From: Marek Olšák Date: Sat, 22 Apr 2017 19:12:08 +0000 (+0200) Subject: radeonsi: fix tess offchip offset for per-patch attributes X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=0490074cab0393b1128d1f316948c7bdd6d45e57;p=mesa.git radeonsi: fix tess offchip offset for per-patch attributes We need 4 more bits there. I don't know what is fixed by this. Reviewed-by: Nicolai Hähnle --- diff --git a/src/gallium/drivers/radeonsi/si_shader.c b/src/gallium/drivers/radeonsi/si_shader.c index f917cf50522..a330b3043c5 100644 --- a/src/gallium/drivers/radeonsi/si_shader.c +++ b/src/gallium/drivers/radeonsi/si_shader.c @@ -715,8 +715,8 @@ static LLVMValueRef get_tcs_tes_buffer_address(struct si_shader_context *ctx, LLVMValueRef base_addr, vertices_per_patch, num_patches, total_vertices; LLVMValueRef param_stride, constant16; - vertices_per_patch = unpack_param(ctx, ctx->param_tcs_offchip_layout, 9, 6); - num_patches = unpack_param(ctx, ctx->param_tcs_offchip_layout, 0, 9); + vertices_per_patch = unpack_param(ctx, ctx->param_tcs_offchip_layout, 6, 6); + num_patches = unpack_param(ctx, ctx->param_tcs_offchip_layout, 0, 6); total_vertices = LLVMBuildMul(gallivm->builder, vertices_per_patch, num_patches, ""); @@ -742,7 +742,7 @@ static LLVMValueRef get_tcs_tes_buffer_address(struct si_shader_context *ctx, if (!vertex_index) { LLVMValueRef patch_data_offset = - unpack_param(ctx, ctx->param_tcs_offchip_layout, 16, 16); + unpack_param(ctx, ctx->param_tcs_offchip_layout, 12, 20); base_addr = LLVMBuildAdd(gallivm->builder, base_addr, patch_data_offset, ""); @@ -1546,7 +1546,7 @@ static void declare_system_value(struct si_shader_context *ctx, if (ctx->type == PIPE_SHADER_TESS_CTRL) value = unpack_param(ctx, ctx->param_tcs_out_lds_layout, 26, 6); else if (ctx->type == PIPE_SHADER_TESS_EVAL) - value = unpack_param(ctx, ctx->param_tcs_offchip_layout, 9, 7); + value = unpack_param(ctx, ctx->param_tcs_offchip_layout, 6, 6); else assert(!"invalid shader stage for TGSI_SEMANTIC_VERTICESIN"); break; diff --git a/src/gallium/drivers/radeonsi/si_shader_internal.h b/src/gallium/drivers/radeonsi/si_shader_internal.h index 8a9bf109af7..954b83db201 100644 --- a/src/gallium/drivers/radeonsi/si_shader_internal.h +++ b/src/gallium/drivers/radeonsi/si_shader_internal.h @@ -130,7 +130,7 @@ struct si_shader_context { * [0] = clamp vertex color * [1] = indexed * [8:20] = stride between patches in DW = num_inputs * num_vertices * 4 - * max = 32*32*4 + * max = 32*32*4 + 32*4 * [24:31] = stride between vertices in DW = num_inputs * 4 * max = 32*4 */ @@ -142,21 +142,26 @@ struct si_shader_context { /* API TCS & TES */ /* Layout of TCS outputs in the offchip buffer - * [0:8] = the number of patches per threadgroup. - * [9:15] = the number of output vertices per patch. - * [16:31] = the offset of per patch attributes in the buffer in bytes. */ + * # 6 bits + * [0:5] = the number of patches per threadgroup, max = NUM_PATCHES (40) + * # 6 bits + * [6:11] = the number of output vertices per patch, max = 32 + * # 20 bits + * [12:31] = the offset of per patch attributes in the buffer in bytes. + * max = NUM_PATCHES*32*32*16 + */ int param_tcs_offchip_layout; /* API TCS */ /* Offsets where TCS outputs and TCS patch outputs live in LDS: * [0:15] = TCS output patch0 offset / 16, max = NUM_PATCHES * 32 * 32 * [16:31] = TCS output patch0 offset for per-patch / 16 - * max = NUM_PATCHES*32*32* + 32*32 + * max = (NUM_PATCHES + 1) * 32*32 */ int param_tcs_out_lds_offsets; /* Layout of TCS outputs / TES inputs: * [0:12] = stride between output patches in DW, num_outputs * num_vertices * 4 - * max = 32*32*4 + * max = 32*32*4 + 32*4 * [13:20] = stride between output vertices in DW = num_inputs * 4 * max = 32*4 * [26:31] = gl_PatchVerticesIn, max = 32 diff --git a/src/gallium/drivers/radeonsi/si_state_draw.c b/src/gallium/drivers/radeonsi/si_state_draw.c index 6a756c4bc5b..2d5a08e878a 100644 --- a/src/gallium/drivers/radeonsi/si_state_draw.c +++ b/src/gallium/drivers/radeonsi/si_state_draw.c @@ -211,8 +211,9 @@ static void si_emit_derived_tess_state(struct si_context *sctx, ((output_vertex_size / 4) << 13); tcs_out_offsets = (output_patch0_offset / 16) | ((perpatch_output_offset / 16) << 16); - offchip_layout = (pervertex_output_patch_size * *num_patches << 16) | - (num_tcs_output_cp << 9) | *num_patches; + offchip_layout = *num_patches | + (num_tcs_output_cp << 6) | + (pervertex_output_patch_size * *num_patches << 12); /* Compute the LDS size. */ lds_size = output_patch0_offset + output_patch_size * *num_patches;