qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
* [Qemu-devel] [PATCH v3 0/4] target-arm: add SHA-3, SM3 and SHA512 instruction support
@ 2018-01-19 14:17 Ard Biesheuvel
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 1/4] target/arm: implement SHA-512 instructions Ard Biesheuvel
                   ` (3 more replies)
  0 siblings, 4 replies; 7+ messages in thread
From: Ard Biesheuvel @ 2018-01-19 14:17 UTC (permalink / raw)
  To: qemu-devel; +Cc: peter.maydell, Ard Biesheuvel

This v3 is a follow-up to the single patch 'target/arm: implement SHA-512
instructions' sent out on the 9th of January.

Changes since v2:
- fix thinko in big-endian aware handling of 64-bit quantities: this is not
  needed given that the NEON registers are represented as arrays of uint64_t
  so they always appear in the correct order.
- add support for SM3 instructions (Chinese SHA derivative)

Changes since v1:
- update SHA512 patch to adhere more closely to the existing style, and to
  the way the instruction encodings are classified in the ARM ARM (#1)
- add patch implementing the new SHA3 instructions EOR3/RAX1/XAR/BCAX (#2)
- enable support for these instructions in user mode emulation (#3)

Ard Biesheuvel (4):
  target/arm: implement SHA-512 instructions
  target/arm: implement SHA-3 instructions
  target/arm: implement SM3 instructions
  target/arm: enable user-mode SHA-3, SM3 and SHA-512 instruction
    support

 linux-user/elfload.c       |  18 ++
 target/arm/cpu.h           |   3 +
 target/arm/cpu64.c         |   3 +
 target/arm/crypto_helper.c | 311 +++++++++++++++++++-
 target/arm/helper.h        |  15 +
 target/arm/translate-a64.c | 255 ++++++++++++++++
 6 files changed, 604 insertions(+), 1 deletion(-)

-- 
2.11.0

^ permalink raw reply	[flat|nested] 7+ messages in thread

* [Qemu-devel] [PATCH v3 1/4] target/arm: implement SHA-512 instructions
  2018-01-19 14:17 [Qemu-devel] [PATCH v3 0/4] target-arm: add SHA-3, SM3 and SHA512 instruction support Ard Biesheuvel
@ 2018-01-19 14:17 ` Ard Biesheuvel
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 2/4] target/arm: implement SHA-3 instructions Ard Biesheuvel
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 7+ messages in thread
From: Ard Biesheuvel @ 2018-01-19 14:17 UTC (permalink / raw)
  To: qemu-devel; +Cc: peter.maydell, Ard Biesheuvel

This implements emulation of the new SHA-512 instructions that have
been added as an optional extensions to the ARMv8 Crypto Extensions
in ARM v8.2.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
 target/arm/cpu.h           |  1 +
 target/arm/crypto_helper.c | 99 +++++++++++++++++++-
 target/arm/helper.h        |  5 +
 target/arm/translate-a64.c | 99 ++++++++++++++++++++
 4 files changed, 203 insertions(+), 1 deletion(-)

diff --git a/target/arm/cpu.h b/target/arm/cpu.h
index 96316700dd6f..295529366c0a 100644
--- a/target/arm/cpu.h
+++ b/target/arm/cpu.h
@@ -1340,6 +1340,7 @@ enum arm_features {
     ARM_FEATURE_VBAR, /* has cp15 VBAR */
     ARM_FEATURE_M_SECURITY, /* M profile Security Extension */
     ARM_FEATURE_JAZELLE, /* has (trivial) Jazelle implementation */
+    ARM_FEATURE_V8_SHA512, /* implements SHA512 part of v8 Crypto Extensions */
 };
 
 static inline int arm_feature(CPUARMState *env, int feature)
diff --git a/target/arm/crypto_helper.c b/target/arm/crypto_helper.c
index 3b6df3f41a42..15c22c82ca5e 100644
--- a/target/arm/crypto_helper.c
+++ b/target/arm/crypto_helper.c
@@ -1,7 +1,7 @@
 /*
  * crypto_helper.c - emulate v8 Crypto Extensions instructions
  *
- * Copyright (C) 2013 - 2014 Linaro Ltd <ard.biesheuvel@linaro.org>
+ * Copyright (C) 2013 - 2018 Linaro Ltd <ard.biesheuvel@linaro.org>
  *
  * This library is free software; you can redistribute it and/or
  * modify it under the terms of the GNU Lesser General Public
@@ -463,3 +463,100 @@ void HELPER(crypto_sha256su1)(CPUARMState *env, uint32_t rd, uint32_t rn,
     env->vfp.regs[rd] = make_float64(d.l[0]);
     env->vfp.regs[rd + 1] = make_float64(d.l[1]);
 }
+
+/*
+ * The SHA-512 logical functions (same as above but using 64-bit operands)
+ */
+
+static uint64_t cho512(uint64_t x, uint64_t y, uint64_t z)
+{
+    return (x & (y ^ z)) ^ z;
+}
+
+static uint64_t maj512(uint64_t x, uint64_t y, uint64_t z)
+{
+    return (x & y) | ((x | y) & z);
+}
+
+static uint64_t S0_512(uint64_t x)
+{
+    return ror64(x, 28) ^ ror64(x, 34) ^ ror64(x, 39);
+}
+
+static uint64_t S1_512(uint64_t x)
+{
+    return ror64(x, 14) ^ ror64(x, 18) ^ ror64(x, 41);
+}
+
+static uint64_t s0_512(uint64_t x)
+{
+    return ror64(x, 1) ^ ror64(x, 8) ^ (x >> 7);
+}
+
+static uint64_t s1_512(uint64_t x)
+{
+    return ror64(x, 19) ^ ror64(x, 61) ^ (x >> 6);
+}
+
+void HELPER(crypto_sha512h)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                            uint32_t rm)
+{
+    uint64_t d0 = float64_val(env->vfp.regs[rd]);
+    uint64_t d1 = float64_val(env->vfp.regs[rd + 1]);
+    uint64_t n0 = float64_val(env->vfp.regs[rn]);
+    uint64_t n1 = float64_val(env->vfp.regs[rn + 1]);
+    uint64_t m0 = float64_val(env->vfp.regs[rm]);
+    uint64_t m1 = float64_val(env->vfp.regs[rm + 1]);
+
+    d1 += S1_512(m1) + cho512(m1, n0, n1);
+    d0 += S1_512(d1 + m0) + cho512(d1 + m0, m1, n0);
+
+    env->vfp.regs[rd] = make_float64(d0);
+    env->vfp.regs[rd + 1] = make_float64(d1);
+}
+
+void HELPER(crypto_sha512h2)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                             uint32_t rm)
+{
+    uint64_t d0 = float64_val(env->vfp.regs[rd]);
+    uint64_t d1 = float64_val(env->vfp.regs[rd + 1]);
+    uint64_t n0 = float64_val(env->vfp.regs[rn]);
+    uint64_t m0 = float64_val(env->vfp.regs[rm]);
+    uint64_t m1 = float64_val(env->vfp.regs[rm + 1]);
+
+    d1 += S0_512(m0) + maj512(n0, m1, m0);
+    d0 += S0_512(d1) + maj512(d1, m0, m1);
+
+    env->vfp.regs[rd] = make_float64(d0);
+    env->vfp.regs[rd + 1] = make_float64(d1);
+}
+
+void HELPER(crypto_sha512su0)(CPUARMState *env, uint32_t rd, uint32_t rn)
+{
+    uint64_t d0 = float64_val(env->vfp.regs[rd]);
+    uint64_t d1 = float64_val(env->vfp.regs[rd + 1]);
+    uint64_t n0 = float64_val(env->vfp.regs[rn]);
+
+    d0 += s0_512(d1);
+    d1 += s0_512(n0);
+
+    env->vfp.regs[rd] = make_float64(d0);
+    env->vfp.regs[rd + 1] = make_float64(d1);
+}
+
+void HELPER(crypto_sha512su1)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                              uint32_t rm)
+{
+    uint64_t d0 = float64_val(env->vfp.regs[rd]);
+    uint64_t d1 = float64_val(env->vfp.regs[rd + 1]);
+    uint64_t n0 = float64_val(env->vfp.regs[rn]);
+    uint64_t n1 = float64_val(env->vfp.regs[rn + 1]);
+    uint64_t m0 = float64_val(env->vfp.regs[rm]);
+    uint64_t m1 = float64_val(env->vfp.regs[rm + 1]);
+
+    d0 += s1_512(n0) + m0;
+    d1 += s1_512(n1) + m1;
+
+    env->vfp.regs[rd] = make_float64(d0);
+    env->vfp.regs[rd + 1] = make_float64(d1);
+}
diff --git a/target/arm/helper.h b/target/arm/helper.h
index 066729e8ad8e..efa75440ce81 100644
--- a/target/arm/helper.h
+++ b/target/arm/helper.h
@@ -534,6 +534,11 @@ DEF_HELPER_4(crypto_sha256h2, void, env, i32, i32, i32)
 DEF_HELPER_3(crypto_sha256su0, void, env, i32, i32)
 DEF_HELPER_4(crypto_sha256su1, void, env, i32, i32, i32)
 
