unsigned int level, unsigned int layer,
enum isl_format format, bool is_render_target)
{
+ assert(level < surf->surf->levels);
+ assert(layer < MAX2(surf->surf->logical_level0_px.depth >> level,
+ surf->surf->logical_level0_px.array_len));
+
info->enabled = true;
if (format == ISL_FORMAT_UNSUPPORTED)
format = surf->surf->format;
- if (format == ISL_FORMAT_R24_UNORM_X8_TYPELESS) {
- /* Unfortunately, ISL_FORMAT_R24_UNORM_X8_TYPELESS it isn't supported as
- * a render target, which would prevent us from blitting to 24-bit
- * depth. The miptree consists of 32 bits per pixel, arranged as 24-bit
- * depth values interleaved with 8 "don't care" bits. Since depth
- * values don't require any blending, it doesn't matter how we interpret
- * the bit pattern as long as we copy the right amount of data, so just
- * map it as 8-bit BGRA.
- */
- format = ISL_FORMAT_B8G8R8A8_UNORM;
- }
-
info->surf = *surf->surf;
info->addr = surf->addr;
if (info->aux_usage != ISL_AUX_USAGE_NONE) {
info->aux_surf = *surf->aux_surf;
info->aux_addr = surf->aux_addr;
+ assert(level < info->aux_surf.levels);
+ assert(layer < MAX2(info->aux_surf.logical_level0_px.depth >> level,
+ info->aux_surf.logical_level0_px.array_len));
}
info->clear_color = surf->clear_color;
+ info->clear_color_addr = surf->clear_color_addr;
info->view = (struct isl_view) {
.usage = is_render_target ? ISL_SURF_USAGE_RENDER_TARGET_BIT :
info->z_offset = 0;
}
- /* Sandy Bridge has a limit of a maximum of 512 layers for layered
- * rendering.
+ /* Sandy Bridge and earlier have a limit of a maximum of 512 layers for
+ * layered rendering.
*/
- if (is_render_target && blorp->isl_dev->info->gen == 6)
+ if (is_render_target && blorp->isl_dev->info->gen <= 6)
info->view.array_len = MIN2(info->view.array_len, 512);
+
+ if (surf->tile_x_sa || surf->tile_y_sa) {
+ /* This is only allowed on simple 2D surfaces without MSAA */
+ assert(info->surf.dim == ISL_SURF_DIM_2D);
+ assert(info->surf.samples == 1);
+ assert(info->surf.levels == 1);
+ assert(info->surf.logical_level0_px.array_len == 1);
+ assert(info->aux_usage == ISL_AUX_USAGE_NONE);
+
+ info->tile_x_sa = surf->tile_x_sa;
+ info->tile_y_sa = surf->tile_y_sa;
+
+ /* Instead of using the X/Y Offset fields in RENDER_SURFACE_STATE, we
+ * place the image at the tile boundary and offset our sampling or
+ * rendering. For this reason, we need to grow the image by the offset
+ * to ensure that the hardware doesn't think we've gone past the edge.
