Skip to contentSkip to conversation

Add RISC-V vector cryptographic instruction set support

PATCH In review Lawrence Hunter 3 participants 50 messages Mar 10, 2023 – Mar 23, 2023

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

cover Add RISC-V vector cryptographic instruction set support

This patchset provides an implementation for Zvkb, Zvkned, Zvknh, Zvksh, Zvkg, and Zvksed of the draft RISC-V vector cryptography extensions as per the 20230303 version of the specification(1) (1fcbb30). Please note that the Zvkt data-independent execution latency extension has not been implemented, and we would recommend not using these patches in an environment where timing attacks are an issue.

Work performed by Dickon, Lawrence, Nazar, Kiran, and William from Codethink sponsored by SiFive, as well as Max Chou and Frank Chang from SiFive.

For convenience we have created a git repo with our patches on top of a recent master. https://github.com/CodethinkLabs/qemu-ct

1. https://github.com/riscv/riscv-crypto/releases

Dickon Hood (2):
qemu/bitops.h: Limit rotate amounts
target/riscv: Add vrol.[vv,vx] and vror.[vv,vx,vi] decoding,
translation and execution support

Kiran Ostrolenk (8):
target/riscv: Refactor some of the generic vector functionality target/riscv: Refactor some of the generic vector functionality target/riscv: Refactor some of the generic vector functionality target/riscv: Refactor some of the generic vector functionality target/riscv: Add vsha2ms.vv decoding, translation and execution
support
target/riscv: Add zvksh cpu property
target/riscv: Add vsm3c.vi decoding, translation and execution support target/riscv: Expose zvksh cpu property

Lawrence Hunter (17):
target/riscv: Add vclmul.vv decoding, translation and execution
support
target/riscv: Add vclmul.vx decoding, translation and execution
support
target/riscv: Add vclmulh.vv decoding, translation and execution
support
target/riscv: Add vclmulh.vx decoding, translation and execution
support
target/riscv: Add vaesef.vv decoding, translation and execution
support
target/riscv: Add vaesef.vs decoding, translation and execution
support
target/riscv: Add vaesdf.vv decoding, translation and execution
support
target/riscv: Add vaesdf.vs decoding, translation and execution
support
target/riscv: Add vaesdm.vv decoding, translation and execution
support
target/riscv: Add vaesdm.vs decoding, translation and execution
support
target/riscv: Add vaesz.vs decoding, translation and execution support target/riscv: Add vsha2c[hl].vv decoding, translation and execution
support
target/riscv: Add vsm3me.vv decoding, translation and execution
support
target/riscv: Add zvkg cpu property
target/riscv: Add vgmul.vv decoding, translation and execution support target/riscv: Add vghsh.vv decoding, translation and execution support target/riscv: Expose zvkg cpu property

Max Chou (5):
crypto: Create sm4_subword
crypto: Add SM4 constant parameter CK
target/riscv: Add zvksed cfg property
target/riscv: Add Zvksed support
target/riscv: Expose Zvksed property

Nazar Kazakov (10):
target/riscv: Add zvkb cpu property
target/riscv: Add vrev8.v decoding, translation and execution support target/riscv: Add vandn.[vv,vx] decoding, translation and execution
support
target/riscv: Expose zvkb cpu property
target/riscv: Add zvkned cpu property
target/riscv: Add vaeskf1.vi decoding, translation and execution
support
target/riscv: Add vaeskf2.vi decoding, translation and execution
support
target/riscv: Expose zvkned cpu property
target/riscv: Add zvknh cpu properties
target/riscv: Expose zvknh cpu properties

William Salmon (3):
target/riscv: Add vbrev8.v decoding, translation and execution support target/riscv: Add vaesem.vv decoding, translation and execution
support
target/riscv: Add vaesem.vs decoding, translation and execution
support

 accel/tcg/tcg-runtime-gvec.c                 |   11 +
 accel/tcg/tcg-runtime.h                      |    1 +
 crypto/sm4.c                                 |   10 +
 include/crypto/sm4.h                         |    9 +
 include/qemu/bitops.h                        |   24 +-
 target/arm/tcg/crypto_helper.c               |   10 +-
 target/riscv/cpu.c                           |   36 +
 target/riscv/cpu.h                           |    7 +
 target/riscv/helper.h                        |   71 ++
 target/riscv/insn32.decode                   |   49 +
 target/riscv/insn_trans/trans_rvv.c.inc      |   93 +-
 target/riscv/insn_trans/trans_rvzvkb.c.inc   |  220 ++++
 target/riscv/insn_trans/trans_rvzvkg.c.inc   |   40 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc |  170 +++
 target/riscv/insn_trans/trans_rvzvknh.c.inc  |   84 ++
 target/riscv/insn_trans/trans_rvzvksed.c.inc |   57 +
 target/riscv/insn_trans/trans_rvzvksh.c.inc  |   43 +
 target/riscv/meson.build                     |    4 +-
 target/riscv/op_helper.c                     |    5 +
 target/riscv/translate.c                     |    6 +
 target/riscv/vcrypto_helper.c                | 1001 ++++++++++++++++++
 target/riscv/vector_helper.c                 |  240 +----
 target/riscv/vector_internals.c              |   81 ++
 target/riscv/vector_internals.h              |  222 ++++
 24 files changed, 2192 insertions(+), 302 deletions(-)
 create mode 100644 target/riscv/insn_trans/trans_rvzvkb.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvkg.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvkned.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvknh.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvksed.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvksh.c.inc
 create mode 100644 target/riscv/vcrypto_helper.c
 create mode 100644 target/riscv/vector_internals.c
 create mode 100644 target/riscv/vector_internals.h
Signature · 1 line
2.39.2
#1 lore raw

PATCH 1/45 target/riscv: Add zvkb cpu property

From: Nazar Kazakov <[email protected]>

---
 target/riscv/cpu.c | 13 +++++++++++++
 target/riscv/cpu.h |  1 +
 2 files changed, 14 insertions(+)
target/riscv/cpu.c +13 -0
index 1e97473af2..69611408f9 100644
@@ -109,6 +109,7 @@ static const struct isa_ext_data isa_edata_arr[] = {
     ISA_EXT_DATA_ENTRY(zve64d, true, PRIV_VERSION_1_12_0, ext_zve64d),
     ISA_EXT_DATA_ENTRY(zvfh, true, PRIV_VERSION_1_12_0, ext_zvfh),
     ISA_EXT_DATA_ENTRY(zvfhmin, true, PRIV_VERSION_1_12_0, ext_zvfhmin),
+    ISA_EXT_DATA_ENTRY(zvkb, true, PRIV_VERSION_1_12_0, ext_zvkb),
     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),
@@ -1211,6 +1212,18 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
         return;
     }
 
+    /*
+    * In principle zve*x would also suffice here, were they supported
+    * in qemu
+    */
+    if (cpu->cfg.ext_zvkb &&
+        !(cpu->cfg.ext_zve32f || cpu->cfg.ext_zve64f ||
+            cpu->cfg.ext_zve64d || cpu->cfg.ext_v)) {
+        error_setg(
+            errp, "Vector crypto extensions require V or Zve* extensions");
+        return;
+    }
+
 #ifndef CONFIG_USER_ONLY
     if (cpu->cfg.pmu_num) {
         if (!riscv_pmu_init(cpu, cpu->cfg.pmu_num) && cpu->cfg.ext_sscofpmf) {
target/riscv/cpu.h +1 -0
index 638e47c75a..7d6699f718 100644
@@ -470,6 +470,7 @@ struct RISCVCPUConfig {
     bool ext_zve32f;
     bool ext_zve64f;
     bool ext_zve64d;
+    bool ext_zvkb;
     bool ext_zmmul;
     bool ext_zvfh;
     bool ext_zvfhmin;
Signature · 1 line
2.39.2
#2 lore raw

PATCH 2/45 target/riscv: Refactor some of the generic vector functionality

In reply to Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

Summary of refactoring:

* take some functions/macros out of vector_helper and put them in a new module called vector_internals

* factor the non SEW-specific stuff out of GEN_OPIVV_TRANS into function opivv_trans (similar to opivi_trans)

All this refactoring ensures more functions/macros can be used by both vector and vector-crypto helpers (latter implemented in proceeding commit).

---
 target/riscv/insn_trans/trans_rvv.c.inc |  62 ++++++------
 target/riscv/meson.build                |   1 +
 target/riscv/vector_helper.c            | 120 +-----------------------
 target/riscv/vector_internals.c         |  57 +++++++++++
 target/riscv/vector_internals.h         | 118 +++++++++++++++++++++++
 5 files changed, 209 insertions(+), 149 deletions(-)
 create mode 100644 target/riscv/vector_internals.c
 create mode 100644 target/riscv/vector_internals.h
target/riscv/insn_trans/trans_rvv.c.inc +32 -30
index f2e3d38515..4106bd6994 100644
@@ -1643,38 +1643,40 @@ GEN_OPIWX_WIDEN_TRANS(vwadd_wx)
 GEN_OPIWX_WIDEN_TRANS(vwsubu_wx)
 GEN_OPIWX_WIDEN_TRANS(vwsub_wx)
 
+static bool opivv_trans(uint32_t vd, uint32_t vs1, uint32_t vs2, uint32_t vm,
+                        gen_helper_gvec_4_ptr *fn, DisasContext *s)
+{
+    uint32_t data = 0;
+    TCGLabel *over = gen_new_label();
+    tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
+    tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
+
+    data = FIELD_DP32(data, VDATA, VM, 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_4_ptr(vreg_ofs(s, vd), vreg_ofs(s, 0), vreg_ofs(s, vs1),
+                       vreg_ofs(s, vs2), cpu_env, s->cfg_ptr->vlen / 8,
+                       s->cfg_ptr->vlen / 8, data, fn);
+    mark_vs_dirty(s);
+    gen_set_label(over);
+    return true;
+}
+
 /* Vector Integer Add-with-Carry / Subtract-with-Borrow Instructions */
 /* OPIVV without GVEC IR */
-#define GEN_OPIVV_TRANS(NAME, CHECK)                               \
-static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
-{                                                                  \
-    if (CHECK(s, a)) {                                             \
-        uint32_t data = 0;                                         \
-        static gen_helper_gvec_4_ptr * const fns[4] = {            \
-            gen_helper_##NAME##_b, gen_helper_##NAME##_h,          \
-            gen_helper_##NAME##_w, gen_helper_##NAME##_d,          \
-        };                                                         \
-        TCGLabel *over = gen_new_label();                          \
-        tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);          \
-        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_4_ptr(vreg_ofs(s, a->rd), vreg_ofs(s, 0),     \
-                           vreg_ofs(s, a->rs1),                    \
-                           vreg_ofs(s, a->rs2), cpu_env,           \
-                           s->cfg_ptr->vlen / 8,                   \
-                           s->cfg_ptr->vlen / 8, data,             \
-                           fns[s->sew]);                           \
-        mark_vs_dirty(s);                                          \
-        gen_set_label(over);                                       \
-        return true;                                               \
-    }                                                              \
-    return false;                                                  \
+#define GEN_OPIVV_TRANS(NAME, CHECK)                                     \
+static bool trans_##NAME(DisasContext *s, arg_rmrr *a)                   \
+{                                                                        \
+    if (CHECK(s, a)) {                                                   \
+        static gen_helper_gvec_4_ptr * const fns[4] = {                  \
+            gen_helper_##NAME##_b, gen_helper_##NAME##_h,                \
+            gen_helper_##NAME##_w, gen_helper_##NAME##_d,                \
+        };                                                               \
+        return opivv_trans(a->rd, a->rs1, a->rs2, a->vm, fns[s->sew], s);\
+    }                                                                    \
+    return false;                                                        \
 }
 
 /*
target/riscv/meson.build +1 -0
index 5dee37a242..a94fc3f598 100644
@@ -16,6 +16,7 @@ riscv_ss.add(files(
   'gdbstub.c',
   'op_helper.c',
   'vector_helper.c',
+  'vector_internals.c',
   'bitmanip_helper.c',
   'translate.c',
   'm128_helper.c',
target/riscv/vector_helper.c +1 -119

162 lines not on this page · raw message

target/riscv/vector_internals.c new file +57 -0

57 lines not on this page · raw message

target/riscv/vector_internals.h new file +118 -0

118 lines not on this page · raw message

Signature · 1 line
2.39.2
#3 lore raw

PATCH 3/45 target/riscv: Add vclmul.vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                      |  3 ++
 target/riscv/insn32.decode                 |  3 ++
 target/riscv/insn_trans/trans_rvzvkb.c.inc | 40 +++++++++++++++++++++
 target/riscv/meson.build                   |  3 +-
 target/riscv/translate.c                   |  1 +
 target/riscv/vcrypto_helper.c              | 42 ++++++++++++++++++++++
 6 files changed, 91 insertions(+), 1 deletion(-)
 create mode 100644 target/riscv/insn_trans/trans_rvzvkb.c.inc
 create mode 100644 target/riscv/vcrypto_helper.c
target/riscv/helper.h +3 -0
index 37b54e0991..33060c3e2f 100644
@@ -1142,3 +1142,6 @@ DEF_HELPER_FLAGS_1(aes64im, TCG_CALL_NO_RWG_SE, tl, tl)
 
 DEF_HELPER_FLAGS_3(sm4ed, TCG_CALL_NO_RWG_SE, tl, tl, tl, tl)
 DEF_HELPER_FLAGS_3(sm4ks, TCG_CALL_NO_RWG_SE, tl, tl, tl, tl)
+
+/* Vector crypto functions */
+DEF_HELPER_6(vclmul_vv, void, ptr, ptr, ptr, ptr, env, i32)
target/riscv/insn32.decode +3 -0
index 73d5d1b045..03a0057d71 100644
@@ -908,3 +908,6 @@ sm4ks       .. 11010 ..... ..... 000 ..... 0110011 @k_aes
 # *** RV32 Zicond Standard Extension ***
 czero_eqz   0000111  ..... ..... 101 ..... 0110011 @r
 czero_nez   0000111  ..... ..... 111 ..... 0110011 @r
+
+# *** RV64 Zvkb vector crypto extension ***
+vclmul_vv       001100 . ..... ..... 010 ..... 1010111 @r_vm
target/riscv/insn_trans/trans_rvzvkb.c.inc new file +40 -0
new file mode 100644
index 0000000000..7cd920e76d
@@ -0,0 +1,40 @@
+/*
+ * RISC-V translation routines for the Zvkb Extension.
+ *
+ * Copyright (C) 2023 SiFive, Inc.
+ * Written by Codethink Ltd and SiFive.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2 or later, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#define GEN_VV_MASKED_TRANS(NAME, CHECK)                 \
+static bool trans_##NAME(DisasContext *s, arg_rmrr *a)   \
+{                                                        \
+    if (CHECK(s, a)) {                                   \
+        return opivv_trans(a->rd, a->rs1, a->rs2, a->vm, \
+                           gen_helper_##NAME, s);        \
+    }                                                    \
+    return false;                                        \
+}
+
+static bool zvkb_vv_check(DisasContext *s, arg_rmrr *a)
+{
+    return opivv_check(s, a) && s->cfg_ptr->ext_zvkb == true;
+}
+
+static bool vclmul_vv_check(DisasContext *s, arg_rmrr *a)
+{
+    return zvkb_vv_check(s, a) && s->sew == MO_64;
+}
+
+GEN_VV_MASKED_TRANS(vclmul_vv, vclmul_vv_check)
target/riscv/meson.build +2 -1
index a94fc3f598..52a61dd66e 100644
@@ -20,7 +20,8 @@ riscv_ss.add(files(
   'bitmanip_helper.c',
   'translate.c',
   'm128_helper.c',
-  'crypto_helper.c'
+  'crypto_helper.c',
+  'vcrypto_helper.c'
 ))
 riscv_ss.add(when: 'CONFIG_KVM', if_true: files('kvm.c'), if_false: files('kvm-stub.c'))
 
target/riscv/translate.c +1 -0
index 0ee8ee147d..939f5aeb1c 100644
@@ -1083,6 +1083,7 @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc)
 #include "insn_trans/trans_rvzicbo.c.inc"
 #include "insn_trans/trans_rvzfh.c.inc"
 #include "insn_trans/trans_rvk.c.inc"
+#include "insn_trans/trans_rvzvkb.c.inc"
 #include "insn_trans/trans_privileged.c.inc"
 #include "insn_trans/trans_svinval.c.inc"
 #include "decode-xthead.c.inc"
target/riscv/vcrypto_helper.c new file +42 -0

42 lines not on this page · raw message

Signature · 1 line
2.39.2
#4 lore raw

PATCH 4/45 target/riscv: Refactor some of the generic vector functionality

In reply to Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

This refactoring ensures these functions/macros can be used by both vector and vector-crypto helpers (latter implemented in proceeding commit).

---
 target/riscv/vector_helper.c    | 36 ---------------------------------
 target/riscv/vector_internals.c | 24 ++++++++++++++++++++++
 target/riscv/vector_internals.h | 16 +++++++++++++++
 3 files changed, 40 insertions(+), 36 deletions(-)
target/riscv/vector_helper.c +0 -36
index 823aa8eb08..09b790653e 100644
@@ -721,8 +721,6 @@ GEN_VEXT_VV(vsub_vv_h, 2)
 GEN_VEXT_VV(vsub_vv_w, 4)
 GEN_VEXT_VV(vsub_vv_d, 8)
 
-typedef void opivx2_fn(void *vd, target_long s1, void *vs2, int i);
-
 /*
  * (T1)s1 gives the real operator type.
  * (TX1)(T1)s1 expands the operator type of widen or narrow operations.
@@ -747,40 +745,6 @@ RVVCALL(OPIVX2, vrsub_vx_h, OP_SSS_H, H2, H2, DO_RSUB)
 RVVCALL(OPIVX2, vrsub_vx_w, OP_SSS_W, H4, H4, DO_RSUB)
 RVVCALL(OPIVX2, vrsub_vx_d, OP_SSS_D, H8, H8, DO_RSUB)
 
-static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
-                       CPURISCVState *env, uint32_t desc,
-                       opivx2_fn fn, uint32_t esz)
-{
-    uint32_t vm = vext_vm(desc);
-    uint32_t vl = env->vl;
-    uint32_t total_elems = vext_get_total_elems(env, desc, esz);
-    uint32_t vta = vext_vta(desc);
-    uint32_t vma = vext_vma(desc);
-    uint32_t i;
-
-    for (i = env->vstart; i < vl; i++) {
-        if (!vm && !vext_elem_mask(v0, i)) {
-            /* set masked-off elements to 1s */
-            vext_set_elems_1s(vd, vma, i * esz, (i + 1) * esz);
-            continue;
-        }
-        fn(vd, s1, vs2, i);
-    }
-    env->vstart = 0;
-    /* set tail elements to 1s */
-    vext_set_elems_1s(vd, vta, vl * esz, total_elems * esz);
-}
-
-/* generate the helpers for OPIVX */
-#define GEN_VEXT_VX(NAME, ESZ)                            \
-void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
-                  void *vs2, CPURISCVState *env,          \
-                  uint32_t desc)                          \
-{                                                         \
-    do_vext_vx(vd, v0, s1, vs2, env, desc,                \
-               do_##NAME, ESZ);                           \
-}
-
 GEN_VEXT_VX(vadd_vx_b, 1)
 GEN_VEXT_VX(vadd_vx_h, 2)
 GEN_VEXT_VX(vadd_vx_w, 4)
target/riscv/vector_internals.c +24 -0
index 95efaa79cb..9cf5c17cde 100644
@@ -55,3 +55,27 @@ void do_vext_vv(void *vd, void *v0, void *vs1, void *vs2,
     /* set tail elements to 1s */
     vext_set_elems_1s(vd, vta, vl * esz, total_elems * esz);
 }
+
+void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
+                CPURISCVState *env, uint32_t desc,
+                opivx2_fn fn, uint32_t esz)
+{
+    uint32_t vm = vext_vm(desc);
+    uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env, desc, esz);
+    uint32_t vta = vext_vta(desc);
+    uint32_t vma = vext_vma(desc);
+    uint32_t i;
+
+    for (i = env->vstart; i < vl; i++) {
+        if (!vm && !vext_elem_mask(v0, i)) {
+            /* set masked-off elements to 1s */
+            vext_set_elems_1s(vd, vma, i * esz, (i + 1) * esz);
+            continue;
+        }
+        fn(vd, s1, vs2, i);
+    }
+    env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s(vd, vta, vl * esz, total_elems * esz);
+}
target/riscv/vector_internals.h +16 -0
index 1d26ff9514..90500e5df6 100644
@@ -115,4 +115,20 @@ void HELPER(NAME)(void *vd, void *v0, void *vs1,          \
                do_##NAME, ESZ);                           \
 }
 
+typedef void opivx2_fn(void *vd, target_long s1, void *vs2, int i);
+
+void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
+                CPURISCVState *env, uint32_t desc,
+                opivx2_fn fn, uint32_t esz);
+
+/* generate the helpers for OPIVX */
+#define GEN_VEXT_VX(NAME, ESZ)                            \
+void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
+                  void *vs2, CPURISCVState *env,          \
+                  uint32_t desc)                          \
+{                                                         \
+    do_vext_vx(vd, v0, s1, vs2, env, desc,                \
+               do_##NAME, ESZ);                           \
+}
+
 #endif /* TARGET_RISCV_VECTOR_INTERNALS_H */
Signature · 1 line
2.39.2
Reviewed-by on v1 4/45
#5 lore raw

Re: [PATCH 04/45] target/riscv: Refactor some of the generic vector functionality

Quoted 6 lines · Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

This refactoring ensures these functions/macros can be used by both vector and vector-crypto helpers (latter implemented in proceeding commit).

