On 2/10/19, H.J. Lu <[email protected]> wrote:
> Emulate MMX ssse3_ph<plusminus_mnemonic>dv2si3 with SSE by moving bits
> 64:95 to bits 32:63 in SSE register. Only SSE register source operand
> is allowed.
>
> PR target/89021
> * config/i386/sse.md (ssse3_ph<plusminus_mnemonic>dv2si3):
> Changed to define_insn_and_split to support SSE emulation.
> ---
> gcc/config/i386/sse.md | 32 ++++++++++++++++++++++++--------
> 1 file changed, 24 insertions(+), 8 deletions(-)
>
> diff --git a/gcc/config/i386/sse.md b/gcc/config/i386/sse.md
> index 6138f59f267..4bcfd3fc272 100644
> --- a/gcc/config/i386/sse.md
> +++ b/gcc/config/i386/sse.md
> @@ -15480,26 +15480,42 @@
> (set_attr "prefix" "orig,vex")
> (set_attr "mode" "TI")])
>
> -(define_insn "ssse3_ph<plusminus_mnemonic>dv2si3"
> - [(set (match_operand:V2SI 0 "register_operand" "=y")
> +(define_insn_and_split "ssse3_ph<plusminus_mnemonic>dv2si3"
> + [(set (match_operand:V2SI 0 "register_operand" "=y,x,Yv")
> (vec_concat:V2SI
> (plusminus:SI
> (vec_select:SI
> - (match_operand:V2SI 1 "register_operand" "0")
> + (match_operand:V2SI 1 "register_operand" "0,0,Yv")
> (parallel [(const_int 0)]))
> (vec_select:SI (match_dup 1) (parallel [(const_int 1)])))
> (plusminus:SI
> (vec_select:SI
> - (match_operand:V2SI 2 "nonimmediate_operand" "ym")
> + (match_operand:V2SI 2 "nonimmediate_operand" "ym,x,Yv")
> (parallel [(const_int 0)]))
> (vec_select:SI (match_dup 2) (parallel [(const_int 1)])))))]
> - "TARGET_SSSE3"
> - "ph<plusminus_mnemonic>d\t{%2, %0|%0, %2}"
> - [(set_attr "type" "sseiadd")
> + "(TARGET_MMX || TARGET_MMX_WITH_SSE) && TARGET_SSSE3"
> + "@
> + ph<plusminus_mnemonic>d\t{%2, %0|%0, %2}
> + #
> + #"
> + "&& reload_completed && TARGET_MMX_WITH_SSE"
Please fix split condition.
Uros.
> + [(const_int 0)]
> +{
> + /* Generate SSE version of the operation. */
> + rtx op0 = gen_rtx_REG (V4SImode, REGNO (operands[0]));
> + rtx op1 = gen_rtx_REG (V4SImode, REGNO (operands[1]));
> + rtx op2 = gen_rtx_REG (V4SImode, REGNO (operands[2]));
> + rtx insn = gen_ssse3_ph<plusminus_mnemonic>dv4si3 (op0, op1, op2);
> + emit_insn (insn);
> + ix86_move_vector_high_sse_to_mmx (op0);
> + DONE;
> +}
> + [(set_attr "mmx_isa" "native,x64_noavx,x64_avx")
> + (set_attr "type" "sseiadd")
> (set_attr "atom_unit" "complex")
> (set_attr "prefix_extra" "1")
> (set (attr "prefix_rex") (symbol_ref "x86_extended_reg_mentioned_p
> (insn)"))
> - (set_attr "mode" "DI")])
> + (set_attr "mode" "DI,TI,TI")])
>
> (define_insn "avx2_pmaddubsw256"
> [(set (match_operand:V16HI 0 "register_operand" "=x,v")
> --
> 2.20.1
>
>