+ */
+ info->surf.logical_level0_px.w += surf->tile_x_sa;
+ info->surf.logical_level0_px.h += surf->tile_y_sa;
+ info->surf.phys_level0_sa.w += surf->tile_x_sa;
+ info->surf.phys_level0_sa.h += surf->tile_y_sa;
+ }
}
const unsigned *
blorp_compile_fs(struct blorp_context *blorp, void *mem_ctx,
struct nir_shader *nir,
- const struct brw_wm_prog_key *wm_key,
+ struct brw_wm_prog_key *wm_key,
bool use_repclear,
- struct brw_wm_prog_data *wm_prog_data,
- unsigned *program_size)
+ struct brw_wm_prog_data *wm_prog_data)
{
const struct brw_compiler *compiler = blorp->compiler;
wm_prog_data->base.nr_params = 0;
wm_prog_data->base.param = NULL;
- /* BLORP always just uses the first two binding table entries */
- wm_prog_data->binding_table.render_target_start = BLORP_RENDERBUFFER_BT_INDEX;
+ /* BLORP always uses the first two binding table entries:
+ * - Surface 0 is the render target (which always start from 0)
+ * - Surface 1 is the source texture
+ */
wm_prog_data->base.binding_table.texture_start = BLORP_TEXTURE_BT_INDEX;
- nir = brw_preprocess_nir(compiler, nir);
+ brw_preprocess_nir(compiler, nir, NULL);
nir_remove_dead_variables(nir, nir_var_shader_in);
nir_shader_gather_info(nir, nir_shader_get_entrypoint(nir));
+ if (blorp->compiler->devinfo->gen < 6) {
+ if (nir->info.fs.uses_discard)
+ wm_key->iz_lookup |= BRW_WM_IZ_PS_KILL_ALPHATEST_BIT;
+
+ wm_key->input_slots_valid = nir->info.inputs_read | VARYING_BIT_POS;
+ }
+
const unsigned *program =
brw_compile_fs(compiler, blorp->driver_ctx, mem_ctx, wm_key,
- wm_prog_data, nir, NULL, -1, -1, false, use_repclear,
- NULL, program_size, NULL);
+ wm_prog_data, nir, NULL, -1, -1, -1, false, use_repclear,
+ NULL, NULL);
return program;
}
const unsigned *
blorp_compile_vs(struct blorp_context *blorp, void *mem_ctx,
struct nir_shader *nir,
- struct brw_vs_prog_data *vs_prog_data,
- unsigned *program_size)
+ struct brw_vs_prog_data *vs_prog_data)
{
const struct brw_compiler *compiler = blorp->compiler;
nir->options =
compiler->glsl_compiler_options[MESA_SHADER_VERTEX].NirOptions;
- nir = brw_preprocess_nir(compiler, nir);
+ brw_preprocess_nir(compiler, nir, NULL);
nir_shader_gather_info(nir, nir_shader_get_entrypoint(nir));
vs_prog_data->inputs_read = nir->info.inputs_read;
const unsigned *program =
brw_compile_vs(compiler, blorp->driver_ctx, mem_ctx,
- &vs_key, vs_prog_data, nir,
- NULL, false, -1, program_size, NULL);
+ &vs_key, vs_prog_data, nir, -1, NULL);
return program;
}
};
bool
-blorp_ensure_sf_program(struct blorp_context *blorp,
+blorp_ensure_sf_program(struct blorp_batch *batch,
struct blorp_params *params)
{
+ struct blorp_context *blorp = batch->blorp;
const struct brw_wm_prog_data *wm_prog_data = params->wm_prog_data;
assert(params->wm_prog_data);
memcpy(key.key.interp_mode, wm_prog_data->interp_mode,
sizeof(key.key.interp_mode));
- if (blorp->lookup_shader(blorp, &key, sizeof(key),
+ if (blorp->lookup_shader(batch, &key, sizeof(key),
¶ms->sf_prog_kernel, ¶ms->sf_prog_data))
return true;
&prog_data_tmp, &vue_map, &program_size);
bool result =
- blorp->upload_shader(blorp, &key, sizeof(key), program, program_size,
+ blorp->upload_shader(batch, &key, sizeof(key), program, program_size,
(void *)&prog_data_tmp, sizeof(prog_data_tmp),
¶ms->sf_prog_kernel, ¶ms->sf_prog_data);
}
void
-blorp_gen6_hiz_op(struct blorp_batch *batch,
- struct blorp_surf *surf, unsigned level, unsigned layer,
- enum blorp_hiz_op op)
+blorp_hiz_op(struct blorp_batch *batch, struct blorp_surf *surf,
+ uint32_t level, uint32_t start_layer, uint32_t num_layers,
+ enum isl_aux_op op)
{
struct blorp_params params;
blorp_params_init(¶ms);
params.hiz_op = op;
-
- brw_blorp_surface_info_init(batch->blorp, ¶ms.depth, surf, level, layer,
- surf->surf->format, true);
-
- /* Align the rectangle primitive to 8x4 pixels.
- *
- * During fast depth clears, the emitted rectangle primitive must be
- * aligned to 8x4 pixels. From the Ivybridge PRM, Vol 2 Part 1 Section
- * 11.5.3.1 Depth Buffer Clear (and the matching section in the Sandybridge
- * PRM):
- * If Number of Multisamples is NUMSAMPLES_1, the rectangle must be
- * aligned to an 8x4 pixel block relative to the upper left corner
- * of the depth buffer [...]
