<div dir="ltr"> <div class="gmail_extra"> <div class="gmail_quote">On Wed, May 11, 2016 at 7:42 PM, Jason Ekstrand <<a href="mailto:jason@jlekstrand.net" target="_blank">jason@jlekstrand.net</a>> wrote: <blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">The current MSAA resolve code has a special-case for if the MCS value is 0. In this case we can only sample once because we know that all values are in slice 0. This commit adds a second optimization that detecs the magic MCS value that indicates the clear color and grabs the color from a push constant and avoids sampling altogether. On a microbenchmark written by Neil Roberts that tests resolving surfaces with just clear color, this improves performance by 60% for 8x, 40% for 4x, and 28% for 2x MSAA on my SKL gte3 laptop. The benchmark can be found on the ML archive: <a href="https://lists.freedesktop.org/archives/mesa-dev/2016-February/108077.html" rel="noreferrer" target="_blank">https://lists.freedesktop.org/archives/mesa-dev/2016-February/108077.html</a> --- src/mesa/drivers/dri/i965/brw_blorp.h | 4 +- src/mesa/drivers/dri/i965/brw_blorp_blit.cpp | 101 +++++++++++++++++++++++++-- 2 files changed, 100 insertions(+), 5 deletions(-) diff --git a/src/mesa/drivers/dri/i965/brw_blorp.h b/src/mesa/drivers/dri/i965/brw_blorp.h index 15114d0..9d71ca4 100644 --- a/src/mesa/drivers/dri/i965/brw_blorp.h +++ b/src/mesa/drivers/dri/i965/brw_blorp.h @@ -197,7 +197,9 @@ struct brw_blorp_wm_push_constants uint32_t src_z; /* Pad out to an integral number of registers */ - uint32_t pad[5]; + uint32_t pad; + + union gl_color_union clear_color; }; #define BRW_BLORP_NUM_PUSH_CONSTANT_DWORDS \ diff --git a/src/mesa/drivers/dri/i965/brw_blorp_blit.cpp b/src/mesa/drivers/dri/i965/brw_blorp_blit.cpp index 514a316..45b696d 100644 <div><div class="h5">--- a/src/mesa/drivers/dri/i965/brw_blorp_blit.cpp +++ b/src/mesa/drivers/dri/i965/brw_blorp_blit.cpp @@ -346,6 +346,7 @@ struct brw_blorp_blit_vars { nir_variable *offset; } u_x_transform, u_y_transform; nir_variable *u_src_z; + nir_variable *u_clear_color; /* gl_FragCoord */ nir_variable *frag_coord; @@ -374,6 +375,7 @@ brw_blorp_blit_vars_init(nir_builder *b, struct brw_blorp_blit_vars *v, LOAD_UNIFORM(y_transform.multiplier, glsl_float_type()) LOAD_UNIFORM(y_transform.offset, glsl_float_type()) LOAD_UNIFORM(src_z, glsl_uint_type()) + LOAD_UNIFORM(clear_color, glsl_vec4_type()) #undef DECL_UNIFORM @@ -858,7 +860,8 @@ static nir_ssa_def * blorp_nir_manual_blend_average(nir_builder *b, nir_ssa_def *pos, unsigned tex_samples, enum intel_msaa_layout tex_layout, - enum brw_reg_type dst_type) + enum brw_reg_type dst_type, + struct brw_blorp_blit_vars *v) { /* If non-null, this is the outer-most if statement */ nir_if *outer_if = NULL; @@ -867,9 +870,53 @@ blorp_nir_manual_blend_average(nir_builder *b, nir_ssa_def *pos, nir_local_variable_create(b->impl, glsl_vec4_type(), "color"); nir_ssa_def *mcs = NULL; - if (tex_layout == INTEL_MSAA_LAYOUT_CMS) + if (tex_layout == INTEL_MSAA_LAYOUT_CMS) { mcs = blorp_nir_txf_ms_mcs(b, pos); + /* The MCS buffer stores a packed value that provides a mapping from + * samples to array slices. The magic value of all ones means that all + * samples have the clear color. In this case, we can short-circuit the + * sampling process and just use the clear color that we pushed into the + * shader. + */ + nir_ssa_def *is_clear_color; + switch (tex_samples) { + case 2: + /* Empirical evidence suggests that the value returned from the + * sampler is not always 0x3 for clear color so we need to mask it. + */ + is_clear_color = + nir_ieq(b, nir_iand(b, nir_channel(b, mcs, 0), nir_imm_int(b, 0x3)), + nir_imm_int(b, 0x3)); + break; + case 4: + is_clear_color = + nir_ieq(b, nir_channel(b, mcs, 