+DEF_HELPER_4(crypto_sha512h, void, env, i32, i32, i32)
+DEF_HELPER_4(crypto_sha512h2, void, env, i32, i32, i32)
+DEF_HELPER_3(crypto_sha512su0, void, env, i32, i32)
+DEF_HELPER_4(crypto_sha512su1, void, env, i32, i32, i32)
+
 DEF_HELPER_FLAGS_3(crc32, TCG_CALL_NO_RWG_SE, i32, i32, i32, i32)
 DEF_HELPER_FLAGS_3(crc32c, TCG_CALL_NO_RWG_SE, i32, i32, i32, i32)
 DEF_HELPER_2(dc_zva, void, env, i64)
diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
index ba94f7d0456a..ec17391f4ff4 100644
--- a/target/arm/translate-a64.c
+++ b/target/arm/translate-a64.c
@@ -11111,6 +11111,103 @@ static void disas_crypto_two_reg_sha(DisasContext *s, uint32_t insn)
     tcg_temp_free_i32(tcg_rn_regno);
 }
 
+/* Crypto three-reg SHA512
+ *  31                   21 20  16 15  14  13 12  11  10  9    5 4    0
+ * +-----------------------+------+---+---+-----+--------+------+------+
+ * | 1 1 0 0 1 1 1 0 0 1 1 |  Rm  | 1 | O | 0 0 | opcode |  Rn  |  Rd  |
+ * +-----------------------+------+---+---+-----+--------+------+------+
+ */
+static void disas_crypto_three_reg_sha512(DisasContext *s, uint32_t insn)
+{
+    int opcode = extract32(insn, 10, 2);
+    int o =  extract32(insn, 14, 1);
+    int rm = extract32(insn, 16, 5);
+    int rn = extract32(insn, 5, 5);
+    int rd = extract32(insn, 0, 5);
+    TCGv_i32 tcg_rd_regno, tcg_rn_regno, tcg_rm_regno;
+    CryptoThreeOpEnvFn *genfn;
+
+    if (o != 0) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    switch (opcode) {
+    case 0: /* SHA512H */
+        genfn = gen_helper_crypto_sha512h;
+        break;
+    case 1: /* SHA512H2 */
+        genfn = gen_helper_crypto_sha512h2;
+        break;
+    case 2: /* SHA512SU1 */
+        genfn = gen_helper_crypto_sha512su1;
+        break;
+    default:
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (!arm_dc_feature(s, ARM_FEATURE_V8_SHA512)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (!fp_access_check(s)) {
+        return;
+    }
+
+    tcg_rd_regno = tcg_const_i32(rd << 1);
+    tcg_rn_regno = tcg_const_i32(rn << 1);
+    tcg_rm_regno = tcg_const_i32(rm << 1);
+
+    genfn(cpu_env, tcg_rd_regno, tcg_rn_regno, tcg_rm_regno);
+
+    tcg_temp_free_i32(tcg_rd_regno);
+    tcg_temp_free_i32(tcg_rn_regno);
+    tcg_temp_free_i32(tcg_rm_regno);
+}
+
+/* Crypto two-reg SHA512
+ *  31                                     12  11  10  9    5 4    0
+ * +-----------------------------------------+--------+------+------+
+ * | 1 1 0 0 1 1 1 0 1 1 0 0 0 0 0 0 1 0 0 0 | opcode |  Rn  |  Rd  |
+ * +-----------------------------------------+--------+------+------+
+ */
+static void disas_crypto_two_reg_sha512(DisasContext *s, uint32_t insn)
+{
+    int opcode = extract32(insn, 10, 2);
+    int rn = extract32(insn, 5, 5);
+    int rd = extract32(insn, 0, 5);
+    TCGv_i32 tcg_rd_regno, tcg_rn_regno;
+    CryptoTwoOpEnvFn *genfn;
+
+    switch (opcode) {
+    case 0: /* SHA512SU0 */
+        genfn = gen_helper_crypto_sha512su0;
+        break;
+    default:
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (!arm_dc_feature(s, ARM_FEATURE_V8_SHA512)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (!fp_access_check(s)) {
+        return;
+    }
+
+    tcg_rd_regno = tcg_const_i32(rd << 1);
+    tcg_rn_regno = tcg_const_i32(rn << 1);
+
+    genfn(cpu_env, tcg_rd_regno, tcg_rn_regno);
+
+    tcg_temp_free_i32(tcg_rd_regno);
+    tcg_temp_free_i32(tcg_rn_regno);
+}
+
 /* C3.6 Data processing - SIMD, inc Crypto
  *
  * As the decode gets a little complex we are using a table based
@@ -11140,6 +11237,8 @@ static const AArch64DecodeTable data_proc_simd[] = {
     { 0x4e280800, 0xff3e0c00, disas_crypto_aes },
     { 0x5e000000, 0xff208c00, disas_crypto_three_reg_sha },
     { 0x5e280800, 0xff3e0c00, disas_crypto_two_reg_sha },
+    { 0xce608000, 0xffe0b000, disas_crypto_three_reg_sha512 },
+    { 0xcec08000, 0xfffff000, disas_crypto_two_reg_sha512 },
     { 0x00000000, 0x00000000, NULL }
 };
 
