radeonsi: don't preload constants at the beginning of shaders

LLVM can CSE the loads, thus we can always re-load constants before each
use. The decrease in SGPR spilling is huge.

The best improvements are the dumbest ones.

26011 shaders in 14651 tests
Totals:
SGPRS: 1453346 -> 1251920 (-13.86 %)
VGPRS: 742576 -> 728421 (-1.91 %)
Spilled SGPRs: 52298 -> 16644 (-68.17 %)
Spilled VGPRs: 397 -> 369 (-7.05 %)
Scratch VGPRs: 1372 -> 1344 (-2.04 %) dwords per thread
Code Size: 36136488 -> 36001064 (-0.37 %) bytes
LDS: 767 -> 767 (0.00 %) blocks
Max Waves: 219315 -> 222221 (1.33 %)

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>

diff --git a/src/gallium/drivers/radeonsi/si_shader.c b/src/gallium/drivers/radeonsi/si_shader.c
index 0b7de182cdb939a67b6524ff0e7f5544320abef6..68017221391d1bd5b64815ac9eb6bb26fef6f43c 100644 (file)
--- a/src/gallium/drivers/radeonsi/si_shader.c
+++ b/src/gallium/drivers/radeonsi/si_shader.c
@@ -107,7 +107,6 @@ struct si_shader_context
 
        LLVMValueRef const_buffers[SI_NUM_CONST_BUFFERS];
        LLVMValueRef lds;
-       LLVMValueRef *constants[SI_NUM_CONST_BUFFERS];
        LLVMValueRef shader_buffers[SI_NUM_SHADER_BUFFERS];
        LLVMValueRef sampler_views[SI_NUM_SAMPLERS];
        LLVMValueRef sampler_states[SI_NUM_SAMPLERS];
@@ -1881,12 +1880,19 @@ static LLVMValueRef fetch_constant(
        idx = reg->Register.Index * 4 + swizzle;
 
        if (!reg->Register.Indirect && !reg->Dimension.Indirect) {
+               LLVMValueRef c0, c1;
+
+               c0 = buffer_load_const(ctx, ctx->const_buffers[buf],
+                                      LLVMConstInt(ctx->i32, idx * 4, 0));
+
                if (!tgsi_type_is_64bit(type))
-                       return bitcast(bld_base, type, ctx->constants[buf][idx]);
+                       return bitcast(bld_base, type, c0);
                else {
+                       c1 = buffer_load_const(ctx, ctx->const_buffers[buf],
+                                              LLVMConstInt(ctx->i32,
+                                                           (idx + 1) * 4, 0));
                        return radeon_llvm_emit_fetch_64bit(bld_base, type,
-                                                           ctx->constants[buf][idx],
-                                                           ctx->constants[buf][idx + 1]);
+                                                           c0, c1);
                }
        }
 
@@ -5796,25 +5802,12 @@ static void preload_constants(struct si_shader_context *ctx)
        LLVMValueRef ptr = LLVMGetParam(ctx->radeon_bld.main_fn, SI_PARAM_CONST_BUFFERS);
 
        for (buf = 0; buf < SI_NUM_CONST_BUFFERS; buf++) {
-               unsigned i, num_const = info->const_file_max[buf] + 1;
-
-               if (num_const == 0)
+               if (info->const_file_max[buf] == -1)
                        continue;
 
-               /* Allocate space for the constant values */
-               ctx->constants[buf] = CALLOC(num_const * 4, sizeof(LLVMValueRef));
-
                /* Load the resource descriptor */
                ctx->const_buffers[buf] =
                        build_indexed_load_const(ctx, ptr, lp_build_const_int32(gallivm, buf));
-
-               /* Load the constants, we rely on the code sinking to do the rest */
-               for (i = 0; i < num_const * 4; ++i) {
-                       ctx->constants[buf][i] =
-                               buffer_load_const(ctx,
-                                       ctx->const_buffers[buf],
-                                       lp_build_const_int32(gallivm, i * 4));
-               }
        }
 }
 
@@ -6905,8 +6898,6 @@ int si_compile_tgsi_shader(struct si_screen *sscreen,
        }
 
 out:
-       for (int i = 0; i < SI_NUM_CONST_BUFFERS; i++)
-               FREE(ctx.constants[i]);
        return r;
 }