Re: [PATCH v6 58/61] target/riscv: vector slide instructions

qemu-devel

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH v6 58/61] target/riscv: vector slide instructions

From:	LIU Zhiwei
Subject:	Re: [PATCH v6 58/61] target/riscv: vector slide instructions
Date:	Sat, 28 Mar 2020 21:40:09 +0800
User-agent:	Mozilla/5.0 (Windows NT 10.0; WOW64; rv:68.0) Gecko/20100101 Thunderbird/68.6.0

On 2020/3/17 23:06, LIU Zhiwei wrote:

Signed-off-by: LIU Zhiwei <address@hidden>
---
 target/riscv/helper.h                   |  17 ++++
 target/riscv/insn32.decode              |   7 ++
 target/riscv/insn_trans/trans_rvv.inc.c |  17 ++++
 target/riscv/vector_helper.c            | 128 ++++++++++++++++++++++++
 4 files changed, 169 insertions(+)

diff --git a/target/riscv/helper.h b/target/riscv/helper.h
index 044538aef9..3b1612012c 100644
--- a/target/riscv/helper.h
+++ b/target/riscv/helper.h
@@ -1118,3 +1118,20 @@ DEF_HELPER_3(vmv_s_x_d, void, ptr, tl, env)
 DEF_HELPER_3(vfmv_s_f_h, void, ptr, i64, env)
 DEF_HELPER_3(vfmv_s_f_w, void, ptr, i64, env)
 DEF_HELPER_3(vfmv_s_f_d, void, ptr, i64, env)
+
+DEF_HELPER_6(vslideup_vx_b, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslideup_vx_h, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslideup_vx_w, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslideup_vx_d, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslidedown_vx_b, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslidedown_vx_h, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslidedown_vx_w, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslidedown_vx_d, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslide1up_vx_b, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslide1up_vx_h, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslide1up_vx_w, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslide1up_vx_d, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslide1down_vx_b, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslide1down_vx_h, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslide1down_vx_w, void, ptr, ptr, tl, ptr, env, i32)
+DEF_HELPER_6(vslide1down_vx_d, void, ptr, ptr, tl, ptr, env, i32)
diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 79f9b37b29..34ccad53a9 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -72,6 +72,7 @@
 @r2_vm   ...... vm:1 ..... ..... ... ..... ....... &rmr %rs2 %rd
 @r1_vm   ...... vm:1 ..... ..... ... ..... ....... %rd
 @r_nfvm  ... ... vm:1 ..... ..... ... ..... ....... &rnfvm %nf %rs2 %rs1 %rd
+@r2rd    .......   ..... ..... ... ..... ....... %rs2 %rd
 @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
@@ -565,6 +566,12 @@ vext_x_v        001100 1 ..... ..... 010 ..... 1010111 @r
 vmv_s_x         001101 1 00000 ..... 110 ..... 1010111 @r2
 vfmv_f_s        001100 1 ..... 00000 001 ..... 1010111 @r2rd
 vfmv_s_f        001101 1 00000 ..... 101 ..... 1010111 @r2
+vslideup_vx     001110 . ..... ..... 100 ..... 1010111 @r_vm
+vslideup_vi     001110 . ..... ..... 011 ..... 1010111 @r_vm
+vslide1up_vx    001110 . ..... ..... 110 ..... 1010111 @r_vm
+vslidedown_vx   001111 . ..... ..... 100 ..... 1010111 @r_vm
+vslidedown_vi   001111 . ..... ..... 011 ..... 1010111 @r_vm
+vslide1down_vx  001111 . ..... ..... 110 ..... 1010111 @r_vm
 