-- 
2.11.0

^ permalink raw reply related	[flat|nested] 7+ messages in thread

* [Qemu-devel] [PATCH v3 2/4] target/arm: implement SHA-3 instructions
  2018-01-19 14:17 [Qemu-devel] [PATCH v3 0/4] target-arm: add SHA-3, SM3 and SHA512 instruction support Ard Biesheuvel
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 1/4] target/arm: implement SHA-512 instructions Ard Biesheuvel
@ 2018-01-19 14:17 ` Ard Biesheuvel
  2018-01-22 16:47   ` Peter Maydell
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 3/4] target/arm: implement SM3 instructions Ard Biesheuvel
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 4/4] target/arm: enable user-mode SHA-3, SM3 and SHA-512 instruction support Ard Biesheuvel
  3 siblings, 1 reply; 7+ messages in thread
From: Ard Biesheuvel @ 2018-01-19 14:17 UTC (permalink / raw)
  To: qemu-devel; +Cc: peter.maydell, Ard Biesheuvel

This implements emulation of the new SHA-3 instructions that have
been added as an optional extensions to the ARMv8 Crypto Extensions
in ARM v8.2.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
 target/arm/cpu.h           |   1 +
 target/arm/crypto_helper.c |  69 +++++++++++++
 target/arm/helper.h        |   5 +
 target/arm/translate-a64.c | 108 +++++++++++++++++++-
 4 files changed, 179 insertions(+), 4 deletions(-)

diff --git a/target/arm/cpu.h b/target/arm/cpu.h
index 295529366c0a..8e355398e3e0 100644
--- a/target/arm/cpu.h
+++ b/target/arm/cpu.h
@@ -1341,6 +1341,7 @@ enum arm_features {
     ARM_FEATURE_M_SECURITY, /* M profile Security Extension */
     ARM_FEATURE_JAZELLE, /* has (trivial) Jazelle implementation */
     ARM_FEATURE_V8_SHA512, /* implements SHA512 part of v8 Crypto Extensions */
+    ARM_FEATURE_V8_SHA3, /* implements SHA3 part of v8 Crypto Extensions */
 };
 
 static inline int arm_feature(CPUARMState *env, int feature)
diff --git a/target/arm/crypto_helper.c b/target/arm/crypto_helper.c
index 15c22c82ca5e..b51fb0bd2897 100644
--- a/target/arm/crypto_helper.c
+++ b/target/arm/crypto_helper.c
@@ -560,3 +560,72 @@ void HELPER(crypto_sha512su1)(CPUARMState *env, uint32_t rd, uint32_t rn,
     env->vfp.regs[rd] = make_float64(d0);
     env->vfp.regs[rd + 1] = make_float64(d1);
 }
+
+void HELPER(crypto_rax1)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                         uint32_t rm)
+{
+    uint64_t d0 = float64_val(env->vfp.regs[rd]);
+    uint64_t d1 = float64_val(env->vfp.regs[rd + 1]);
+    uint64_t n0 = float64_val(env->vfp.regs[rn]);
+    uint64_t n1 = float64_val(env->vfp.regs[rn + 1]);
+    uint64_t m0 = float64_val(env->vfp.regs[rm]);
+    uint64_t m1 = float64_val(env->vfp.regs[rm + 1]);
+
+    d0 = n0 ^ rol64(m0, 1);
+    d1 = n1 ^ rol64(m1, 1);
+
+    env->vfp.regs[rd] = make_float64(d0);
+    env->vfp.regs[rd + 1] = make_float64(d1);
+}
+
+void HELPER(crypto_eor3)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                         uint32_t ra, uint32_t rm)
+{
+    uint64_t d0, d1;
+    uint64_t n0 = float64_val(env->vfp.regs[rn]);
+    uint64_t n1 = float64_val(env->vfp.regs[rn + 1]);
+    uint64_t a0 = float64_val(env->vfp.regs[ra]);
+    uint64_t a1 = float64_val(env->vfp.regs[ra + 1]);
+    uint64_t m0 = float64_val(env->vfp.regs[rm]);
+    uint64_t m1 = float64_val(env->vfp.regs[rm + 1]);
+
+    d0 = n0 ^ a0 ^ m0;
+    d1 = n1 ^ a1 ^ m1;
+
+    env->vfp.regs[rd] = make_float64(d0);
+    env->vfp.regs[rd + 1] = make_float64(d1);
+}
+
+void HELPER(crypto_bcax)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                         uint32_t ra, uint32_t rm)
+{
+    uint64_t d0, d1;
+    uint64_t n0 = float64_val(env->vfp.regs[rn]);
+    uint64_t n1 = float64_val(env->vfp.regs[rn + 1]);
+    uint64_t a0 = float64_val(env->vfp.regs[ra]);
+    uint64_t a1 = float64_val(env->vfp.regs[ra + 1]);
+    uint64_t m0 = float64_val(env->vfp.regs[rm]);
+    uint64_t m1 = float64_val(env->vfp.regs[rm + 1]);
+
+    d0 = n0 ^ (~a0 & m0);
+    d1 = n1 ^ (~a1 & m1);
+
+    env->vfp.regs[rd] = make_float64(d0);
+    env->vfp.regs[rd + 1] = make_float64(d1);
+}
+
+void HELPER(crypto_xar)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                        uint32_t imm6, uint32_t rm)
+{
+    uint64_t d0, d1;
+    uint64_t n0 = float64_val(env->vfp.regs[rn]);
+    uint64_t n1 = float64_val(env->vfp.regs[rn + 1]);
+    uint64_t m0 = float64_val(env->vfp.regs[rm]);
+    uint64_t m1 = float64_val(env->vfp.regs[rm + 1]);
+
+    d0 = ror64(n0 ^ m0, imm6);
+    d1 = ror64(n1 ^ m1, imm6);
+
+    env->vfp.regs[rd] = make_float64(d0);
+    env->vfp.regs[rd + 1] = make_float64(d1);
+}
diff --git a/target/arm/helper.h b/target/arm/helper.h
index efa75440ce81..2053ea10b2a5 100644
--- a/target/arm/helper.h
+++ b/target/arm/helper.h
@@ -539,6 +539,11 @@ DEF_HELPER_4(crypto_sha512h2, void, env, i32, i32, i32)
 DEF_HELPER_3(crypto_sha512su0, void, env, i32, i32)
 DEF_HELPER_4(crypto_sha512su1, void, env, i32, i32, i32)
 
+DEF_HELPER_4(crypto_rax1, void, env, i32, i32, i32)
+DEF_HELPER_5(crypto_eor3, void, env, i32, i32, i32, i32)
+DEF_HELPER_5(crypto_bcax, void, env, i32, i32, i32, i32)
+DEF_HELPER_5(crypto_xar, void, env, i32, i32, i32, i32)
+
 DEF_HELPER_FLAGS_3(crc32, TCG_CALL_NO_RWG_SE, i32, i32, i32, i32)
 DEF_HELPER_FLAGS_3(crc32c, TCG_CALL_NO_RWG_SE, i32, i32, i32, i32)
 DEF_HELPER_2(dc_zva, void, env, i64)
diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
index ec17391f4ff4..c632f584aad6 100644
--- a/target/arm/translate-a64.c
+++ b/target/arm/translate-a64.c
@@ -82,6 +82,8 @@ typedef void NeonGenTwoDoubleOPFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr);
 typedef void NeonGenOneOpFn(TCGv_i64, TCGv_i64);
 typedef void CryptoTwoOpEnvFn(TCGv_ptr, TCGv_i32, TCGv_i32);
 typedef void CryptoThreeOpEnvFn(TCGv_ptr, TCGv_i32, TCGv_i32, TCGv_i32);
+typedef void CryptoFourOpEnvFn(TCGv_ptr, TCGv_i32, TCGv_i32, TCGv_i32,
+                               TCGv_i32);
 
