature(s) work.
+         */
+        if (i == XSTATE_CET_U_BIT || i == XSTATE_CET_S_BIT) {
+            uint64_t ecx = env->features[FEAT_7_0_ECX];
+            uint64_t edx = env->features[FEAT_7_0_EDX];
+
+            if ((ecx & CPUID_7_0_ECX_CET_SHSTK) ||
+                (edx & CPUID_7_0_EDX_CET_IBT)) {
+                mask |= (1ULL << i);
+            }
+        }
     }
 
-    /* Only request permission for first vcpu */
+    /* Only request permission from fisrt vcpu. */
     if (kvm_enabled() && !request_perm) {
         kvm_request_xsave_components(cpu, mask);
         request_perm = true;
target/i386/cpu.h +23 -0
index d243e290d3..06855e0926 100644
@@ -554,6 +554,8 @@ typedef enum X86Seg {
 #define XSTATE_ZMM_Hi256_BIT            6
 #define XSTATE_Hi16_ZMM_BIT             7
 #define XSTATE_PKRU_BIT                 9
+#define XSTATE_CET_U_BIT                11
+#define XSTATE_CET_S_BIT                12
 #define XSTATE_ARCH_LBR_BIT             15
 #define XSTATE_XTILE_CFG_BIT            17
 #define XSTATE_XTILE_DATA_BIT           18
@@ -567,6 +569,8 @@ typedef enum X86Seg {
 #define XSTATE_ZMM_Hi256_MASK           (1ULL << XSTATE_ZMM_Hi256_BIT)
 #define XSTATE_Hi16_ZMM_MASK            (1ULL << XSTATE_Hi16_ZMM_BIT)
 #define XSTATE_PKRU_MASK                (1ULL << XSTATE_PKRU_BIT)
+#define XSTATE_CET_U_MASK               (1ULL << XSTATE_CET_U_BIT)
+#define XSTATE_CET_S_MASK               (1ULL << XSTATE_CET_S_BIT)
 #define XSTATE_ARCH_LBR_MASK            (1ULL << XSTATE_ARCH_LBR_BIT)
 #define XSTATE_XTILE_CFG_MASK           (1ULL << XSTATE_XTILE_CFG_BIT)
 #define XSTATE_XTILE_DATA_MASK          (1ULL << XSTATE_XTILE_DATA_BIT)
@@ -841,6 +845,8 @@ uint64_t x86_cpu_get_supported_feature_word(FeatureWord w,
 #define CPUID_7_0_ECX_WAITPKG           (1U << 5)
 /* Additional AVX-512 Vector Byte Manipulation Instruction */
 #define CPUID_7_0_ECX_AVX512_VBMI2      (1U << 6)
+/* CET SHSTK feature */
+#define CPUID_7_0_ECX_CET_SHSTK         (1U << 7)
 /* Galois Field New Instructions */
 #define CPUID_7_0_ECX_GFNI              (1U << 8)
 /* Vector AES Instructions */
@@ -884,6 +890,8 @@ uint64_t x86_cpu_get_supported_feature_word(FeatureWord w,
 #define CPUID_7_0_EDX_TSX_LDTRK         (1U << 16)
 /* Architectural LBRs */
 #define CPUID_7_0_EDX_ARCH_LBR          (1U << 19)
+/* CET IBT feature */
+#define CPUID_7_0_EDX_CET_IBT           (1U << 20)
 /* AMX_BF16 instruction */
 #define CPUID_7_0_EDX_AMX_BF16          (1U << 22)
 /* AVX512_FP16 instruction */
@@ -1428,6 +1436,19 @@ typedef struct XSavePKRU {
     uint32_t padding;
 } XSavePKRU;
 
