From b207e88b25e526d0f1ada7b19605b880a27866dc Mon Sep 17 00:00:00 2001 From: Kenneth Graunke Date: Fri, 7 Mar 2014 16:10:50 -0800 Subject: [PATCH] i965/fs: Support pull parameters in SIMD16 mode. This is just a matter of reusing the pull/push constant information set up by the SIMD8 compile. This gains us 78 SIMD16 programs in shader-db. Signed-off-by: Kenneth Graunke Reviewed-by: Topi Pohjolainen Reviewed-by: Eric Anholt --- src/mesa/drivers/dri/i965/brw_fs.cpp | 21 ++++++++++---------- src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 3 ++- 2 files changed, 13 insertions(+), 11 deletions(-) diff --git a/src/mesa/drivers/dri/i965/brw_fs.cpp b/src/mesa/drivers/dri/i965/brw_fs.cpp index 054bed544b6..1f8590143fe 100644 --- a/src/mesa/drivers/dri/i965/brw_fs.cpp +++ b/src/mesa/drivers/dri/i965/brw_fs.cpp @@ -886,7 +886,9 @@ fs_visitor::import_uniforms(fs_visitor *v) import_uniforms_callback, variable_ht); this->push_constant_loc = v->push_constant_loc; + this->pull_constant_loc = v->pull_constant_loc; this->uniforms = v->uniforms; + this->param_size = v->param_size; } /* Our support for uniforms is piggy-backed on the struct @@ -1748,6 +1750,9 @@ fs_visitor::compact_virtual_grfs() void fs_visitor::move_uniform_array_access_to_pull_constants() { + if (dispatch_width != 8) + return; + pull_constant_loc = ralloc_array(mem_ctx, int, uniforms); for (unsigned int i = 0; i < uniforms; i++) { @@ -3500,17 +3505,13 @@ brw_wm_fs_emit(struct brw_context *brw, struct brw_wm_compile *c, exec_list *simd16_instructions = NULL; fs_visitor v2(brw, c, prog, fp, 16); if (brw->gen >= 5 && likely(!(INTEL_DEBUG & DEBUG_NO16))) { - if (c->prog_data.base.nr_pull_params == 0) { - /* Try a SIMD16 compile */ - v2.import_uniforms(&v); - if (!v2.run()) { - perf_debug("SIMD16 shader failed to compile, falling back to " - "SIMD8 at a 10-20%% performance cost: %s", v2.fail_msg); - } else { - simd16_instructions = &v2.instructions; - } + /* Try a SIMD16 compile */ + v2.import_uniforms(&v); + if (!v2.run()) { + perf_debug("SIMD16 shader failed to compile, falling back to " + "SIMD8 at a 10-20%% performance cost: %s", v2.fail_msg); } else { - perf_debug("Skipping SIMD16 due to pull parameters.\n"); + simd16_instructions = &v2.instructions; } } diff --git a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp index 7088502340b..cd90e23202c 100644 --- a/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp +++ b/src/mesa/drivers/dri/i965/brw_fs_visitor.cpp @@ -2978,7 +2978,8 @@ fs_visitor::fs_visitor(struct brw_context *brw, this->spilled_any_registers = false; - this->param_size = rzalloc_array(mem_ctx, int, stage_prog_data->nr_params); + if (dispatch_width == 8) + this->param_size = rzalloc_array(mem_ctx, int, stage_prog_data->nr_params); } fs_visitor::~fs_visitor() -- 2.30.2