Message ID | 1395912558-1041-1-git-send-email-ard.biesheuvel@linaro.org |
---|---|
State | New |
Headers | show |
On 27 March 2014 09:29, Ard Biesheuvel <ard.biesheuvel@linaro.org> wrote: > This adds support for the VMULL.P64 polynomial 64x64 to 128 bit multiplication > instruction, which is an optional feature that is part of the v8 Crypto > Extensions. > +void HELPER(crypto_pmull)(CPUARMState *env, uint32_t rd, uint32_t rn, > + uint32_t rm) > +{ > + uint64_t n = float64_val(env->vfp.regs[rn]); > + uint64_t m = float64_val(env->vfp.regs[rm]); > + uint64_t d0 = (n & 1) ? m : 0; > + uint64_t d1 = 0; > + int shift; > + > + for (shift = 1; (n >>= 1); shift++) { > + if (n & 1) { > + d0 ^= m << shift; > + d1 ^= m >> (64 - shift); > + } > + } > + env->vfp.regs[rd] = make_float64(d0); > + env->vfp.regs[rd + 1] = make_float64(d1); > +} Surely we can reuse the helper we already have for implementing the A64 version of this instruction ? thanks -- PMM
On 27 March 2014 11:55, Peter Maydell <peter.maydell@linaro.org> wrote: > On 27 March 2014 09:29, Ard Biesheuvel <ard.biesheuvel@linaro.org> wrote: >> This adds support for the VMULL.P64 polynomial 64x64 to 128 bit multiplication >> instruction, which is an optional feature that is part of the v8 Crypto >> Extensions. > >> +void HELPER(crypto_pmull)(CPUARMState *env, uint32_t rd, uint32_t rn, >> + uint32_t rm) >> +{ >> + uint64_t n = float64_val(env->vfp.regs[rn]); >> + uint64_t m = float64_val(env->vfp.regs[rm]); >> + uint64_t d0 = (n & 1) ? m : 0; >> + uint64_t d1 = 0; >> + int shift; >> + >> + for (shift = 1; (n >>= 1); shift++) { >> + if (n & 1) { >> + d0 ^= m << shift; >> + d1 ^= m >> (64 - shift); >> + } >> + } >> + env->vfp.regs[rd] = make_float64(d0); >> + env->vfp.regs[rd + 1] = make_float64(d1); >> +} > > Surely we can reuse the helper we already have for implementing > the A64 version of this instruction ? > Absolutely! I just wasn't aware there was one :-) Regards, Ard.
On 27 March 2014 09:29, Ard Biesheuvel <ard.biesheuvel@linaro.org> wrote: > This adds support for the VMULL.P64 polynomial 64x64 to 128 bit multiplication > instruction, which is an optional feature that is part of the v8 Crypto > Extensions. > > Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> > --- I've reimplemented this to use the existing A64 helpers, and will post a patch shortly. thanks -- PMM
diff --git a/target-arm/cpu.c b/target-arm/cpu.c index 58c4584ac3bc..60244c7ffc82 100644 --- a/target-arm/cpu.c +++ b/target-arm/cpu.c @@ -293,6 +293,7 @@ static void arm_cpu_realizefn(DeviceState *dev, Error **errp) set_feature(env, ARM_FEATURE_V8_AES); set_feature(env, ARM_FEATURE_V8_SHA1); set_feature(env, ARM_FEATURE_V8_SHA256); + set_feature(env, ARM_FEATURE_V8_PMULL); } if (arm_feature(env, ARM_FEATURE_V7)) { set_feature(env, ARM_FEATURE_VAPA); diff --git a/target-arm/cpu.h b/target-arm/cpu.h index f5039d8b0177..d8add6d565a6 100644 --- a/target-arm/cpu.h +++ b/target-arm/cpu.h @@ -632,6 +632,7 @@ enum arm_features { ARM_FEATURE_CRC, /* ARMv8 CRC instructions */ ARM_FEATURE_V8_SHA1, /* implements SHA1 part of v8 Crypto Extensions */ ARM_FEATURE_V8_SHA256, /* implements SHA256 part of v8 Crypto Extensions */ + ARM_FEATURE_V8_PMULL, /* implements PMULL 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 211be36ebda8..b56a767b527e 100644 --- a/target-arm/crypto_helper.c +++ b/target-arm/crypto_helper.c @@ -522,3 +522,22 @@ 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]); } + +void HELPER(crypto_pmull)(CPUARMState *env, uint32_t rd, uint32_t rn, + uint32_t rm) +{ + uint64_t n = float64_val(env->vfp.regs[rn]); + uint64_t m = float64_val(env->vfp.regs[rm]); + uint64_t d0 = (n & 1) ? m : 0; + uint64_t d1 = 0; + int shift; + + for (shift = 1; (n >>= 1); shift++) { + if (n & 1) { + d0 ^= m << shift; + d1 ^= m >> (64 - shift); + } + } + 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 9024aef75157..8333f7dd0be2 100644 --- a/target-arm/helper.h +++ b/target-arm/helper.h @@ -521,6 +521,8 @@ 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_pmull, 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) diff --git a/target-arm/translate.c b/target-arm/translate.c index e79241402da8..576cdc24b530 100644 --- a/target-arm/translate.c +++ b/target-arm/translate.c @@ -5917,7 +5917,7 @@ static int disas_neon_data_insn(CPUARMState * env, DisasContext *s, uint32_t ins {0, 0, 0, 6}, /* VQDMLSL */ {0, 0, 0, 0}, /* Integer VMULL */ {0, 0, 0, 2}, /* VQDMULL */ - {0, 0, 0, 5}, /* Polynomial VMULL */ + {0, 0, 0, 4}, /* Polynomial VMULL */ {0, 0, 0, 3}, /* Reserved: always UNDEF */ }; @@ -5937,6 +5937,22 @@ static int disas_neon_data_insn(CPUARMState * env, DisasContext *s, uint32_t ins return 1; } + /* Handle VMULL.P64 (Polynomial 64x64 to 128 bit multiply) + outside the loop below as it only performs a single pass. */ + if (op == 14 && size == 2) { + if (!arm_feature(env, ARM_FEATURE_V8_PMULL)) { + return 1; + } + tmp = tcg_const_i32(rd); + tmp2 = tcg_const_i32(rn); + tmp3 = tcg_const_i32(rm); + gen_helper_crypto_pmull(cpu_env, tmp, tmp2, tmp3); + tcg_temp_free_i32(tmp); + tcg_temp_free_i32(tmp2); + tcg_temp_free_i32(tmp3); + return 0; + } + /* Avoid overlapping operands. Wide source operands are always aligned so will never overlap with wide destinations in problematic ways. */
This adds support for the VMULL.P64 polynomial 64x64 to 128 bit multiplication instruction, which is an optional feature that is part of the v8 Crypto Extensions. Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> --- This is an incremental patch on top of the SHA-1/SHA-256 patch I sent earlier this week. target-arm/cpu.c | 1 + target-arm/cpu.h | 1 + target-arm/crypto_helper.c | 19 +++++++++++++++++++ target-arm/helper.h | 2 ++ target-arm/translate.c | 18 +++++++++++++++++- 5 files changed, 40 insertions(+), 1 deletion(-)