From: Kiran Ostrolenk <kiran.ostrolenk@xxxxxxxxxxxxxxx>
This commit adds support for the Zvknh vector-crypto extension, which
consists of the following instructions:
* vsha2ms.vv
* vsha2c[hl].vv
Translation functions are defined in
`target/riscv/insn_trans/trans_rvvk.c.inc` and helpers are defined in
`target/riscv/vcrypto_helper.c`.
Co-authored-by: Nazar Kazakov <nazar.kazakov@xxxxxxxxxxxxxxx>
Co-authored-by: Lawrence Hunter <lawrence.hunter@xxxxxxxxxxxxxxx>
Signed-off-by: Nazar Kazakov <nazar.kazakov@xxxxxxxxxxxxxxx>
Signed-off-by: Lawrence Hunter <lawrence.hunter@xxxxxxxxxxxxxxx>
Signed-off-by: Kiran Ostrolenk <kiran.ostrolenk@xxxxxxxxxxxxxxx>
---
target/riscv/cpu.c | 10 +-
target/riscv/cpu.h | 2 +
target/riscv/helper.h | 4 +
target/riscv/insn32.decode | 5 +
target/riscv/insn_trans/trans_rvvk.c.inc | 70 ++++++++
target/riscv/vcrypto_helper.c | 214 +++++++++++++++++++++++
6 files changed, 302 insertions(+), 3 deletions(-)
diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 54d27301bce..dd8573bb02f 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -112,6 +112,8 @@ static const struct isa_ext_data isa_edata_arr[] = {
ISA_EXT_DATA_ENTRY(zvbb, true, PRIV_VERSION_1_12_0, ext_zvbb),
ISA_EXT_DATA_ENTRY(zvbc, true, PRIV_VERSION_1_12_0, ext_zvbc),
ISA_EXT_DATA_ENTRY(zvkned, true, PRIV_VERSION_1_12_0, ext_zvkned),
+ ISA_EXT_DATA_ENTRY(zvknha, true, PRIV_VERSION_1_12_0, ext_zvknha),
+ ISA_EXT_DATA_ENTRY(zvknhb, true, PRIV_VERSION_1_12_0, ext_zvknhb),
ISA_EXT_DATA_ENTRY(zhinx, true, PRIV_VERSION_1_12_0, ext_zhinx),
ISA_EXT_DATA_ENTRY(zhinxmin, true, PRIV_VERSION_1_12_0, ext_zhinxmin),
ISA_EXT_DATA_ENTRY(smaia, true, PRIV_VERSION_1_12_0, ext_smaia),
@@ -1218,7 +1220,7 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
* In principle Zve*x would also suffice here, were they supported
* in qemu
*/
- if ((cpu->cfg.ext_zvbb || cpu->cfg.ext_zvkned) &&
+ if ((cpu->cfg.ext_zvbb || cpu->cfg.ext_zvkned || cpu->cfg.ext_zvknha) &&
!(cpu->cfg.ext_zve32f || cpu->cfg.ext_zve64f || cpu->cfg.ext_zve64d ||
cpu->cfg.ext_v)) {
error_setg(errp,
@@ -1226,9 +1228,11 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
return;
}
- if (cpu->cfg.ext_zvbc &&
+ if ((cpu->cfg.ext_zvbc || cpu->cfg.ext_zvknhb) &&
!(cpu->cfg.ext_zve64f || cpu->cfg.ext_zve64d || cpu->cfg.ext_v)) {
- error_setg(errp, "Zvbc extension requires V or Zve64{f,d} extensions");
+ error_setg(
+ errp,
+ "Zvbc and Zvknhb extensions require V or Zve64{f,d} extensions");
return;
}
diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index 0d6c216572e..4c44088c28b 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -473,6 +473,8 @@ struct RISCVCPUConfig {
bool ext_zvbb;
bool ext_zvbc;
bool ext_zvkned;
+ bool ext_zvknha;
+ bool ext_zvknhb;
bool ext_zmmul;
bool ext_zvfh;
bool ext_zvfhmin;
diff --git a/target/riscv/helper.h b/target/riscv/helper.h
index db629cf6a89..a60129983be 100644
--- a/target/riscv/helper.h
+++ b/target/riscv/helper.h
@@ -1223,3 +1223,7 @@ DEF_HELPER_4(vaesdm_vs, void, ptr, ptr, env, i32)
DEF_HELPER_4(vaesz_vs, void, ptr, ptr, env, i32)
DEF_HELPER_5(vaeskf1_vi, void, ptr, ptr, i32, env, i32)
DEF_HELPER_5(vaeskf2_vi, void, ptr, ptr, i32, env, i32)
+
+DEF_HELPER_5(vsha2ms_vv, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vsha2ch_vv, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vsha2cl_vv, void, ptr, ptr, ptr, env, i32)
diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 7e0295d4935..d2cfb2729c2 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -948,3 +948,8 @@ vaesdm_vs 101001 1 ..... 00000 010 ..... 1110111 @r2_vm_1
vaesz_vs 101001 1 ..... 00111 010 ..... 1110111 @r2_vm_1
vaeskf1_vi 100010 1 ..... ..... 010 ..... 1110111 @r_vm_1
vaeskf2_vi 101010 1 ..... ..... 010 ..... 1110111 @r_vm_1
+
+# *** Zvknh vector crypto extension ***
+vsha2ms_vv 101101 1 ..... ..... 010 ..... 1110111 @r_vm_1
+vsha2ch_vv 101110 1 ..... ..... 010 ..... 1110111 @r_vm_1
+vsha2cl_vv 101111 1 ..... ..... 010 ..... 1110111 @r_vm_1
diff --git a/target/riscv/insn_trans/trans_rvvk.c.inc b/target/riscv/insn_trans/trans_rvvk.c.inc
index 9a36a731e5e..fd4e498b838 100644
--- a/target/riscv/insn_trans/trans_rvvk.c.inc
+++ b/target/riscv/insn_trans/trans_rvvk.c.inc
@@ -437,3 +437,73 @@ static bool vaeskf2_check(DisasContext *s, arg_vaeskf2_vi *a)
GEN_VI_UNMASKED_TRANS(vaeskf1_vi, vaeskf1_check, ZVKNED_EGS)
GEN_VI_UNMASKED_TRANS(vaeskf2_vi, vaeskf2_check, ZVKNED_EGS)
+
+/*
+ * Zvknh
+ */
+
+#define ZVKNH_EGS 4
+
+#define GEN_VV_UNMASKED_TRANS(NAME, CHECK, VL_MULTIPLE) \
+ static bool trans_##NAME(DisasContext *s, arg_rmrr *a) \
+ { \
+ if (CHECK(s, a)) { \
+ uint32_t data = 0; \
+ TCGLabel *over = gen_new_label(); \
+ TCGLabel *vl_ok = gen_new_label(); \
+ TCGv_i32 tmp = tcg_temp_new_i32(); \
+ \
+ /* save opcode for unwinding in case we throw an exception */ \
+ decode_save_opc(s); \
+ \
+ /* check (vl % VL_MULTIPLE == 0) assuming it's power of 2 */ \
+ tcg_gen_trunc_tl_i32(tmp, cpu_vl); \
