From 788c63998ce6e17b7d7f2e30d28ee5979a6d6cec Mon Sep 17 00:00:00 2001 From: Matheus Ferst Date: Thu, 4 Nov 2021 09:37:17 -0300 Subject: [PATCH] target/ppc: Implement xxblendvb/xxblendvh/xxblendvw/xxblendvd instructions Reviewed-by: Richard Henderson Signed-off-by: Bruno Larsen (billionai) Signed-off-by: Matheus Ferst Message-Id: <20211104123719.323713-24-matheus.ferst@eldorado.org.br> Signed-off-by: David Gibson --- target/ppc/helper.h | 4 +++ target/ppc/insn64.decode | 19 ++++++++++ target/ppc/int_helper.c | 15 ++++++++ target/ppc/translate/vsx-impl.c.inc | 55 +++++++++++++++++++++++++++++ 4 files changed, 93 insertions(+) diff --git a/target/ppc/helper.h b/target/ppc/helper.h index 7ff1d055c4..627811cefc 100644 --- a/target/ppc/helper.h +++ b/target/ppc/helper.h @@ -520,6 +520,10 @@ DEF_HELPER_4(xxpermr, void, env, vsr, vsr, vsr) DEF_HELPER_4(xxextractuw, void, env, vsr, vsr, i32) DEF_HELPER_4(xxinsertw, void, env, vsr, vsr, i32) DEF_HELPER_3(xvxsigsp, void, env, vsr, vsr) +DEF_HELPER_5(XXBLENDVB, void, vsr, vsr, vsr, vsr, i32) +DEF_HELPER_5(XXBLENDVH, void, vsr, vsr, vsr, vsr, i32) +DEF_HELPER_5(XXBLENDVW, void, vsr, vsr, vsr, vsr, i32) +DEF_HELPER_5(XXBLENDVD, void, vsr, vsr, vsr, vsr, i32) DEF_HELPER_2(efscfsi, i32, env, i32) DEF_HELPER_2(efscfui, i32, env, i32) diff --git a/target/ppc/insn64.decode b/target/ppc/insn64.decode index 20aa2b4615..39e610913d 100644 --- a/target/ppc/insn64.decode +++ b/target/ppc/insn64.decode @@ -44,6 +44,16 @@ ...... ..... .... . ................ \ &8RR_D si=%8rr_si xt=%8rr_xt +# Format XX4 +&XX4 xt xa xb xc +%xx4_xt 0:1 21:5 +%xx4_xa 2:1 16:5 +%xx4_xb 1:1 11:5 +%xx4_xc 3:1 6:5 +@XX4 ........ ........ ........ ........ \ + ...... ..... ..... ..... ..... .. .... \ + &XX4 xt=%xx4_xt xa=%xx4_xa xb=%xx4_xb xc=%xx4_xc + ### Fixed-Point Load Instructions PLBZ 000001 10 0--.-- .................. \ @@ -175,3 +185,12 @@ XXSPLTIW 000001 01 0000 -- -- ................ \ 100000 ..... 0011 . ................ @8RR_D XXSPLTI32DX 000001 01 0000 -- -- ................ \ 100000 ..... 000 .. ................ @8RR_D_IX + +XXBLENDVD 000001 01 0000 -- ------------------ \ + 100001 ..... ..... ..... ..... 11 .... @XX4 +XXBLENDVW 000001 01 0000 -- ------------------ \ + 100001 ..... ..... ..... ..... 10 .... @XX4 +XXBLENDVH 000001 01 0000 -- ------------------ \ + 100001 ..... ..... ..... ..... 01 .... @XX4 +XXBLENDVB 000001 01 0000 -- ------------------ \ + 100001 ..... ..... ..... ..... 00 .... @XX4 diff --git a/target/ppc/int_helper.c b/target/ppc/int_helper.c index b7861776c2..9bc327bcba 100644 --- a/target/ppc/int_helper.c +++ b/target/ppc/int_helper.c @@ -1737,6 +1737,21 @@ void helper_xxinsertw(CPUPPCState *env, ppc_vsr_t *xt, *xt = t; } +#define XXBLEND(name, sz) \ +void glue(helper_XXBLENDV, name)(ppc_avr_t *t, ppc_avr_t *a, ppc_avr_t *b, \ + ppc_avr_t *c, uint32_t desc) \ +{ \ + for (int i = 0; i < ARRAY_SIZE(t->glue(u, sz)); i++) { \ + t->glue(u, sz)[i] = (c->glue(s, sz)[i] >> (sz - 1)) ? \ + b->glue(u, sz)[i] : a->glue(u, sz)[i]; \ + } \ +} +XXBLEND(B, 8) +XXBLEND(H, 16) +XXBLEND(W, 32) +XXBLEND(D, 64) +#undef XXBLEND + #define VEXT_SIGNED(name, element, cast) \ void helper_##name(ppc_avr_t *r, ppc_avr_t *b) \ { \ diff --git a/target/ppc/translate/vsx-impl.c.inc b/target/ppc/translate/vsx-impl.c.inc index 180d329f1a..d1de0da877 100644 --- a/target/ppc/translate/vsx-impl.c.inc +++ b/target/ppc/translate/vsx-impl.c.inc @@ -2087,6 +2087,61 @@ TRANS64(PLXV, do_lstxv_PLS_D, false, false) TRANS64(PSTXVP, do_lstxv_PLS_D, true, true) TRANS64(PLXVP, do_lstxv_PLS_D, false, true) +static void gen_xxblendv_vec(unsigned vece, TCGv_vec t, TCGv_vec a, TCGv_vec b, + TCGv_vec c) +{ + TCGv_vec tmp = tcg_temp_new_vec_matching(c); + tcg_gen_sari_vec(vece, tmp, c, (8 << vece) - 1); + tcg_gen_bitsel_vec(vece, t, tmp, b, a); + tcg_temp_free_vec(tmp); +} + +static bool do_xxblendv(DisasContext *ctx, arg_XX4 *a, unsigned vece) +{ + static const TCGOpcode vecop_list[] = { + INDEX_op_sari_vec, 0 + }; + static const GVecGen4 ops[4] = { + { + .fniv = gen_xxblendv_vec, + .fno = gen_helper_XXBLENDVB, + .opt_opc = vecop_list, + .vece = MO_8 + }, + { + .fniv = gen_xxblendv_vec, + .fno = gen_helper_XXBLENDVH, + .opt_opc = vecop_list, + .vece = MO_16 + }, + { + .fniv = gen_xxblendv_vec, + .fno = gen_helper_XXBLENDVW, + .opt_opc = vecop_list, + .vece = MO_32 + }, + { + .fniv = gen_xxblendv_vec, + .fno = gen_helper_XXBLENDVD, + .opt_opc = vecop_list, + .vece = MO_64 + } + }; + + REQUIRE_VSX(ctx); + + tcg_gen_gvec_4(vsr_full_offset(a->xt), vsr_full_offset(a->xa), + vsr_full_offset(a->xb), vsr_full_offset(a->xc), + 16, 16, &ops[vece]); + + return true; +} + +TRANS(XXBLENDVB, do_xxblendv, MO_8) +TRANS(XXBLENDVH, do_xxblendv, MO_16) +TRANS(XXBLENDVW, do_xxblendv, MO_32) +TRANS(XXBLENDVD, do_xxblendv, MO_64) + #undef GEN_XX2FORM #undef GEN_XX3FORM #undef GEN_XX2IFORM