0), nir_imm_int(b, 0xff)); + break; + case 8: + is_clear_color = + nir_ieq(b, nir_channel(b, mcs, 0), nir_imm_int(b, ~0)); + break; + case 16: + is_clear_color = + nir_ior(b, nir_ieq(b, nir_channel(b, mcs, 0), nir_imm_int(b, ~0)), </div></div></blockquote><div> </div><div>This needs to be nir_iand. Fixed locally... </div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div><div class="h5"> + nir_ieq(b, nir_channel(b, mcs, 1), nir_imm_int(b, ~0))); + break; + default: + unreachable("Invalid sample count"); + } + + nir_if *if_stmt = nir_if_create(b->shader); + if_stmt->condition = nir_src_for_ssa(is_clear_color); + nir_cf_node_insert(b->cursor, &if_stmt->cf_node); + + b->cursor = nir_after_cf_list(&if_stmt->then_list); + nir_store_var(b, color, nir_load_var(b, v->u_clear_color), 0xf); + + b->cursor = nir_after_cf_list(&if_stmt->else_list); + outer_if = if_stmt; + } + /* We add together samples using a binary tree structure, e.g. for 4x MSAA: * * result = ((sample[0] + sample[1]) + (sample[2] + sample[3])) / 4 @@ -937,7 +984,8 @@ blorp_nir_manual_blend_average(nir_builder *b, nir_ssa_def *pos, nir_store_var(b, color, texture_data[0], 0xf); b->cursor = nir_after_cf_list(&if_stmt->else_list); - outer_if = if_stmt; + if (!outer_if) + outer_if = if_stmt; } for (int j = 0; j < count_trailing_one_bits(i); j++) { </div></div>@@ -1341,7 +1389,7 @@ brw_blorp_build_nir_shader(struct brw_context *brw, /* Gen7+ hardware doesn't automaticaly blend. */ color = blorp_nir_manual_blend_average(&b, src_pos, key->src_samples, key->src_layout, - key->texture_data_type); + key->texture_data_type, &v); } } else if (key->blend && key->blit_scaled) { color = blorp_nir_manual_blend_bilinear(&b, src_pos, key->src_samples, key, &v); @@ -1493,6 +1541,48 @@ compute_msaa_layout_for_pipeline(struct brw_context *brw, unsigned num_samples, } +static union gl_color_union +brw_blorp_get_clear_color_for_mt(struct brw_context *brw, + struct intel_mipmap_tree *mt, + unsigned swizzle) +{ + union gl_color_union color; + if (brw->gen >= 9) { + color = mt->gen9_fast_clear_color; + } else if (_mesa_is_format_integer(mt->format)) { + color.i[0] = (mt->fast_clear_color_value & (1 << 31)) != 0; + color.i[1] = (mt->fast_clear_color_value & (1 << 30)) != 0; + color.i[2] = (mt->fast_clear_color_value & (1 << 29)) != 0; + color.i[3] = (mt->fast_clear_color_value & (1 << 28)) != 0; + } else { + color.f[0] = (mt->fast_clear_color_value & (1 << 31)) != 0; + color.f[1] = (mt->fast_clear_color_value & (1 << 30)) != 0; + color.f[2] = (mt->fast_clear_color_value & (1 << 29)) != 0; + color.f[3] = (mt->fast_clear_color_value & (1 << 28)) != 0; + } + + if (swizzle != SWIZZLE_NOOP) { + union gl_color_union orig_color = color; + for (unsigned i = 0; i < 4; i++) { + unsigned s = GET_SWZ(swizzle, i); + if (s <= SWIZZLE_W) { + color.i[i] = orig_color.i[s]; + } else if (s == SWIZZLE_ZERO) { + color.i[i] = 0; + } else { + assert(s == SWIZZLE_ONE); + if (_mesa_is_format_integer(mt->format)) + color.i[i] = 1; + else + color.f[i] = 1; + } + } + } + + return color; +} + + /** * Note: if the src (or dst) is a 2D multisample array texture on Gen7+ using * INTEL_MSAA_LAYOUT_UMS or INTEL_MSAA_LAYOUT_CMS, src_layer (dst_layer) is @@ -1665,6 +1755,9 @@ brw_blorp_blit_miptrees(struct brw_context *brw, params.src.num_samples <= 1 && params.dst.num_samples <= 1) wm_prog_key.bilinear_filter = true; + params.wm_push_consts.clear_color = + brw_blorp_get_clear_color_for_mt(brw, src_mt, src_swizzle); <div class="HOEnZb"><div class="h5">+ GLenum base_format = _mesa_get_format_base_format(src_mt->format); if (base_format != GL_DEPTH_COMPONENT && /* TODO: what about depth/stencil? */ base_format != GL_STENCIL_INDEX && -- 2.5.0.400.gff86faf </div></div></blockquote></div> </div></div>