+ tcg_gen_andi_i32(tmp, tmp, VL_MULTIPLE - 1); \
+ tcg_gen_brcondi_i32(TCG_COND_EQ, tmp, 0, vl_ok); \
+ gen_helper_restore_cpu_and_raise_exception( \
+ cpu_env, tcg_constant_i32(RISCV_EXCP_ILLEGAL_INST)); \
+ gen_set_label(vl_ok); \
+ \
+ tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over); \
+ \
+ data = FIELD_DP32(data, VDATA, VM, a->vm); \
+ data = FIELD_DP32(data, VDATA, LMUL, s->lmul); \
+ data = FIELD_DP32(data, VDATA, VTA, s->vta); \
+ data = FIELD_DP32(data, VDATA, VTA_ALL_1S, s->cfg_vta_all_1s); \
+ data = FIELD_DP32(data, VDATA, VMA, s->vma); \
+ \
+ tcg_gen_gvec_3_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, a->rs1), \
+ vreg_ofs(s, a->rs2), cpu_env, \
+ s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, \
+ data, gen_helper_##NAME); \
+ \
+ mark_vs_dirty(s); \
+ gen_set_label(over); \
+ return true; \
+ } \
+ return false; \
+ }
+
+static bool vsha_check_sew(DisasContext *s)
+{
+ return (s->cfg_ptr->ext_zvknha == true && s->sew == MO_32) ||
+ (s->cfg_ptr->ext_zvknhb == true &&
+ (s->sew == MO_32 || s->sew == MO_64));
+}
+
+static bool vsha_check(DisasContext *s, arg_rmrr *a)
+{
+ int egw_bytes = ZVKNH_EGS << s->sew;
+ int mult = 1 << MAX(s->lmul, 0);
+ return opivv_check(s, a) &&
+ vsha_check_sew(s) &&
+ MAXSZ(s) >= egw_bytes &&
+ !is_overlapped(a->rd, mult, a->rs1, mult) &&
+ !is_overlapped(a->rd, mult, a->rs2, mult) &&
+ s->vstart % ZVKNH_EGS == 0 &&
+ s->lmul >= 0;
+}
+
+GEN_VV_UNMASKED_TRANS(vsha2ms_vv, vsha_check, ZVKNH_EGS)
+GEN_VV_UNMASKED_TRANS(vsha2cl_vv, vsha_check, ZVKNH_EGS)
+GEN_VV_UNMASKED_TRANS(vsha2ch_vv, vsha_check, ZVKNH_EGS)
diff --git a/target/riscv/vcrypto_helper.c b/target/riscv/vcrypto_helper.c
index 0988eb74c81..ca09062c6c1 100644
--- a/target/riscv/vcrypto_helper.c
+++ b/target/riscv/vcrypto_helper.c
@@ -503,3 +503,217 @@ void HELPER(vaeskf2_vi)(void *vd_vptr, void *vs2_vptr, uint32_t uimm,
/* set tail elements to 1s */
vext_set_elems_1s(vd, vta, vl * 4, total_elems * 4);
}
+
+static inline uint32_t sig0_sha256(uint32_t x)
+{
+ return ror32(x, 7) ^ ror32(x, 18) ^ (x >> 3);
+}
+
+static inline uint32_t sig1_sha256(uint32_t x)
+{
+ return ror32(x, 17) ^ ror32(x, 19) ^ (x >> 10);
+}
+
+static inline uint64_t sig0_sha512(uint64_t x)
+{
+ return ror64(x, 1) ^ ror64(x, 8) ^ (x >> 7);
+}
+
+static inline uint64_t sig1_sha512(uint64_t x)
+{
+ return ror64(x, 19) ^ ror64(x, 61) ^ (x >> 6);
+}
+
+static inline void vsha2ms_e32(uint32_t *vd, uint32_t *vs1, uint32_t *vs2)
+{
+ uint32_t res[4];