- *
- * For hiz resolves, the rectangle must also be 8x4 aligned. Item
- * WaHizAmbiguate8x4Aligned from the Haswell workarounds page and the
- * Ivybridge simulator require the alignment.
- *
- * To be safe, let's just align the rect for all hiz operations and all
- * hardware generations.
- *
- * However, for some miptree slices of a Z24 texture, emitting an 8x4
- * aligned rectangle that covers the slice may clobber adjacent slices if
- * we strictly adhered to the texture alignments specified in the PRM. The
- * Ivybridge PRM, Section "Alignment Unit Size", states that
- * SURFACE_STATE.Surface_Horizontal_Alignment should be 4 for Z24 surfaces,
- * not 8. But commit 1f112cc increased the alignment from 4 to 8, which
- * prevents the clobbering.
- */
- params.x1 = minify(params.depth.surf.logical_level0_px.width,
- params.depth.view.base_level);
- params.y1 = minify(params.depth.surf.logical_level0_px.height,
- params.depth.view.base_level);
- params.x1 = ALIGN(params.x1, 8);
- params.y1 = ALIGN(params.y1, 4);
-
- if (params.depth.view.base_level == 0) {
- /* TODO: What about MSAA? */
- params.depth.surf.logical_level0_px.width = params.x1;
- params.depth.surf.logical_level0_px.height = params.y1;
+ params.full_surface_hiz_op = true;
+
+ for (uint32_t a = 0; a < num_layers; a++) {
+ const uint32_t layer = start_layer + a;
+
+ brw_blorp_surface_info_init(batch->blorp, ¶ms.depth, surf, level,
+ layer, surf->surf->format, true);
+
+ /* Align the rectangle primitive to 8x4 pixels.
+ *
+ * During fast depth clears, the emitted rectangle primitive must be
+ * aligned to 8x4 pixels. From the Ivybridge PRM, Vol 2 Part 1 Section
+ * 11.5.3.1 Depth Buffer Clear (and the matching section in the
+ * Sandybridge PRM):
+ *
+ * If Number of Multisamples is NUMSAMPLES_1, the rectangle must be
+ * aligned to an 8x4 pixel block relative to the upper left corner
+ * of the depth buffer [...]
+ *
+ * For hiz resolves, the rectangle must also be 8x4 aligned. Item
+ * WaHizAmbiguate8x4Aligned from the Haswell workarounds page and the
+ * Ivybridge simulator require the alignment.
+ *
+ * To be safe, let's just align the rect for all hiz operations and all
+ * hardware generations.
+ *
+ * However, for some miptree slices of a Z24 texture, emitting an 8x4
+ * aligned rectangle that covers the slice may clobber adjacent slices
+ * if we strictly adhered to the texture alignments specified in the
+ * PRM. The Ivybridge PRM, Section "Alignment Unit Size", states that
+ * SURFACE_STATE.Surface_Horizontal_Alignment should be 4 for Z24
+ * surfaces, not 8. But commit 1f112cc increased the alignment from 4 to
+ * 8, which prevents the clobbering.
+ */
+ params.x1 = minify(params.depth.surf.logical_level0_px.width,
+ params.depth.view.base_level);
+ params.y1 = minify(params.depth.surf.logical_level0_px.height,
+ params.depth.view.base_level);
+ params.x1 = ALIGN(params.x1, 8);
+ params.y1 = ALIGN(params.y1, 4);
+
+ if (params.depth.view.base_level == 0) {
+ /* TODO: What about MSAA? */
+ params.depth.surf.logical_level0_px.width = params.x1;
+ params.depth.surf.logical_level0_px.height = params.y1;
+ }
+
+ params.dst.surf.samples = params.depth.surf.samples;
+ params.dst.surf.logical_level0_px = params.depth.surf.logical_level0_px;
+ params.depth_format =
+ isl_format_get_depth_format(surf->surf->format, false);
+ params.num_samples = params.depth.surf.samples;
+
+ batch->blorp->exec(batch, ¶ms);
}
-
- params.dst.surf.samples = params.depth.surf.samples;
- params.dst.surf.logical_level0_px = params.depth.surf.logical_level0_px;
- params.depth_format = isl_format_get_depth_format(surf->surf->format, false);
- params.num_samples = params.depth.surf.samples;
-
- batch->blorp->exec(batch, ¶ms);
}