Quoted 8 lines

---
 target/riscv/vector_helper.c    | 36 ------------------------------- --
 target/riscv/vector_internals.c | 24 ++++++++++++++++++++++ target/riscv/vector_internals.h | 16 +++++++++++++++

 3 files changed, 40 insertions(+), 36 deletions(-)
Quoted 119 lines
target/riscv/vector_helper.c +0 -0
index 823aa8eb08..09b790653e 100644
@@ -721,8 +721,6 @@ GEN_VEXT_VV(vsub_vv_h, 2)
  GEN_VEXT_VV(vsub_vv_w, 4)
  GEN_VEXT_VV(vsub_vv_d, 8)

target/riscv/vector_helper.c +0 -2
-typedef void opivx2_fn(void *vd, target_long s1, void *vs2, int i);
-

 /*
  * (T1)s1 gives the real operator type.
  * (TX1)(T1)s1 expands the operator type of widen or narrow operations.

target/riscv/vector_helper.c +0 -0
@@ -747,40 +745,6 @@ RVVCALL(OPIVX2, vrsub_vx_h, OP_SSS_H, H2, H2,
 DO_RSUB)
  RVVCALL(OPIVX2, vrsub_vx_w, OP_SSS_W, H4, H4, DO_RSUB)
  RVVCALL(OPIVX2, vrsub_vx_d, OP_SSS_D, H8, H8, DO_RSUB)


-static void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
-                       CPURISCVState *env, uint32_t desc, -                       opivx2_fn fn, uint32_t esz)

target/riscv/vector_helper.c +0 -31
-{
-    uint32_t vm = vext_vm(desc);
-    uint32_t vl = env->vl;
-    uint32_t total_elems = vext_get_total_elems(env, desc, esz);
-    uint32_t vta = vext_vta(desc);
-    uint32_t vma = vext_vma(desc);
-    uint32_t i;
-
-    for (i = env->vstart; i < vl; i++) {
-        if (!vm && !vext_elem_mask(v0, i)) {
-            /* set masked-off elements to 1s */
-            vext_set_elems_1s(vd, vma, i * esz, (i + 1) * esz);
-            continue;
-        }
-        fn(vd, s1, vs2, i);
-    }
-    env->vstart = 0;
-    /* set tail elements to 1s */
-    vext_set_elems_1s(vd, vta, vl * esz, total_elems * esz);
-}
-
-/* generate the helpers for OPIVX */
-#define GEN_VEXT_VX(NAME, ESZ)                            \
-void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
-                  void *vs2, CPURISCVState *env,          \
-                  uint32_t desc)                          \
-{                                                         \
-    do_vext_vx(vd, v0, s1, vs2, env, desc,                \
-               do_##NAME, ESZ);                           \
-}
-

 GEN_VEXT_VX(vadd_vx_b, 1)
 GEN_VEXT_VX(vadd_vx_h, 2)
 GEN_VEXT_VX(vadd_vx_w, 4)

target/riscv/vector_internals.c +0 -0
index 95efaa79cb..9cf5c17cde 100644
@@ -55,3 +55,27 @@ void do_vext_vv(void *vd, void *v0, void *vs1,
 void *vs2,
      /* set tail elements to 1s */
      vext_set_elems_1s(vd, vta, vl * esz, total_elems * esz);

 }

target/riscv/vector_internals.c +24 -0
+
+void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
+                CPURISCVState *env, uint32_t desc,
+                opivx2_fn fn, uint32_t esz)
+{
+    uint32_t vm = vext_vm(desc);
+    uint32_t vl = env->vl;
+    uint32_t total_elems = vext_get_total_elems(env, desc, esz);
+    uint32_t vta = vext_vta(desc);
+    uint32_t vma = vext_vma(desc);
+    uint32_t i;
+
+    for (i = env->vstart; i < vl; i++) {
+        if (!vm && !vext_elem_mask(v0, i)) {
+            /* set masked-off elements to 1s */
+            vext_set_elems_1s(vd, vma, i * esz, (i + 1) * esz);
+            continue;
+        }
+        fn(vd, s1, vs2, i);
+    }
+    env->vstart = 0;
+    /* set tail elements to 1s */
+    vext_set_elems_1s(vd, vta, vl * esz, total_elems * esz);
+}
target/riscv/vector_internals.h +0 -0
index 1d26ff9514..90500e5df6 100644
@@ -115,4 +115,20 @@ void HELPER(NAME)(void *vd, void *v0, void
 *vs1,          \
                 do_##NAME, ESZ);                           \
  }

target/riscv/vector_internals.h +16 -0
+typedef void opivx2_fn(void *vd, target_long s1, void *vs2, int i);
+
+void do_vext_vx(void *vd, void *v0, target_long s1, void *vs2,
+                CPURISCVState *env, uint32_t desc,
+                opivx2_fn fn, uint32_t esz);
+
+/* generate the helpers for OPIVX */
+#define GEN_VEXT_VX(NAME, ESZ)                            \
+void HELPER(NAME)(void *vd, void *v0, target_ulong s1,    \
+                  void *vs2, CPURISCVState *env,          \
+                  uint32_t desc)                          \
+{                                                         \
+    do_vext_vx(vd, v0, s1, vs2, env, desc,                \
+               do_##NAME, ESZ);                           \
+}
+

 #endif /* TARGET_RISCV_VECTOR_INTERNALS_H */

#6 lore raw

PATCH 5/45 target/riscv: Add vclmul.vx decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                      |  1 +
 target/riscv/insn32.decode                 |  1 +
 target/riscv/insn_trans/trans_rvzvkb.c.inc | 48 ++++++++++++++++++++++
 target/riscv/vcrypto_helper.c              |  2 +
 4 files changed, 52 insertions(+)
target/riscv/helper.h +1 -0
index 33060c3e2f..352921ead6 100644
@@ -1145,3 +1145,4 @@ DEF_HELPER_FLAGS_3(sm4ks, TCG_CALL_NO_RWG_SE, tl, tl, tl, tl)
 
 /* Vector crypto functions */
 DEF_HELPER_6(vclmul_vv, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vclmul_vx, void, ptr, ptr, tl, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index 03a0057d71..6b8466424d 100644
@@ -911,3 +911,4 @@ czero_nez   0000111  ..... ..... 111 ..... 0110011 @r
 
 # *** RV64 Zvkb vector crypto extension ***
 vclmul_vv       001100 . ..... ..... 010 ..... 1010111 @r_vm
+vclmul_vx       001100 . ..... ..... 110 ..... 1010111 @r_vm
target/riscv/insn_trans/trans_rvzvkb.c.inc +48 -0

51 lines not on this page · raw message

target/riscv/vcrypto_helper.c +2 -0
index b4983886bd..749a9cb30b 100644
@@ -40,3 +40,5 @@ static uint64_t clmul64(uint64_t y, uint64_t x)
 
 RVVCALL(OPIVV2, vclmul_vv, OP_UUU_D, H8, H8, H8, clmul64)
 GEN_VEXT_VV(vclmul_vv, 8)
+RVVCALL(OPIVX2, vclmul_vx, OP_UUU_D, H8, H8, clmul64)
+GEN_VEXT_VX(vclmul_vx, 8)
Signature · 1 line
2.39.2
#7 lore raw

PATCH 6/45 target/riscv: Add vclmulh.vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                      |  1 +
 target/riscv/insn32.decode                 |  1 +
 target/riscv/insn_trans/trans_rvzvkb.c.inc |  1 +
 target/riscv/vcrypto_helper.c              | 13 +++++++++++++
 4 files changed, 16 insertions(+)
target/riscv/helper.h +1 -0
index 352921ead6..1c69c34a78 100644
@@ -1146,3 +1146,4 @@ DEF_HELPER_FLAGS_3(sm4ks, TCG_CALL_NO_RWG_SE, tl, tl, tl, tl)
 /* Vector crypto functions */
 DEF_HELPER_6(vclmul_vv, void, ptr, ptr, ptr, ptr, env, i32)
 DEF_HELPER_6(vclmul_vx, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vclmulh_vv, void, ptr, ptr, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index 6b8466424d..3ad8e2055b 100644
@@ -912,3 +912,4 @@ czero_nez   0000111  ..... ..... 111 ..... 0110011 @r
 # *** RV64 Zvkb vector crypto extension ***
 vclmul_vv       001100 . ..... ..... 010 ..... 1010111 @r_vm
 vclmul_vx       001100 . ..... ..... 110 ..... 1010111 @r_vm
+vclmulh_vv      001101 . ..... ..... 010 ..... 1010111 @r_vm
target/riscv/insn_trans/trans_rvzvkb.c.inc +1 -0
index 76efade1b6..63a8778acc 100644
@@ -38,6 +38,7 @@ static bool vclmul_vv_check(DisasContext *s, arg_rmrr *a)
 }
 
 GEN_VV_MASKED_TRANS(vclmul_vv, vclmul_vv_check)
+GEN_VV_MASKED_TRANS(vclmulh_vv, vclmul_vv_check)
 
 #define GEN_VX_MASKED_TRANS(NAME, CHECK)                                \
 static bool trans_##NAME(DisasContext *s, arg_rmrr *a)                  \
target/riscv/vcrypto_helper.c +13 -0
index 749a9cb30b..1891c29767 100644
@@ -38,7 +38,20 @@ static uint64_t clmul64(uint64_t y, uint64_t x)
     return result;
 }
 
+static uint64_t clmulh64(uint64_t y, uint64_t x)
+{
+    uint64_t result = 0;
+    for (int j = 63; j >= 1; j--) {
+        if ((y >> j) & 1) {
+            result ^= (x >> (64 - j));
+        }
+    }
+    return result;
+}
+
 RVVCALL(OPIVV2, vclmul_vv, OP_UUU_D, H8, H8, H8, clmul64)
 GEN_VEXT_VV(vclmul_vv, 8)
 RVVCALL(OPIVX2, vclmul_vx, OP_UUU_D, H8, H8, clmul64)
 GEN_VEXT_VX(vclmul_vx, 8)
+RVVCALL(OPIVV2, vclmulh_vv, OP_UUU_D, H8, H8, H8, clmulh64)
+GEN_VEXT_VV(vclmulh_vv, 8)
Signature · 1 line
2.39.2
#8 lore raw

PATCH 7/45 target/riscv: Add vclmulh.vx decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                      | 1 +
 target/riscv/insn32.decode                 | 1 +
 target/riscv/insn_trans/trans_rvzvkb.c.inc | 1 +
 target/riscv/vcrypto_helper.c              | 2 ++
 4 files changed, 5 insertions(+)
target/riscv/helper.h +1 -0
index 1c69c34a78..37f2e162f6 100644
@@ -1147,3 +1147,4 @@ DEF_HELPER_FLAGS_3(sm4ks, TCG_CALL_NO_RWG_SE, tl, tl, tl, tl)
 DEF_HELPER_6(vclmul_vv, void, ptr, ptr, ptr, ptr, env, i32)
 DEF_HELPER_6(vclmul_vx, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vclmulh_vv, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vclmulh_vx, void, ptr, ptr, tl, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index 3ad8e2055b..488e01ca59 100644
@@ -913,3 +913,4 @@ czero_nez   0000111  ..... ..... 111 ..... 0110011 @r
 vclmul_vv       001100 . ..... ..... 010 ..... 1010111 @r_vm
 vclmul_vx       001100 . ..... ..... 110 ..... 1010111 @r_vm
 vclmulh_vv      001101 . ..... ..... 010 ..... 1010111 @r_vm
+vclmulh_vx      001101 . ..... ..... 110 ..... 1010111 @r_vm
target/riscv/insn_trans/trans_rvzvkb.c.inc +1 -0
index 63a8778acc..810e469e13 100644
@@ -87,3 +87,4 @@ static bool vclmul_vx_check(DisasContext *s, arg_rmrr *a)
 }
 
 GEN_VX_MASKED_TRANS(vclmul_vx, vclmul_vx_check)
+GEN_VX_MASKED_TRANS(vclmulh_vx, vclmul_vx_check)
target/riscv/vcrypto_helper.c +2 -0
index 1891c29767..8b7c63d499 100644
@@ -55,3 +55,5 @@ RVVCALL(OPIVX2, vclmul_vx, OP_UUU_D, H8, H8, clmul64)
 GEN_VEXT_VX(vclmul_vx, 8)
 RVVCALL(OPIVV2, vclmulh_vv, OP_UUU_D, H8, H8, H8, clmulh64)
 GEN_VEXT_VV(vclmulh_vv, 8)
+RVVCALL(OPIVX2, vclmulh_vx, OP_UUU_D, H8, H8, clmulh64)
+GEN_VEXT_VX(vclmulh_vx, 8)
Signature · 1 line
2.39.2
#9 lore raw

PATCH 8/45 target/riscv: Refactor some of the generic vector functionality

In reply to Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

This refactoring ensures these functions/macros can be used by both vector and vector-crypto helpers (latter implemented in proceeding commit).

Also moves the checks out of do_opiv{v,x,i}_gvec{,_shift} functions and into the corresponding macros. This enables the functions to be reused in proceeding commit without check duplication.

---
 target/riscv/insn_trans/trans_rvv.c.inc | 28 +++++++--------
 target/riscv/vector_helper.c            | 45 -----------------------
 target/riscv/vector_internals.h         | 48 +++++++++++++++++++++++++
 3 files changed, 60 insertions(+), 61 deletions(-)
target/riscv/insn_trans/trans_rvv.c.inc +12 -16
index 4106bd6994..bb5e2c5407 100644
@@ -1187,9 +1187,6 @@ do_opivv_gvec(DisasContext *s, arg_rmrr *a, GVecGen3Fn *gvec_fn,
               gen_helper_gvec_4_ptr *fn)
 {
     TCGLabel *over = gen_new_label();
-    if (!opivv_check(s, a)) {
-        return false;
-    }
 
     tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_vl, 0, over);
     tcg_gen_brcond_tl(TCG_COND_GEU, cpu_vstart, cpu_vl, over);
@@ -1223,6 +1220,9 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         gen_helper_##NAME##_b, gen_helper_##NAME##_h,              \
         gen_helper_##NAME##_w, gen_helper_##NAME##_d,              \
     };                                                             \
+    if (!opivv_check(s, a)) {                                      \
+        return false;                                              \
+    }                                                              \
     return do_opivv_gvec(s, a, tcg_gen_gvec_##SUF, fns[s->sew]);   \
 }
 
@@ -1282,10 +1282,6 @@ static inline bool
 do_opivx_gvec(DisasContext *s, arg_rmrr *a, GVecGen2sFn *gvec_fn,
               gen_helper_opivx *fn)
 {
-    if (!opivx_check(s, a)) {
-        return false;
-    }
-
     if (a->vm && s->vl_eq_vlmax && !(s->vta && s->lmul < 0)) {
         TCGv_i64 src1 = tcg_temp_new_i64();
 
@@ -1307,6 +1303,9 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         gen_helper_##NAME##_b, gen_helper_##NAME##_h,              \
         gen_helper_##NAME##_w, gen_helper_##NAME##_d,              \
     };                                                             \
+    if (!opivx_check(s, a)) {                                      \
+        return false;                                              \
+    }                                                              \
     return do_opivx_gvec(s, a, tcg_gen_gvec_##SUF, fns[s->sew]);   \
 }
 
@@ -1439,10 +1438,6 @@ static inline bool
 do_opivi_gvec(DisasContext *s, arg_rmrr *a, GVecGen2iFn *gvec_fn,
               gen_helper_opivx *fn, imm_mode_t imm_mode)
 {
-    if (!opivx_check(s, a)) {
-        return false;
-    }
-
     if (a->vm && s->vl_eq_vlmax && !(s->vta && s->lmul < 0)) {
         gvec_fn(s->sew, vreg_ofs(s, a->rd), vreg_ofs(s, a->rs2),
                 extract_imm(s, a->rs1, imm_mode), MAXSZ(s), MAXSZ(s));
@@ -1460,6 +1455,9 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)             \
         gen_helper_##OPIVX##_b, gen_helper_##OPIVX##_h,            \
         gen_helper_##OPIVX##_w, gen_helper_##OPIVX##_d,            \
     };                                                             \
+    if (!opivx_check(s, a)) {                                      \
+        return false;                                              \
+    }                                                              \
     return do_opivi_gvec(s, a, tcg_gen_gvec_##SUF,                 \
                          fns[s->sew], IMM_MODE);                   \
 }
@@ -1785,10 +1783,6 @@ static inline bool
 do_opivx_gvec_shift(DisasContext *s, arg_rmrr *a, GVecGen2sFn32 *gvec_fn,
                     gen_helper_opivx *fn)
 {
-    if (!opivx_check(s, a)) {
-        return false;
-    }
-
     if (a->vm && s->vl_eq_vlmax && !(s->vta && s->lmul < 0)) {
         TCGv_i32 src1 = tcg_temp_new_i32();
 
@@ -1810,7 +1804,9 @@ static bool trans_##NAME(DisasContext *s, arg_rmrr *a)                    \
         gen_helper_##NAME##_b, gen_helper_##NAME##_h,                     \
         gen_helper_##NAME##_w, gen_helper_##NAME##_d,                     \
     };                                                                    \
-                                                                          \
+    if (!opivx_check(s, a)) {                                             \
+        return false;                                                     \
+    }                                                                     \
     return do_opivx_gvec_shift(s, a, tcg_gen_gvec_##SUF, fns[s->sew]);    \
 }
 
target/riscv/vector_helper.c +0 -45

74 lines not on this page · raw message

target/riscv/vector_internals.h +48 -0

69 lines not on this page · raw message

Signature · 1 line
2.39.2
#10 lore raw

PATCH 9/45 qemu/bitops.h: Limit rotate amounts

In reply to Lawrence Hunter

From: Dickon Hood <[email protected]>

Rotates have been fixed up to only allow for reasonable rotate amounts (ie, no rotates >7 on an 8b value etc.) This fixes a problem with riscv vector rotate instructions.