+ res[0] = sig1_sha256(vs1[H4(2)]) + vs2[H4(1)] + sig0_sha256(vd[H4(1)]) +
+ vd[H4(0)];
+ res[1] = sig1_sha256(vs1[H4(3)]) + vs2[H4(2)] + sig0_sha256(vd[H4(2)]) +
+ vd[H4(1)];
+ res[2] =
+ sig1_sha256(res[0]) + vs2[H4(3)] + sig0_sha256(vd[H4(3)]) + vd[H4(2)];
+ res[3] =
+ sig1_sha256(res[1]) + vs1[H4(0)] + sig0_sha256(vs2[H4(0)]) + vd[H4(3)];
+ vd[H4(3)] = res[3];
+ vd[H4(2)] = res[2];
+ vd[H4(1)] = res[1];
+ vd[H4(0)] = res[0];
+}
+
+static inline void vsha2ms_e64(uint64_t *vd, uint64_t *vs1, uint64_t *vs2)
+{
+ uint64_t res[4];
+ res[0] = sig1_sha512(vs1[2]) + vs2[1] + sig0_sha512(vd[1]) + vd[0];
+ res[1] = sig1_sha512(vs1[3]) + vs2[2] + sig0_sha512(vd[2]) + vd[1];
+ res[2] = sig1_sha512(res[0]) + vs2[3] + sig0_sha512(vd[3]) + vd[2];
+ res[3] = sig1_sha512(res[1]) + vs1[0] + sig0_sha512(vs2[0]) + vd[3];
+ vd[3] = res[3];
+ vd[2] = res[2];
+ vd[1] = res[1];
+ vd[0] = res[0];
+}
+
+void HELPER(vsha2ms_vv)(void *vd, void *vs1, void *vs2, CPURISCVState *env,
+ uint32_t desc)
+{
+ uint32_t sew = FIELD_EX64(env->vtype, VTYPE, VSEW);
+ uint32_t esz = sew == MO_32 ? 4 : 8;
+ uint32_t total_elems;
+ uint32_t vta = vext_vta(desc);
+
+ for (uint32_t i = env->vstart / 4; i < env->vl / 4; i++) {
+ if (sew == MO_32) {
+ vsha2ms_e32(((uint32_t *)vd) + i * 4, ((uint32_t *)vs1) + i * 4,
+ ((uint32_t *)vs2) + i * 4);
+ } else {
+ /* If not 32 then SEW should be 64 */
+ vsha2ms_e64(((uint64_t *)vd) + i * 4, ((uint64_t *)vs1) + i * 4,
+ ((uint64_t *)vs2) + i * 4);
+ }
+ }
+ /* set tail elements to 1s */
+ total_elems = vext_get_total_elems(env, desc, esz);
+ vext_set_elems_1s(vd, vta, env->vl * esz, total_elems * esz);
+ env->vstart = 0;
+}
+
+static inline uint64_t sum0_64(uint64_t x)
+{
+ return ror64(x, 28) ^ ror64(x, 34) ^ ror64(x, 39);
+}
+
+static inline uint32_t sum0_32(uint32_t x)
+{
+ return ror32(x, 2) ^ ror32(x, 13) ^ ror32(x, 22);
+}
+
+static inline uint64_t sum1_64(uint64_t x)
+{
+ return ror64(x, 14) ^ ror64(x, 18) ^ ror64(x, 41);
+}
+
+static inline uint32_t sum1_32(uint32_t x)
+{
+ return ror32(x, 6) ^ ror32(x, 11) ^ ror32(x, 25);
+}
+
+#define ch(x, y, z) ((x & y) ^ ((~x) & z))
+
+#define maj(x, y, z) ((x & y) ^ (x & z) ^ (y & z))
+
+static void vsha2c_64(uint64_t *vs2, uint64_t *vd, uint64_t *vs1)
+{
+ uint64_t a = vs2[3], b = vs2[2], e = vs2[1], f = vs2[0];
+ uint64_t c = vd[3], d = vd[2], g = vd[1], h = vd[0];
+ uint64_t W0 = vs1[0], W1 = vs1[1];
+ uint64_t T1 = h + sum1_64(e) + ch(e, f, g) + W0;