 /* initialize TCG globals.  */
 void a64_translate_init(void)
@@ -11125,6 +11127,7 @@ static void disas_crypto_three_reg_sha512(DisasContext *s, uint32_t insn)
     int rn = extract32(insn, 5, 5);
     int rd = extract32(insn, 0, 5);
     TCGv_i32 tcg_rd_regno, tcg_rn_regno, tcg_rm_regno;
+    int feature;
     CryptoThreeOpEnvFn *genfn;
 
     if (o != 0) {
@@ -11134,20 +11137,24 @@ static void disas_crypto_three_reg_sha512(DisasContext *s, uint32_t insn)
 
     switch (opcode) {
     case 0: /* SHA512H */
+        feature = ARM_FEATURE_V8_SHA512;
         genfn = gen_helper_crypto_sha512h;
         break;
     case 1: /* SHA512H2 */
+        feature = ARM_FEATURE_V8_SHA512;
         genfn = gen_helper_crypto_sha512h2;
         break;
     case 2: /* SHA512SU1 */
+        feature = ARM_FEATURE_V8_SHA512;
         genfn = gen_helper_crypto_sha512su1;
         break;
-    default:
-        unallocated_encoding(s);
-        return;
+    case 3: /* RAX1 */
+        feature = ARM_FEATURE_V8_SHA3;
+        genfn = gen_helper_crypto_rax1;
+        break;
     }
 
-    if (!arm_dc_feature(s, ARM_FEATURE_V8_SHA512)) {
+    if (!arm_dc_feature(s, feature)) {
         unallocated_encoding(s);
         return;
     }
@@ -11208,6 +11215,97 @@ static void disas_crypto_two_reg_sha512(DisasContext *s, uint32_t insn)
     tcg_temp_free_i32(tcg_rn_regno);
 }
 
+/* Crypto four-register
+ *  31               23 22 21 20  16 15  14  10 9    5 4    0
+ * +-------------------+-----+------+---+------+------+------+
+ * | 1 1 0 0 1 1 1 0 0 | Op0 |  Rm  | 0 |  Ra  |  Rn  |  Rd  |
+ * +-------------------+-----+------+---+------+------+------+
+ */
+static void disas_crypto_four_reg(DisasContext *s, uint32_t insn)
+{
+    int op0 = extract32(insn, 21, 2);
+    int rm = extract32(insn, 16, 5);
+    int ra = extract32(insn, 10, 5);
+    int rn = extract32(insn, 5, 5);
+    int rd = extract32(insn, 0, 5);
+    TCGv_i32 tcg_rd_regno, tcg_rn_regno, tcg_ra_regno, tcg_rm_regno;
+    int feature;
+    CryptoFourOpEnvFn *genfn;
+
+    switch (op0) {
+    case 0: /* EOR3 */
+        feature = ARM_FEATURE_V8_SHA3;
+        genfn = gen_helper_crypto_eor3;
+        break;
+    case 1: /* BCAX */
+        feature = ARM_FEATURE_V8_SHA3;
+        genfn = gen_helper_crypto_bcax;
+        break;
+    default:
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (!arm_dc_feature(s, feature)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (!fp_access_check(s)) {
+        return;
+    }
+
+    tcg_rd_regno = tcg_const_i32(rd << 1);
+    tcg_rn_regno = tcg_const_i32(rn << 1);
+    tcg_ra_regno = tcg_const_i32(ra << 1);
+    tcg_rm_regno = tcg_const_i32(rm << 1);
+
+    genfn(cpu_env, tcg_rd_regno, tcg_rn_regno, tcg_ra_regno, tcg_rm_regno);
+
+    tcg_temp_free_i32(tcg_rd_regno);
+    tcg_temp_free_i32(tcg_rn_regno);
+    tcg_temp_free_i32(tcg_ra_regno);
+    tcg_temp_free_i32(tcg_rm_regno);
+}
+
+/* Crypto XAR
+ *  31                   21 20  16 15    10 9    5 4    0
+ * +-----------------------+------+--------+------+------+
+ * | 1 1 0 0 1 1 1 0 1 0 0 |  Rm  |  imm6  |  Rn  |  Rd  |
+ * +-----------------------+------+--------+------+------+
+ */
+static void disas_crypto_xar(DisasContext *s, uint32_t insn)
+{
+    int rm = extract32(insn, 16, 5);
+    int imm6 = extract32(insn, 10, 6);
+    int rn = extract32(insn, 5, 5);
+    int rd = extract32(insn, 0, 5);
+    TCGv_i32 tcg_rd_regno, tcg_rn_regno, tcg_imm6, tcg_rm_regno;
+
+    if (!arm_dc_feature(s, ARM_FEATURE_V8_SHA3)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (!fp_access_check(s)) {
+        return;
+    }
+
+    tcg_rd_regno = tcg_const_i32(rd << 1);
+    tcg_rn_regno = tcg_const_i32(rn << 1);
+    tcg_imm6 = tcg_const_i32(imm6);
+    tcg_rm_regno = tcg_const_i32(rm << 1);
+
+    gen_helper_crypto_xar(cpu_env, tcg_rd_regno, tcg_rn_regno, tcg_imm6,
+                          tcg_rm_regno);
+
+    tcg_temp_free_i32(tcg_rd_regno);
+    tcg_temp_free_i32(tcg_rn_regno);
+    tcg_temp_free_i32(tcg_imm6);
+    tcg_temp_free_i32(tcg_rm_regno);
+
+}
+
 /* C3.6 Data processing - SIMD, inc Crypto
  *
  * As the decode gets a little complex we are using a table based
@@ -11239,6 +11337,8 @@ static const AArch64DecodeTable data_proc_simd[] = {
     { 0x5e280800, 0xff3e0c00, disas_crypto_two_reg_sha },
     { 0xce608000, 0xffe0b000, disas_crypto_three_reg_sha512 },
     { 0xcec08000, 0xfffff000, disas_crypto_two_reg_sha512 },
+    { 0xce000000, 0xff808000, disas_crypto_four_reg },
+    { 0xce800000, 0xffe00000, disas_crypto_xar },
     { 0x00000000, 0x00000000, NULL }
 };
 
-- 
2.11.0

^ permalink raw reply related	[flat|nested] 7+ messages in thread

* [Qemu-devel] [PATCH v3 3/4] target/arm: implement SM3 instructions
  2018-01-19 14:17 [Qemu-devel] [PATCH v3 0/4] target-arm: add SHA-3, SM3 and SHA512 instruction support Ard Biesheuvel
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 1/4] target/arm: implement SHA-512 instructions Ard Biesheuvel
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 2/4] target/arm: implement SHA-3 instructions Ard Biesheuvel
@ 2018-01-19 14:17 ` Ard Biesheuvel
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 4/4] target/arm: enable user-mode SHA-3, SM3 and SHA-512 instruction support Ard Biesheuvel
  3 siblings, 0 replies; 7+ messages in thread
From: Ard Biesheuvel @ 2018-01-19 14:17 UTC (permalink / raw)
  To: qemu-devel; +Cc: peter.maydell, Ard Biesheuvel