---
 include/qemu/bitops.h | 24 ++++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)
include/qemu/bitops.h +16 -8
index 03213ce952..c443995b3b 100644
@@ -218,7 +218,8 @@ static inline unsigned long find_first_zero_bit(const unsigned long *addr,
  */
 static inline uint8_t rol8(uint8_t word, unsigned int shift)
 {
-    return (word << shift) | (word >> ((8 - shift) & 7));
+    shift &= 7;
+    return (word << shift) | (word >> (8 - shift));
 }
 
 /**
@@ -228,7 +229,8 @@ static inline uint8_t rol8(uint8_t word, unsigned int shift)
  */
 static inline uint8_t ror8(uint8_t word, unsigned int shift)
 {
-    return (word >> shift) | (word << ((8 - shift) & 7));
+    shift &= 7;
+    return (word >> shift) | (word << (8 - shift));
 }
 
 /**
@@ -238,7 +240,8 @@ static inline uint8_t ror8(uint8_t word, unsigned int shift)
  */
 static inline uint16_t rol16(uint16_t word, unsigned int shift)
 {
-    return (word << shift) | (word >> ((16 - shift) & 15));
+    shift &= 15;
+    return (word << shift) | (word >> (16 - shift));
 }
 
 /**
@@ -248,7 +251,8 @@ static inline uint16_t rol16(uint16_t word, unsigned int shift)
  */
 static inline uint16_t ror16(uint16_t word, unsigned int shift)
 {
-    return (word >> shift) | (word << ((16 - shift) & 15));
+    shift &= 15;
+    return (word >> shift) | (word << (16 - shift));
 }
 
 /**
@@ -258,7 +262,8 @@ static inline uint16_t ror16(uint16_t word, unsigned int shift)
  */
 static inline uint32_t rol32(uint32_t word, unsigned int shift)
 {
-    return (word << shift) | (word >> ((32 - shift) & 31));
+    shift &= 31;
+    return (word << shift) | (word >> (32 - shift));
 }
 
 /**
@@ -268,7 +273,8 @@ static inline uint32_t rol32(uint32_t word, unsigned int shift)
  */
 static inline uint32_t ror32(uint32_t word, unsigned int shift)
 {
-    return (word >> shift) | (word << ((32 - shift) & 31));
+    shift &= 31;
+    return (word >> shift) | (word << (32 - shift));
 }
 
 /**
@@ -278,7 +284,8 @@ static inline uint32_t ror32(uint32_t word, unsigned int shift)
  */
 static inline uint64_t rol64(uint64_t word, unsigned int shift)
 {
-    return (word << shift) | (word >> ((64 - shift) & 63));
+    shift &= 63;
+    return (word << shift) | (word >> (64 - shift));
 }
 
 /**
@@ -288,7 +295,8 @@ static inline uint64_t rol64(uint64_t word, unsigned int shift)
  */
 static inline uint64_t ror64(uint64_t word, unsigned int shift)
 {
-    return (word >> shift) | (word << ((64 - shift) & 63));
+    shift &= 63;
+    return (word >> shift) | (word << (64 - shift));
 }
 
 /**
Signature · 1 line
2.39.2
#11 lore raw

PATCH 10/45 target/riscv: Add vrol.[vv,vx] and vror.[vv,vx,vi] decoding, translation and execution support

In reply to Lawrence Hunter

From: Dickon Hood <[email protected]>

Add an implementation of the vrol.* and vror.* instructions, with mappings between the RISC-V instructions and their internal TCG accelerated implmentations.

There are some missing ror helpers, so I've bodged it by converting them to rols.

---
 target/riscv/helper.h                      | 20 ++++++++
 target/riscv/insn32.decode                 |  7 +++
 target/riscv/insn_trans/trans_rvv.c.inc    |  3 ++
 target/riscv/insn_trans/trans_rvzvkb.c.inc | 57 ++++++++++++++++++++++
 target/riscv/vcrypto_helper.c              | 36 ++++++++++++++
 5 files changed, 123 insertions(+)
target/riscv/helper.h +20 -0
index 37f2e162f6..841cb43f04 100644
@@ -1148,3 +1148,23 @@ DEF_HELPER_6(vclmul_vv, void, ptr, ptr, ptr, ptr, env, i32)
 DEF_HELPER_6(vclmul_vx, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vclmulh_vv, void, ptr, ptr, ptr, ptr, env, i32)
 DEF_HELPER_6(vclmulh_vx, void, ptr, ptr, tl, ptr, env, i32)
+
+DEF_HELPER_6(vror_vv_b, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vror_vv_h, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vror_vv_w, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vror_vv_d, void, ptr, ptr, ptr, ptr, env, i32)
+
+DEF_HELPER_6(vror_vx_b, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vror_vx_h, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vror_vx_w, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vror_vx_d, void, ptr, ptr, tl, ptr, env, i32)
+
+DEF_HELPER_6(vrol_vv_b, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vrol_vv_h, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vrol_vv_w, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vrol_vv_d, void, ptr, ptr, ptr, ptr, env, i32)
+
+DEF_HELPER_6(vrol_vx_b, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vrol_vx_h, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vrol_vx_w, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vrol_vx_d, void, ptr, ptr, tl, ptr, env, i32)
target/riscv/insn32.decode +7 -0
index 488e01ca59..c557c063df 100644
@@ -37,6 +37,7 @@
 %imm_u    12:s20                 !function=ex_shift_12
 %imm_bs   30:2                   !function=ex_shift_3
 %imm_rnum 20:4
+%imm_z6   26:1 15:5
 
 # Argument sets:
 &empty
@@ -82,6 +83,7 @@
 @r_vm    ...... vm:1 ..... ..... ... ..... ....... &rmrr %rs2 %rs1 %rd
 @r_vm_1  ...... . ..... ..... ... ..... .......    &rmrr vm=1 %rs2 %rs1 %rd
 @r_vm_0  ...... . ..... ..... ... ..... .......    &rmrr vm=0 %rs2 %rs1 %rd
+@r2_zimm6  ..... . vm:1 ..... ..... ... ..... .......  &rmrr %rs2 rs1=%imm_z6 %rd
 @r2_zimm11 . zimm:11  ..... ... ..... ....... %rs1 %rd
 @r2_zimm10 .. zimm:10  ..... ... ..... ....... %rs1 %rd
 @r2_s    .......   ..... ..... ... ..... ....... %rs2 %rs1
@@ -914,3 +916,8 @@ vclmul_vv       001100 . ..... ..... 010 ..... 1010111 @r_vm
 vclmul_vx       001100 . ..... ..... 110 ..... 1010111 @r_vm
 vclmulh_vv      001101 . ..... ..... 010 ..... 1010111 @r_vm
 vclmulh_vx      001101 . ..... ..... 110 ..... 1010111 @r_vm
+vrol_vv         010101 . ..... ..... 000 ..... 1010111 @r_vm
+vrol_vx         010101 . ..... ..... 100 ..... 1010111 @r_vm
+vror_vv         010100 . ..... ..... 000 ..... 1010111 @r_vm
+vror_vx         010100 . ..... ..... 100 ..... 1010111 @r_vm
+vror_vi         01010. . ..... ..... 011 ..... 1010111 @r2_zimm6
target/riscv/insn_trans/trans_rvv.c.inc +3 -0
index bb5e2c5407..fa89a2f466 100644
@@ -1374,6 +1374,7 @@ GEN_OPIVX_GVEC_TRANS(vrsub_vx, rsubs)
 typedef enum {
     IMM_ZX,         /* Zero-extended */
     IMM_SX,         /* Sign-extended */
+    IMM_ZIMM6,      /* Truncate to 6 bits */
     IMM_TRUNC_SEW,  /* Truncate to log(SEW) bits */
     IMM_TRUNC_2SEW, /* Truncate to log(2*SEW) bits */
 } imm_mode_t;
@@ -1389,6 +1390,8 @@ static int64_t extract_imm(DisasContext *s, uint32_t imm, imm_mode_t imm_mode)
         return extract64(imm, 0, s->sew + 3);
     case IMM_TRUNC_2SEW:
         return extract64(imm, 0, s->sew + 4);
+    case IMM_ZIMM6:
+        return extract64(imm, 0, 6);
     default:
         g_assert_not_reached();
     }
target/riscv/insn_trans/trans_rvzvkb.c.inc +57 -0

60 lines not on this page · raw message

target/riscv/vcrypto_helper.c +36 -0
index 8b7c63d499..30ed9b1270 100644
@@ -57,3 +57,39 @@ RVVCALL(OPIVV2, vclmulh_vv, OP_UUU_D, H8, H8, H8, clmulh64)
 GEN_VEXT_VV(vclmulh_vv, 8)
 RVVCALL(OPIVX2, vclmulh_vx, OP_UUU_D, H8, H8, clmulh64)
 GEN_VEXT_VX(vclmulh_vx, 8)
+
+RVVCALL(OPIVV2, vror_vv_b, OP_UUU_B, H1, H1, H1, ror8)
+RVVCALL(OPIVV2, vror_vv_h, OP_UUU_H, H2, H2, H2, ror16)
+RVVCALL(OPIVV2, vror_vv_w, OP_UUU_W, H4, H4, H4, ror32)
+RVVCALL(OPIVV2, vror_vv_d, OP_UUU_D, H8, H8, H8, ror64)
+GEN_VEXT_VV(vror_vv_b, 1)
+GEN_VEXT_VV(vror_vv_h, 2)
+GEN_VEXT_VV(vror_vv_w, 4)
+GEN_VEXT_VV(vror_vv_d, 8)
+
+RVVCALL(OPIVX2, vror_vx_b, OP_UUU_B, H1, H1, ror8)
+RVVCALL(OPIVX2, vror_vx_h, OP_UUU_H, H2, H2, ror16)
+RVVCALL(OPIVX2, vror_vx_w, OP_UUU_W, H4, H4, ror32)
+RVVCALL(OPIVX2, vror_vx_d, OP_UUU_D, H8, H8, ror64)
+GEN_VEXT_VX(vror_vx_b, 1)
+GEN_VEXT_VX(vror_vx_h, 2)
+GEN_VEXT_VX(vror_vx_w, 4)
+GEN_VEXT_VX(vror_vx_d, 8)
+
+RVVCALL(OPIVV2, vrol_vv_b, OP_UUU_B, H1, H1, H1, rol8)
+RVVCALL(OPIVV2, vrol_vv_h, OP_UUU_H, H2, H2, H2, rol16)
+RVVCALL(OPIVV2, vrol_vv_w, OP_UUU_W, H4, H4, H4, rol32)
+RVVCALL(OPIVV2, vrol_vv_d, OP_UUU_D, H8, H8, H8, rol64)
+GEN_VEXT_VV(vrol_vv_b, 1)
+GEN_VEXT_VV(vrol_vv_h, 2)
+GEN_VEXT_VV(vrol_vv_w, 4)
+GEN_VEXT_VV(vrol_vv_d, 8)
+
+RVVCALL(OPIVX2, vrol_vx_b, OP_UUU_B, H1, H1, rol8)
+RVVCALL(OPIVX2, vrol_vx_h, OP_UUU_H, H2, H2, rol16)
+RVVCALL(OPIVX2, vrol_vx_w, OP_UUU_W, H4, H4, rol32)
+RVVCALL(OPIVX2, vrol_vx_d, OP_UUU_D, H8, H8, rol64)
+GEN_VEXT_VX(vrol_vx_b, 1)
+GEN_VEXT_VX(vrol_vx_h, 2)
+GEN_VEXT_VX(vrol_vx_w, 4)
+GEN_VEXT_VX(vrol_vx_d, 8)
Signature · 1 line
2.39.2
#12 lore raw

PATCH 11/45 target/riscv: Refactor some of the generic vector functionality

In reply to Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

This refactoring ensures these functions/macros can be used by both vector and vector-crypto helpers (latter implemented in proceeding commit).

---
 target/riscv/vector_helper.c    | 39 --------------------------------
 target/riscv/vector_internals.h | 40 +++++++++++++++++++++++++++++++++
 2 files changed, 40 insertions(+), 39 deletions(-)
target/riscv/vector_helper.c +0 -39
index 27fefef10e..7c8775fd7b 100644
@@ -3412,11 +3412,6 @@ GEN_VEXT_VF(vfwnmsac_vf_h, 4)
 GEN_VEXT_VF(vfwnmsac_vf_w, 8)
 
 /* Vector Floating-Point Square-Root Instruction */
-/* (TD, T2, TX2) */
-#define OP_UU_H uint16_t, uint16_t, uint16_t
-#define OP_UU_W uint32_t, uint32_t, uint32_t
-#define OP_UU_D uint64_t, uint64_t, uint64_t
-
 #define OPFVV1(NAME, TD, T2, TX2, HD, HS2, OP)        \
 static void do_##NAME(void *vd, void *vs2, int i,      \
         CPURISCVState *env)                            \
@@ -4109,40 +4104,6 @@ GEN_VEXT_CMP_VF(vmfge_vf_w, uint32_t, H4, vmfge32)
 GEN_VEXT_CMP_VF(vmfge_vf_d, uint64_t, H8, vmfge64)
 
 /* Vector Floating-Point Classify Instruction */
-#define OPIVV1(NAME, TD, T2, TX2, HD, HS2, OP)         \
-static void do_##NAME(void *vd, void *vs2, int i)      \
-{                                                      \
-    TX2 s2 = *((T2 *)vs2 + HS2(i));                    \
-    *((TD *)vd + HD(i)) = OP(s2);                      \
-}
-
-#define GEN_VEXT_V(NAME, ESZ)                          \
-void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
-                  CPURISCVState *env, uint32_t desc)   \
-{                                                      \
-    uint32_t vm = vext_vm(desc);                       \
-    uint32_t vl = env->vl;                             \
-    uint32_t total_elems =                             \
-        vext_get_total_elems(env, desc, ESZ);          \
-    uint32_t vta = vext_vta(desc);                     \
-    uint32_t vma = vext_vma(desc);                     \
-    uint32_t i;                                        \
-                                                       \
-    for (i = env->vstart; i < vl; i++) {               \
-        if (!vm && !vext_elem_mask(v0, i)) {           \
-            /* set masked-off elements to 1s */        \
-            vext_set_elems_1s(vd, vma, i * ESZ,        \
-                              (i + 1) * ESZ);          \
-            continue;                                  \
-        }                                              \
-        do_##NAME(vd, vs2, i);                         \
-    }                                                  \
-    env->vstart = 0;                                   \
-    /* set tail elements to 1s */                      \
-    vext_set_elems_1s(vd, vta, vl * ESZ,               \
-                      total_elems * ESZ);              \
-}
-
 target_ulong fclass_h(uint64_t frs1)
 {
     float16 f = frs1;
target/riscv/vector_internals.h +40 -0
index 749d138beb..15cbc5ddac 100644
@@ -121,12 +121,52 @@ void vext_set_elems_1s(void *base, uint32_t is_agnostic, uint32_t cnt,
 /* expand macro args before macro */
 #define RVVCALL(macro, ...)  macro(__VA_ARGS__)
 
+/* (TD, T2, TX2) */
+#define OP_UU_B uint8_t, uint8_t, uint8_t
+#define OP_UU_H uint16_t, uint16_t, uint16_t
+#define OP_UU_W uint32_t, uint32_t, uint32_t
+#define OP_UU_D uint64_t, uint64_t, uint64_t
+
 /* (TD, T1, T2, TX1, TX2) */
 #define OP_UUU_B uint8_t, uint8_t, uint8_t, uint8_t, uint8_t
 #define OP_UUU_H uint16_t, uint16_t, uint16_t, uint16_t, uint16_t
 #define OP_UUU_W uint32_t, uint32_t, uint32_t, uint32_t, uint32_t
 #define OP_UUU_D uint64_t, uint64_t, uint64_t, uint64_t, uint64_t
 
+#define OPIVV1(NAME, TD, T2, TX2, HD, HS2, OP)         \
+static void do_##NAME(void *vd, void *vs2, int i)      \
+{                                                      \
+    TX2 s2 = *((T2 *)vs2 + HS2(i));                    \
+    *((TD *)vd + HD(i)) = OP(s2);                      \
+}
+
+#define GEN_VEXT_V(NAME, ESZ)                          \
+void HELPER(NAME)(void *vd, void *v0, void *vs2,       \
+                  CPURISCVState *env, uint32_t desc)   \
+{                                                      \
+    uint32_t vm = vext_vm(desc);                       \
+    uint32_t vl = env->vl;                             \
+    uint32_t total_elems =                             \
+        vext_get_total_elems(env, desc, ESZ);          \
+    uint32_t vta = vext_vta(desc);                     \
+    uint32_t vma = vext_vma(desc);                     \
+    uint32_t i;                                        \
+                                                       \
+    for (i = env->vstart; i < vl; i++) {               \
+        if (!vm && !vext_elem_mask(v0, i)) {           \
+            /* set masked-off elements to 1s */        \
+            vext_set_elems_1s(vd, vma, i * ESZ,        \
+                              (i + 1) * ESZ);          \
+            continue;                                  \
+        }                                              \
+        do_##NAME(vd, vs2, i);                         \
+    }                                                  \
+    env->vstart = 0;                                   \
+    /* set tail elements to 1s */                      \
+    vext_set_elems_1s(vd, vta, vl * ESZ,               \
+                      total_elems * ESZ);              \
+}
+
 /* operation of two vector elements */
 typedef void opivv2_fn(void *vd, void *vs1, void *vs2, int i);
 
Signature · 1 line
2.39.2
#13 lore raw

PATCH 12/45 target/riscv: Add vbrev8.v decoding, translation and execution support

In reply to Lawrence Hunter

From: William Salmon <[email protected]>

---
 target/riscv/helper.h                      |  5 +++
 target/riscv/insn32.decode                 |  1 +
 target/riscv/insn_trans/trans_rvzvkb.c.inc | 38 ++++++++++++++++++++++
 target/riscv/vcrypto_helper.c              | 21 ++++++++++++
 4 files changed, 65 insertions(+)
target/riscv/helper.h +5 -0
index 841cb43f04..625f9872d0 100644
@@ -1168,3 +1168,8 @@ DEF_HELPER_6(vrol_vx_b, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vrol_vx_h, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vrol_vx_w, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vrol_vx_d, void, ptr, ptr, tl, ptr, env, i32)
+
+DEF_HELPER_5(vbrev8_v_b, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vbrev8_v_h, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vbrev8_v_w, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vbrev8_v_d, void, ptr, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index c557c063df..87473a77c0 100644
@@ -921,3 +921,4 @@ vrol_vx         010101 . ..... ..... 100 ..... 1010111 @r_vm
 vror_vv         010100 . ..... ..... 000 ..... 1010111 @r_vm
 vror_vx         010100 . ..... ..... 100 ..... 1010111 @r_vm
 vror_vi         01010. . ..... ..... 011 ..... 1010111 @r2_zimm6
+vbrev8_v        010010 . ..... 01000 010 ..... 1010111 @r2_vm
target/riscv/insn_trans/trans_rvzvkb.c.inc +38 -0
index f71383e482..7cd114ae71 100644
@@ -145,3 +145,41 @@ GEN_OPIVX_GVEC_SHIFT_TRANS_CHECK(vrol_vx, rotls, zvkb_vx_check)
 GEN_OPIVV_GVEC_TRANS_CHECK(vror_vv, rotrv, zvkb_vv_check)
 GEN_OPIVX_GVEC_SHIFT_TRANS_CHECK(vror_vx, rotrs, zvkb_vx_check)
 GEN_OPIVI_GVEC_TRANS_CHECK(vror_vi, IMM_ZIMM6, vror_vx, rotri, zvkb_vx_check)
+
+#define GEN_OPIV_TRANS(NAME, CHECK)                                    \
+static bool trans_##NAME(DisasContext *s, arg_rmr * a)                 \
+{                                                                      \
+    if (CHECK(s, a)) {                                                 \
+        uint32_t data = 0;                                             \
+        static gen_helper_gvec_3_ptr * const fns[4] = {                \
+            gen_helper_##NAME##_b, gen_helper_##NAME##_h,              \
+            gen_helper_##NAME##_w, gen_helper_##NAME##_d,              \
+        };                                                             \
+        TCGLabel *over = gen_new_label();                              \
+        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, 0),         \
+                           vreg_ofs(s, a->rs2), cpu_env,               \
+                           s->cfg_ptr->vlen / 8, s->cfg_ptr->vlen / 8, \
+                           data, fns[s->sew]);                         \
+        mark_vs_dirty(s);                                              \
+        gen_set_label(over);                                           \
+        return true;                                                   \
+    }                                                                  \
+    return false;                                                      \
+}
+
+static bool vxrev8_check(DisasContext *s, arg_rmr *a)
+{
+    return s->cfg_ptr->ext_zvkb == true &&
+           require_rvv(s) &&
+           vext_check_isa_ill(s) &&
+           vext_check_ss(s, a->rd, a->rs2, a->vm);
+}
+
+GEN_OPIV_TRANS(vbrev8_v, vxrev8_check)
target/riscv/vcrypto_helper.c +21 -0
index 30ed9b1270..5d2a995de6 100644
@@ -93,3 +93,24 @@ GEN_VEXT_VX(vrol_vx_b, 1)
 GEN_VEXT_VX(vrol_vx_h, 2)
 GEN_VEXT_VX(vrol_vx_w, 4)
 GEN_VEXT_VX(vrol_vx_d, 8)
+
+static uint64_t brev8(uint64_t val)
+{
+    val = ((val & 0x5555555555555555ull) << 1)
+        | ((val & 0xAAAAAAAAAAAAAAAAull) >> 1);
+    val = ((val & 0x3333333333333333ull) << 2)
+        | ((val & 0xCCCCCCCCCCCCCCCCull) >> 2);
+    val = ((val & 0x0F0F0F0F0F0F0F0Full) << 4)
+        | ((val & 0xF0F0F0F0F0F0F0F0ull) >> 4);
+
+    return val;
+}
+
+RVVCALL(OPIVV1, vbrev8_v_b, OP_UU_B, H1, H1, brev8)
+RVVCALL(OPIVV1, vbrev8_v_h, OP_UU_H, H2, H2, brev8)
+RVVCALL(OPIVV1, vbrev8_v_w, OP_UU_W, H4, H4, brev8)
+RVVCALL(OPIVV1, vbrev8_v_d, OP_UU_D, H8, H8, brev8)
+GEN_VEXT_V(vbrev8_v_b, 1)
+GEN_VEXT_V(vbrev8_v_h, 2)
+GEN_VEXT_V(vbrev8_v_w, 4)
+GEN_VEXT_V(vbrev8_v_d, 8)
Signature · 1 line
2.39.2
#14 lore raw