+/* Ext. save area 11: User mode CET state */
+typedef struct XSavesCETU {
+    uint64_t u_cet;
+    uint64_t user_ssp;
+} XSavesCETU;
+
+/* Ext. save area 12: Supervisor mode CET state */
+typedef struct XSavesCETS {
+    uint64_t pl0_ssp;
+    uint64_t pl1_ssp;
+    uint64_t pl2_ssp;
+} XSavesCETS;
+
 /* Ext. save area 17: AMX XTILECFG state */
 typedef struct XSaveXTILECFG {
     uint8_t xtilecfg[64];
@@ -1463,6 +1484,8 @@ QEMU_BUILD_BUG_ON(sizeof(XSaveOpmask) != 0x40);
 QEMU_BUILD_BUG_ON(sizeof(XSaveZMM_Hi256) != 0x200);
 QEMU_BUILD_BUG_ON(sizeof(XSaveHi16_ZMM) != 0x400);
 QEMU_BUILD_BUG_ON(sizeof(XSavePKRU) != 0x8);
+QEMU_BUILD_BUG_ON(sizeof(XSavesCETU) != 0x10);
+QEMU_BUILD_BUG_ON(sizeof(XSavesCETS) != 0x18);
 QEMU_BUILD_BUG_ON(sizeof(XSaveXTILECFG) != 0x40);
 QEMU_BUILD_BUG_ON(sizeof(XSaveXTILEDATA) != 0x2000);
 QEMU_BUILD_BUG_ON(sizeof(XSavesArchLBR) != 0x328);
Signature · 1 line
2.27.0
#2 lore raw

PATCH v2 2/4 target/i386: Add CET MSRs access interface

In reply to Yang Weijiang

CET MSRs include:
MSR_IA32_U_CET - user mode CET control bits.
MSR_IA32_S_CET - supervisor mode CET control bits. MSR_IA32_PL{0,1,2,3}_SSP - linear addresses of SSPs for user/kernel modes. MSR_IA32_SSP_TBL_ADDR - linear address of interrupt SSP table MSR_KVM_GUEST_SSP - current shadow stack pointer

---
 target/i386/cpu.h     | 18 +++++++++++++
 target/i386/kvm/kvm.c | 59 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 77 insertions(+)
target/i386/cpu.h +18 -0
index 06855e0926..ef1f3d6138 100644
@@ -545,6 +545,15 @@ typedef enum X86Seg {
 #define MSR_IA32_VMX_TRUE_ENTRY_CTLS     0x00000490
 #define MSR_IA32_VMX_VMFUNC             0x00000491
 
+#define MSR_IA32_U_CET                  0x000006a0
+#define MSR_IA32_S_CET                  0x000006a2
+#define MSR_IA32_PL0_SSP                0x000006a4
+#define MSR_IA32_PL1_SSP                0x000006a5
+#define MSR_IA32_PL2_SSP                0x000006a6
+#define MSR_IA32_PL3_SSP                0x000006a7
+#define MSR_IA32_SSP_TBL_ADDR           0x000006a8
+#define MSR_KVM_GUEST_SSP               0x4b564d09
+
 #define XSTATE_FP_BIT                   0
 #define XSTATE_SSE_BIT                  1
 #define XSTATE_YMM_BIT                  2
@@ -1766,6 +1775,15 @@ typedef struct CPUArchState {
 
     uintptr_t retaddr;
 
+    uint64_t u_cet;
+    uint64_t s_cet;
+    uint64_t pl0_ssp;
+    uint64_t pl1_ssp;
+    uint64_t pl2_ssp;
+    uint64_t pl3_ssp;
+    uint64_t ssp_table_addr;
+    uint64_t guest_ssp;
+
     /* Fields up to this point are cleared by a CPU reset */
     struct {} end_reset_fields;
 
target/i386/kvm/kvm.c +59 -0

77 lines not on this page · raw message

Signature · 1 line
2.27.0
#3 lore raw

PATCH v2 3/4 target/i386: Add CET states to vmstate

In reply to Yang Weijiang

Add CET states in vmstate if the feature is enabled.

---
 target/i386/machine.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
target/i386/machine.c +28 -0
index c7ac8084b2..6d42f6dc7e 100644
@@ -1586,6 +1586,33 @@ static const VMStateDescription vmstate_arch_lbr = {
     }
 };
 