+ uint64_t T2 = sum0_64(a) + maj(a, b, c);
+
+ h = g;
+ g = f;
+ f = e;
+ e = d + T1;
+ d = c;
+ c = b;
+ b = a;
+ a = T1 + T2;
+
+ T1 = h + sum1_64(e) + ch(e, f, g) + W1;
+ T2 = sum0_64(a) + maj(a, b, c);
+ h = g;
+ g = f;
+ f = e;
+ e = d + T1;
+ d = c;
+ c = b;
+ b = a;
+ a = T1 + T2;
+
+ vd[0] = f;
+ vd[1] = e;
+ vd[2] = b;
+ vd[3] = a;
+}
+
+static void vsha2c_32(uint32_t *vs2, uint32_t *vd, uint32_t *vs1)
+{
+ uint32_t a = vs2[H4(3)], b = vs2[H4(2)], e = vs2[H4(1)], f = vs2[H4(0)];
+ uint32_t c = vd[H4(3)], d = vd[H4(2)], g = vd[H4(1)], h = vd[H4(0)];
+ uint32_t W0 = vs1[H4(0)], W1 = vs1[H4(1)];
+ uint32_t T1 = h + sum1_32(e) + ch(e, f, g) + W0;
+ uint32_t T2 = sum0_32(a) + maj(a, b, c);
+
+ h = g;
+ g = f;
+ f = e;
+ e = d + T1;
+ d = c;
+ c = b;
+ b = a;
+ a = T1 + T2;
+
+ T1 = h + sum1_32(e) + ch(e, f, g) + W1;
+ T2 = sum0_32(a) + maj(a, b, c);
+ h = g;
+ g = f;
+ f = e;
+ e = d + T1;
+ d = c;
+ c = b;
+ b = a;
+ a = T1 + T2;
+
+ vd[H4(0)] = f;
+ vd[H4(1)] = e;
+ vd[H4(2)] = b;
+ vd[H4(3)] = a;
+}
+
+void HELPER(vsha2ch_vv)(void *vd, void *vs1, void *vs2, CPURISCVState *env,
+ uint32_t desc)
+{
+ uint32_t sew = FIELD_EX64(env->vtype, VTYPE, VSEW);
+ uint32_t esz = sew == MO_64 ? 8 : 4;
+ uint32_t total_elems;
+ uint32_t vta = vext_vta(desc);
+
+ for (uint32_t i = env->vstart / 4; i < env->vl / 4; i++) {
+ if (sew == MO_64) {
+ vsha2c_64(((uint64_t *)vs2) + 4 * i, ((uint64_t *)vd) + 4 * i,
+ ((uint64_t *)vs1) + 4 * i + 2);
+ } else {
+ vsha2c_32(((uint32_t *)vs2) + 4 * i, ((uint32_t *)vd) + 4 * i,
+ ((uint32_t *)vs1) + 4 * i + 2);
+ }
+ }
+
+ /* set tail elements to 1s */
+ total_elems = vext_get_total_elems(env, desc, esz);
+ vext_set_elems_1s(vd, vta, env->vl * esz, total_elems * esz);
+ env->vstart = 0;
+}
+
+void HELPER(vsha2cl_vv)(void *vd, void *vs1, void *vs2, CPURISCVState *env,
+ uint32_t desc)
+{
+ uint32_t sew = FIELD_EX64(env->vtype, VTYPE, VSEW);
+ uint32_t esz = sew == MO_64 ? 8 : 4;
+ uint32_t total_elems;
+ uint32_t vta = vext_vta(desc);
+
+ for (uint32_t i = env->vstart / 4; i < env->vl / 4; i++) {
+ if (sew == MO_64) {
+ vsha2c_64(((uint64_t *)vs2) + 4 * i, ((uint64_t *)vd) + 4 * i,
+ (((uint64_t *)vs1) + 4 * i));
+ } else {
+ vsha2c_32(((uint32_t *)vs2) + 4 * i, ((uint32_t *)vd) + 4 * i,
+ (((uint32_t *)vs1) + 4 * i));
+ }
+ }
+
+ /* set tail elements to 1s */
+ total_elems = vext_get_total_elems(env, desc, esz);
+ vext_set_elems_1s(vd, vta, env->vl * esz, total_elems * esz);
+ env->vstart = 0;
+}
--
2.40.0