PATCH 13/45 target/riscv: Add vrev8.v decoding, translation and execution support

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 target/riscv/helper.h                      |  4 ++++
 target/riscv/insn32.decode                 |  1 +
 target/riscv/insn_trans/trans_rvzvkb.c.inc |  1 +
 target/riscv/vcrypto_helper.c              | 11 +++++++++++
 4 files changed, 17 insertions(+)
target/riscv/helper.h +4 -0
index 625f9872d0..b4baa22692 100644
@@ -1169,6 +1169,10 @@ DEF_HELPER_6(vrol_vx_h, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vrol_vx_w, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vrol_vx_d, void, ptr, ptr, tl, ptr, env, i32)
 
+DEF_HELPER_5(vrev8_v_b, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vrev8_v_h, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vrev8_v_w, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vrev8_v_d, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vbrev8_v_b, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vbrev8_v_h, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vbrev8_v_w, void, ptr, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index 87473a77c0..bdefcd3fa2 100644
@@ -922,3 +922,4 @@ vror_vv         010100 . ..... ..... 000 ..... 1010111 @r_vm
 vror_vx         010100 . ..... ..... 100 ..... 1010111 @r_vm
 vror_vi         01010. . ..... ..... 011 ..... 1010111 @r2_zimm6
 vbrev8_v        010010 . ..... 01000 010 ..... 1010111 @r2_vm
+vrev8_v         010010 . ..... 01001 010 ..... 1010111 @r2_vm
target/riscv/insn_trans/trans_rvzvkb.c.inc +1 -0
index 7cd114ae71..77ba8bc713 100644
@@ -183,3 +183,4 @@ static bool vxrev8_check(DisasContext *s, arg_rmr *a)
 }
 
 GEN_OPIV_TRANS(vbrev8_v, vxrev8_check)
+GEN_OPIV_TRANS(vrev8_v, vxrev8_check)
target/riscv/vcrypto_helper.c +11 -0
index 5d2a995de6..ecf21c50f8 100644
@@ -20,6 +20,7 @@
 #include "qemu/osdep.h"
 #include "qemu/host-utils.h"
 #include "qemu/bitops.h"
+#include "qemu/bswap.h"
 #include "cpu.h"
 #include "exec/memop.h"
 #include "exec/exec-all.h"
@@ -114,3 +115,13 @@ GEN_VEXT_V(vbrev8_v_b, 1)
 GEN_VEXT_V(vbrev8_v_h, 2)
 GEN_VEXT_V(vbrev8_v_w, 4)
 GEN_VEXT_V(vbrev8_v_d, 8)
+
+#define DO_IDENTITY(a) (a)
+RVVCALL(OPIVV1, vrev8_v_b, OP_UU_B, H1, H1, DO_IDENTITY)
+RVVCALL(OPIVV1, vrev8_v_h, OP_UU_H, H2, H2, bswap16)
+RVVCALL(OPIVV1, vrev8_v_w, OP_UU_W, H4, H4, bswap32)
+RVVCALL(OPIVV1, vrev8_v_d, OP_UU_D, H8, H8, bswap64)
+GEN_VEXT_V(vrev8_v_b, 1)
+GEN_VEXT_V(vrev8_v_h, 2)
+GEN_VEXT_V(vrev8_v_w, 4)
+GEN_VEXT_V(vrev8_v_d, 8)
Signature · 1 line
2.39.2
#15 lore raw

PATCH 14/45 target/riscv: Add vandn.[vv,vx] decoding, translation and execution support

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 accel/tcg/tcg-runtime-gvec.c               | 11 +++++++
 accel/tcg/tcg-runtime.h                    |  1 +
 target/riscv/helper.h                      |  9 ++++++
 target/riscv/insn32.decode                 |  2 ++
 target/riscv/insn_trans/trans_rvzvkb.c.inc | 34 ++++++++++++++++++++++
 target/riscv/vcrypto_helper.c              | 19 ++++++++++++
 6 files changed, 76 insertions(+)
accel/tcg/tcg-runtime-gvec.c +11 -0
index ac7d28c251..322dcc0687 100644
@@ -550,6 +550,17 @@ void HELPER(gvec_ands)(void *d, void *a, uint64_t b, uint32_t desc)
     clear_high(d, oprsz, desc);
 }
 
+void HELPER(gvec_andsc)(void *d, void *a, uint64_t b, uint32_t desc)
+{
+    intptr_t oprsz = simd_oprsz(desc);
+    intptr_t i;
+
+    for (i = 0; i < oprsz; i += sizeof(uint64_t)) {
+        *(uint64_t *)(d + i) = *(uint64_t *)(a + i) & ~b;
+    }
+    clear_high(d, oprsz, desc);
+}
+
 void HELPER(gvec_xors)(void *d, void *a, uint64_t b, uint32_t desc)
 {
     intptr_t oprsz = simd_oprsz(desc);
accel/tcg/tcg-runtime.h +1 -0
index e141a6ab24..d086200483 100644
@@ -217,6 +217,7 @@ DEF_HELPER_FLAGS_4(gvec_nor, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32)
 DEF_HELPER_FLAGS_4(gvec_eqv, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32)
 
 DEF_HELPER_FLAGS_4(gvec_ands, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32)
+DEF_HELPER_FLAGS_4(gvec_andsc, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32)
 DEF_HELPER_FLAGS_4(gvec_xors, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32)
 DEF_HELPER_FLAGS_4(gvec_ors, TCG_CALL_NO_RWG, void, ptr, ptr, i64, i32)
 
target/riscv/helper.h +9 -0
index b4baa22692..1bc10ca5be 100644
@@ -1177,3 +1177,12 @@ DEF_HELPER_5(vbrev8_v_b, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vbrev8_v_h, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vbrev8_v_w, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vbrev8_v_d, void, ptr, ptr, ptr, env, i32)
+
+DEF_HELPER_6(vandn_vv_b, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vandn_vv_h, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vandn_vv_w, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vandn_vv_d, void, ptr, ptr, ptr, ptr, env, i32)
+DEF_HELPER_6(vandn_vx_b, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vandn_vx_h, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vandn_vx_w, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vandn_vx_d, void, ptr, ptr, tl, ptr, env, i32)
target/riscv/insn32.decode +2 -0
index bdefcd3fa2..bbf128e4dc 100644
@@ -923,3 +923,5 @@ vror_vx         010100 . ..... ..... 100 ..... 1010111 @r_vm
 vror_vi         01010. . ..... ..... 011 ..... 1010111 @r2_zimm6
 vbrev8_v        010010 . ..... 01000 010 ..... 1010111 @r2_vm
 vrev8_v         010010 . ..... 01001 010 ..... 1010111 @r2_vm
+vandn_vv        000001 . ..... ..... 000 ..... 1010111 @r_vm
+vandn_vx        000001 . ..... ..... 100 ..... 1010111 @r_vm
target/riscv/insn_trans/trans_rvzvkb.c.inc +34 -0
index 77ba8bc713..a9a820ec0a 100644
@@ -146,6 +146,40 @@ GEN_OPIVV_GVEC_TRANS_CHECK(vror_vv, rotrv, zvkb_vv_check)
 GEN_OPIVX_GVEC_SHIFT_TRANS_CHECK(vror_vx, rotrs, zvkb_vx_check)
 GEN_OPIVI_GVEC_TRANS_CHECK(vror_vi, IMM_ZIMM6, vror_vx, rotri, zvkb_vx_check)
 
+
+static void tcg_gen_gvec_andsc(unsigned vece, uint32_t dofs, uint32_t aofs,
+                               TCGv_i64 c, uint32_t oprsz, uint32_t maxsz)
+{
+    static GVecGen2s g = {
+        .fni8 = tcg_gen_andc_i64,
+        .fniv = tcg_gen_andc_vec,
+        .fno = gen_helper_gvec_andsc,
+        .prefer_i64 = TCG_TARGET_REG_BITS == 64,
+    };
+
+    g.vece = vece;
+
+    tcg_gen_dup_i64(vece, c, c);
+    tcg_gen_gvec_2s(dofs, aofs, oprsz, maxsz, c, &g);
+}
+
+#define GEN_OPIVX_GVEC_TRANS_CHECK(NAME, SUF, CHECK)                 \
+static bool trans_##NAME(DisasContext *s, arg_rmrr *a)               \
+{                                                                    \
+    if (CHECK(s, a)) {                                               \
+        static gen_helper_opivx * const fns[4] = {                   \
+            gen_helper_##NAME##_b, gen_helper_##NAME##_h,            \
+            gen_helper_##NAME##_w, gen_helper_##NAME##_d,            \
+        };                                                           \
+        return do_opivx_gvec(s, a, tcg_gen_gvec_##SUF, fns[s->sew]); \
+    }                                                                \
+    return false;                                                    \
+}
+
+/* vandn.v[vx] */
+GEN_OPIVV_GVEC_TRANS_CHECK(vandn_vv, andc, zvkb_vv_check)
+GEN_OPIVX_GVEC_TRANS_CHECK(vandn_vx, andsc, zvkb_vx_check)
+
 #define GEN_OPIV_TRANS(NAME, CHECK)                                    \
 static bool trans_##NAME(DisasContext *s, arg_rmr * a)                 \
 {                                                                      \
target/riscv/vcrypto_helper.c +19 -0
index ecf21c50f8..7d2a355418 100644
@@ -125,3 +125,22 @@ GEN_VEXT_V(vrev8_v_b, 1)
 GEN_VEXT_V(vrev8_v_h, 2)
 GEN_VEXT_V(vrev8_v_w, 4)
 GEN_VEXT_V(vrev8_v_d, 8)
+
+#define DO_ANDN(a, b) ((a) & ~(b))
+RVVCALL(OPIVV2, vandn_vv_b, OP_UUU_B, H1, H1, H1, DO_ANDN)
+RVVCALL(OPIVV2, vandn_vv_h, OP_UUU_H, H2, H2, H2, DO_ANDN)
+RVVCALL(OPIVV2, vandn_vv_w, OP_UUU_W, H4, H4, H4, DO_ANDN)
+RVVCALL(OPIVV2, vandn_vv_d, OP_UUU_D, H8, H8, H8, DO_ANDN)
+GEN_VEXT_VV(vandn_vv_b, 1)
+GEN_VEXT_VV(vandn_vv_h, 2)
+GEN_VEXT_VV(vandn_vv_w, 4)
+GEN_VEXT_VV(vandn_vv_d, 8)
+
+RVVCALL(OPIVX2, vandn_vx_b, OP_UUU_B, H1, H1, DO_ANDN)
+RVVCALL(OPIVX2, vandn_vx_h, OP_UUU_H, H2, H2, DO_ANDN)
+RVVCALL(OPIVX2, vandn_vx_w, OP_UUU_W, H4, H4, DO_ANDN)
+RVVCALL(OPIVX2, vandn_vx_d, OP_UUU_D, H8, H8, DO_ANDN)
+GEN_VEXT_VX(vandn_vx_b, 1)
+GEN_VEXT_VX(vandn_vx_h, 2)
+GEN_VEXT_VX(vandn_vx_w, 4)
+GEN_VEXT_VX(vandn_vx_d, 8)
Signature · 1 line
2.39.2
#16 lore raw

PATCH 15/45 target/riscv: Expose zvkb cpu property

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 target/riscv/cpu.c | 3 +++
 1 file changed, 3 insertions(+)
target/riscv/cpu.c +3 -0
index 69611408f9..462615140c 100644
@@ -1469,6 +1469,9 @@ static Property riscv_cpu_extensions[] = {
     DEFINE_PROP_BOOL("x-zvfh", RISCVCPU, cfg.ext_zvfh, false),
     DEFINE_PROP_BOOL("x-zvfhmin", RISCVCPU, cfg.ext_zvfhmin, false),
 
+    /* Vector cryptography extensions */
+    DEFINE_PROP_BOOL("x-zvkb", RISCVCPU, cfg.ext_zvkb, false),
+
     DEFINE_PROP_END_OF_LIST(),
 };
 
Signature · 1 line
2.39.2
#17 lore raw

PATCH 16/45 target/riscv: Add zvkned cpu property

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 target/riscv/cpu.c | 3 ++-
 target/riscv/cpu.h | 1 +
 2 files changed, 3 insertions(+), 1 deletion(-)
target/riscv/cpu.c +2 -1
index 462615140c..00e1d007a4 100644
@@ -110,6 +110,7 @@ static const struct isa_ext_data isa_edata_arr[] = {
     ISA_EXT_DATA_ENTRY(zvfh, true, PRIV_VERSION_1_12_0, ext_zvfh),
     ISA_EXT_DATA_ENTRY(zvfhmin, true, PRIV_VERSION_1_12_0, ext_zvfhmin),
     ISA_EXT_DATA_ENTRY(zvkb, true, PRIV_VERSION_1_12_0, ext_zvkb),
+    ISA_EXT_DATA_ENTRY(zvkned, true, PRIV_VERSION_1_12_0, ext_zvkned),
     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),
@@ -1216,7 +1217,7 @@ 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_zvkb &&
+    if ((cpu->cfg.ext_zvkb || cpu->cfg.ext_zvkned) &&
         !(cpu->cfg.ext_zve32f || cpu->cfg.ext_zve64f ||
             cpu->cfg.ext_zve64d || cpu->cfg.ext_v)) {
         error_setg(
target/riscv/cpu.h +1 -0
index 7d6699f718..4f3b97e0f1 100644
@@ -471,6 +471,7 @@ struct RISCVCPUConfig {
     bool ext_zve64f;
     bool ext_zve64d;
     bool ext_zvkb;
+    bool ext_zvkned;
     bool ext_zmmul;
     bool ext_zvfh;
     bool ext_zvfhmin;
Signature · 1 line
2.39.2
#18 lore raw

PATCH 17/45 target/riscv: Add vaesef.vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                        |  3 +
 target/riscv/insn32.decode                   |  4 ++
 target/riscv/insn_trans/trans_rvzvkned.c.inc | 72 +++++++++++++++++++
 target/riscv/op_helper.c                     |  5 ++
 target/riscv/translate.c                     |  1 +
 target/riscv/vcrypto_helper.c                | 73 ++++++++++++++++++++
 6 files changed, 158 insertions(+)
 create mode 100644 target/riscv/insn_trans/trans_rvzvkned.c.inc
target/riscv/helper.h +3 -0
index 1bc10ca5be..a402bb5d40 100644
@@ -1,5 +1,6 @@
 /* Exceptions */
 DEF_HELPER_2(raise_exception, noreturn, env, i32)
+DEF_HELPER_2(restore_cpu_and_raise_exception, noreturn, env, i32)
 
 /* Floating Point - rounding mode */
 DEF_HELPER_FLAGS_2(set_rounding_mode, TCG_CALL_NO_WG, void, env, i32)
@@ -1186,3 +1187,5 @@ DEF_HELPER_6(vandn_vx_b, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vandn_vx_h, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vandn_vx_w, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vandn_vx_d, void, ptr, ptr, tl, ptr, env, i32)
+
+DEF_HELPER_4(vaesef_vv, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +4 -0
index bbf128e4dc..f68025755a 100644
@@ -75,6 +75,7 @@
 @r_rm    .......   ..... ..... ... ..... ....... %rs2 %rs1 %rm %rd
 @r2_rm   .......   ..... ..... ... ..... ....... %rs1 %rm %rd
 @r2      .......   ..... ..... ... ..... ....... &r2 %rs1 %rd
+@r2_vm_1 ...... . ..... ..... ... ..... ....... &rmr vm=1 %rs2 %rd
 @r2_nfvm ... ... vm:1 ..... ..... ... ..... ....... &r2nfvm %nf %rs1 %rd
 @r2_vm   ...... vm:1 ..... ..... ... ..... ....... &rmr %rs2 %rd
 @r1_vm   ...... vm:1 ..... ..... ... ..... ....... %rd
@@ -925,3 +926,6 @@ vbrev8_v        010010 . ..... 01000 010 ..... 1010111 @r2_vm
 vrev8_v         010010 . ..... 01001 010 ..... 1010111 @r2_vm
 vandn_vv        000001 . ..... ..... 000 ..... 1010111 @r_vm
 vandn_vx        000001 . ..... ..... 100 ..... 1010111 @r_vm
+
+# *** RV64 Zvkned vector crypto extension ***
+vaesef_vv       101000 1 ..... 00011 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc new file +72 -0

72 lines not on this page · raw message

target/riscv/op_helper.c +5 -0
index 84ee018f7d..f26c6396a5 100644
@@ -38,6 +38,11 @@ void helper_raise_exception(CPURISCVState *env, uint32_t exception)
     riscv_raise_exception(env, exception, 0);
 }
 
+void helper_restore_cpu_and_raise_exception(CPURISCVState *env, uint32_t exception)
+{
+    riscv_raise_exception(env, exception, GETPC());
+}
+
 target_ulong helper_csrr(CPURISCVState *env, int csr)
 {
     /*
target/riscv/translate.c +1 -0
index 939f5aeb1c..e3561b0bbd 100644
@@ -1084,6 +1084,7 @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc)
 #include "insn_trans/trans_rvzfh.c.inc"
 #include "insn_trans/trans_rvk.c.inc"
 #include "insn_trans/trans_rvzvkb.c.inc"
+#include "insn_trans/trans_rvzvkned.c.inc"
 #include "insn_trans/trans_privileged.c.inc"
 #include "insn_trans/trans_svinval.c.inc"
 #include "decode-xthead.c.inc"
target/riscv/vcrypto_helper.c +73 -0

82 lines not on this page · raw message

Signature · 1 line
2.39.2
#19 lore raw

PATCH 18/45 target/riscv: Add vaesef.vs decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                        |  1 +
 target/riscv/insn32.decode                   |  1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc | 20 +++++++++++
 target/riscv/vcrypto_helper.c                | 36 ++++++++++++++++++++
 4 files changed, 58 insertions(+)
target/riscv/helper.h +1 -0
index a402bb5d40..fb30b4d13e 100644
@@ -1189,3 +1189,4 @@ DEF_HELPER_6(vandn_vx_w, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_6(vandn_vx_d, void, ptr, ptr, tl, ptr, env, i32)
 
 DEF_HELPER_4(vaesef_vv, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vaesef_vs, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index f68025755a..5d1bb6ccc6 100644
@@ -929,3 +929,4 @@ vandn_vx        000001 . ..... ..... 100 ..... 1010111 @r_vm
 