 vsetvli         0 ........... ..... 111 ..... 1010111  @r2_zimm
 vsetvl          1000000 ..... ..... 111 ..... 1010111  @r
diff --git a/target/riscv/insn_trans/trans_rvv.inc.c b/target/riscv/insn_trans/trans_rvv.inc.c
index 07033662c3..10482fd1d4 100644
--- a/target/riscv/insn_trans/trans_rvv.inc.c
+++ b/target/riscv/insn_trans/trans_rvv.inc.c
@@ -2536,3 +2536,20 @@ static bool trans_vfmv_s_f(DisasContext *s, arg_vfmv_s_f *a)
     }
     return false;
 }
+
+/* Vector Slide Instructions */
+static bool slideup_check(DisasContext *s, arg_rmrr *a)
+{
+    return (vext_check_isa_ill(s) &&
+            vext_check_overlap_mask(s, a->rd, a->vm, true) &&
+            vext_check_reg(s, a->rd, false) &&
+            vext_check_reg(s, a->rs2, false) &&
+            (a->rd != a->rs2));
+}
+GEN_OPIVX_TRANS(vslideup_vx, slideup_check)
+GEN_OPIVX_TRANS(vslide1up_vx, slideup_check)
+GEN_OPIVI_TRANS(vslideup_vi, 1, vslideup_vx, slideup_check)
+
+GEN_OPIVX_TRANS(vslidedown_vx, opivx_check)
+GEN_OPIVX_TRANS(vslide1down_vx, opivx_check)
+GEN_OPIVI_TRANS(vslidedown_vi, 1, vslidedown_vx, opivx_check)
diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
index 723e15a670..b0439ac3d1 100644
--- a/target/riscv/vector_helper.c
+++ b/target/riscv/vector_helper.c
@@ -4706,3 +4706,131 @@ void HELPER(NAME)(void *vd, uint64_t s1, CPURISCVState *env)            \
 GEN_VEXT_VFMV_S_F(vfmv_s_f_h, uint16_t, H2, clearh)
 GEN_VEXT_VFMV_S_F(vfmv_s_f_w, uint32_t, H4, clearl)
 GEN_VEXT_VFMV_S_F(vfmv_s_f_d, uint64_t, H8, clearq)
+
+/* Vector Slide Instructions */
+#define GEN_VEXT_VSLIDEUP_VX(NAME, ETYPE, H, CLEAR_FN)                    \
+void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
+        CPURISCVState *env, uint32_t desc)                                \
+{                                                                         \
+    uint32_t mlen = vext_mlen(desc);                                      \
+    uint32_t vlmax = env_archcpu(env)->cfg.vlen / mlen;                   \
+    uint32_t vm = vext_vm(desc);                                          \
+    uint32_t vl = env->vl;                                                \
+    target_ulong offset = s1, i;                                          \
+                                                                          \
+    if (vl == 0) {                                                        \
+        return;                                                           \
+    }                                                                     \
+    for (i = offset; i < vl; i++) {                                       \
+        if (!vm && !vext_elem_mask(v0, mlen, i)) {                        \
+            continue;                                                     \
+        }                                                                 \
+        *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i - offset));          \
+    }                                                                     \
+    CLEAR_FN(vd, vl, vl * sizeof(ETYPE), vlmax * sizeof(ETYPE));          \
+}
+
+/* vslideup.vx vd, vs2, rs1, vm # vd[i+rs1] = vs2[i] */
+GEN_VEXT_VSLIDEUP_VX(vslideup_vx_b, uint8_t, H1, clearb)
+GEN_VEXT_VSLIDEUP_VX(vslideup_vx_h, uint16_t, H2, clearh)
+GEN_VEXT_VSLIDEUP_VX(vslideup_vx_w, uint32_t, H4, clearl)
+GEN_VEXT_VSLIDEUP_VX(vslideup_vx_d, uint64_t, H8, clearq)
+
+#define GEN_VEXT_VSLIDEDOWN_VX(NAME, ETYPE, H, CLEAR_FN)                  \
+void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
+        CPURISCVState *env, uint32_t desc)                                \
+{                                                                         \
+    uint32_t mlen = vext_mlen(desc);                                      \
+    uint32_t vlmax = env_archcpu(env)->cfg.vlen / mlen;                   \
+    uint32_t vm = vext_vm(desc);                                          \
+    uint32_t vl = env->vl;                                                \
+    target_ulong offset = s1, i, max;                                     \
+                                                                          \
+    if (vl == 0) {                                                        \
+        return;                                                           \
+    }                                                                     \
+    if (offset >= vlmax) {                                                \
+        max = 0;                                                          \
+    } else {                                                              \
+        max = MIN(vl, vlmax - offset);                                    \
+    }                                                                     \
+    for (i = 0; i < max; ++i) {                                           \
+        if (!vm && !vext_elem_mask(v0, mlen, i)) {                        \
+            continue;                                                     \
+        }                                                                 \
+        *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i + offset));          \
+    }                                                                     \
+    CLEAR_FN(vd, max, max * sizeof(ETYPE), vlmax * sizeof(ETYPE));        \
+}
+

There is a mistake when vlmax - offset < vl.