This implements emulation of the new SM3 instructions that have
been added as an optional extension to the ARMv8 Crypto Extensions
in ARM v8.2.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
 target/arm/cpu.h           |   1 +
 target/arm/crypto_helper.c | 143 ++++++++++++++++++++
 target/arm/helper.h        |   5 +
 target/arm/translate-a64.c | 100 +++++++++++---
 4 files changed, 227 insertions(+), 22 deletions(-)

diff --git a/target/arm/cpu.h b/target/arm/cpu.h
index 8e355398e3e0..917d15fa346c 100644
--- a/target/arm/cpu.h
+++ b/target/arm/cpu.h
@@ -1342,6 +1342,7 @@ enum arm_features {
     ARM_FEATURE_JAZELLE, /* has (trivial) Jazelle implementation */
     ARM_FEATURE_V8_SHA512, /* implements SHA512 part of v8 Crypto Extensions */
     ARM_FEATURE_V8_SHA3, /* implements SHA3 part of v8 Crypto Extensions */
+    ARM_FEATURE_V8_SM3, /* implements SM3 part of v8 Crypto Extensions */
 };
 
 static inline int arm_feature(CPUARMState *env, int feature)
diff --git a/target/arm/crypto_helper.c b/target/arm/crypto_helper.c
index b51fb0bd2897..6c601351b177 100644
--- a/target/arm/crypto_helper.c
+++ b/target/arm/crypto_helper.c
@@ -629,3 +629,146 @@ void HELPER(crypto_xar)(CPUARMState *env, uint32_t rd, uint32_t rn,
     env->vfp.regs[rd] = make_float64(d0);
     env->vfp.regs[rd + 1] = make_float64(d1);
 }
+
+void HELPER(crypto_sm3partw1)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                              uint32_t rm)
+{
+    union CRYPTO_STATE d = { .l = {
+        float64_val(env->vfp.regs[rd]),
+        float64_val(env->vfp.regs[rd + 1])
+    } };
+    union CRYPTO_STATE n = { .l = {
+        float64_val(env->vfp.regs[rn]),
+        float64_val(env->vfp.regs[rn + 1])
+    } };
+    union CRYPTO_STATE m = { .l = {
+        float64_val(env->vfp.regs[rm]),
+        float64_val(env->vfp.regs[rm + 1])
+    } };
+    uint32_t t;
+
+    t = CR_ST_WORD(d, 0) ^ CR_ST_WORD(n, 0) ^ ror32(CR_ST_WORD(m, 1), 17);
+    CR_ST_WORD(d, 0) = t ^ ror32(t, 17) ^ ror32(t, 9);
+
+    t = CR_ST_WORD(d, 1) ^ CR_ST_WORD(n, 1) ^ ror32(CR_ST_WORD(m, 2), 17);
+    CR_ST_WORD(d, 1) = t ^ ror32(t, 17) ^ ror32(t, 9);
+
+    t = CR_ST_WORD(d, 2) ^ CR_ST_WORD(n, 2) ^ ror32(CR_ST_WORD(m, 3), 17);
+    CR_ST_WORD(d, 2) = t ^ ror32(t, 17) ^ ror32(t, 9);
+
+    t = CR_ST_WORD(d, 3) ^ CR_ST_WORD(n, 3) ^ ror32(CR_ST_WORD(d, 0), 17);
+    CR_ST_WORD(d, 3) = t ^ ror32(t, 17) ^ ror32(t, 9);
+
+    env->vfp.regs[rd] = make_float64(d.l[0]);
+    env->vfp.regs[rd + 1] = make_float64(d.l[1]);
+}
+
+void HELPER(crypto_sm3partw2)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                              uint32_t rm)
+{
+    union CRYPTO_STATE d = { .l = {
+        float64_val(env->vfp.regs[rd]),
+        float64_val(env->vfp.regs[rd + 1])
+    } };
+    union CRYPTO_STATE n = { .l = {
+        float64_val(env->vfp.regs[rn]),
+        float64_val(env->vfp.regs[rn + 1])
+    } };
+    union CRYPTO_STATE m = { .l = {
+        float64_val(env->vfp.regs[rm]),
+        float64_val(env->vfp.regs[rm + 1])
+    } };
+    uint32_t t = CR_ST_WORD(n, 0) ^ ror32(CR_ST_WORD(m, 0), 25);
+
+    CR_ST_WORD(d, 0) ^= t;
+    CR_ST_WORD(d, 1) ^= CR_ST_WORD(n, 1) ^ ror32(CR_ST_WORD(m, 1), 25);
+    CR_ST_WORD(d, 2) ^= CR_ST_WORD(n, 2) ^ ror32(CR_ST_WORD(m, 2), 25);
+    CR_ST_WORD(d, 3) ^= CR_ST_WORD(n, 3) ^ ror32(CR_ST_WORD(m, 3), 25) ^
+                        ror32(t, 17) ^ ror32(t, 2) ^ ror32(t, 26);
+
+    env->vfp.regs[rd] = make_float64(d.l[0]);
+    env->vfp.regs[rd + 1] = make_float64(d.l[1]);
+}
+
+void HELPER(crypto_sm3ss1)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                           uint32_t ra, uint32_t rm)
+{
+    union CRYPTO_STATE d;
+    union CRYPTO_STATE a = { .l = {
+        float64_val(env->vfp.regs[ra]),
+        float64_val(env->vfp.regs[ra + 1])
+    } };
+    union CRYPTO_STATE n = { .l = {
+        float64_val(env->vfp.regs[rn]),
+        float64_val(env->vfp.regs[rn + 1])
+    } };
+    union CRYPTO_STATE m = { .l = {
+        float64_val(env->vfp.regs[rm]),
+        float64_val(env->vfp.regs[rm + 1])
+    } };
+
+    CR_ST_WORD(d, 0) = 0;
+    CR_ST_WORD(d, 1) = 0;
+    CR_ST_WORD(d, 2) = 0;
+    CR_ST_WORD(d, 3) = ror32(ror32(CR_ST_WORD(n, 3), 20) + CR_ST_WORD(m, 3) +
+                             CR_ST_WORD(a, 3), 25);
+
+    env->vfp.regs[rd] = make_float64(d.l[0]);
+    env->vfp.regs[rd + 1] = make_float64(d.l[1]);
+}
+
+void HELPER(crypto_sm3tt)(CPUARMState *env, uint32_t rd, uint32_t rn,
+                          uint32_t rm, uint32_t imm2, uint32_t opcode)
+{
+    union CRYPTO_STATE d = { .l = {
+        float64_val(env->vfp.regs[rd]),
+        float64_val(env->vfp.regs[rd + 1])
+    } };
+    union CRYPTO_STATE n = { .l = {
+        float64_val(env->vfp.regs[rn]),
+        float64_val(env->vfp.regs[rn + 1])
+    } };
+    union CRYPTO_STATE m = { .l = {
+        float64_val(env->vfp.regs[rm]),
+        float64_val(env->vfp.regs[rm + 1])
+    } };
+    uint32_t t;
+
+    assert(imm2 < 4);
+
+    if (opcode == 0 || opcode == 2) {
+        /* SM3TT1A, SM3TT2A */
+        t = par(CR_ST_WORD(d, 3), CR_ST_WORD(d, 2), CR_ST_WORD(d, 1));
+    } else if (opcode == 1) {
+        /* SM3TT1B */
+        t = maj(CR_ST_WORD(d, 3), CR_ST_WORD(d, 2), CR_ST_WORD(d, 1));
+    } else if (opcode == 3) {
+        /* SM3TT2B */
+        t = cho(CR_ST_WORD(d, 3), CR_ST_WORD(d, 2), CR_ST_WORD(d, 1));
+    } else {
+        g_assert_not_reached();
+    }
+
+    t += CR_ST_WORD(d, 0) + CR_ST_WORD(m, imm2);
+
+    CR_ST_WORD(d, 0) = CR_ST_WORD(d, 1);
+
+    if (opcode < 2) {
+        /* SM3TT1A, SM3TT1B */
+        t += CR_ST_WORD(n, 3) ^ ror32(CR_ST_WORD(d, 3), 20);
+
+        CR_ST_WORD(d, 1) = ror32(CR_ST_WORD(d, 2), 23);
+    } else {
+        /* SM3TT2A, SM3TT2B */
+        t += CR_ST_WORD(n, 3);
+        t ^= rol32(t, 9) ^ rol32(t, 17);
+
+        CR_ST_WORD(d, 1) = ror32(CR_ST_WORD(d, 2), 13);
+    }
+
+    CR_ST_WORD(d, 2) = CR_ST_WORD(d, 3);
+    CR_ST_WORD(d, 3) = t;
+
+    env->vfp.regs[rd] = make_float64(d.l[0]);
+    env->vfp.regs[rd + 1] = make_float64(d.l[1]);
+}
diff --git a/target/arm/helper.h b/target/arm/helper.h
index 2053ea10b2a5..913c551bf14b 100644
--- a/target/arm/helper.h
+++ b/target/arm/helper.h
@@ -544,6 +544,11 @@ DEF_HELPER_5(crypto_eor3, void, env, i32, i32, i32, i32)
 DEF_HELPER_5(crypto_bcax, void, env, i32, i32, i32, i32)
 DEF_HELPER_5(crypto_xar, void, env, i32, i32, i32, i32)
 
