diff options
| author | Ian Romanick <[email protected]> | 2018-08-13 18:08:23 -0700 |
|---|---|---|
| committer | Ian Romanick <[email protected]> | 2019-05-14 11:25:02 -0700 |
| commit | 7b4ff6a1af27af458446640c41c61cdee6c960f8 (patch) | |
| tree | ee901b5609751b261bb1b50d5e1ecfd32ab8b6fa /src/compiler/nir/nir_opcodes.py | |
| parent | e049a9c92b3048f2d28d5a36f0dd780b19fe4b2a (diff) | |
nir: Mark ffma as 2src_commutative
This doesn't make any real difference now, but future work (not in this
series) will add a LOT of ffma patterns. Having to duplicate all of
them for ffma(a, b, c) and ffma(b, a, c) is just terrible.
No shader-db changes on any Intel platform.
Reviewed-by: Jason Ekstrand <[email protected]>
Diffstat (limited to 'src/compiler/nir/nir_opcodes.py')
| -rw-r--r-- | src/compiler/nir/nir_opcodes.py | 30 |
1 files changed, 15 insertions, 15 deletions
diff --git a/src/compiler/nir/nir_opcodes.py b/src/compiler/nir/nir_opcodes.py index fee9c221c16..3ea25f31d4a 100644 --- a/src/compiler/nir/nir_opcodes.py +++ b/src/compiler/nir/nir_opcodes.py @@ -816,16 +816,16 @@ binop("extract_u16", tuint, "", "(uint16_t)(src0 >> (src1 * 16))") binop("extract_i16", tint, "", "(int16_t)(src0 >> (src1 * 16))") -def triop(name, ty, const_expr): - opcode(name, 0, ty, [0, 0, 0], [ty, ty, ty], False, "", const_expr) +def triop(name, ty, alg_props, const_expr): + opcode(name, 0, ty, [0, 0, 0], [ty, ty, ty], False, alg_props, const_expr) def triop_horiz(name, output_size, src1_size, src2_size, src3_size, const_expr): opcode(name, output_size, tuint, [src1_size, src2_size, src3_size], [tuint, tuint, tuint], False, "", const_expr) -triop("ffma", tfloat, "src0 * src1 + src2") +triop("ffma", tfloat, _2src_commutative, "src0 * src1 + src2") -triop("flrp", tfloat, "src0 * (1 - src2) + src1 * src2") +triop("flrp", tfloat, "", "src0 * (1 - src2) + src1 * src2") # Conditional Select # @@ -834,20 +834,20 @@ triop("flrp", tfloat, "src0 * (1 - src2) + src1 * src2") # bools (0.0 vs 1.0) and one for integer bools (0 vs ~0). -triop("fcsel", tfloat32, "(src0 != 0.0f) ? src1 : src2") +triop("fcsel", tfloat32, "", "(src0 != 0.0f) ? src1 : src2") # 3 way min/max/med -triop("fmin3", tfloat, "fminf(src0, fminf(src1, src2))") -triop("imin3", tint, "MIN2(src0, MIN2(src1, src2))") -triop("umin3", tuint, "MIN2(src0, MIN2(src1, src2))") +triop("fmin3", tfloat, "", "fminf(src0, fminf(src1, src2))") +triop("imin3", tint, "", "MIN2(src0, MIN2(src1, src2))") +triop("umin3", tuint, "", "MIN2(src0, MIN2(src1, src2))") -triop("fmax3", tfloat, "fmaxf(src0, fmaxf(src1, src2))") -triop("imax3", tint, "MAX2(src0, MAX2(src1, src2))") -triop("umax3", tuint, "MAX2(src0, MAX2(src1, src2))") +triop("fmax3", tfloat, "", "fmaxf(src0, fmaxf(src1, src2))") +triop("imax3", tint, "", "MAX2(src0, MAX2(src1, src2))") +triop("umax3", tuint, "", "MAX2(src0, MAX2(src1, src2))") -triop("fmed3", tfloat, "fmaxf(fminf(fmaxf(src0, src1), src2), fminf(src0, src1))") -triop("imed3", tint, "MAX2(MIN2(MAX2(src0, src1), src2), MIN2(src0, src1))") -triop("umed3", tuint, "MAX2(MIN2(MAX2(src0, src1), src2), MIN2(src0, src1))") +triop("fmed3", tfloat, "", "fmaxf(fminf(fmaxf(src0, src1), src2), fminf(src0, src1))") +triop("imed3", tint, "", "MAX2(MIN2(MAX2(src0, src1), src2), MIN2(src0, src1))") +triop("umed3", tuint, "", "MAX2(MIN2(MAX2(src0, src1), src2), MIN2(src0, src1))") opcode("bcsel", 0, tuint, [0, 0, 0], [tbool1, tuint, tuint], False, "", "src0 ? src1 : src2") @@ -855,7 +855,7 @@ opcode("b32csel", 0, tuint, [0, 0, 0], [tbool32, tuint, tuint], False, "", "src0 ? src1 : src2") # SM5 bfi assembly -triop("bfi", tuint32, """ +triop("bfi", tuint32, "", """ unsigned mask = src0, insert = src1, base = src2; if (mask == 0) { dst = base; |