 # *** RV64 Zvkned vector crypto extension ***
 vaesef_vv       101000 1 ..... 00011 010 ..... 1110111 @r2_vm_1
+vaesef_vs       101001 1 ..... 00011 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc +20 -0
index 6f3d62bef1..69bf7f9fee 100644
@@ -69,4 +69,24 @@ static bool vaes_check_vv(DisasContext *s, arg_rmr *a)
            require_align(a->rs2, s->lmul) &&
            s->vstart % 4 == 0 && s->sew == MO_32;
 }
+
+static bool vaes_check_overlap(DisasContext *s, int vd, int vs2)
+{
+    int8_t op_size = s->lmul <= 0 ? 1 : 1 << s->lmul;
+    return !is_overlapped(vd, op_size, vs2, 1);
+}
+
+static bool vaes_check_vs(DisasContext *s, arg_rmr *a)
+{
+    return vaes_check_overlap(s, a->rd, a->rs2) &&
+           MAXSZ(s) >= (128 / 8) && /* EGW in bytes */
+           s->cfg_ptr->ext_zvkned == true &&
+           require_rvv(s) &&
+           vext_check_isa_ill(s) &&
+           require_align(a->rd, s->lmul) &&
+           s->vstart % 4 == 0 &&
+           s->sew == MO_32;
+}
+
 GEN_V_UNMASKED_TRANS(vaesef_vv, vaes_check_vv)
+GEN_V_UNMASKED_TRANS(vaesef_vs, vaes_check_vs)
target/riscv/vcrypto_helper.c +36 -0
index bb5ce5b50a..b079b543c7 100644
@@ -214,6 +214,42 @@ void HELPER(NAME)(void *vd_vptr, void *vs2_vptr, CPURISCVState *env,      \
     vext_set_elems_1s(vd, vta, vl * 4, total_elems * 4);                  \
 }
 
+#define GEN_ZVKNED_HELPER_VS(NAME, ...)                                   \
+void HELPER(NAME)(void *vd_vptr, void *vs2_vptr, CPURISCVState *env,      \
+                  uint32_t desc)                                          \
+{                                                                         \
+    uint64_t *vd = vd_vptr;                                               \
+    uint64_t *vs2 = vs2_vptr;                                             \
+    uint32_t vl = env->vl;                                                \
+    uint32_t total_elems = vext_get_total_elems(env, desc, 4);            \
+    uint32_t vta = vext_vta(desc);                                        \
+                                                                          \
+    for (uint32_t i = env->vstart / 4; i < env->vl / 4; i++) {            \
+        uint64_t round_key[2] = {                                         \
+            cpu_to_le64(vs2[0]),                                          \
+            cpu_to_le64(vs2[1]),                                          \
+        };                                                                \
+        uint8_t round_state[4][4];                                        \
+        cpu_to_le64s(vd + i * 2 + 0);                                     \
+        cpu_to_le64s(vd + i * 2 + 1);                                     \
+        for (int j = 0; j < 16; j++) {                                    \
+            round_state[j / 4][j % 4] = ((uint8_t *)(vd + i * 2))[j];     \
+        }                                                                 \
+        __VA_ARGS__;                                                      \
+        for (int j = 0; j < 16; j++) {                                    \
+            ((uint8_t *)(vd + i * 2))[j] = round_state[j / 4][j % 4];     \
+        }                                                                 \
+        le64_to_cpus(vd + i * 2 + 0);                                     \
+        le64_to_cpus(vd + i * 2 + 1);                                     \
+    }                                                                     \
+    env->vstart = 0;                                                      \
+    /* set tail elements to 1s */                                         \
+    vext_set_elems_1s(vd, vta, vl * 4, total_elems * 4);                  \
+}
+
 GEN_ZVKNED_HELPER_VV(vaesef_vv, aes_sub_bytes(round_state);
                     aes_shift_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);)
+GEN_ZVKNED_HELPER_VS(vaesef_vs, aes_sub_bytes(round_state);
+                    aes_shift_bytes(round_state);
+                    xor_round_key(round_state, (uint8_t *)round_key);)
Signature · 1 line
2.39.2
#20 lore raw

PATCH 19/45 target/riscv: Add vaesdf.vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                        |  1 +
 target/riscv/insn32.decode                   |  1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc |  1 +
 target/riscv/vcrypto_helper.c                | 31 ++++++++++++++++++++
 4 files changed, 34 insertions(+)
target/riscv/helper.h +1 -0
index fb30b4d13e..c626ddcee1 100644
@@ -1190,3 +1190,4 @@ DEF_HELPER_6(vandn_vx_d, void, ptr, ptr, tl, ptr, env, i32)
 
 DEF_HELPER_4(vaesef_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesef_vs, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vaesdf_vv, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index 5d1bb6ccc6..e9ccc56915 100644
@@ -930,3 +930,4 @@ vandn_vx        000001 . ..... ..... 100 ..... 1010111 @r_vm
 # *** RV64 Zvkned vector crypto extension ***
 vaesef_vv       101000 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesef_vs       101001 1 ..... 00011 010 ..... 1110111 @r2_vm_1
+vaesdf_vv       101000 1 ..... 00001 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc +1 -0
index 69bf7f9fee..134fd59df8 100644
@@ -90,3 +90,4 @@ static bool vaes_check_vs(DisasContext *s, arg_rmr *a)
 
 GEN_V_UNMASKED_TRANS(vaesef_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesef_vs, vaes_check_vs)
+GEN_V_UNMASKED_TRANS(vaesdf_vv, vaes_check_vv)
target/riscv/vcrypto_helper.c +31 -0
index b079b543c7..8f448de86e 100644
@@ -181,6 +181,34 @@ static inline void xor_round_key(uint8_t round_state[4][4], uint8_t *round_key)
     }
 }
 
+static inline void aes_inv_sub_bytes(uint8_t round_state[4][4])
+{
+    for (int j = 0; j < 16; j++) {
+        round_state[j / 4][j % 4] = AES_isbox[round_state[j / 4][j % 4]];
+    }
+}
+
+static inline void aes_inv_shift_bytes(uint8_t round_state[4][4])
+{
+    uint8_t temp;
+    temp = round_state[3][1];
+    round_state[3][1] = round_state[2][1];
+    round_state[2][1] = round_state[1][1];
+    round_state[1][1] = round_state[0][1];
+    round_state[0][1] = temp;
+    temp = round_state[0][2];
+    round_state[0][2] = round_state[2][2];
+    round_state[2][2] = temp;
+    temp = round_state[1][2];
+    round_state[1][2] = round_state[3][2];
+    round_state[3][2] = temp;
+    temp = round_state[0][3];
+    round_state[0][3] = round_state[1][3];
+    round_state[1][3] = round_state[2][3];
+    round_state[2][3] = round_state[3][3];
+    round_state[3][3] = temp;
+}
+
 #define GEN_ZVKNED_HELPER_VV(NAME, ...)                                   \
 void HELPER(NAME)(void *vd_vptr, void *vs2_vptr, CPURISCVState *env,      \
                   uint32_t desc)                                          \
@@ -253,3 +281,6 @@ GEN_ZVKNED_HELPER_VV(vaesef_vv, aes_sub_bytes(round_state);
 GEN_ZVKNED_HELPER_VS(vaesef_vs, aes_sub_bytes(round_state);
                     aes_shift_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);)
+GEN_ZVKNED_HELPER_VV(vaesdf_vv, aes_inv_shift_bytes(round_state);
+                    aes_inv_sub_bytes(round_state);
+                    xor_round_key(round_state, (uint8_t *)round_key);)
Signature · 1 line
2.39.2
#21 lore raw

PATCH 20/45 target/riscv: Add vaesdf.vs decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                        | 1 +
 target/riscv/insn32.decode                   | 1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc | 1 +
 target/riscv/vcrypto_helper.c                | 3 +++
 4 files changed, 6 insertions(+)
target/riscv/helper.h +1 -0
index c626ddcee1..059d63b0ea 100644
@@ -1191,3 +1191,4 @@ DEF_HELPER_6(vandn_vx_d, void, ptr, ptr, tl, ptr, env, i32)
 DEF_HELPER_4(vaesef_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesef_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vv, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vaesdf_vs, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index e9ccc56915..df1ae7425d 100644
@@ -931,3 +931,4 @@ vandn_vx        000001 . ..... ..... 100 ..... 1010111 @r_vm
 vaesef_vv       101000 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesef_vs       101001 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesdf_vv       101000 1 ..... 00001 010 ..... 1110111 @r2_vm_1
+vaesdf_vs       101001 1 ..... 00001 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc +1 -0
index 134fd59df8..7a14c33e01 100644
@@ -91,3 +91,4 @@ static bool vaes_check_vs(DisasContext *s, arg_rmr *a)
 GEN_V_UNMASKED_TRANS(vaesef_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesef_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesdf_vv, vaes_check_vv)
+GEN_V_UNMASKED_TRANS(vaesdf_vs, vaes_check_vs)
target/riscv/vcrypto_helper.c +3 -0
index 8f448de86e..69e0843ae4 100644
@@ -284,3 +284,6 @@ GEN_ZVKNED_HELPER_VS(vaesef_vs, aes_sub_bytes(round_state);
 GEN_ZVKNED_HELPER_VV(vaesdf_vv, aes_inv_shift_bytes(round_state);
                     aes_inv_sub_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);)
+GEN_ZVKNED_HELPER_VS(vaesdf_vs, aes_inv_shift_bytes(round_state);
+                    aes_inv_sub_bytes(round_state);
+                    xor_round_key(round_state, (uint8_t *)round_key);)
Signature · 1 line
2.39.2
#22 lore raw

PATCH 21/45 target/riscv: Add vaesdm.vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                        |  1 +
 target/riscv/insn32.decode                   |  1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc |  1 +
 target/riscv/vcrypto_helper.c                | 36 ++++++++++++++++++++
 4 files changed, 39 insertions(+)
target/riscv/helper.h +1 -0
index 059d63b0ea..9b04f90240 100644
@@ -1192,3 +1192,4 @@ DEF_HELPER_4(vaesef_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesef_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vs, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vaesdm_vv, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index df1ae7425d..cdaa320f19 100644
@@ -932,3 +932,4 @@ vaesef_vv       101000 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesef_vs       101001 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesdf_vv       101000 1 ..... 00001 010 ..... 1110111 @r2_vm_1
 vaesdf_vs       101001 1 ..... 00001 010 ..... 1110111 @r2_vm_1
+vaesdm_vv       101000 1 ..... 00000 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc +1 -0
index 7a14c33e01..4f1105ce1d 100644
@@ -92,3 +92,4 @@ GEN_V_UNMASKED_TRANS(vaesef_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesef_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesdf_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesdf_vs, vaes_check_vs)
+GEN_V_UNMASKED_TRANS(vaesdm_vv, vaes_check_vv)
target/riscv/vcrypto_helper.c +36 -0
index 69e0843ae4..f3b9070768 100644
@@ -209,6 +209,38 @@ static inline void aes_inv_shift_bytes(uint8_t round_state[4][4])
     round_state[3][3] = temp;
 }
 
+static inline uint8_t xtime(uint8_t x)
+{
+    return (x << 1) ^ (((x >> 7) & 1) * 0x1b);
+}
+
+static inline uint8_t multiply(uint8_t x, uint8_t y)
+{
+    return (((y & 1) * x) ^ ((y >> 1 & 1) * xtime(x)) ^
+            ((y >> 2 & 1) * xtime(xtime(x))) ^
+            ((y >> 3 & 1) * xtime(xtime(xtime(x)))) ^
+            ((y >> 4 & 1) * xtime(xtime(xtime(xtime(x))))));
+}
+
+static inline void aes_inv_mix_cols(uint8_t round_state[4][4])
+{
+    uint8_t a, b, c, d;
+    for (int j = 0; j < 4; ++j) {
+        a = round_state[j][0];
+        b = round_state[j][1];
+        c = round_state[j][2];
+        d = round_state[j][3];
+        round_state[j][0] = multiply(a, 0x0e) ^ multiply(b, 0x0b) ^
+                            multiply(c, 0x0d) ^ multiply(d, 0x09);
+        round_state[j][1] = multiply(a, 0x09) ^ multiply(b, 0x0e) ^
+                            multiply(c, 0x0b) ^ multiply(d, 0x0d);
+        round_state[j][2] = multiply(a, 0x0d) ^ multiply(b, 0x09) ^
+                            multiply(c, 0x0e) ^ multiply(d, 0x0b);
+        round_state[j][3] = multiply(a, 0x0b) ^ multiply(b, 0x0d) ^
+                            multiply(c, 0x09) ^ multiply(d, 0x0e);
+    }
+}
+
 #define GEN_ZVKNED_HELPER_VV(NAME, ...)                                   \
 void HELPER(NAME)(void *vd_vptr, void *vs2_vptr, CPURISCVState *env,      \
                   uint32_t desc)                                          \
@@ -287,3 +319,7 @@ GEN_ZVKNED_HELPER_VV(vaesdf_vv, aes_inv_shift_bytes(round_state);
 GEN_ZVKNED_HELPER_VS(vaesdf_vs, aes_inv_shift_bytes(round_state);
                     aes_inv_sub_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);)
+GEN_ZVKNED_HELPER_VV(vaesdm_vv, aes_inv_shift_bytes(round_state);
+                    aes_inv_sub_bytes(round_state);
+                    xor_round_key(round_state, (uint8_t *)round_key);
+                    aes_inv_mix_cols(round_state);)
Signature · 1 line
2.39.2
#23 lore raw

PATCH 22/45 target/riscv: Add vaesdm.vs decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                        | 1 +
 target/riscv/insn32.decode                   | 1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc | 1 +
 target/riscv/vcrypto_helper.c                | 4 ++++
 4 files changed, 7 insertions(+)
target/riscv/helper.h +1 -0
index 9b04f90240..ab0d2a4225 100644
@@ -1193,3 +1193,4 @@ DEF_HELPER_4(vaesef_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdm_vv, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vaesdm_vs, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index cdaa320f19..61f6b81644 100644
@@ -933,3 +933,4 @@ vaesef_vs       101001 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesdf_vv       101000 1 ..... 00001 010 ..... 1110111 @r2_vm_1
 vaesdf_vs       101001 1 ..... 00001 010 ..... 1110111 @r2_vm_1
 vaesdm_vv       101000 1 ..... 00000 010 ..... 1110111 @r2_vm_1
+vaesdm_vs       101001 1 ..... 00000 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc +1 -0
index 4f1105ce1d..93f0df6b78 100644
@@ -93,3 +93,4 @@ GEN_V_UNMASKED_TRANS(vaesef_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesdf_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesdf_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesdm_vv, vaes_check_vv)
+GEN_V_UNMASKED_TRANS(vaesdm_vs, vaes_check_vs)
target/riscv/vcrypto_helper.c +4 -0
index f3b9070768..04d5ce5dc0 100644
@@ -323,3 +323,7 @@ GEN_ZVKNED_HELPER_VV(vaesdm_vv, aes_inv_shift_bytes(round_state);
                     aes_inv_sub_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);
                     aes_inv_mix_cols(round_state);)
+GEN_ZVKNED_HELPER_VS(vaesdm_vs, aes_inv_shift_bytes(round_state);
+                    aes_inv_sub_bytes(round_state);
+                    xor_round_key(round_state, (uint8_t *)round_key);
+                    aes_inv_mix_cols(round_state);)
Signature · 1 line
2.39.2
#24 lore raw

PATCH 23/45 target/riscv: Add vaesz.vs decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                        | 1 +
 target/riscv/insn32.decode                   | 1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc | 1 +
 target/riscv/vcrypto_helper.c                | 2 ++
 4 files changed, 5 insertions(+)
target/riscv/helper.h +1 -0
index ab0d2a4225..58121ba8ad 100644
@@ -1194,3 +1194,4 @@ DEF_HELPER_4(vaesdf_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdm_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdm_vs, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vaesz_vs, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index 61f6b81644..22059ef95b 100644
@@ -934,3 +934,4 @@ vaesdf_vv       101000 1 ..... 00001 010 ..... 1110111 @r2_vm_1
 vaesdf_vs       101001 1 ..... 00001 010 ..... 1110111 @r2_vm_1
 vaesdm_vv       101000 1 ..... 00000 010 ..... 1110111 @r2_vm_1
 vaesdm_vs       101001 1 ..... 00000 010 ..... 1110111 @r2_vm_1
+vaesz_vs        101001 1 ..... 00111 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc +1 -0
index 93f0df6b78..1f59dbcc68 100644
@@ -94,3 +94,4 @@ GEN_V_UNMASKED_TRANS(vaesdf_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesdf_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesdm_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesdm_vs, vaes_check_vs)
+GEN_V_UNMASKED_TRANS(vaesz_vs, vaes_check_vs)
target/riscv/vcrypto_helper.c +2 -0
index 04d5ce5dc0..41e138ece5 100644
@@ -327,3 +327,5 @@ GEN_ZVKNED_HELPER_VS(vaesdm_vs, aes_inv_shift_bytes(round_state);
                     aes_inv_sub_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);
                     aes_inv_mix_cols(round_state);)
+GEN_ZVKNED_HELPER_VS(vaesz_vs,
+                    xor_round_key(round_state, (uint8_t *)round_key);)
Signature · 1 line
2.39.2
#25 lore raw

PATCH 24/45 target/riscv: Add vaesem.vv decoding, translation and execution support

In reply to Lawrence Hunter

From: William Salmon <[email protected]>

---
 target/riscv/helper.h                        |  1 +
 target/riscv/insn32.decode                   |  1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc |  1 +
 target/riscv/vcrypto_helper.c                | 17 +++++++++++++++++
 4 files changed, 20 insertions(+)
target/riscv/helper.h +1 -0
index 58121ba8ad..c55d59dc5e 100644
@@ -1192,6 +1192,7 @@ DEF_HELPER_4(vaesef_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesef_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vs, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vaesem_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdm_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdm_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesz_vs, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index 22059ef95b..f03b41f9e2 100644
@@ -932,6 +932,7 @@ vaesef_vv       101000 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesef_vs       101001 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesdf_vv       101000 1 ..... 00001 010 ..... 1110111 @r2_vm_1
 vaesdf_vs       101001 1 ..... 00001 010 ..... 1110111 @r2_vm_1
+vaesem_vv       101000 1 ..... 00010 010 ..... 1110111 @r2_vm_1
 vaesdm_vv       101000 1 ..... 00000 010 ..... 1110111 @r2_vm_1
 vaesdm_vs       101001 1 ..... 00000 010 ..... 1110111 @r2_vm_1
 vaesz_vs        101001 1 ..... 00111 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc +1 -0
index 1f59dbcc68..0b54d6e9d3 100644
@@ -95,3 +95,4 @@ GEN_V_UNMASKED_TRANS(vaesdf_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesdm_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesdm_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesz_vs, vaes_check_vs)
+GEN_V_UNMASKED_TRANS(vaesem_vv, vaes_check_vv)
target/riscv/vcrypto_helper.c +17 -0
index 41e138ece5..beef7699db 100644
@@ -241,6 +241,20 @@ static inline void aes_inv_mix_cols(uint8_t round_state[4][4])
     }
 }
 
+static inline void aes_mix_cols(uint8_t round_state[4][4])
+{
+    uint8_t a, b;
+    for (int j = 0; j < 4; ++j) {
+        a = round_state[j][0];
+        b = round_state[j][0] ^ round_state[j][1] ^ round_state[j][2] ^
+            round_state[j][3];
+        round_state[j][0] ^= xtime(round_state[j][0] ^ round_state[j][1]) ^ b;
+        round_state[j][1] ^= xtime(round_state[j][1] ^ round_state[j][2]) ^ b;
+        round_state[j][2] ^= xtime(round_state[j][2] ^ round_state[j][3]) ^ b;
+        round_state[j][3] ^= xtime(round_state[j][3] ^ a) ^ b;
+    }
+}
+
 #define GEN_ZVKNED_HELPER_VV(NAME, ...)                                   \
 void HELPER(NAME)(void *vd_vptr, void *vs2_vptr, CPURISCVState *env,      \
                   uint32_t desc)                                          \
@@ -319,6 +333,9 @@ GEN_ZVKNED_HELPER_VV(vaesdf_vv, aes_inv_shift_bytes(round_state);
 GEN_ZVKNED_HELPER_VS(vaesdf_vs, aes_inv_shift_bytes(round_state);
                     aes_inv_sub_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);)
+GEN_ZVKNED_HELPER_VV(vaesem_vv, aes_shift_bytes(round_state);
+                    aes_sub_bytes(round_state); aes_mix_cols(round_state);
+                    xor_round_key(round_state, (uint8_t *)round_key);)
 GEN_ZVKNED_HELPER_VV(vaesdm_vv, aes_inv_shift_bytes(round_state);
                     aes_inv_sub_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);
Signature · 1 line
2.39.2
#26 lore raw

PATCH 25/45 target/riscv: Add vaesem.vs decoding, translation and execution support