+DEF_HELPER_5(crypto_sm3ss1, void, env, i32, i32, i32, i32)
+DEF_HELPER_6(crypto_sm3tt, void, env, i32, i32, i32, i32, i32)
+DEF_HELPER_4(crypto_sm3partw1, void, env, i32, i32, i32)
+DEF_HELPER_4(crypto_sm3partw2, void, env, i32, i32, i32)
+
 DEF_HELPER_FLAGS_3(crc32, TCG_CALL_NO_RWG_SE, i32, i32, i32, i32)
 DEF_HELPER_FLAGS_3(crc32c, TCG_CALL_NO_RWG_SE, i32, i32, i32, i32)
 DEF_HELPER_2(dc_zva, void, env, i64)
diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
index c632f584aad6..9d805d4a7d26 100644
--- a/target/arm/translate-a64.c
+++ b/target/arm/translate-a64.c
@@ -11130,28 +11130,39 @@ static void disas_crypto_three_reg_sha512(DisasContext *s, uint32_t insn)
     int feature;
     CryptoThreeOpEnvFn *genfn;
 
-    if (o != 0) {
-        unallocated_encoding(s);
-        return;
-    }
-
-    switch (opcode) {
-    case 0: /* SHA512H */
-        feature = ARM_FEATURE_V8_SHA512;
-        genfn = gen_helper_crypto_sha512h;
-        break;
-    case 1: /* SHA512H2 */
-        feature = ARM_FEATURE_V8_SHA512;
-        genfn = gen_helper_crypto_sha512h2;
-        break;
-    case 2: /* SHA512SU1 */
-        feature = ARM_FEATURE_V8_SHA512;
-        genfn = gen_helper_crypto_sha512su1;
-        break;
-    case 3: /* RAX1 */
-        feature = ARM_FEATURE_V8_SHA3;
-        genfn = gen_helper_crypto_rax1;
-        break;
+    if (o == 0) {
+        switch (opcode) {
+        case 0: /* SHA512H */
+            feature = ARM_FEATURE_V8_SHA512;
+            genfn = gen_helper_crypto_sha512h;
+            break;
+        case 1: /* SHA512H2 */
+            feature = ARM_FEATURE_V8_SHA512;
+            genfn = gen_helper_crypto_sha512h2;
+            break;
+        case 2: /* SHA512SU1 */
+            feature = ARM_FEATURE_V8_SHA512;
+            genfn = gen_helper_crypto_sha512su1;
+            break;
+        case 3: /* RAX1 */
+            feature = ARM_FEATURE_V8_SHA3;
+            genfn = gen_helper_crypto_rax1;
+            break;
+        }
+    } else {
+        switch (opcode) {
+        case 0: /* SM3PARTW1 */
+            feature = ARM_FEATURE_V8_SM3;
+            genfn = gen_helper_crypto_sm3partw1;
+            break;
+        case 1: /* SM3PARTW2 */
+            feature = ARM_FEATURE_V8_SM3;
+            genfn = gen_helper_crypto_sm3partw2;
+            break;
+        default:
+            unallocated_encoding(s);
+            return;
+        }
     }
 
     if (!arm_dc_feature(s, feature)) {
@@ -11241,6 +11252,10 @@ static void disas_crypto_four_reg(DisasContext *s, uint32_t insn)
         feature = ARM_FEATURE_V8_SHA3;
         genfn = gen_helper_crypto_bcax;
         break;
+    case 2: /* SM3SS1 */
+        feature = ARM_FEATURE_V8_SM3;
+        genfn = gen_helper_crypto_sm3ss1;
+        break;
     default:
         unallocated_encoding(s);
         return;
@@ -11306,6 +11321,46 @@ static void disas_crypto_xar(DisasContext *s, uint32_t insn)
 
