diff options
| -rw-r--r-- | src/gallium/drivers/radeonsi/si_shader.c | 8 | ||||
| -rw-r--r-- | src/gallium/drivers/radeonsi/si_shader.h | 3 | ||||
| -rw-r--r-- | src/gallium/drivers/radeonsi/si_state_shaders.c | 5 |
3 files changed, 10 insertions, 6 deletions
diff --git a/src/gallium/drivers/radeonsi/si_shader.c b/src/gallium/drivers/radeonsi/si_shader.c index a6b7e5ebf53..4ee4a64f531 100644 --- a/src/gallium/drivers/radeonsi/si_shader.c +++ b/src/gallium/drivers/radeonsi/si_shader.c @@ -2276,6 +2276,7 @@ static void si_llvm_export_vs(struct lp_build_tgsi_context *bld_base, semantic_name = outputs[i].semantic_name; semantic_index = outputs[i].semantic_index; bool export_param = true; + unsigned id; switch (semantic_name) { case TGSI_SEMANTIC_POSITION: /* ignore these */ @@ -2289,8 +2290,8 @@ static void si_llvm_export_vs(struct lp_build_tgsi_context *bld_base, break; /* fall through */ default: - if (shader->key.opt.kill_outputs & - (1ull << si_shader_io_get_unique_index(semantic_name, semantic_index))) + id = si_shader_io_get_unique_index(semantic_name, semantic_index); + if (shader->key.opt.kill_outputs[id / 32] & (1u << (id % 32))) export_param = false; } @@ -5335,7 +5336,8 @@ static void si_dump_shader_key(unsigned processor, const struct si_shader *shade processor == PIPE_SHADER_TESS_EVAL || processor == PIPE_SHADER_VERTEX) && !key->as_es && !key->as_ls) { - fprintf(f, " opt.kill_outputs = 0x%"PRIx64"\n", key->opt.kill_outputs); + fprintf(f, " opt.kill_outputs[0] = 0x%x\n", key->opt.kill_outputs[0]); + fprintf(f, " opt.kill_outputs[1] = 0x%x\n", key->opt.kill_outputs[1]); fprintf(f, " opt.clip_disable = %u\n", key->opt.clip_disable); } } diff --git a/src/gallium/drivers/radeonsi/si_shader.h b/src/gallium/drivers/radeonsi/si_shader.h index de520a27363..76e09b27e7c 100644 --- a/src/gallium/drivers/radeonsi/si_shader.h +++ b/src/gallium/drivers/radeonsi/si_shader.h @@ -501,7 +501,8 @@ struct si_shader_key { /* Optimization flags for asynchronous compilation only. */ struct { /* For HW VS (it can be VS, TES, GS) */ - uint64_t kill_outputs; /* "get_unique_index" bits */ + /* Don't use "uint64_t" in order to get 32-bit alignment. */ + uint32_t kill_outputs[2]; /* "get_unique_index" bits */ unsigned clip_disable:1; /* For shaders where monolithic variants have better code. diff --git a/src/gallium/drivers/radeonsi/si_state_shaders.c b/src/gallium/drivers/radeonsi/si_state_shaders.c index 07e6a428575..15e46b5a5e8 100644 --- a/src/gallium/drivers/radeonsi/si_state_shaders.c +++ b/src/gallium/drivers/radeonsi/si_state_shaders.c @@ -1241,9 +1241,10 @@ static void si_shader_selector_key_hw_vs(struct si_context *sctx, inputs_read = ps->inputs_read; } - uint64_t linked = outputs_written & inputs_read; + uint64_t kill_outputs = ~(outputs_written & inputs_read) & outputs_written; - key->opt.kill_outputs = ~linked & outputs_written; + key->opt.kill_outputs[0] = kill_outputs; + key->opt.kill_outputs[1] = kill_outputs >> 32; } /* Compute the key for the hw shader variant */ |