+static bool cet_needed(void *opaque)
+{
+    X86CPU *cpu = opaque;
+    CPUX86State *env = &cpu->env;
+
+    return !!((env->features[FEAT_7_0_ECX] & CPUID_7_0_ECX_CET_SHSTK) ||
+              (env->features[FEAT_7_0_EDX] & CPUID_7_0_EDX_CET_IBT));
+}
+
+static const VMStateDescription vmstate_cet = {
+    .name = "cpu/cet",
+    .version_id = 1,
+    .minimum_version_id = 1,
+    .needed = cet_needed,
+    .fields = (VMStateField[]) {
+        VMSTATE_UINT64(env.u_cet, X86CPU),
+        VMSTATE_UINT64(env.s_cet, X86CPU),
+        VMSTATE_UINT64(env.guest_ssp, X86CPU),
+        VMSTATE_UINT64(env.pl0_ssp, X86CPU),
+        VMSTATE_UINT64(env.pl1_ssp, X86CPU),
+        VMSTATE_UINT64(env.pl2_ssp, X86CPU),
+        VMSTATE_UINT64(env.pl3_ssp, X86CPU),
+        VMSTATE_UINT64(env.ssp_table_addr, X86CPU),
+        VMSTATE_END_OF_LIST()
+    }
+};
+
 static bool triple_fault_needed(void *opaque)
 {
     X86CPU *cpu = opaque;
@@ -1745,6 +1772,7 @@ const VMStateDescription vmstate_x86_cpu = {
         &vmstate_msr_tsx_ctrl,
         &vmstate_msr_intel_sgx,
         &vmstate_pdptrs,
+        &vmstate_cet,
         &vmstate_msr_xfd,
 #ifdef TARGET_X86_64
         &vmstate_amx_xtile,
Signature · 1 line
2.27.0
#4 lore raw

PATCH v2 4/4 target/i386: Advertise CET related flags in feature words

In reply to Yang Weijiang

Add SHSTK and IBT flags in feature words with entry/exit control flags.

CET SHSTK and IBT feature are enumerated via CPUID(EAX=7,ECX=0) ECX[bit 7] and EDX[bit 20]. CET states load/restore at vmentry/ vmexit are controlled by VMX_ENTRY_CTLS[bit 20] and VMX_EXIT_CTLS[bit 28]. Enable these flags so that KVM can enumerate the features properly.

CET feature is only available on platforms with IA32_VMX_BASIC[bit 56] set, otherwise, CET features are disabled in KVM.

---
 target/i386/cpu.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
target/i386/cpu.c +5 -4
index ea11b589e3..79a90e35bf 100644
@@ -835,7 +835,7 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
         .type = CPUID_FEATURE_WORD,
         .feat_names = {
             NULL, "avx512vbmi", "umip", "pku",
-            NULL /* ospke */, "waitpkg", "avx512vbmi2", NULL,
+            NULL /* ospke */, "waitpkg", "avx512vbmi2", "shstk",
             "gfni", "vaes", "vpclmulqdq", "avx512vnni",
             "avx512bitalg", NULL, "avx512-vpopcntdq", NULL,
             "la57", NULL, NULL, NULL,
@@ -858,7 +858,7 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
             "avx512-vp2intersect", NULL, "md-clear", NULL,
             NULL, NULL, "serialize", NULL,
             "tsx-ldtrk", NULL, NULL /* pconfig */, "arch-lbr",
-            NULL, NULL, "amx-bf16", "avx512-fp16",
+            "ibt", NULL, "amx-bf16", "avx512-fp16",
             "amx-tile", "amx-int8", "spec-ctrl", "stibp",
             NULL, "arch-capabilities", "core-capability", "ssbd",
         },
@@ -1120,7 +1120,7 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
             "vmx-exit-save-efer", "vmx-exit-load-efer",
                 "vmx-exit-save-preemption-timer", "vmx-exit-clear-bndcfgs",
             NULL, "vmx-exit-clear-rtit-ctl", NULL, NULL,
-            NULL, "vmx-exit-load-pkrs", NULL, NULL,
+            "vmx-exit-save-cet-ctl", "vmx-exit-load-pkrs", NULL, NULL,
         },
         .msr = {
             .index = MSR_IA32_VMX_TRUE_EXIT_CTLS,
@@ -1135,7 +1135,7 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
             NULL, "vmx-entry-ia32e-mode", NULL, NULL,
             NULL, "vmx-entry-load-perf-global-ctrl", "vmx-entry-load-pat", "vmx-entry-load-efer",
             "vmx-entry-load-bndcfgs", NULL, "vmx-entry-load-rtit-ctl", NULL,
-            NULL, NULL, "vmx-entry-load-pkrs", NULL,
+            "vmx-entry-load-cet-ctl", NULL, "vmx-entry-load-pkrs", NULL,
             NULL, NULL, NULL, NULL,
             NULL, NULL, NULL, NULL,
         },
@@ -1192,6 +1192,7 @@ FeatureWordInfo feature_word_info[FEATURE_WORDS] = {
         .feat_names = {
             [54] = "vmx-ins-outs",
             [55] = "vmx-true-ctls",
+            [56] = "vmx-hw-no-errcode",
         },
         .msr = {
             .index = MSR_IA32_VMX_BASIC,
Signature · 1 line
2.27.0