In reply to Lawrence Hunter

From: William Salmon <[email protected]>

---
 target/riscv/helper.h                        | 1 +
 target/riscv/insn32.decode                   | 1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc | 1 +
 target/riscv/vcrypto_helper.c                | 3 +++
 4 files changed, 6 insertions(+)
target/riscv/helper.h +1 -0
index c55d59dc5e..946ae8c51d 100644
@@ -1193,6 +1193,7 @@ DEF_HELPER_4(vaesef_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdf_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesem_vv, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vaesem_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdm_vv, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdm_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesz_vs, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index f03b41f9e2..3187c5cc64 100644
@@ -933,6 +933,7 @@ vaesef_vs       101001 1 ..... 00011 010 ..... 1110111 @r2_vm_1
 vaesdf_vv       101000 1 ..... 00001 010 ..... 1110111 @r2_vm_1
 vaesdf_vs       101001 1 ..... 00001 010 ..... 1110111 @r2_vm_1
 vaesem_vv       101000 1 ..... 00010 010 ..... 1110111 @r2_vm_1
+vaesem_vs       101001 1 ..... 00010 010 ..... 1110111 @r2_vm_1
 vaesdm_vv       101000 1 ..... 00000 010 ..... 1110111 @r2_vm_1
 vaesdm_vs       101001 1 ..... 00000 010 ..... 1110111 @r2_vm_1
 vaesz_vs        101001 1 ..... 00111 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkned.c.inc +1 -0
index 0b54d6e9d3..028f04a4d7 100644
@@ -96,3 +96,4 @@ GEN_V_UNMASKED_TRANS(vaesdm_vv, vaes_check_vv)
 GEN_V_UNMASKED_TRANS(vaesdm_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesz_vs, vaes_check_vs)
 GEN_V_UNMASKED_TRANS(vaesem_vv, vaes_check_vv)
+GEN_V_UNMASKED_TRANS(vaesem_vs, vaes_check_vs)
target/riscv/vcrypto_helper.c +3 -0
index beef7699db..600069adb1 100644
@@ -336,6 +336,9 @@ GEN_ZVKNED_HELPER_VS(vaesdf_vs, aes_inv_shift_bytes(round_state);
 GEN_ZVKNED_HELPER_VV(vaesem_vv, aes_shift_bytes(round_state);
                     aes_sub_bytes(round_state); aes_mix_cols(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);)
+GEN_ZVKNED_HELPER_VS(vaesem_vs, aes_shift_bytes(round_state);
+                    aes_sub_bytes(round_state); aes_mix_cols(round_state);
+                    xor_round_key(round_state, (uint8_t *)round_key);)
 GEN_ZVKNED_HELPER_VV(vaesdm_vv, aes_inv_shift_bytes(round_state);
                     aes_inv_sub_bytes(round_state);
                     xor_round_key(round_state, (uint8_t *)round_key);
Signature · 1 line
2.39.2
#27 lore raw

PATCH 26/45 target/riscv: Add vaeskf1.vi decoding, translation and execution support

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 target/riscv/helper.h                        |  1 +
 target/riscv/insn32.decode                   |  1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc | 58 ++++++++++++++++++++
 target/riscv/vcrypto_helper.c                | 44 +++++++++++++++
 4 files changed, 104 insertions(+)
target/riscv/helper.h +1 -0
index 946ae8c51d..e68ced7796 100644
@@ -1197,3 +1197,4 @@ DEF_HELPER_4(vaesem_vs, void, ptr, ptr, env, i32)
 DEF_HELPER_4(vaesdm_vv, void, ptr, ptr, env, i32)
 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)
target/riscv/insn32.decode +1 -0
index 3187c5cc64..0b3146c4f4 100644
@@ -937,3 +937,4 @@ vaesem_vs       101001 1 ..... 00010 010 ..... 1110111 @r2_vm_1
 vaesdm_vv       101000 1 ..... 00000 010 ..... 1110111 @r2_vm_1
 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
target/riscv/insn_trans/trans_rvzvkned.c.inc +58 -0

61 lines not on this page · raw message

target/riscv/vcrypto_helper.c +44 -0

47 lines not on this page · raw message

Signature · 1 line
2.39.2
#28 lore raw

PATCH 27/45 target/riscv: Add vaeskf2.vi decoding, translation and execution support

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 target/riscv/helper.h                        |  1 +
 target/riscv/insn32.decode                   |  1 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc | 13 +++++
 target/riscv/vcrypto_helper.c                | 59 ++++++++++++++++++++
 4 files changed, 74 insertions(+)
target/riscv/helper.h +1 -0
index e68ced7796..f07f261b7b 100644
@@ -1198,3 +1198,4 @@ DEF_HELPER_4(vaesdm_vv, void, ptr, ptr, env, i32)
 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)
target/riscv/insn32.decode +1 -0
index 0b3146c4f4..43dfd63e0d 100644
@@ -938,3 +938,4 @@ vaesdm_vv       101000 1 ..... 00000 010 ..... 1110111 @r2_vm_1
 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
target/riscv/insn_trans/trans_rvzvkned.c.inc +13 -0
index c97780f468..c135b8f768 100644
@@ -154,4 +154,17 @@ static bool vaeskf1_check(DisasContext *s, arg_vaeskf1_vi * a)
            require_align(a->rs2, s->lmul);
 }
 
+static bool vaeskf2_check(DisasContext *s, arg_vaeskf2_vi *a)
+{
+    return s->cfg_ptr->ext_zvkned == true &&
+           require_rvv(s) &&
+           vext_check_isa_ill(s) &&
+           MAXSZ(s) >= (128 / 8) && /* EGW in bytes */
+           s->vstart % 4 == 0 &&
+           s->sew == MO_32 &&
+           require_align(a->rd, s->lmul) &&
+           require_align(a->rs2, s->lmul);
+}
+
 GEN_VI_UNMASKED_TRANS(vaeskf1_vi, vaeskf1_check, 4)
+GEN_VI_UNMASKED_TRANS(vaeskf2_vi, vaeskf2_check, 4)
target/riscv/vcrypto_helper.c +59 -0

62 lines not on this page · raw message

Signature · 1 line
2.39.2
#29 lore raw

PATCH 28/45 target/riscv: Expose zvkned cpu property

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 target/riscv/cpu.c | 1 +
 1 file changed, 1 insertion(+)
target/riscv/cpu.c +1 -0
index 00e1d007a4..cd87eec919 100644
@@ -1472,6 +1472,7 @@ static Property riscv_cpu_extensions[] = {
 
     /* Vector cryptography extensions */
     DEFINE_PROP_BOOL("x-zvkb", RISCVCPU, cfg.ext_zvkb, false),
+    DEFINE_PROP_BOOL("x-zvkned", RISCVCPU, cfg.ext_zvkned, false),
 
     DEFINE_PROP_END_OF_LIST(),
 };
Signature · 1 line
2.39.2
#30 lore raw

PATCH 29/45 target/riscv: Add zvknh cpu properties

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 target/riscv/cpu.c | 11 ++++++++++-
 target/riscv/cpu.h |  2 ++
 2 files changed, 12 insertions(+), 1 deletion(-)
target/riscv/cpu.c +10 -1
index cd87eec919..3ffbdd53cc 100644
@@ -111,6 +111,8 @@ static const struct isa_ext_data isa_edata_arr[] = {
     ISA_EXT_DATA_ENTRY(zvfhmin, true, PRIV_VERSION_1_12_0, ext_zvfhmin),
     ISA_EXT_DATA_ENTRY(zvkb, true, PRIV_VERSION_1_12_0, ext_zvkb),
     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),
@@ -1217,7 +1219,7 @@ 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_zvkb || cpu->cfg.ext_zvkned) &&
+    if ((cpu->cfg.ext_zvkb || 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(
@@ -1225,6 +1227,13 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
         return;
     }
 
+    if (cpu->cfg.ext_zvknhb &&
+        !(cpu->cfg.ext_zve64f || cpu->cfg.ext_zve64d || cpu->cfg.ext_v)) {
+            error_setg(errp,
+                       "Zvknhb extension requires V or Zve64{f,d} extensions");
+            return;
+    }
+
 #ifndef CONFIG_USER_ONLY
     if (cpu->cfg.pmu_num) {
         if (!riscv_pmu_init(cpu, cpu->cfg.pmu_num) && cpu->cfg.ext_sscofpmf) {
target/riscv/cpu.h +2 -0
index 4f3b97e0f1..5d101fc405 100644
@@ -472,6 +472,8 @@ struct RISCVCPUConfig {
     bool ext_zve64d;
     bool ext_zvkb;
     bool ext_zvkned;
+    bool ext_zvknha;
+    bool ext_zvknhb;
     bool ext_zmmul;
     bool ext_zvfh;
     bool ext_zvfhmin;
Signature · 1 line
2.39.2
#31 lore raw

PATCH 30/45 target/riscv: Add vsha2ms.vv decoding, translation and execution support

In reply to Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

---
 target/riscv/helper.h                       |  2 +
 target/riscv/insn32.decode                  |  3 +
 target/riscv/insn_trans/trans_rvzvknh.c.inc | 82 +++++++++++++++++++++
 target/riscv/translate.c                    |  1 +
 target/riscv/vcrypto_helper.c               | 74 +++++++++++++++++++
 5 files changed, 162 insertions(+)
 create mode 100644 target/riscv/insn_trans/trans_rvzvknh.c.inc
target/riscv/helper.h +2 -0
index f07f261b7b..76ea2ff49b 100644
@@ -1199,3 +1199,5 @@ 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)
target/riscv/insn32.decode +3 -0
index 43dfd63e0d..aef4d0b476 100644
@@ -939,3 +939,6 @@ 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
+
+# *** RV64 Zvknh vector crypto extension ***
+vsha2ms_vv      101101 1 ..... ..... 010 ..... 1110111 @r_vm_1
target/riscv/insn_trans/trans_rvzvknh.c.inc new file +82 -0

82 lines not on this page · raw message

target/riscv/translate.c +1 -0
index e3561b0bbd..1c1e36c10a 100644
@@ -1085,6 +1085,7 @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc)
 #include "insn_trans/trans_rvk.c.inc"
 #include "insn_trans/trans_rvzvkb.c.inc"
 #include "insn_trans/trans_rvzvkned.c.inc"
+#include "insn_trans/trans_rvzvknh.c.inc"
 #include "insn_trans/trans_privileged.c.inc"
 #include "insn_trans/trans_svinval.c.inc"
 #include "decode-xthead.c.inc"
target/riscv/vcrypto_helper.c +74 -0

77 lines not on this page · raw message

Signature · 1 line
2.39.2
#32 lore raw

PATCH 31/45 target/riscv: Add vsha2c[hl].vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                       |   2 +
 target/riscv/insn32.decode                  |   2 +
 target/riscv/insn_trans/trans_rvzvknh.c.inc |   2 +
 target/riscv/vcrypto_helper.c               | 140 ++++++++++++++++++++
 4 files changed, 146 insertions(+)
target/riscv/helper.h +2 -0
index 76ea2ff49b..77bbd9db56 100644
@@ -1201,3 +1201,5 @@ 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)
target/riscv/insn32.decode +2 -0
index aef4d0b476..c95886040b 100644
@@ -942,3 +942,5 @@ vaeskf2_vi      101010 1 ..... ..... 010 ..... 1110111 @r_vm_1
 
 # *** RV64 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
target/riscv/insn_trans/trans_rvzvknh.c.inc +2 -0
index 97bdf4d72f..3cf3ceaf3a 100644
@@ -80,3 +80,5 @@ static bool vsha_check(DisasContext *s, arg_rmrr *a)
 }
 
 GEN_VV_UNMASKED_TRANS(vsha2ms_vv, vsha_check, 4)
+GEN_VV_UNMASKED_TRANS(vsha2cl_vv, vsha_check, 4)
+GEN_VV_UNMASKED_TRANS(vsha2ch_vv, vsha_check, 4)
target/riscv/vcrypto_helper.c +140 -0

143 lines not on this page · raw message

Signature · 1 line
2.39.2
#33 lore raw

PATCH 32/45 target/riscv: Expose zvknh cpu properties

In reply to Lawrence Hunter

From: Nazar Kazakov <[email protected]>

---
 target/riscv/cpu.c | 2 ++
 1 file changed, 2 insertions(+)
target/riscv/cpu.c +2 -0
index 3ffbdd53cc..b3f9638067 100644
@@ -1482,6 +1482,8 @@ static Property riscv_cpu_extensions[] = {
     /* Vector cryptography extensions */
     DEFINE_PROP_BOOL("x-zvkb", RISCVCPU, cfg.ext_zvkb, false),
     DEFINE_PROP_BOOL("x-zvkned", RISCVCPU, cfg.ext_zvkned, false),
+    DEFINE_PROP_BOOL("x-zvknha", RISCVCPU, cfg.ext_zvknha, false),
+    DEFINE_PROP_BOOL("x-zvknhb", RISCVCPU, cfg.ext_zvknhb, false),
 
     DEFINE_PROP_END_OF_LIST(),
 };
Signature · 1 line
2.39.2
#34 lore raw

PATCH 33/45 target/riscv: Add zvksh cpu property

In reply to Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

---
 target/riscv/cpu.c | 4 +++-
 target/riscv/cpu.h | 1 +
 2 files changed, 4 insertions(+), 1 deletion(-)
target/riscv/cpu.c +3 -1
index b3f9638067..e218a00a2d 100644
@@ -113,6 +113,7 @@ static const struct isa_ext_data isa_edata_arr[] = {
     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(zvksh, true, PRIV_VERSION_1_12_0, ext_zvksh),
     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),
@@ -1219,7 +1220,8 @@ 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_zvkb || cpu->cfg.ext_zvkned || cpu->cfg.ext_zvknha) &&
+    if ((cpu->cfg.ext_zvkb || cpu->cfg.ext_zvkned || cpu->cfg.ext_zvknha ||
+             cpu->cfg.ext_zvksh) &&
         !(cpu->cfg.ext_zve32f || cpu->cfg.ext_zve64f ||
             cpu->cfg.ext_zve64d || cpu->cfg.ext_v)) {
         error_setg(
target/riscv/cpu.h +1 -0
index 5d101fc405..3b0f322f3e 100644
@@ -474,6 +474,7 @@ struct RISCVCPUConfig {
     bool ext_zvkned;
     bool ext_zvknha;
     bool ext_zvknhb;
+    bool ext_zvksh;
     bool ext_zmmul;
     bool ext_zvfh;
     bool ext_zvfhmin;
Signature · 1 line
2.39.2
#35 lore raw

PATCH 34/45 target/riscv: Add vsm3me.vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                       |  2 ++
 target/riscv/insn32.decode                  |  3 ++
 target/riscv/insn_trans/trans_rvzvksh.c.inc | 37 +++++++++++++++++++
 target/riscv/translate.c                    |  1 +
 target/riscv/vcrypto_helper.c               | 39 +++++++++++++++++++++
 5 files changed, 82 insertions(+)
 create mode 100644 target/riscv/insn_trans/trans_rvzvksh.c.inc
target/riscv/helper.h +2 -0
index 77bbd9db56..d8f67b924e 100644
@@ -1203,3 +1203,5 @@ 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)
+
+DEF_HELPER_5(vsm3me_vv, void, ptr, ptr, ptr, env, i32)
target/riscv/insn32.decode +3 -0
index c95886040b..588907dd4d 100644
@@ -944,3 +944,6 @@ vaeskf2_vi      101010 1 ..... ..... 010 ..... 1110111 @r_vm_1
 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
+
+# *** RV64 Zvksh vector crypto extension ***
+vsm3me_vv       100000 1 ..... ..... 010 ..... 1110111 @r_vm_1
target/riscv/insn_trans/trans_rvzvksh.c.inc new file +37 -0
new file mode 100644
index 0000000000..a0b3de1b21
@@ -0,0 +1,37 @@
+/*
+ * RISC-V translation routines for the Zvksh Extension.
+ *
+ * Copyright (C) 2023 SiFive, Inc.
+ * Written by Codethink Ltd and SiFive.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2 or later, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+static inline bool vsm3_check(DisasContext *s, arg_rmrr *a)
+{
+    int mult = 1 << MAX(s->lmul, 0);
+    return s->cfg_ptr->ext_zvksh == true &&
+           require_rvv(s) &&
+           vext_check_isa_ill(s) &&
+           !is_overlapped(a->rd, mult, a->rs2, mult) &&
+           MAXSZ(s) >= (256 / 8) && /* EGW in bytes */
+           s->vstart % 8 == 0 &&
+           s->sew == MO_32;
+}
+
+static inline bool vsm3me_check(DisasContext *s, arg_rmrr *a)
+{
+    return vsm3_check(s, a) && vext_check_sss(s, a->rd, a->rs1, a->rs2, a->vm);
+}
+
+GEN_VV_UNMASKED_TRANS(vsm3me_vv, vsm3me_check, 8)
target/riscv/translate.c +1 -0
index 1c1e36c10a..256872ec28 100644
@@ -1086,6 +1086,7 @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc)
 #include "insn_trans/trans_rvzvkb.c.inc"
 #include "insn_trans/trans_rvzvkned.c.inc"
 #include "insn_trans/trans_rvzvknh.c.inc"
+#include "insn_trans/trans_rvzvksh.c.inc"
 #include "insn_trans/trans_privileged.c.inc"
 #include "insn_trans/trans_svinval.c.inc"
 #include "decode-xthead.c.inc"
target/riscv/vcrypto_helper.c +39 -0
index bf0455f8e0..20c4ed8c4a 100644
@@ -666,3 +666,42 @@ void HELPER(vsha2cl_vv)(void *vd, void *vs1, void *vs2, CPURISCVState *env,
     vext_set_elems_1s(vd, vta, env->vl * esz, total_elems * esz);
     env->vstart = 0;
 }
+
+static inline uint32_t p1(uint32_t x)
+{
+    return x ^ rol32(x, 15) ^ rol32(x, 23);
+}
+
+static inline uint32_t zvksh_w(uint32_t m16, uint32_t m9, uint32_t m3,
+                               uint32_t m13, uint32_t m6)
+{
+    return p1(m16 ^ m9 ^ rol32(m3, 15)) ^ rol32(m13, 7) ^ m6;
+}
+
+void HELPER(vsm3me_vv)(void *vd_vptr, void *vs1_vptr, void *vs2_vptr,
+                       CPURISCVState *env, uint32_t desc)
+{
+    uint32_t esz = memop_size(FIELD_EX64(env->vtype, VTYPE, VSEW));
+    uint32_t total_elems = vext_get_total_elems(env, desc, esz);
+    uint32_t vta = vext_vta(desc);
+    uint32_t *vd = vd_vptr;
+    uint32_t *vs1 = vs1_vptr;
+    uint32_t *vs2 = vs2_vptr;
+
+    for (int i = env->vstart / 8; i < env->vl / 8; i++) {
+        uint32_t w[24];
+        for (int j = 0; j < 8; j++) {
+            w[j] = bswap32(vs1[H4((i * 8) + j)]);
+            w[j + 8] = bswap32(vs2[H4((i * 8) + j)]);
+        }
+        for (int j = 0; j < 8; j++) {
+            w[j + 16] =
+                zvksh_w(w[j], w[j + 7], w[j + 13], w[j + 3], w[j + 10]);
+        }
+        for (int j = 0; j < 8; j++) {
+            vd[(i * 8) + j] = bswap32(w[H4(j + 16)]);
+        }
+    }
+    vext_set_elems_1s(vd_vptr, vta, env->vl * esz, total_elems * esz);
+    env->vstart = 0;
+}
Signature · 1 line
2.39.2
#36 lore raw

PATCH 35/45 target/riscv: Add vsm3c.vi decoding, translation and execution support

In reply to Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

---
 target/riscv/helper.h                       |  1 +
 target/riscv/insn32.decode                  |  1 +
 target/riscv/insn_trans/trans_rvzvksh.c.inc |  6 ++
 target/riscv/vcrypto_helper.c               | 95 +++++++++++++++++++++
 4 files changed, 103 insertions(+)
target/riscv/helper.h +1 -0
index d8f67b924e..494ebf8cda 100644
@@ -1205,3 +1205,4 @@ DEF_HELPER_5(vsha2ch_vv, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vsha2cl_vv, void, ptr, ptr, ptr, env, i32)
 
 DEF_HELPER_5(vsm3me_vv, void, ptr, ptr, ptr, env, i32)
+DEF_HELPER_5(vsm3c_vi, void, ptr, ptr, i32, env, i32)
target/riscv/insn32.decode +1 -0
index 588907dd4d..76802f37db 100644
@@ -947,3 +947,4 @@ vsha2cl_vv      101111 1 ..... ..... 010 ..... 1110111 @r_vm_1
 
