From 2dbba8b024720c11cb2d812b5ccb61ecb9887faa Mon Sep 17 00:00:00 2001 From: Zack Rusin Date: Fri, 12 Oct 2007 13:57:53 -0400 Subject: [PATCH] Implement the conversion and do the initial execution pass. --- configs/default | 3 +- src/mesa/pipe/draw/draw_private.h | 1 + src/mesa/pipe/draw/draw_vertex_fetch.c | 4 +- src/mesa/pipe/draw/draw_vertex_shader.c | 13 +- src/mesa/pipe/draw/draw_vertex_shader_llvm.c | 129 +++++++++++++++++++ src/mesa/pipe/llvm/llvm_builtins.c | 90 +++++++++++++ src/mesa/pipe/llvm/llvmtgsi.cpp | 6 +- src/mesa/pipe/llvm/llvmtgsi.h | 6 +- src/mesa/pipe/softpipe/sp_quad_fs.c | 2 +- src/mesa/sources | 1 + 10 files changed, 244 insertions(+), 11 deletions(-) create mode 100644 src/mesa/pipe/draw/draw_vertex_shader_llvm.c create mode 100644 src/mesa/pipe/llvm/llvm_builtins.c diff --git a/configs/default b/configs/default index ac8de653ec3..bb8f9a18082 100644 --- a/configs/default +++ b/configs/default @@ -76,7 +76,8 @@ else endif ifeq ($(MESA_NO_LLVM),0) - LLVM_CFLAGS=`llvm-config --cflags` -DMESA_NO_LLVM=0 +# LLVM_CFLAGS=`llvm-config --cflags` -DMESA_NO_LLVM=0 + LLVM_CFLAGS=-DMESA_NO_LLVM=0 LLVM_CXXFLAGS=`llvm-config --cxxflags` -DMESA_NO_LLVM=0 -Wno-long-long LLVM_LDFLAGS=`llvm-config --ldflags` LLVM_LIBS=`llvm-config --libs` diff --git a/src/mesa/pipe/draw/draw_private.h b/src/mesa/pipe/draw/draw_private.h index b7811e891e7..b672cc2e8b5 100644 --- a/src/mesa/pipe/draw/draw_private.h +++ b/src/mesa/pipe/draw/draw_private.h @@ -251,6 +251,7 @@ extern void draw_vertex_cache_reset_vertex_ids( struct draw_context *draw ); extern void draw_vertex_shader_queue_flush( struct draw_context *draw ); +extern void draw_vertex_shader_queue_flush_llvm( struct draw_context *draw ); struct tgsi_exec_machine; diff --git a/src/mesa/pipe/draw/draw_vertex_fetch.c b/src/mesa/pipe/draw/draw_vertex_fetch.c index de1cd06da6d..1b093b73420 100644 --- a/src/mesa/pipe/draw/draw_vertex_fetch.c +++ b/src/mesa/pipe/draw/draw_vertex_fetch.c @@ -96,7 +96,7 @@ void draw_vertex_fetch( struct draw_context *draw, for (j = 0; j < count; j++) { uint attr; - /*printf("fetch vertex %u: \n", j);*/ + printf("fetch vertex %u: \n", j); /* loop over vertex attributes (vertex shader inputs) */ for (attr = 0; attr < draw->vertex_shader->state->num_inputs; attr++) { @@ -111,7 +111,7 @@ void draw_vertex_fetch( struct draw_context *draw, fetch_attrib4(src, draw->vertex_element[attr].src_format, p); - /*printf(" %u: %f %f %f %f\n", attr, p[0], p[1], p[2], p[3]);*/ + printf("> %u: %f %f %f %f\n", attr, p[0], p[1], p[2], p[3]); /* Transform to AoS xxxx/yyyy/zzzz/wwww representation: */ diff --git a/src/mesa/pipe/draw/draw_vertex_shader.c b/src/mesa/pipe/draw/draw_vertex_shader.c index fd0183f8cd4..874259e727f 100644 --- a/src/mesa/pipe/draw/draw_vertex_shader.c +++ b/src/mesa/pipe/draw/draw_vertex_shader.c @@ -91,6 +91,8 @@ run_vertex_program(struct draw_context *draw, assert(draw->vertex_shader->state->output_semantic_name[0] == TGSI_SEMANTIC_POSITION); + fprintf(stderr, "------ run_vertex\n"); + /* Consts does not require 16 byte alignment. */ machine->Consts = (float (*)[4]) draw->mapped_constants; @@ -161,10 +163,6 @@ run_vertex_program(struct draw_context *draw, vOut[j]->data[slot][3] = machine->Outputs[slot].xyzw[3].f[j]; #if DBG printf("output[%d][%d]: %f %f %f %f\n", j, slot, - vOut[j]->data[slot][0], - vOut[j]->data[slot][1], - vOut[j]->data[slot][2], - vOut[j]->data[slot][3]); #endif } } /* loop over vertices */ @@ -179,7 +177,12 @@ void draw_vertex_shader_queue_flush( struct draw_context *draw ) { unsigned i, j; -// fprintf(stderr, " q(%d) ", draw->vs.queue_nr ); + fprintf(stderr, "XX q(%d) ", draw->vs.queue_nr ); + + if (draw->vertex_shader->state->llvm_prog) { + draw_vertex_shader_queue_flush_llvm(draw); + return; + } /* run vertex shader on vertex cache entries, four per invokation */ for (i = 0; i < draw->vs.queue_nr; i += 4) { diff --git a/src/mesa/pipe/draw/draw_vertex_shader_llvm.c b/src/mesa/pipe/draw/draw_vertex_shader_llvm.c new file mode 100644 index 00000000000..b38498efc57 --- /dev/null +++ b/src/mesa/pipe/draw/draw_vertex_shader_llvm.c @@ -0,0 +1,129 @@ +/************************************************************************** + * + * Copyright 2007 Tungsten Graphics, Inc., Cedar Park, Texas. + * All Rights Reserved. + * + * 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, sub license, 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 NON-INFRINGEMENT. + * IN NO EVENT SHALL TUNGSTEN GRAPHICS AND/OR ITS SUPPLIERS 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. + * + **************************************************************************/ + + /* + * Authors: + * Zack Rusin zack@tungstengraphics.com + */ + +#include "pipe/p_util.h" +#include "draw_private.h" +#include "draw_context.h" +#include "draw_vertex.h" + +#include "pipe/llvm/llvmtgsi.h" +#include "pipe/tgsi/exec/tgsi_core.h" + +static INLINE void +fetch_attrib4(const void *ptr, unsigned format, float attrib[4]) +{ + /* defaults */ + attrib[1] = 0.0; + attrib[2] = 0.0; + attrib[3] = 1.0; + switch (format) { + case PIPE_FORMAT_R32G32B32A32_FLOAT: + attrib[3] = ((float *) ptr)[3]; + /* fall-through */ + case PIPE_FORMAT_R32G32B32_FLOAT: + attrib[2] = ((float *) ptr)[2]; + /* fall-through */ + case PIPE_FORMAT_R32G32_FLOAT: + attrib[1] = ((float *) ptr)[1]; + /* fall-through */ + case PIPE_FORMAT_R32_FLOAT: + attrib[0] = ((float *) ptr)[0]; + break; + default: + assert(0); + } +} + + +/** + * Fetch vertex attributes for 'count' vertices. + */ +static INLINE +void vertex_fetch(struct draw_context *draw, + const unsigned elt, + float (*inputs)[4]) +{ + uint attr; + + printf("fetch vertex %u: \n", elt); + + /* loop over vertex attributes (vertex shader inputs) */ + for (attr = 0; attr < draw->vertex_shader->state->num_inputs; attr++) { + + unsigned buf = draw->vertex_element[attr].vertex_buffer_index; + const void *src + = (const void *) ((const ubyte *) draw->mapped_vbuffer[buf] + + draw->vertex_buffer[buf].buffer_offset + + draw->vertex_element[attr].src_offset + + elt * draw->vertex_buffer[buf].pitch); + float p[4]; + + fetch_attrib4(src, draw->vertex_element[attr].src_format, p); + + printf("> %u: %f %f %f %f\n", attr, p[0], p[1], p[2], p[3]); + + inputs[attr][0] = p[0]; /*X*/ + inputs[attr][1] = p[1]; /*Y*/ + inputs[attr][2] = p[2]; /*Z*/ + inputs[attr][3] = p[3]; /*W*/ + } +} + + +/** + * Called by the draw module when the vertx cache needs to be flushed. + * This involves running the vertex shader. + */ +void draw_vertex_shader_queue_flush_llvm(struct draw_context *draw) +{ + unsigned i; + + struct vertex_header *dests[VS_QUEUE_LENGTH]; + float inputs[VS_QUEUE_LENGTH][PIPE_MAX_SHADER_INPUTS][4]; + float (*consts)[4] = (float (*)[4]) draw->mapped_constants; + struct ga_llvm_prog *prog = draw->vertex_shader->state->llvm_prog; + + fprintf(stderr, "XX q(%d) ", draw->vs.queue_nr); + + /* fetch the inputs */ + for (i = 0; i < draw->vs.queue_nr; ++i) { + unsigned elt = draw->vs.queue[i].elt; + dests[i] = draw->vs.queue[i].dest; + vertex_fetch(draw, elt, inputs[i]); + } + + /* batch execute the shaders on all the vertices */ + ga_llvm_prog_exec(prog, inputs, dests, consts, + draw->vs.queue_nr); + + draw->vs.queue_nr = 0; +} diff --git a/src/mesa/pipe/llvm/llvm_builtins.c b/src/mesa/pipe/llvm/llvm_builtins.c new file mode 