[v2,14/34] target/ppc: Implement vsldbi/vsrdbi instructions

Commit Message

Matheus K. Ferst Oct. 29, 2021, 8:24 p.m. UTC

From: Matheus Ferst <[e-mail]>

Suggested-by: Richard Henderson <[e-mail]>
Signed-off-by: Matheus Ferst <[e-mail]>
---
v2:
- tcg_gen_extract2_i64 instead of shifts and or
---
 target/ppc/insn32.decode            |  8 ++++
 target/ppc/translate/vmx-impl.c.inc | 66 +++++++++++++++++++++++++++++
 2 files changed, 74 insertions(+)

Comments

Richard Henderson Oct. 30, 2021, 9:42 p.m. UTC | #1

On 10/29/21 1:24 PM, [e-mail] wrote:
> From: Matheus Ferst<[e-mail]>
> 
> Suggested-by: Richard Henderson<[e-mail]>
> Signed-off-by: Matheus Ferst<[e-mail]>
> ---
> v2:
> - tcg_gen_extract2_i64 instead of shifts and or
> ---
>   target/ppc/insn32.decode            |  8 ++++
>   target/ppc/translate/vmx-impl.c.inc | 66 +++++++++++++++++++++++++++++
>   2 files changed, 74 insertions(+)

Reviewed-by: Richard Henderson <[e-mail]>

r~

Patch

diff --git a/target/ppc/insn32.decode b/target/ppc/insn32.decode
index 4666c06f55..257b11113d 100644
--- a/target/ppc/insn32.decode
+++ b/target/ppc/insn32.decode
@@ -38,6 +38,9 @@ 
 %dx_d           6:s10 16:5 0:1
 @DX             ...... rt:5  ..... .......... ..... .   &DX d=%dx_d
 
+&VN             vrt vra vrb sh
+@VN             ...... vrt:5 vra:5 vrb:5 .. sh:3 ......         &VN
+
 &VX             vrt vra vrb
 @VX             ...... vrt:5 vra:5 vrb:5 .......... .   &VX
 
@@ -338,3 +341,8 @@  VCLZDM          000100 ..... ..... ..... 11110000100    @VX
 VCTZDM          000100 ..... ..... ..... 11111000100    @VX
 VPDEPD          000100 ..... ..... ..... 10111001101    @VX
 VPEXTD          000100 ..... ..... ..... 10110001101    @VX
+
+## Vector Permute and Formatting Instruction
+
+VSLDBI          000100 ..... ..... ..... 00 ... 010110  @VN
+VSRDBI          000100 ..... ..... ..... 01 ... 010110  @VN
diff --git a/target/ppc/translate/vmx-impl.c.inc b/target/ppc/translate/vmx-impl.c.inc
index cddb3848ab..6edffd5637 100644
--- a/target/ppc/translate/vmx-impl.c.inc
+++ b/target/ppc/translate/vmx-impl.c.inc
@@ -1257,6 +1257,72 @@  static void gen_vsldoi(DisasContext *ctx)
     tcg_temp_free_i32(sh);
 }
 
+static bool trans_VSLDBI(DisasContext *ctx, arg_VN *a)
+{
+    TCGv_i64 t0, t1, t2;
+
+    REQUIRE_INSNS_FLAGS2(ctx, ISA310);
+    REQUIRE_VECTOR(ctx);
+
+    t0 = tcg_temp_new_i64();
+    t1 = tcg_temp_new_i64();
+
+    get_avr64(t0, a->vra, true);
+    get_avr64(t1, a->vra, false);
+
+    if (a->sh != 0) {
+        t2 = tcg_temp_new_i64();
+
+        get_avr64(t2, a->vrb, true);
+
+        tcg_gen_extract2_i64(t0, t1, t0, 64 - a->sh);
+        tcg_gen_extract2_i64(t1, t2, t1, 64 - a->sh);
+
+        tcg_temp_free_i64(t2);
+    }
+
+    set_avr64(a->vrt, t0, true);
+    set_avr64(a->vrt, t1, false);
+
+    tcg_temp_free_i64(t0);
+    tcg_temp_free_i64(t1);
+
+    return true;
+}
+
+static bool trans_VSRDBI(DisasContext *ctx, arg_VN *a)
+{
+    TCGv_i64 t2, t1, t0;
+
+    REQUIRE_INSNS_FLAGS2(ctx, ISA310);
+    REQUIRE_VECTOR(ctx);
+
+    t0 = tcg_temp_new_i64();
+    t1 = tcg_temp_new_i64();
+
+    get_avr64(t0, a->vrb, false);
+    get_avr64(t1, a->vrb, true);
+
+    if (a->sh != 0) {
+        t2 = tcg_temp_new_i64();
+
+        get_avr64(t2, a->vra, false);
+
+        tcg_gen_extract2_i64(t0, t0, t1, a->sh);
+        tcg_gen_extract2_i64(t1, t1, t2, a->sh);
+
+        tcg_temp_free_i64(t2);
+    }
+
+    set_avr64(a->vrt, t0, false);
+    set_avr64(a->vrt, t1, true);
+
+    tcg_temp_free_i64(t0);
+    tcg_temp_free_i64(t1);
+
+    return true;
+}
+
 #define GEN_VAFORM_PAIRED(name0, name1, opc2)                           \
 static void glue(gen_, name0##_##name1)(DisasContext *ctx)              \
     {                                                                   \