1# Copyright (C) 2021 Collabora, Ltd. 2# Copyright (C) 2016 Intel Corporation 3# 4# Permission is hereby granted, free of charge, to any person obtaining a 5# copy of this software and associated documentation files (the "Software"), 6# to deal in the Software without restriction, including without limitation 7# the rights to use, copy, modify, merge, publish, distribute, sublicense, 8# and/or sell copies of the Software, and to permit persons to whom the 9# Software is furnished to do so, subject to the following conditions: 10# 11# The above copyright notice and this permission notice (including the next 12# paragraph) shall be included in all copies or substantial portions of the 13# Software. 14# 15# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR 16# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, 17# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL 18# THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER 19# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING 20# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS 21# IN THE SOFTWARE. 22 23import argparse 24import sys 25import math 26 27a = 'a' 28b = 'b' 29c = 'c' 30 31algebraic_late = [ 32 # Canonical form. The scheduler will convert back if it makes sense. 33 (('fmul', a, 2.0), ('fadd', a, a)), 34 35 # Fuse Mali-specific clamps 36 (('fmin', ('fmax', a, -1.0), 1.0), ('fsat_signed_mali', a)), 37 (('fmax', ('fmin', a, 1.0), -1.0), ('fsat_signed_mali', a)), 38 (('fmax', a, 0.0), ('fclamp_pos_mali', a)), 39 40 (('fabs', ('fddx', a)), ('fabs', ('fddx_must_abs_mali', a))), 41 (('fabs', ('fddy', b)), ('fabs', ('fddy_must_abs_mali', b))), 42] 43 44# Handling all combinations of boolean and float sizes for b2f is nontrivial. 45# bcsel has the same problem in more generality; lower b2f to bcsel in NIR to 46# reuse the efficient implementations of bcsel. This includes special handling 47# to allow vectorization in places the hardware does not directly. 48# 49# Because this lowering must happen late, NIR won't squash inot in 50# automatically. Do so explicitly. (The more specific pattern must be first.) 51for bsz in [8, 16, 32]: 52 for fsz in [16, 32]: 53 algebraic_late += [ 54 ((f'b2f{fsz}', ('inot', f'a@{bsz}')), (f'b{bsz}csel', a, 0.0, 1.0)), 55 ((f'b2f{fsz}', f'a@{bsz}'), (f'b{bsz}csel', a, 1.0, 0.0)), 56 ] 57 58 59def main(): 60 parser = argparse.ArgumentParser() 61 parser.add_argument('-p', '--import-path', required=True) 62 args = parser.parse_args() 63 sys.path.insert(0, args.import_path) 64 run() 65 66 67def run(): 68 import nir_algebraic # pylint: disable=import-error 69 70 print('#include "bifrost_nir.h"') 71 72 print(nir_algebraic.AlgebraicPass("bifrost_nir_lower_algebraic_late", 73 algebraic_late).render()) 74 75 76if __name__ == '__main__': 77 main() 78