100644 index 00000000000..0f0efeb3039 --- /dev/null +++ b/src/mesa/pipe/llvm/llvm_builtins.c @@ -0,0 +1,90 @@ + + +inline float4 compute_clip(float4 vec, float4 scale, float4 trans) +{ + return vec*scale + trans; +} + + +inline float +dot4(const float4 a, const float4 b) +{ + float4 c = a*b; + return c.x + c.y + c.z + c.w; +} + +inline unsigned +compute_clipmask(float4 clip, const float4 (*plane), unsigned nr) +{ + unsigned mask = 0; + unsigned i; + + for (i = 0; i < nr; i++) { + if (dot4(clip, plane[i]) < 0) + mask |= (1<clip[0] = clip.x; + y = vOut->clip[1] = clip.y; + z = vOut->clip[2] = clip.z; + w = vOut->clip[3] = clip.w; + + vOut[i]->clipmask = compute_clipmask(res0, planes, nr_planes); + vOut[i]->edgeflag = 1; + + /* divide by w */ + w = 1.0f / w; + x *= w; + y *= w; + z *= w; + res0.x = x; res0.y = y; res0.z = z; res0.w = 1; + + /* Viewport mapping */ + res = res * scale + trans; + vOut->data[0][0] = res.x; + vOut->data[0][1] = res.y; + vOut->data[0][2] = res.z; + vOut->data[0][3] = w; + + /* Remaining attributes are packed into sequential post-transform + * vertex attrib slots. + * Skip 0 since we just did it above. + * Subtract two because of the VERTEX_HEADER, CLIP_POS attribs. + */ + for (slot = 1; slot < draw->vertex_info.num_attribs - 2; slot++) { + float4 vec = results[slot]; + vOut->data[slot][0] = vec.x; + vOut->data[slot][1] = vec.y; + vOut->data[slot][2] = vec.z; + vOut->data[slot][3] = vec.w; + + printf("output %d: %f %f %f %f\n", slot, + vOut->data[slot][0], + vOut->data[slot][1], + vOut->data[slot][2], + vOut->data[slot][3]); + } +} + +void run_vertex_shader(float ainputs[VS_QUEUE_LENGTH][PIPE_MAX_SHADER_INPUTS][4], + struct vertex_header *dests[VS_QUEUE_LENGTH], + float *aconsts[4] + int count) +{ + float4 inputs[VS_QUEUE_LENGTH][PIPE_MAX_SHADER_INPUTS]; + float4 *consts; +} diff --git a/src/mesa/pipe/llvm/llvmtgsi.cpp b/src/mesa/pipe/llvm/llvmtgsi.cpp index 2221a2c353d..46b7561b5e8 100644 --- a/src/mesa/pipe/llvm/llvmtgsi.cpp +++ b/src/mesa/pipe/llvm/llvmtgsi.cpp @@ -426,7 +426,11 @@ void ga_llvm_prog_delete(struct ga_llvm_prog *prog) free(prog); } -int ga_llvm_prog_exec(struct ga_llvm_prog *prog) +int ga_llvm_prog_exec(struct ga_llvm_prog *prog, + float (*inputs)[32][4], + void *dests[16*32*4], + float (*consts)[4], + int count) { //std::cout << "START "<(prog->module); diff --git a/src/mesa/pipe/llvm/llvmtgsi.h b/src/mesa/pipe/llvm/llvmtgsi.h index c0cee915b92..9fbb0ea8f9d 100644 --- a/src/mesa/pipe/llvm/llvmtgsi.h +++ b/src/mesa/pipe/llvm/llvmtgsi.h @@ -18,7 +18,11 @@ ga_llvm_from_tgsi(const struct tgsi_token *tokens); void ga_llvm_prog_delete(struct ga_llvm_prog *prog); -int ga_llvm_prog_exec(struct ga_llvm_prog *prog); +int ga_llvm_prog_exec(struct ga_llvm_prog *prog, + float (*inputs)[32][4], + void *dests[16*32*4], + float (*consts)[4], + int count); #if defined __cplusplus } // extern "C" diff --git a/src/mesa/pipe/softpipe/sp_quad_fs.c b/src/mesa/pipe/softpipe/sp_quad_fs.c index 035f05170bc..00afcd930ad 100644 --- a/src/mesa/pipe/softpipe/sp_quad_fs.c +++ b/src/mesa/pipe/softpipe/sp_quad_fs.c @@ -109,7 +109,7 @@ shade_quad( machine->InterpCoefs ); } else { - ga_llvm_prog_exec(softpipe->fs->llvm_prog); + //ga_llvm_prog_exec(softpipe->fs->llvm_prog); quad->mask &= tgsi_exec_machine_run( machine ); } diff --git a/src/mesa/sources b/src/mesa/sources index 41c3ae387ed..c2544ed322d 100644 --- a/src/mesa/sources +++ b/src/mesa/sources @@ -172,6 +172,7 @@ DRAW_SOURCES = \ pipe/draw/draw_vertex_cache.c \ pipe/draw/draw_vertex_fetch.c \ pipe/draw/draw_vertex_shader.c \ + pipe/draw/draw_vertex_shader_llvm.c \ pipe/draw/draw_wide_prims.c TGSIEXEC_SOURCES = \ -- 2.30.2