 # *** RV64 Zvksh vector crypto extension ***
 vsm3me_vv       100000 1 ..... ..... 010 ..... 1110111 @r_vm_1
+vsm3c_vi        101011 1 ..... ..... 010 ..... 1110111 @r_vm_1
target/riscv/insn_trans/trans_rvzvksh.c.inc +6 -0
index a0b3de1b21..e15cb828b7 100644
@@ -34,4 +34,10 @@ static inline bool vsm3me_check(DisasContext *s, arg_rmrr *a)
     return vsm3_check(s, a) && vext_check_sss(s, a->rd, a->rs1, a->rs2, a->vm);
 }
 
+static inline bool vsm3c_check(DisasContext *s, arg_rmrr *a)
+{
+    return vsm3_check(s, a) && vext_check_ss(s, a->rd, a->rs2, a->vm);
+}
+
 GEN_VV_UNMASKED_TRANS(vsm3me_vv, vsm3me_check, 8)
+GEN_VI_UNMASKED_TRANS(vsm3c_vi, vsm3c_check, 8)
target/riscv/vcrypto_helper.c +95 -0

98 lines not on this page · raw message

Signature · 1 line
2.39.2
#37 lore raw

PATCH 36/45 target/riscv: Expose zvksh cpu property

In reply to Lawrence Hunter

From: Kiran Ostrolenk <[email protected]>

---
 target/riscv/cpu.c | 1 +
 1 file changed, 1 insertion(+)
target/riscv/cpu.c +1 -0
index e218a00a2d..c136a17112 100644
@@ -1486,6 +1486,7 @@ static Property riscv_cpu_extensions[] = {
     DEFINE_PROP_BOOL("x-zvkned", RISCVCPU, cfg.ext_zvkned, false),
     DEFINE_PROP_BOOL("x-zvknha", RISCVCPU, cfg.ext_zvknha, false),
     DEFINE_PROP_BOOL("x-zvknhb", RISCVCPU, cfg.ext_zvknhb, false),
+    DEFINE_PROP_BOOL("x-zvksh", RISCVCPU, cfg.ext_zvksh, false),
 
     DEFINE_PROP_END_OF_LIST(),
 };
Signature · 1 line
2.39.2
#38 lore raw

PATCH 37/45 target/riscv: Add zvkg cpu property

In reply to Lawrence Hunter
---
 target/riscv/cpu.c | 5 +++--
 target/riscv/cpu.h | 1 +
 2 files changed, 4 insertions(+), 2 deletions(-)
target/riscv/cpu.c +3 -2
index c136a17112..79079d517d 100644
@@ -110,6 +110,7 @@ static const struct isa_ext_data isa_edata_arr[] = {
     ISA_EXT_DATA_ENTRY(zvfh, true, PRIV_VERSION_1_12_0, ext_zvfh),
     ISA_EXT_DATA_ENTRY(zvfhmin, true, PRIV_VERSION_1_12_0, ext_zvfhmin),
     ISA_EXT_DATA_ENTRY(zvkb, true, PRIV_VERSION_1_12_0, ext_zvkb),
+    ISA_EXT_DATA_ENTRY(zvkg, true, PRIV_VERSION_1_12_0, ext_zvkg),
     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),
@@ -1220,8 +1221,8 @@ 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_zvkb || cpu->cfg.ext_zvkned || cpu->cfg.ext_zvknha ||
-             cpu->cfg.ext_zvksh) &&
+    if ((cpu->cfg.ext_zvkb || cpu->cfg.ext_zvkg || cpu->cfg.ext_zvkned ||
+             cpu->cfg.ext_zvknha || cpu->cfg.ext_zvksh) &&
         !(cpu->cfg.ext_zve32f || cpu->cfg.ext_zve64f ||
             cpu->cfg.ext_zve64d || cpu->cfg.ext_v)) {
         error_setg(
target/riscv/cpu.h +1 -0
index 3b0f322f3e..40c4e23209 100644
@@ -471,6 +471,7 @@ struct RISCVCPUConfig {
     bool ext_zve64f;
     bool ext_zve64d;
     bool ext_zvkb;
+    bool ext_zvkg;
     bool ext_zvkned;
     bool ext_zvknha;
     bool ext_zvknhb;
Signature · 1 line
2.39.2
#39 lore raw

PATCH 38/45 target/riscv: Add vgmul.vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                      |  2 ++
 target/riscv/insn32.decode                 |  3 ++
 target/riscv/insn_trans/trans_rvzvkg.c.inc | 30 +++++++++++++++++++
 target/riscv/translate.c                   |  1 +
 target/riscv/vcrypto_helper.c              | 34 ++++++++++++++++++++++
 5 files changed, 70 insertions(+)
 create mode 100644 target/riscv/insn_trans/trans_rvzvkg.c.inc
target/riscv/helper.h +2 -0
index 494ebf8cda..680f695e75 100644
@@ -1206,3 +1206,5 @@ DEF_HELPER_5(vsha2cl_vv, void, ptr, ptr, ptr, env, i32)
 
 DEF_HELPER_5(vsm3me_vv, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vsm3c_vi, void, ptr, ptr, i32, env, i32)
+
+DEF_HELPER_4(vgmul_vv, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +3 -0
index 76802f37db..fdb535906f 100644
@@ -948,3 +948,6 @@ vsha2cl_vv      101111 1 ..... ..... 010 ..... 1110111 @r_vm_1
 # *** RV64 Zvksh vector crypto extension ***
 vsm3me_vv       100000 1 ..... ..... 010 ..... 1110111 @r_vm_1
 vsm3c_vi        101011 1 ..... ..... 010 ..... 1110111 @r_vm_1
+
+# *** RV64 Zvkg vector crypto extension ***
+vgmul_vv        101000 1 ..... 10001 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkg.c.inc new file +30 -0
new file mode 100644
index 0000000000..f1e4ea1381
@@ -0,0 +1,30 @@
+/*
+ * RISC-V translation routines for the Zvkg Extension.
+ *
+ * Copyright (C) 2023 SiFive, Inc.
+ * Written by Codethink Ltd and SiFive.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2 or later, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+static bool vgmul_check(DisasContext *s, arg_rmr *a)
+{
+    return s->cfg_ptr->ext_zvkg == true &&
+            vext_check_isa_ill(s) &&
+            require_rvv(s) &&
+            MAXSZ(s) >= (128 / 8) && /* EGW in bytes */
+            vext_check_ss(s, a->rd, a->rs2, a->vm) &&
+            s->vstart % 4 == 0 && s->sew == MO_32;
+}
+
+GEN_V_UNMASKED_TRANS(vgmul_vv, vgmul_check)
target/riscv/translate.c +1 -0
index 256872ec28..fdb5c3364e 100644
@@ -1087,6 +1087,7 @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc)
 #include "insn_trans/trans_rvzvkned.c.inc"
 #include "insn_trans/trans_rvzvknh.c.inc"
 #include "insn_trans/trans_rvzvksh.c.inc"
+#include "insn_trans/trans_rvzvkg.c.inc"
 #include "insn_trans/trans_privileged.c.inc"
 #include "insn_trans/trans_svinval.c.inc"
 #include "decode-xthead.c.inc"
target/riscv/vcrypto_helper.c +34 -0
index 2d8db45740..eb70e2e26c 100644
@@ -800,3 +800,37 @@ void HELPER(vsm3c_vi)(void *vd_vptr, void *vs2_vptr, uint32_t uimm,
     vext_set_elems_1s(vd_vptr, vta, env->vl * esz, total_elems * esz);
     env->vstart = 0;
 }
+
+void HELPER(vgmul_vv)(void *vd_vptr, void *vs2_vptr,
+                       CPURISCVState *env, uint32_t desc)
+{
+    uint64_t *vd = vd_vptr;
+    uint64_t *vs2 = vs2_vptr;
+    uint32_t vta = vext_vta(desc);
+    uint32_t total_elems = vext_get_total_elems(env, desc, 4);
+
+    for (uint32_t i = env->vstart / 4; i < env->vl / 4; i++) {
+        uint64_t Y[2] = {brev8(vd[i * 2 + 0]), brev8(vd[i * 2 + 1])};
+        uint64_t H[2] = {brev8(vs2[i * 2 + 0]), brev8(vs2[i * 2 + 1])};
+        uint64_t Z[2] = {0, 0};
+
+        for (uint j = 0; j < 128; j++) {
+            if ((Y[j / 64] >> (j % 64)) & 1) {
+                Z[0] ^= H[0];
+                Z[1] ^= H[1];
+            }
+            bool reduce = ((H[1] >> 63) & 1);
+            H[1] = H[1] << 1 | H[0] >> 63;
+            H[0] = H[0] << 1;
+            if (reduce) {
+                H[0] ^= 0x87;
+            }
+        }
+
+        vd[i * 2 + 0] = brev8(Z[0]);
+        vd[i * 2 + 1] = brev8(Z[1]);
+    }
+    /* set tail elements to 1s */
+    vext_set_elems_1s(vd, vta, env->vl * 4, total_elems * 4);
+    env->vstart = 0;
+}
Signature · 1 line
2.39.2
#40 lore raw

PATCH 39/45 target/riscv: Add vghsh.vv decoding, translation and execution support

In reply to Lawrence Hunter
---
 target/riscv/helper.h                      |  1 +
 target/riscv/insn32.decode                 |  1 +
 target/riscv/insn_trans/trans_rvzvkg.c.inc | 10 ++++++
 target/riscv/vcrypto_helper.c              | 38 ++++++++++++++++++++++
 4 files changed, 50 insertions(+)
target/riscv/helper.h +1 -0
index 680f695e75..3c4aa4b5df 100644
@@ -1207,4 +1207,5 @@ DEF_HELPER_5(vsha2cl_vv, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vsm3me_vv, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_5(vsm3c_vi, void, ptr, ptr, i32, env, i32)
 
+DEF_HELPER_5(vghsh_vv, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_4(vgmul_vv, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +1 -0
index fdb535906f..856e088bad 100644
@@ -950,4 +950,5 @@ vsm3me_vv       100000 1 ..... ..... 010 ..... 1110111 @r_vm_1
 vsm3c_vi        101011 1 ..... ..... 010 ..... 1110111 @r_vm_1
 
 # *** RV64 Zvkg vector crypto extension ***
+vghsh_vv        101100 1 ..... ..... 010 ..... 1110111 @r_vm_1
 vgmul_vv        101000 1 ..... 10001 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvkg.c.inc +10 -0
index f1e4ea1381..9280300ce0 100644
@@ -28,3 +28,13 @@ static bool vgmul_check(DisasContext *s, arg_rmr *a)
 }
 
 GEN_V_UNMASKED_TRANS(vgmul_vv, vgmul_check)
+
+static bool vghsh_check(DisasContext *s, arg_rmrr *a)
+{
+    return s->cfg_ptr->ext_zvkg == true &&
+            opivv_check(s, a) &&
+            MAXSZ(s) >= (128 / 8) && /* EGW in bytes */
+            s->vstart % 4 == 0 && s->sew == MO_32;
+}
+
+GEN_VV_UNMASKED_TRANS(vghsh_vv, vghsh_check, 4)
target/riscv/vcrypto_helper.c +38 -0
index eb70e2e26c..fe9b05253d 100644
@@ -801,6 +801,44 @@ void HELPER(vsm3c_vi)(void *vd_vptr, void *vs2_vptr, uint32_t uimm,
     env->vstart = 0;
 }
 
+void HELPER(vghsh_vv)(void *vd_vptr, void *vs1_vptr, void *vs2_vptr,
+                       CPURISCVState *env, uint32_t desc)
+{
+    uint64_t *vd = vd_vptr;
+    uint64_t *vs1 = vs1_vptr;
+    uint64_t *vs2 = vs2_vptr;
+    uint32_t vta = vext_vta(desc);
+    uint32_t total_elems = vext_get_total_elems(env, desc, 4);
+
+    for (uint32_t i = env->vstart / 4; i < env->vl / 4; i++) {
+        uint64_t Y[2] = {vd[i * 2 + 0], vd[i * 2 + 1]};
+        uint64_t H[2] = {brev8(vs2[i * 2 + 0]), brev8(vs2[i * 2 + 1])};
+        uint64_t X[2] = {vs1[i * 2 + 0], vs1[i * 2 + 1]};
+        uint64_t Z[2] = {0, 0};
+
+        uint64_t S[2] = {brev8(Y[0] ^ X[0]), brev8(Y[1] ^ X[1])};
+
+        for (uint j = 0; j < 128; j++) {
+            if ((S[j / 64] >> (j % 64)) & 1) {
+                Z[0] ^= H[0];
+                Z[1] ^= H[1];
+            }
+            bool reduce = ((H[1] >> 63) & 1);
+            H[1] = H[1] << 1 | H[0] >> 63;
+            H[0] = H[0] << 1;
+            if (reduce) {
+                H[0] ^= 0x87;
+            }
+        }
+
+        vd[i * 2 + 0] = brev8(Z[0]);
+        vd[i * 2 + 1] = brev8(Z[1]);
+    }
+    /* set tail elements to 1s */
+    vext_set_elems_1s(vd, vta, env->vl * 4, total_elems * 4);
+    env->vstart = 0;
+}
+
 void HELPER(vgmul_vv)(void *vd_vptr, void *vs2_vptr,
                        CPURISCVState *env, uint32_t desc)
 {
Signature · 1 line
2.39.2
#41 lore raw

PATCH 40/45 target/riscv: Expose zvkg cpu property

In reply to Lawrence Hunter
---
 target/riscv/cpu.c | 1 +
 1 file changed, 1 insertion(+)
target/riscv/cpu.c +1 -0
index 79079d517d..323e0c462b 100644
@@ -1484,6 +1484,7 @@ static Property riscv_cpu_extensions[] = {
 
     /* Vector cryptography extensions */
     DEFINE_PROP_BOOL("x-zvkb", RISCVCPU, cfg.ext_zvkb, false),
+    DEFINE_PROP_BOOL("x-zvkg", RISCVCPU, cfg.ext_zvkg, false),
     DEFINE_PROP_BOOL("x-zvkned", RISCVCPU, cfg.ext_zvkned, false),
     DEFINE_PROP_BOOL("x-zvknha", RISCVCPU, cfg.ext_zvknha, false),
     DEFINE_PROP_BOOL("x-zvknhb", RISCVCPU, cfg.ext_zvknhb, false),
Signature · 1 line
2.39.2
#42 lore raw

PATCH 41/45 crypto: Create sm4_subword

In reply to Lawrence Hunter

From: Max Chou <[email protected]>

- Share sm4_subword between different targets.

---
 include/crypto/sm4.h           |  8 ++++++++
 target/arm/tcg/crypto_helper.c | 10 ++--------
 2 files changed, 10 insertions(+), 8 deletions(-)
include/crypto/sm4.h +8 -0
index 9bd3ebc62e..de8245d8a7 100644
@@ -3,4 +3,12 @@
 
 extern const uint8_t sm4_sbox[256];
 
+static inline uint32_t sm4_subword(uint32_t word)
+{
+    return sm4_sbox[word & 0xff] |
+           sm4_sbox[(word >> 8) & 0xff] << 8 |
+           sm4_sbox[(word >> 16) & 0xff] << 16 |
+           sm4_sbox[(word >> 24) & 0xff] << 24;
+}
+
 #endif
target/arm/tcg/crypto_helper.c +2 -8
index d28690321f..58e6c4f779 100644
@@ -707,10 +707,7 @@ static void do_crypto_sm4e(uint64_t *rd, uint64_t *rn, uint64_t *rm)
             CR_ST_WORD(d, (i + 3) % 4) ^
             CR_ST_WORD(n, i);
 
-        t = sm4_sbox[t & 0xff] |
-            sm4_sbox[(t >> 8) & 0xff] << 8 |
-            sm4_sbox[(t >> 16) & 0xff] << 16 |
-            sm4_sbox[(t >> 24) & 0xff] << 24;
+        t = sm4_subword(t);
 
         CR_ST_WORD(d, i) ^= t ^ rol32(t, 2) ^ rol32(t, 10) ^ rol32(t, 18) ^
                             rol32(t, 24);
@@ -744,10 +741,7 @@ static void do_crypto_sm4ekey(uint64_t *rd, uint64_t *rn, uint64_t *rm)
             CR_ST_WORD(d, (i + 3) % 4) ^
             CR_ST_WORD(m, i);
 
-        t = sm4_sbox[t & 0xff] |
-            sm4_sbox[(t >> 8) & 0xff] << 8 |
-            sm4_sbox[(t >> 16) & 0xff] << 16 |
-            sm4_sbox[(t >> 24) & 0xff] << 24;
+        t = sm4_subword(t);
 
         CR_ST_WORD(d, i) ^= t ^ rol32(t, 13) ^ rol32(t, 23);
     }
Signature · 1 line
2.39.2
#43 lore raw

PATCH 42/45 crypto: Add SM4 constant parameter CK

In reply to Lawrence Hunter

From: Max Chou <[email protected]>

---
 crypto/sm4.c         | 10 ++++++++++
 include/crypto/sm4.h |  1 +
 2 files changed, 11 insertions(+)
crypto/sm4.c +10 -0
index 9f0cd452c7..2987306cf7 100644
@@ -47,3 +47,13 @@ uint8_t const sm4_sbox[] = {
     0x79, 0xee, 0x5f, 0x3e, 0xd7, 0xcb, 0x39, 0x48,
 };
 
+uint32_t const sm4_ck[] = {
+    0x00070e15, 0x1c232a31, 0x383f464d, 0x545b6269,
+    0x70777e85, 0x8c939aa1, 0xa8afb6bd, 0xc4cbd2d9,
+    0xe0e7eef5, 0xfc030a11, 0x181f262d, 0x343b4249,
+    0x50575e65, 0x6c737a81, 0x888f969d, 0xa4abb2b9,
+    0xc0c7ced5, 0xdce3eaf1, 0xf8ff060d, 0x141b2229,
+    0x30373e45, 0x4c535a61, 0x686f767d, 0x848b9299,
+    0xa0a7aeb5, 0xbcc3cad1, 0xd8dfe6ed, 0xf4fb0209,
+    0x10171e25, 0x2c333a41, 0x484f565d, 0x646b7279
+};
include/crypto/sm4.h +1 -0
index de8245d8a7..382b26d922 100644
@@ -2,6 +2,7 @@
 #define QEMU_SM4_H
 
 extern const uint8_t sm4_sbox[256];
+extern const uint32_t sm4_ck[32];
 
 static inline uint32_t sm4_subword(uint32_t word)
 {
Signature · 1 line
2.39.2
#44 lore raw

PATCH 43/45 target/riscv: Add zvksed cfg property

In reply to Lawrence Hunter

From: Max Chou <[email protected]>

---
 target/riscv/cpu.c | 3 ++-
 target/riscv/cpu.h | 1 +
 2 files changed, 3 insertions(+), 1 deletion(-)
target/riscv/cpu.c +2 -1
index 323e0c462b..84a225bf5f 100644
@@ -114,6 +114,7 @@ static const struct isa_ext_data isa_edata_arr[] = {
     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(zvksed, true, PRIV_VERSION_1_12_0, ext_zvksed),
     ISA_EXT_DATA_ENTRY(zvksh, true, PRIV_VERSION_1_12_0, ext_zvksh),
     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),
@@ -1222,7 +1223,7 @@ static void riscv_cpu_realize(DeviceState *dev, Error **errp)
     * in qemu
     */
     if ((cpu->cfg.ext_zvkb || cpu->cfg.ext_zvkg || cpu->cfg.ext_zvkned ||
-             cpu->cfg.ext_zvknha || cpu->cfg.ext_zvksh) &&
+             cpu->cfg.ext_zvknha || cpu->cfg.ext_zvksed || cpu->cfg.ext_zvksh) &&
         !(cpu->cfg.ext_zve32f || cpu->cfg.ext_zve64f ||
             cpu->cfg.ext_zve64d || cpu->cfg.ext_v)) {
         error_setg(
target/riscv/cpu.h +1 -0
index 40c4e23209..55bbc4375a 100644
@@ -475,6 +475,7 @@ struct RISCVCPUConfig {
     bool ext_zvkned;
     bool ext_zvknha;
     bool ext_zvknhb;
+    bool ext_zvksed;
     bool ext_zvksh;
     bool ext_zmmul;
     bool ext_zvfh;
Signature · 1 line
2.39.2
#45 lore raw

