<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Mon, Apr 30, 2018 at 7:18 AM, Iago Toral Quiroga <<a href="mailto:itoral@igalia.com" target="_blank">itoral@igalia.com</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Not all bit-sizes may be supported natively in hardware for all operations. This pass allows drivers to lower such operations to a bit-size that is actually supported and then converts the result back to the original bit-size. Compiler backends control which operations and wich bit-sizes require the lowering through a callback function. v2: generalize this pass and make it available in NIR core (Rob, Jason) --- src/compiler/Makefile.sources | 1 + src/compiler/nir/meson.build | 1 + src/compiler/nir/nir.h | 6 ++ src/compiler/nir/nir_lower_bit_size.c | 127 ++++++++++++++++++++++++++++++++++ 4 files changed, 135 insertions(+) create mode 100644 src/compiler/nir/nir_lower_bit_size.c diff --git a/src/compiler/Makefile.sources b/src/compiler/Makefile.sources index aca9dab476..b5321588be 100644 --- a/src/compiler/Makefile.sources +++ b/src/compiler/Makefile.sources @@ -237,6 +237,7 @@ NIR_FILES = \ nir/nir_lower_phis_to_scalar.c \ nir/nir_lower_regs_to_ssa.c \ nir/nir_lower_returns.c \ + nir/nir_lower_bit_size.c \ nir/nir_lower_samplers.c \ nir/nir_lower_samplers_as_deref.c \ nir/nir_lower_subgroups.c \ diff --git a/src/compiler/nir/meson.build b/src/compiler/nir/meson.build index b28a565d0c..62fe5167d3 100644 --- a/src/compiler/nir/meson.build +++ b/src/compiler/nir/meson.build @@ -144,6 +144,7 @@ files_libnir = files( 'nir_lower_vec_to_movs.c', 'nir_lower_wpos_center.c', 'nir_lower_wpos_ytransform.c', + 'nir_lower_bit_size.c', 'nir_metadata.c', 'nir_move_vec_src_uses_to_dest.c', 'nir_normalize_cubemap_coords.c', diff --git a/src/compiler/nir/nir.h b/src/compiler/nir/nir.h index f3326e6df9..e7f2b145b3 100644 --- a/src/compiler/nir/nir.h +++ b/src/compiler/nir/nir.h @@ -2763,6 +2763,12 @@ bool nir_lower_to_source_mods(nir_shader *shader); bool nir_lower_gs_intrinsics(nir_shader *shader); +typedef unsigned (*nir_lower_bit_size_callback)(const nir_alu_instr *, void *); + +bool nir_lower_bit_size(nir_shader *shader, + nir_lower_bit_size_callback callback, + void *callback_data); + typedef enum { nir_lower_imul64 = (1 << 0), nir_lower_isign64 = (1 << 1), diff --git a/src/compiler/nir/nir_lower_bit_size.c b/src/compiler/nir/nir_lower_bit_size.c new file mode 100644 index 0000000000..8e0fd663a1 --- /dev/null +++ b/src/compiler/nir/nir_lower_bit_size.c @@ -0,0 +1,127 @@ +/* + * Copyright © 2018 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + */ + +#include "nir_builder.h" + +/** + * Some ALU operations may not be supported in hardware in specific bit-sizes. + * This pass allows implementations to selectively lower such operations to + * a bit-size that is supported natively and the converts the result back to + * the original bit-size. + */ +static nir_ssa_def * +convert_to_bit_size(nir_builder *bld, + nir_ssa_def *src, + nir_alu_type type, + unsigned bit_size) +{ + nir_alu_type base_type = nir_alu_type_get_base_type(type); + nir_alu_type lowered_type = bit_size | base_type; + + nir_op opcode = + nir_type_conversion_op(type, lowered_type, nir_rounding_mode_undef); + + return nir_build_alu(bld, opcode, src, NULL, NULL, NULL); +} + +static void +lower_instr(nir_builder *bld, nir_alu_instr *alu, unsigned bit_size) +{ + const nir_op op = alu->op; + + bld->cursor = nir_before_instr(&alu->instr); + + /* Convert each source to the requested bit-size */ + nir_ssa_def *srcs[4] = { NULL, NULL, NULL, NULL }; + const unsigned num_inputs = nir_op_infos[op].num_inputs;' </blockquote><div> </div><div>I don't know that we need the num_inputs temporary </div><div> </div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div> + for (unsigned i = 0; i < num_inputs; i++) { + nir_ssa_def *src = nir_ssa_for_alu_src(bld, alu, i); </div> + <div> + nir_alu_type type = nir_op_infos[op].input_types[i]; + unsigned type_size = nir_alu_type_get_type_size(type); </div></blockquote><div> </div><div>Again, I don't think this temporary is needed. </div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> + if (type_size == 0) + srcs[i] = convert_to_bit_size(bld, src, type, bit_size); + else + srcs[i] = src; + } + + /* Emit the lowered ALU instruction */ + nir_ssa_def *lowered_dst = + nir_build_alu(bld, op, srcs[0], srcs[1], srcs[2], srcs[3]); + + /* Convert result back to the original bit-size */ + unsigned dst_bit_size = alu->dest.dest.ssa.bit_size; + nir_alu_type type = nir_op_infos[op].output_type; + nir_ssa_def *dst = convert_to_bit_size(bld, lowered_dst, type, dst_bit_size); + nir_ssa_def_rewrite_uses(&alu->dest.dest.ssa, nir_src_for_ssa(dst)); +} + +static bool +lower_impl(nir_function_impl *impl, + nir_lower_bit_size_callback callback, + void *callback_data) +{ + nir_builder b; + nir_builder_init(&b, impl); + bool progress = false; + + nir_foreach_block(block, impl) { + nir_foreach_instr_safe(instr, block) { + if (instr->type == nir_instr_type_alu) { </blockquote><div> </div><div>I'd make this "if (!= alu) continue;" so we can keep the nesting depth down. </div><div>All my comments are fairly trivial so Reviewed-by: Jason Ekstrand <<a href="mailto:jason@jlekstrand.net">jason@jlekstrand.net</a>> </div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"> + nir_alu_instr *alu = nir_instr_as_alu(instr); + assert(alu->dest.dest.is_ssa); + + unsigned lower_bit_size = callback(alu, callback_data); + if (lower_bit_size == 0) + continue; + + assert(lower_bit_size != alu->dest.dest.ssa.bit_size); + + lower_instr(&b, alu, lower_bit_size); + progress = true; + } + } + } + + if (progress) { + nir_metadata_preserve(impl, nir_metadata_block_index | + nir_metadata_dominance); + } + + return progress; +} + +bool +nir_lower_bit_size(nir_shader *shader, + nir_lower_bit_size_callback callback, + void *callback_data) +{ + bool progress = false; + + nir_foreach_function(function, shader) { + if (function->impl) + progress |= lower_impl(function->impl, callback, callback_data); + } + + return progress; +} -- 2.14.1 </blockquote></div> </div></div>