The elements between (vlmax - offset) and vl should be unchanged(masked) or zeroed(unmasked).
However, in this implementation, these elements are always zeroed.

I will fix it in v7 like

#define GEN_VEXT_VSLIDEDOWN_VX(NAME, ETYPE, H, CLEAR_FN)                  \
void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
        CPURISCVState *env, uint32_t desc)                                \
{                                                                         \
    uint32_t mlen = vext_mlen(desc);                                      \
    uint32_t vlmax = env_archcpu(env)->cfg.vlen / mlen;                   \
    uint32_t vm = vext_vm(desc);                                          \
    uint32_t vl = env->vl;                                                \
    target_ulong offset = s1, i;                                          \
                                                                          \
    if (vl == 0) {                                                        \
        return;                                                           \
    }                                                                     \
    for (i = 0; i < vl; ++i) {                                            \
        target_ulong j = i + offset;                                      \
        if (!vm && !vext_elem_mask(v0, mlen, i)) {                        \
            continue;                                                     \
        }                                                                 \
        *((ETYPE *)vd + H(i)) = j >= vlmax ? 0 : *((ETYPE *)vs2 + H(j));  \
    }                                                                     \
    CLEAR_FN(vd, vl, vl * sizeof(ETYPE), vlmax * sizeof(ETYPE));          \
}

Zhiwei

+/* vslidedown.vx vd, vs2, rs1, vm # vd[i] = vs2[i+rs1] */
+GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_b, uint8_t, H1, clearb)
+GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_h, uint16_t, H2, clearh)
+GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_w, uint32_t, H4, clearl)
+GEN_VEXT_VSLIDEDOWN_VX(vslidedown_vx_d, uint64_t, H8, clearq)
+
+#define GEN_VEXT_VSLIDE1UP_VX(NAME, ETYPE, H, CLEAR_FN)                   \
+void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
+        CPURISCVState *env, uint32_t desc)                                \
+{                                                                         \
+    uint32_t mlen = vext_mlen(desc);                                      \
+    uint32_t vlmax = env_archcpu(env)->cfg.vlen / mlen;                   \
+    uint32_t vm = vext_vm(desc);                                          \
+    uint32_t vl = env->vl;                                                \
+    uint32_t i;                                                           \
+                                                                          \
+    if (vl == 0) {                                                        \
+        return;                                                           \
+    }                                                                     \
+    for (i = 0; i < vl; i++) {                                            \
+        if (!vm && !vext_elem_mask(v0, mlen, i)) {                        \
+            continue;                                                     \
+        }                                                                 \
+        if (i == 0) {                                                     \
+            *((ETYPE *)vd + H(i)) = s1;                                   \
+        } else {                                                          \
+            *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i - 1));           \
+        }                                                                 \
+    }                                                                     \
+    CLEAR_FN(vd, vl, vl * sizeof(ETYPE), vlmax * sizeof(ETYPE));          \
+}
+
+/* vslide1up.vx vd, vs2, rs1, vm # vd[0]=x[rs1], vd[i+1] = vs2[i] */
+GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_b, uint8_t, H1, clearb)
+GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_h, uint16_t, H2, clearh)
+GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_w, uint32_t, H4, clearl)
+GEN_VEXT_VSLIDE1UP_VX(vslide1up_vx_d, uint64_t, H8, clearq)
+
+#define GEN_VEXT_VSLIDE1DOWN_VX(NAME, ETYPE, H, CLEAR_FN)                 \
+void HELPER(NAME)(void *vd, void *v0, target_ulong s1, void *vs2,         \
+        CPURISCVState *env, uint32_t desc)                                \
+{                                                                         \
+    uint32_t mlen = vext_mlen(desc);                                      \
+    uint32_t vlmax = env_archcpu(env)->cfg.vlen / mlen;                   \
+    uint32_t vm = vext_vm(desc);                                          \
+    uint32_t vl = env->vl;                                                \
+    uint32_t i;                                                           \
+                                                                          \
+    if (vl == 0) {                                                        \
+        return;                                                           \
+    }                                                                     \
+    for (i = 0; i < vl; i++) {                                            \
+        if (!vm && !vext_elem_mask(v0, mlen, i)) {                        \
+            continue;                                                     \
+        }                                                                 \
+        if (i == vl - 1) {                                                \
+            *((ETYPE *)vd + H(i)) = s1;                                   \
+        } else {                                                          \
+            *((ETYPE *)vd + H(i)) = *((ETYPE *)vs2 + H(i + 1));           \
+        }                                                                 \
+    }                                                                     \
+    if (i == 0) {                                                         \
+        return;                                                           \
+    }                                                                     \
+    CLEAR_FN(vd, vl, vl * sizeof(ETYPE), vlmax * sizeof(ETYPE));          \
+}
+/* vslide1down.vx vd, vs2, rs1, vm # vd[i] = vs2[i+1], vd[vl-1]=x[rs1] */
+GEN_VEXT_VSLIDE1DOWN_VX(vslide1down_vx_b, uint8_t, H1, clearb)
+GEN_VEXT_VSLIDE1DOWN_VX(vslide1down_vx_h, uint16_t, H2, clearh)
+GEN_VEXT_VSLIDE1DOWN_VX(vslide1down_vx_w, uint32_t, H4, clearl)
+GEN_VEXT_VSLIDE1DOWN_VX(vslide1down_vx_d, uint64_t, H8, clearq)