PATCH 44/45 target/riscv: Add Zvksed support

In reply to Lawrence Hunter

From: Max Chou <[email protected]>

- add vsm4k, vsm4r instructions

[[email protected]: Moved SM4 functions from crypto_helper.c to vcrypto_helper.c] [[email protected]: Added alignment checks, refactored code to use macros, and minor style changes]

---
 target/riscv/helper.h                        |   4 +
 target/riscv/insn32.decode                   |   5 +
 target/riscv/insn_trans/trans_rvzvksed.c.inc |  57 +++++++++
 target/riscv/translate.c                     |   1 +
 target/riscv/vcrypto_helper.c                | 127 +++++++++++++++++++
 5 files changed, 194 insertions(+)
 create mode 100644 target/riscv/insn_trans/trans_rvzvksed.c.inc
target/riscv/helper.h +4 -0
index 3c4aa4b5df..4e71738b38 100644
@@ -1209,3 +1209,7 @@ DEF_HELPER_5(vsm3c_vi, void, ptr, ptr, i32, env, i32)
 
 DEF_HELPER_5(vghsh_vv, void, ptr, ptr, ptr, env, i32)
 DEF_HELPER_4(vgmul_vv, void, ptr, ptr, env, i32)
+
+DEF_HELPER_5(vsm4k_vi, void, ptr, ptr, i32, env, i32)
+DEF_HELPER_4(vsm4r_vv, void, ptr, ptr, env, i32)
+DEF_HELPER_4(vsm4r_vs, void, ptr, ptr, env, i32)
target/riscv/insn32.decode +5 -0
index 856e088bad..543e58ef18 100644
@@ -952,3 +952,8 @@ vsm3c_vi        101011 1 ..... ..... 010 ..... 1110111 @r_vm_1
 # *** RV64 Zvkg vector crypto extension ***
 vghsh_vv        101100 1 ..... ..... 010 ..... 1110111 @r_vm_1
 vgmul_vv        101000 1 ..... 10001 010 ..... 1110111 @r2_vm_1
+
+# *** RV64 Zvksed vector crypto extension ***
+vsm4k_vi        100001 1 ..... ..... 010 ..... 1110111 @r_vm_1
+vsm4r_vv        101000 1 ..... 10000 010 ..... 1110111 @r2_vm_1
+vsm4r_vs        101001 1 ..... 10000 010 ..... 1110111 @r2_vm_1
target/riscv/insn_trans/trans_rvzvksed.c.inc new file +57 -0

57 lines not on this page · raw message

target/riscv/translate.c +1 -0
index fdb5c3364e..521bc2e3a9 100644
@@ -1088,6 +1088,7 @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc)
 #include "insn_trans/trans_rvzvknh.c.inc"
 #include "insn_trans/trans_rvzvksh.c.inc"
 #include "insn_trans/trans_rvzvkg.c.inc"
+#include "insn_trans/trans_rvzvksed.c.inc"
 #include "insn_trans/trans_privileged.c.inc"
 #include "insn_trans/trans_svinval.c.inc"
 #include "decode-xthead.c.inc"
target/riscv/vcrypto_helper.c +127 -0

136 lines not on this page · raw message

Signature · 1 line
2.39.2
#46 lore raw

PATCH 45/45 target/riscv: Expose Zvksed property

In reply to Lawrence Hunter

From: Max Chou <[email protected]>

---
 target/riscv/cpu.c | 1 +
 1 file changed, 1 insertion(+)
target/riscv/cpu.c +1 -0
index 84a225bf5f..8caa485f28 100644
@@ -1489,6 +1489,7 @@ static Property riscv_cpu_extensions[] = {
     DEFINE_PROP_BOOL("x-zvkned", RISCVCPU, cfg.ext_zvkned, false),
     DEFINE_PROP_BOOL("x-zvknha", RISCVCPU, cfg.ext_zvknha, false),
     DEFINE_PROP_BOOL("x-zvknhb", RISCVCPU, cfg.ext_zvknhb, false),
+    DEFINE_PROP_BOOL("x-zvksed", RISCVCPU, cfg.ext_zvksed, false),
     DEFINE_PROP_BOOL("x-zvksh", RISCVCPU, cfg.ext_zvksh, false),
 
     DEFINE_PROP_END_OF_LIST(),
Signature · 1 line
2.39.2
#47 lore raw
In reply to Lawrence Hunter
Quoted 6 lines · Lawrence Hunter

This patchset provides an implementation for Zvkb, Zvkned, Zvknh, Zvksh, Zvkg, and Zvksed of the draft RISC-V vector cryptography extensions as per the 20230303 version of the specification(1) (1fcbb30). Please note that the Zvkt data-independent execution latency extension has not been implemented, and we would recommend not using these patches in an environment where timing attacks are an issue.

Work performed by Dickon, Lawrence, Nazar, Kiran, and William from Codethink sponsored by SiFive, as well as Max Chou and Frank Chang from SiFive.

For convenience we have created a git repo with our patches on top of a recent master. https://github.com/CodethinkLabs/qemu-ct

I did test and review this patchset.
Since most of my comments affect multiple patches I have summarized them here in one email.
Observations that only affect a single patch will be sent in response to the corresponding email.

I have tested this series with the OpenSSL PR for Zvk that can be found here:
https://github.com/openssl/openssl/pull/20149
I ran with all Zvk* extensions enabled (using Zvkg for GCM) and with Zvkb only (using Zvkb for GCM).
All tests succeed. Note, however, that the test coverage is limited (e.g. no .vv instructions, vstart is always zero).

When sending out a follow-up version (even if it just introduces a minimal fix), then consider using patchset versioning (e.g. git format-patch -v2 ...).

It might be a matter of taste, but I would prefer a series that groups and orders the commits differently:
a) independent changes to the existing code (refactoring only, but
no new features) - one commit per topic
b) introduction of new functionality - one commit per extension
A series using such a commit granularity and order would be easier to maintain and review (and not result in 45 patches).
Also, the refactoring changes could land before Zvk freezes if maintainers decide to do so.

So far all translation files in target/riscv/insn_trans/* contain multiple extensions if they are related.
I think we should follow this pattern and use a common trans_zvk.c.inc file.

All patches to insn32.decode have comments of the form "RV64 Zvk* vector crypto extension".
What is the point of the "RV64"? I would simply remove that.

All instructions set "env->vstart = 0;" at the end.
I don't think that this is correct (the specification does not require this).

The tests of the reserved encodings are not consistent:
* Zvknh does a dynamic test (query tcg_gen_*())
* Zvkned does a dynamic test (tcg_gen_*())
* Zvkg does not test for (vl%EGS == 0)
The vl CSR can only be updated by the vset{i}vl{i} instructions. The same applies to the vstart CSR and the vtype CSR that holds vsew, vlmul and other fields.
The current code tests the VSTART/SEW value using "s->vstart % 4 == 0"/"s->sew == MO_32".
Why is it not possible to do the same with VL, i.e. "s->vl % 4 == 0" (after adding it to DisasContext)?
Also, I would introduce named constants or macros for the EGS values to avoid magic constants in the code
(some extensions do that - e.g. ZVKSED_EGS).

BR
Christoph

Quoted 119 lines

1. https://github.com/riscv/riscv-crypto/releases

Dickon Hood (2):
qemu/bitops.h: Limit rotate amounts
target/riscv: Add vrol.[vv,vx] and vror.[vv,vx,vi] decoding,
translation and execution support

Kiran Ostrolenk (8):
target/riscv: Refactor some of the generic vector functionality target/riscv: Refactor some of the generic vector functionality target/riscv: Refactor some of the generic vector functionality target/riscv: Refactor some of the generic vector functionality target/riscv: Add vsha2ms.vv decoding, translation and execution
support
target/riscv: Add zvksh cpu property
target/riscv: Add vsm3c.vi decoding, translation and execution support target/riscv: Expose zvksh cpu property

Lawrence Hunter (17):
target/riscv: Add vclmul.vv decoding, translation and execution
support
target/riscv: Add vclmul.vx decoding, translation and execution
support
target/riscv: Add vclmulh.vv decoding, translation and execution
support
target/riscv: Add vclmulh.vx decoding, translation and execution
support
target/riscv: Add vaesef.vv decoding, translation and execution
support
target/riscv: Add vaesef.vs decoding, translation and execution
support
target/riscv: Add vaesdf.vv decoding, translation and execution
support
target/riscv: Add vaesdf.vs decoding, translation and execution
support
target/riscv: Add vaesdm.vv decoding, translation and execution
support
target/riscv: Add vaesdm.vs decoding, translation and execution
support
target/riscv: Add vaesz.vs decoding, translation and execution support target/riscv: Add vsha2c[hl].vv decoding, translation and execution
support
target/riscv: Add vsm3me.vv decoding, translation and execution
support
target/riscv: Add zvkg cpu property
target/riscv: Add vgmul.vv decoding, translation and execution support target/riscv: Add vghsh.vv decoding, translation and execution support target/riscv: Expose zvkg cpu property

Max Chou (5):
crypto: Create sm4_subword
crypto: Add SM4 constant parameter CK
target/riscv: Add zvksed cfg property
target/riscv: Add Zvksed support
target/riscv: Expose Zvksed property

Nazar Kazakov (10):
target/riscv: Add zvkb cpu property
target/riscv: Add vrev8.v decoding, translation and execution support target/riscv: Add vandn.[vv,vx] decoding, translation and execution
support
target/riscv: Expose zvkb cpu property
target/riscv: Add zvkned cpu property
target/riscv: Add vaeskf1.vi decoding, translation and execution
support
target/riscv: Add vaeskf2.vi decoding, translation and execution
support
target/riscv: Expose zvkned cpu property
target/riscv: Add zvknh cpu properties
target/riscv: Expose zvknh cpu properties

William Salmon (3):
target/riscv: Add vbrev8.v decoding, translation and execution support target/riscv: Add vaesem.vv decoding, translation and execution
support
target/riscv: Add vaesem.vs decoding, translation and execution
support

 accel/tcg/tcg-runtime-gvec.c                 |   11 +
 accel/tcg/tcg-runtime.h                      |    1 +
 crypto/sm4.c                                 |   10 +
 include/crypto/sm4.h                         |    9 +
 include/qemu/bitops.h                        |   24 +-
 target/arm/tcg/crypto_helper.c               |   10 +-
 target/riscv/cpu.c                           |   36 +
 target/riscv/cpu.h                           |    7 +
 target/riscv/helper.h                        |   71 ++
 target/riscv/insn32.decode                   |   49 +
 target/riscv/insn_trans/trans_rvv.c.inc      |   93 +-
 target/riscv/insn_trans/trans_rvzvkb.c.inc   |  220 ++++
 target/riscv/insn_trans/trans_rvzvkg.c.inc   |   40 +
 target/riscv/insn_trans/trans_rvzvkned.c.inc |  170 +++
 target/riscv/insn_trans/trans_rvzvknh.c.inc  |   84 ++
 target/riscv/insn_trans/trans_rvzvksed.c.inc |   57 +
 target/riscv/insn_trans/trans_rvzvksh.c.inc  |   43 +
 target/riscv/meson.build                     |    4 +-
 target/riscv/op_helper.c                     |    5 +
 target/riscv/translate.c                     |    6 +
 target/riscv/vcrypto_helper.c                | 1001 ++++++++++++++++++
 target/riscv/vector_helper.c                 |  240 +----
 target/riscv/vector_internals.c              |   81 ++
 target/riscv/vector_internals.h              |  222 ++++
 24 files changed, 2192 insertions(+), 302 deletions(-)
 create mode 100644 target/riscv/insn_trans/trans_rvzvkb.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvkg.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvkned.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvknh.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvksed.c.inc
 create mode 100644 target/riscv/insn_trans/trans_rvzvksh.c.inc
 create mode 100644 target/riscv/vcrypto_helper.c
 create mode 100644 target/riscv/vector_internals.c
 create mode 100644 target/riscv/vector_internals.h

--
2.39.2

#48 lore raw
Quoted 16 lines · Christoph Müllner
Quoted 6 lines · Lawrence Hunter

This patchset provides an implementation for Zvkb, Zvkned, Zvknh, Zvksh, Zvkg, and Zvksed of the draft RISC-V vector cryptography extensions as per the 20230303 version of the specification(1) (1fcbb30). Please note that the Zvkt data-independent execution latency extension has not been implemented, and we would recommend not using these patches in an environment where timing attacks are an issue.

Work performed by Dickon, Lawrence, Nazar, Kiran, and William from Codethink sponsored by SiFive, as well as Max Chou and Frank Chang from SiFive.

For convenience we have created a git repo with our patches on top of a recent master. https://github.com/CodethinkLabs/qemu-ct

I did test and review this patchset.
Since most of my comments affect multiple patches I have summarized them here in one email.
Observations that only affect a single patch will be sent in response to the corresponding email.

I have tested this series with the OpenSSL PR for Zvk that can be found here:

Quoted 8 lines
   https://github.com/openssl/openssl/pull/20149

I ran with all Zvk* extensions enabled (using Zvkg for GCM) and with Zvkb only (using Zvkb for GCM).
All tests succeed. Note, however, that the test coverage is limited (e.g. no .vv instructions, vstart is always zero).

When sending out a follow-up version (even if it just introduces a minimal fix), then consider using patchset versioning (e.g. git format-patch -v2 ...).

Ok, will do

Quoted 9 lines

It might be a matter of taste, but I would prefer a series that groups and orders the commits differently:
a) independent changes to the existing code (refactoring only, but
no new features) - one commit per topic
b) introduction of new functionality - one commit per extension
A series using such a commit granularity and order would be easier to maintain and review (and not result in 45 patches).
Also, the refactoring changes could land before Zvk freezes if maintainers decide to do so.

Makes sense, will do

Quoted 3 lines

So far all translation files in target/riscv/insn_trans/* contain multiple extensions if they are related.
I think we should follow this pattern and use a common trans_zvk.c.inc file.

Agree, will do

Quoted 3 lines

All patches to insn32.decode have comments of the form "RV64 Zvk* vector crypto extension".
What is the point of the "RV64"? I would simply remove that.

Ok, will remove it

Quoted 2 lines

All instructions set "env->vstart = 0;" at the end.
I don't think that this is correct (the specification does not require this).

That's from vector spec: "All vector instructions are defined to begin execution with the element number given in the vstart CSR, leaving earlier elements in the destination vector undisturbed, and to reset the vstart CSR to zero at the end of execution." - from "3.7. Vector Start Index CSR vstart"

Quoted 4 lines

The tests of the reserved encodings are not consistent:
* Zvknh does a dynamic test (query tcg_gen_*())
* Zvkned does a dynamic test (tcg_gen_*())
* Zvkg does not test for (vl%EGS == 0)

Zvkg also does dynamic test, by calling macros GEN_V_UNMASKED_TRANS and GEN_VV_UNMASKED_TRANS

Quoted 7 lines

The vl CSR can only be updated by the vset{i}vl{i} instructions. The same applies to the vstart CSR and the vtype CSR that holds vsew, vlmul and other fields.
The current code tests the VSTART/SEW value using "s->vstart % 4 == 0"/"s->sew == MO_32".
Why is it not possible to do the same with VL, i.e. "s->vl % 4 == 0" (after adding it to DisasContext)?

vl can also be updated by another instruction - from vector spec "3.5. Vector Length Register vl" - "The XLEN-bit-wide read-only vl CSR can only be updated by the vset{i}vl{i} instructions, and the fault-only-first vector load instruction variants." So just because of that fault-only-first instruction we need dynamic checks.

vstart is just another CSR -- software can write to it, but probably shouldn't. Whether that's ever going to be useful outside testing ISA conformance tests or not I don't know, but it's clearly read-write (also section 3.7).

Quoted 3 lines

Also, I would introduce named constants or macros for the EGS values to avoid magic constants in the code
(some extensions do that - e.g. ZVKSED_EGS).

Makes sense, will do

Best,
Lawrence

#49 lore raw
Quoted 70 lines · Lawrence Hunter
Quoted 32 lines · Christoph Müllner
Quoted 9 lines · Lawrence Hunter

This patchset provides an implementation for Zvkb, Zvkned, Zvknh, Zvksh, Zvkg, and Zvksed of the draft RISC-V vector cryptography extensions as per the 20230303 version of the specification(1) (1fcbb30). Please note that the Zvkt data-independent execution latency extension has not been implemented, and we would recommend not using these patches in an environment where timing attacks are an issue.

Quoted 6 lines

Work performed by Dickon, Lawrence, Nazar, Kiran, and William from Codethink sponsored by SiFive, as well as Max Chou and Frank Chang from SiFive.

For convenience we have created a git repo with our patches on top of a recent master. https://github.com/CodethinkLabs/qemu-ct

I did test and review this patchset.
Since most of my comments affect multiple patches I have summarized them here in one email.
Observations that only affect a single patch will be sent in response to the corresponding email.

I have tested this series with the OpenSSL PR for Zvk that can be found here:

   https://github.com/openssl/openssl/pull/20149

I ran with all Zvk* extensions enabled (using Zvkg for GCM) and with Zvkb only (using Zvkb for GCM).
All tests succeed. Note, however, that the test coverage is limited (e.g. no .vv instructions, vstart is always zero).

Quoted 4 lines

When sending out a follow-up version (even if it just introduces a minimal fix),
then consider using patchset versioning (e.g. git format-patch -v2 ...).

Ok, will do

Quoted 9 lines

It might be a matter of taste, but I would prefer a series that groups and orders the commits differently:
a) independent changes to the existing code (refactoring only, but
no new features) - one commit per topic
b) introduction of new functionality - one commit per extension
A series using such a commit granularity and order would be easier to maintain and review (and not result in 45 patches).
Also, the refactoring changes could land before Zvk freezes if maintainers decide to do so.

Makes sense, will do

Quoted 4 lines

So far all translation files in target/riscv/insn_trans/* contain multiple extensions if they are related.
I think we should follow this pattern and use a common trans_zvk.c.inc file.

Agree, will do

Quoted 3 lines

All patches to insn32.decode have comments of the form "RV64 Zvk* vector crypto extension".
What is the point of the "RV64"? I would simply remove that.

Ok, will remove it

Quoted 3 lines

All instructions set "env->vstart = 0;" at the end.
I don't think that this is correct (the specification does not require this).

Quoted 5 lines

That's from vector spec: "All vector instructions are defined to begin execution with the element number given in the vstart CSR, leaving earlier elements in the destination vector undisturbed, and to reset the vstart CSR to zero at the end of execution." - from "3.7. Vector Start Index CSR vstart"

Yes, you are right.
I just created a PR for the Zvk* spec to clarify this:
https://github.com/riscv/riscv-crypto/pull/308

Quoted 36 lines
Quoted 5 lines

The tests of the reserved encodings are not consistent:
* Zvknh does a dynamic test (query tcg_gen_*())
* Zvkned does a dynamic test (tcg_gen_*())
* Zvkg does not test for (vl%EGS == 0)

Zvkg also does dynamic test, by calling macros GEN_V_UNMASKED_TRANS and GEN_VV_UNMASKED_TRANS

Quoted 7 lines

The vl CSR can only be updated by the vset{i}vl{i} instructions. The same applies to the vstart CSR and the vtype CSR that holds vsew, vlmul and other fields.
The current code tests the VSTART/SEW value using "s->vstart % 4 == 0"/"s->sew == MO_32".
Why is it not possible to do the same with VL, i.e. "s->vl % 4 == 0" (after adding it to DisasContext)?

vl can also be updated by another instruction - from vector spec "3.5. Vector Length Register vl" - "The XLEN-bit-wide read-only vl CSR can only be updated by the vset{i}vl{i} instructions, and the fault-only-first vector load instruction variants." So just because of that fault-only-first instruction we need dynamic checks.

vstart is just another CSR -- software can write to it, but probably shouldn't. Whether that's ever going to be useful outside testing ISA conformance tests or not I don't know, but it's clearly read-write (also section 3.7).

Quoted 3 lines

Also, I would introduce named constants or macros for the EGS values to avoid magic constants in the code
(some extensions do that - e.g. ZVKSED_EGS).

Makes sense, will do

Best,
Lawrence