 }
 
+/* Crypto three-reg imm2
+ *  31                   21 20  16 15  14 13 12  11  10  9    5 4    0
+ * +-----------------------+------+-----+------+--------+------+------+
+ * | 1 1 0 0 1 1 1 0 0 1 0 |  Rm  | 1 0 | imm2 | opcode |  Rn  |  Rd  |
+ * +-----------------------+------+-----+------+--------+------+------+
+ */
+static void disas_crypto_three_reg_imm2(DisasContext *s, uint32_t insn)
+{
+    int opcode = extract32(insn, 10, 2);
+    int imm2 = extract32(insn, 12, 2);
+    int rm = extract32(insn, 16, 5);
+    int rn = extract32(insn, 5, 5);
+    int rd = extract32(insn, 0, 5);
+    TCGv_i32 tcg_rd_regno, tcg_rn_regno, tcg_rm_regno, tcg_imm2, tcg_opcode;
+
+    if (!arm_dc_feature(s, ARM_FEATURE_V8_SM3)) {
+        unallocated_encoding(s);
+        return;
+    }
+
+    if (!fp_access_check(s)) {
+        return;
+    }
+
+    tcg_rd_regno = tcg_const_i32(rd << 1);
+    tcg_rn_regno = tcg_const_i32(rn << 1);
+    tcg_rm_regno = tcg_const_i32(rm << 1);
+    tcg_imm2 = tcg_const_i32(imm2);
+    tcg_opcode = tcg_const_i32(opcode);
+
+    gen_helper_crypto_sm3tt(cpu_env, tcg_rd_regno, tcg_rn_regno, tcg_rm_regno,
+                            tcg_imm2, tcg_opcode);
+
+    tcg_temp_free_i32(tcg_rd_regno);
+    tcg_temp_free_i32(tcg_rn_regno);
+    tcg_temp_free_i32(tcg_rm_regno);
+    tcg_temp_free_i32(tcg_imm2);
+    tcg_temp_free_i32(tcg_opcode);
+}
+
 /* C3.6 Data processing - SIMD, inc Crypto
  *
  * As the decode gets a little complex we are using a table based
@@ -11339,6 +11394,7 @@ static const AArch64DecodeTable data_proc_simd[] = {
     { 0xcec08000, 0xfffff000, disas_crypto_two_reg_sha512 },
     { 0xce000000, 0xff808000, disas_crypto_four_reg },
     { 0xce800000, 0xffe00000, disas_crypto_xar },
+    { 0xce408000, 0xffe0c000, disas_crypto_three_reg_imm2 },
     { 0x00000000, 0x00000000, NULL }
 };
 
-- 
2.11.0

^ permalink raw reply related	[flat|nested] 7+ messages in thread

* [Qemu-devel] [PATCH v3 4/4] target/arm: enable user-mode SHA-3, SM3 and SHA-512 instruction support
  2018-01-19 14:17 [Qemu-devel] [PATCH v3 0/4] target-arm: add SHA-3, SM3 and SHA512 instruction support Ard Biesheuvel
                   ` (2 preceding siblings ...)
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 3/4] target/arm: implement SM3 instructions Ard Biesheuvel
@ 2018-01-19 14:17 ` Ard Biesheuvel
  3 siblings, 0 replies; 7+ messages in thread
From: Ard Biesheuvel @ 2018-01-19 14:17 UTC (permalink / raw)
  To: qemu-devel; +Cc: peter.maydell, Ard Biesheuvel

Add support for the new ARMv8.2 SHA-3, SM3 and SHA-512 instructions to
AArch64 user mode emulation.

Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
---
 linux-user/elfload.c | 18 ++++++++++++++++++
 target/arm/cpu64.c   |  3 +++
 2 files changed, 21 insertions(+)

diff --git a/linux-user/elfload.c b/linux-user/elfload.c
index 20f3d8c2c373..5d5aa26d2710 100644
--- a/linux-user/elfload.c
+++ b/linux-user/elfload.c
@@ -512,6 +512,21 @@ enum {
     ARM_HWCAP_A64_SHA1          = 1 << 5,
     ARM_HWCAP_A64_SHA2          = 1 << 6,
     ARM_HWCAP_A64_CRC32         = 1 << 7,
+    ARM_HWCAP_A64_ATOMICS       = 1 << 8,
+    ARM_HWCAP_A64_FPHP          = 1 << 9,
+    ARM_HWCAP_A64_ASIMDHP       = 1 << 10,
+    ARM_HWCAP_A64_CPUID         = 1 << 11,
+    ARM_HWCAP_A64_ASIMDRDM      = 1 << 12,
+    ARM_HWCAP_A64_JSCVT         = 1 << 13,
+    ARM_HWCAP_A64_FCMA          = 1 << 14,
+    ARM_HWCAP_A64_LRCPC         = 1 << 15,
+    ARM_HWCAP_A64_DCPOP         = 1 << 16,
+    ARM_HWCAP_A64_SHA3          = 1 << 17,
+    ARM_HWCAP_A64_SM3           = 1 << 18,
+    ARM_HWCAP_A64_SM4           = 1 << 19,
+    ARM_HWCAP_A64_ASIMDDP       = 1 << 20,
+    ARM_HWCAP_A64_SHA512        = 1 << 21,
+    ARM_HWCAP_A64_SVE           = 1 << 22,
 };
 
 #define ELF_HWCAP get_elf_hwcap()
@@ -532,6 +547,9 @@ static uint32_t get_elf_hwcap(void)
     GET_FEATURE(ARM_FEATURE_V8_SHA1, ARM_HWCAP_A64_SHA1);
     GET_FEATURE(ARM_FEATURE_V8_SHA256, ARM_HWCAP_A64_SHA2);
     GET_FEATURE(ARM_FEATURE_CRC, ARM_HWCAP_A64_CRC32);
+    GET_FEATURE(ARM_FEATURE_V8_SHA3, ARM_HWCAP_A64_SHA3);
+    GET_FEATURE(ARM_FEATURE_V8_SM3, ARM_HWCAP_A64_SM3);
+    GET_FEATURE(ARM_FEATURE_V8_SHA512, ARM_HWCAP_A64_SHA512);
 #undef GET_FEATURE
 
     return hwcaps;
diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c
index 670c07ab6ed4..56d50ba57194 100644
--- a/target/arm/cpu64.c
+++ b/target/arm/cpu64.c
@@ -224,6 +224,9 @@ static void aarch64_any_initfn(Object *obj)
     set_feature(&cpu->env, ARM_FEATURE_V8_AES);
     set_feature(&cpu->env, ARM_FEATURE_V8_SHA1);
     set_feature(&cpu->env, ARM_FEATURE_V8_SHA256);
+    set_feature(&cpu->env, ARM_FEATURE_V8_SHA3);
+    set_feature(&cpu->env, ARM_FEATURE_V8_SM3);
+    set_feature(&cpu->env, ARM_FEATURE_V8_SHA512);
     set_feature(&cpu->env, ARM_FEATURE_V8_PMULL);
     set_feature(&cpu->env, ARM_FEATURE_CRC);
     cpu->ctr = 0x80038003; /* 32 byte I and D cacheline size, VIPT icache */
-- 
2.11.0

^ permalink raw reply related	[flat|nested] 7+ messages in thread

* Re: [Qemu-devel] [PATCH v3 2/4] target/arm: implement SHA-3 instructions
  2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 2/4] target/arm: implement SHA-3 instructions Ard Biesheuvel
@ 2018-01-22 16:47   ` Peter Maydell
  2018-01-22 16:48     ` Ard Biesheuvel
  0 siblings, 1 reply; 7+ messages in thread
From: Peter Maydell @ 2018-01-22 16:47 UTC (permalink / raw)
  To: Ard Biesheuvel; +Cc: QEMU Developers