[Prev in Thread]

Current Thread

[Next in Thread]

[PATCH v6 52/61] target/riscv: set-X-first mask bit, (continued)
- [PATCH v6 52/61] target/riscv: set-X-first mask bit, LIU Zhiwei, 2020/03/17
- [PATCH v6 53/61] target/riscv: vector iota instruction, LIU Zhiwei, 2020/03/17
- [PATCH v6 54/61] target/riscv: vector element index instruction, LIU Zhiwei, 2020/03/17
- [PATCH v6 55/61] target/riscv: integer extract instruction, LIU Zhiwei, 2020/03/17
  - Re: [PATCH v6 55/61] target/riscv: integer extract instruction, Richard Henderson, 2020/03/27
- [PATCH v6 56/61] target/riscv: integer scalar move instruction, LIU Zhiwei, 2020/03/17
- [PATCH v6 57/61] target/riscv: floating-point scalar move instructions, LIU Zhiwei, 2020/03/17
  - Re: [PATCH v6 57/61] target/riscv: floating-point scalar move instructions, Richard Henderson, 2020/03/27
- [PATCH v6 58/61] target/riscv: vector slide instructions, LIU Zhiwei, 2020/03/17
  - Re: [PATCH v6 58/61] target/riscv: vector slide instructions, Richard Henderson, 2020/03/27
  - Re: [PATCH v6 58/61] target/riscv: vector slide instructions, LIU Zhiwei <=
- [PATCH v6 59/61] target/riscv: vector register gather instruction, LIU Zhiwei, 2020/03/17
  - Re: [PATCH v6 59/61] target/riscv: vector register gather instruction, Richard Henderson, 2020/03/27
- [PATCH v6 60/61] target/riscv: vector compress instruction, LIU Zhiwei, 2020/03/17
- [PATCH v6 61/61] target/riscv: configure and turn on vector extension from command line, LIU Zhiwei, 2020/03/17
  - Re: [PATCH v6 61/61] target/riscv: configure and turn on vector extension from command line, Alistair Francis, 2020/03/25
  - Re: [PATCH v6 61/61] target/riscv: configure and turn on vector extension from command line, Richard Henderson, 2020/03/28
- Re: [PATCH v6 00/61] target/riscv: support vector extension v0.7.1, no-reply, 2020/03/17

Prev by Date: [PATCH v4 2/2] net/colo-compare.c: handling of the full primary or secondary queue
Next by Date: [PATCH] target/arm: fix incorrect current EL bug in aarch32 exception emulation
Previous by thread: Re: [PATCH v6 58/61] target/riscv: vector slide instructions
Next by thread: [PATCH v6 59/61] target/riscv: vector register gather instruction
Index(es):
- Date
- Thread