On 19 January 2018 at 14:17, Ard Biesheuvel <ard.biesheuvel@linaro.org> wrote:
> This implements emulation of the new SHA-3 instructions that have
> been added as an optional extensions to the ARMv8 Crypto Extensions
> in ARM v8.2.
>
> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
> ---
>  target/arm/cpu.h           |   1 +
>  target/arm/crypto_helper.c |  69 +++++++++++++
>  target/arm/helper.h        |   5 +
>  target/arm/translate-a64.c | 108 +++++++++++++++++++-
>  4 files changed, 179 insertions(+), 4 deletions(-)
>
> diff --git a/target/arm/cpu.h b/target/arm/cpu.h
> index 295529366c0a..8e355398e3e0 100644
> --- a/target/arm/cpu.h
> +++ b/target/arm/cpu.h
> @@ -1341,6 +1341,7 @@ enum arm_features {
>      ARM_FEATURE_M_SECURITY, /* M profile Security Extension */
>      ARM_FEATURE_JAZELLE, /* has (trivial) Jazelle implementation */
>      ARM_FEATURE_V8_SHA512, /* implements SHA512 part of v8 Crypto Extensions */
> +    ARM_FEATURE_V8_SHA3, /* implements SHA3 part of v8 Crypto Extensions */
>  };
>

> diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
>  /* initialize TCG globals.  */
>  void a64_translate_init(void)
> @@ -11125,6 +11127,7 @@ static void disas_crypto_three_reg_sha512(DisasContext *s, uint32_t insn)
>      int rn = extract32(insn, 5, 5);
>      int rd = extract32(insn, 0, 5);
>      TCGv_i32 tcg_rd_regno, tcg_rn_regno, tcg_rm_regno;
> +    int feature;
>      CryptoThreeOpEnvFn *genfn;
>
>      if (o != 0) {
> @@ -11134,20 +11137,24 @@ static void disas_crypto_three_reg_sha512(DisasContext *s, uint32_t insn)
>
>      switch (opcode) {
>      case 0: /* SHA512H */
> +        feature = ARM_FEATURE_V8_SHA512;
>          genfn = gen_helper_crypto_sha512h;
>          break;
>      case 1: /* SHA512H2 */
> +        feature = ARM_FEATURE_V8_SHA512;
>          genfn = gen_helper_crypto_sha512h2;
>          break;
>      case 2: /* SHA512SU1 */
> +        feature = ARM_FEATURE_V8_SHA512;
>          genfn = gen_helper_crypto_sha512su1;
>          break;

I think these lines should have been in patch 1, right?

> -    default:
> -        unallocated_encoding(s);
> -        return;
> +    case 3: /* RAX1 */
> +        feature = ARM_FEATURE_V8_SHA3;
> +        genfn = gen_helper_crypto_rax1;
> +        break;
>      }
>
> -    if (!arm_dc_feature(s, ARM_FEATURE_V8_SHA512)) {
> +    if (!arm_dc_feature(s, feature)) {
>          unallocated_encoding(s);
>          return;

thanks
-- PMM

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [Qemu-devel] [PATCH v3 2/4] target/arm: implement SHA-3 instructions
  2018-01-22 16:47   ` Peter Maydell
@ 2018-01-22 16:48     ` Ard Biesheuvel
  0 siblings, 0 replies; 7+ messages in thread
From: Ard Biesheuvel @ 2018-01-22 16:48 UTC (permalink / raw)
  To: Peter Maydell; +Cc: QEMU Developers

On 22 January 2018 at 16:47, Peter Maydell <peter.maydell@linaro.org> wrote:
> On 19 January 2018 at 14:17, Ard Biesheuvel <ard.biesheuvel@linaro.org> wrote:
>> This implements emulation of the new SHA-3 instructions that have
>> been added as an optional extensions to the ARMv8 Crypto Extensions
>> in ARM v8.2.
>>
>> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
>> ---
>>  target/arm/cpu.h           |   1 +
>>  target/arm/crypto_helper.c |  69 +++++++++++++
>>  target/arm/helper.h        |   5 +
>>  target/arm/translate-a64.c | 108 +++++++++++++++++++-
>>  4 files changed, 179 insertions(+), 4 deletions(-)
>>
>> diff --git a/target/arm/cpu.h b/target/arm/cpu.h
>> index 295529366c0a..8e355398e3e0 100644
>> --- a/target/arm/cpu.h
>> +++ b/target/arm/cpu.h
>> @@ -1341,6 +1341,7 @@ enum arm_features {
>>      ARM_FEATURE_M_SECURITY, /* M profile Security Extension */
>>      ARM_FEATURE_JAZELLE, /* has (trivial) Jazelle implementation */
>>      ARM_FEATURE_V8_SHA512, /* implements SHA512 part of v8 Crypto Extensions */
>> +    ARM_FEATURE_V8_SHA3, /* implements SHA3 part of v8 Crypto Extensions */
>>  };
>>
>
>> diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
>>  /* initialize TCG globals.  */
>>  void a64_translate_init(void)
>> @@ -11125,6 +11127,7 @@ static void disas_crypto_three_reg_sha512(DisasContext *s, uint32_t insn)
>>      int rn = extract32(insn, 5, 5);
>>      int rd = extract32(insn, 0, 5);
>>      TCGv_i32 tcg_rd_regno, tcg_rn_regno, tcg_rm_regno;
>> +    int feature;
>>      CryptoThreeOpEnvFn *genfn;
>>
>>      if (o != 0) {
>> @@ -11134,20 +11137,24 @@ static void disas_crypto_three_reg_sha512(DisasContext *s, uint32_t insn)
>>
>>      switch (opcode) {
>>      case 0: /* SHA512H */
>> +        feature = ARM_FEATURE_V8_SHA512;
>>          genfn = gen_helper_crypto_sha512h;
>>          break;
>>      case 1: /* SHA512H2 */
>> +        feature = ARM_FEATURE_V8_SHA512;
>>          genfn = gen_helper_crypto_sha512h2;
>>          break;
>>      case 2: /* SHA512SU1 */
>> +        feature = ARM_FEATURE_V8_SHA512;
>>          genfn = gen_helper_crypto_sha512su1;
>>          break;
>
> I think these lines should have been in patch 1, right?
>
>> -    default:
>> -        unallocated_encoding(s);
>> -        return;
>> +    case 3: /* RAX1 */
>> +        feature = ARM_FEATURE_V8_SHA3;
>> +        genfn = gen_helper_crypto_rax1;
>> +        break;
>>      }
>>
>> -    if (!arm_dc_feature(s, ARM_FEATURE_V8_SHA512)) {
>> +    if (!arm_dc_feature(s, feature)) {

only if I include this line as well

>>          unallocated_encoding(s);
>>          return;
>
> thanks
> -- PMM

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2018-01-22 16:48 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-01-19 14:17 [Qemu-devel] [PATCH v3 0/4] target-arm: add SHA-3, SM3 and SHA512 instruction support Ard Biesheuvel
2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 1/4] target/arm: implement SHA-512 instructions Ard Biesheuvel
2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 2/4] target/arm: implement SHA-3 instructions Ard Biesheuvel
2018-01-22 16:47   ` Peter Maydell
2018-01-22 16:48     ` Ard Biesheuvel
2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 3/4] target/arm: implement SM3 instructions Ard Biesheuvel
2018-01-19 14:17 ` [Qemu-devel] [PATCH v3 4/4] target/arm: enable user-mode SHA-3, SM3 and SHA-512 instruction support Ard Biesheuvel

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).