[Mesa-dev] [PATCH v4 13/34] i965/blorp: Add a generic ISL-based surface state emit path
Chad Versace
chad.versace at intel.com
Thu Jul 14 21:26:21 UTC 2016
On Wed 13 Jul 2016, Jason Ekstrand wrote:
> Reviewed-by: Topi Pohjolainen <topi.pohjolainen at intel.com>
> ---
> src/mesa/drivers/dri/i965/brw_blorp.c | 157 ++++++++++++++++++++++++++++++++++
> src/mesa/drivers/dri/i965/brw_blorp.h | 6 ++
> 2 files changed, 163 insertions(+)
>
> diff --git a/src/mesa/drivers/dri/i965/brw_blorp.c b/src/mesa/drivers/dri/i965/brw_blorp.c
> index 04c10b6..282a5b2 100644
> --- a/src/mesa/drivers/dri/i965/brw_blorp.c
> +++ b/src/mesa/drivers/dri/i965/brw_blorp.c
> @@ -231,6 +231,163 @@ brw_blorp_compile_nir_shader(struct brw_context *brw, struct nir_shader *nir,
> return program;
> }
>
> +static enum isl_msaa_layout
> +get_isl_msaa_layout(enum intel_msaa_layout layout)
> +{
> + switch (layout) {
> + case INTEL_MSAA_LAYOUT_NONE:
> + return ISL_MSAA_LAYOUT_NONE;
> + case INTEL_MSAA_LAYOUT_IMS:
> + return ISL_MSAA_LAYOUT_INTERLEAVED;
> + case INTEL_MSAA_LAYOUT_UMS:
> + case INTEL_MSAA_LAYOUT_CMS:
> + return ISL_MSAA_LAYOUT_ARRAY;
> + default:
> + unreachable("Invalid MSAA layout");
> + }
> +}
Lookup table? With nchery-style validation? Just a suggestion.
> +
> +struct surface_state_info {
> + unsigned num_dwords;
> + unsigned ss_align; /* Required alignment of RENDER_SURFACE_STATE in bytes */
> + unsigned reloc_dw;
> + unsigned aux_reloc_dw;
> + unsigned tex_mocs;
> + unsigned rb_mocs;
> +};
> +
> +static const struct surface_state_info surface_state_infos[] = {
> + [6] = {6, 32, 1, 0},
> + [7] = {8, 32, 1, 6, GEN7_MOCS_L3, GEN7_MOCS_L3},
> + [8] = {13, 64, 8, 10, BDW_MOCS_WB, BDW_MOCS_PTE},
> + [9] = {16, 64, 8, 10, SKL_MOCS_WB, SKL_MOCS_PTE},
> +};
> +
> +uint32_t
> +brw_blorp_emit_surface_state(struct brw_context *brw,
> + const struct brw_blorp_surface_info *surface,
> + uint32_t read_domains, uint32_t write_domain,
> + bool is_render_target)
> +{
> + const struct surface_state_info ss_info = surface_state_infos[brw->gen];
> +
> + struct isl_surf surf;
> + intel_miptree_get_isl_surf(brw, surface->mt, &surf);
> +
> + /* Stomp surface dimensions and tiling (if needed) with info from blorp */
> + surf.dim = ISL_SURF_DIM_2D;
> + surf.dim_layout = ISL_DIM_LAYOUT_GEN4_2D;
> + surf.msaa_layout = get_isl_msaa_layout(surface->msaa_layout);
> + surf.logical_level0_px.width = surface->width;
> + surf.logical_level0_px.height = surface->height;
> + surf.logical_level0_px.depth = 1;
> + surf.logical_level0_px.array_len = 1;
> + surf.levels = 1;
> + surf.samples = MAX2(surface->num_samples, 1);
This scares me. I'm really really scared.
/me closes eyes and continues
> +
> + /* Alignment doesn't matter since we have 1 miplevel and 1 array slice so
> + * just pick something that works for everybody.
> + */
> + surf.image_alignment_el = isl_extent3d(4, 4, 1);
> +
> + if (brw->gen == 6 && surface->num_samples > 1) {
> + /* Since gen6 uses INTEL_MSAA_LAYOUT_IMS, width and height are measured
> + * in samples. But SURFACE_STATE wants them in pixels, so we need to
> + * divide them each by 2.
> + */
> + surf.logical_level0_px.width /= 2;
> + surf.logical_level0_px.height /= 2;
> + }
> +
> + if (brw->gen == 6 && surf.image_alignment_el.height > 4) {
> + /* This can happen on stencil buffers on Sandy Bridge due to the
> + * single-LOD work-around. It's fairly harmless as long as we don't
> + * pass a bogus value into isl_surf_fill_state().
> + */
> + surf.image_alignment_el = isl_extent3d(4, 2, 1);
> + }
> +
> + /* We need to fake W-tiling with Y-tiling */
> + if (surface->map_stencil_as_y_tiled)
> + surf.tiling = ISL_TILING_Y0;
> +
> + union isl_color_value clear_color = { .u32 = { 0, 0, 0, 0 } };
> +
> + struct isl_surf *aux_surf = NULL, aux_surf_s;
> + uint64_t aux_offset = 0;
> + enum isl_aux_usage aux_usage = ISL_AUX_USAGE_NONE;
> + if (surface->mt->mcs_mt) {
> + /* We should probably to similar stomping to above but most of the aux
Correction:
We should probably do similar stomping as above but most of the aux
or
We should probably do stomping similar to the above but most of the aux
> + * surf gets ignored when we fill out the surface state anyway so
> + * there's no point.
> + */
I'm scared again. Lovecraft wrote that comment.
/me shuts eyes tighter
> + intel_miptree_get_aux_isl_surf(brw, surface->mt, &aux_surf_s, &aux_usage);
> + aux_surf = &aux_surf_s;
> + assert(surface->mt->mcs_mt->offset == 0);
> + aux_offset = surface->mt->mcs_mt->bo->offset64/
> +
> + /* We only really need a clear color if we also have an auxiliary
> + * surface. Without one, it does nothing.
> + */
> + clear_color = intel_miptree_get_isl_clear_color(brw, surface->mt);
> + }
> +
> + struct isl_view view = {
> + .format = surface->brw_surfaceformat,
> + .base_level = 0,
> + .levels = 1,
> + .base_array_layer = 0,
> + .array_len = 1,
> + .channel_select = {
> + ISL_CHANNEL_SELECT_RED,
> + ISL_CHANNEL_SELECT_GREEN,
> + ISL_CHANNEL_SELECT_BLUE,
> + ISL_CHANNEL_SELECT_ALPHA,
> + },
> + .usage = is_render_target ? ISL_SURF_USAGE_RENDER_TARGET_BIT :
> + ISL_SURF_USAGE_TEXTURE_BIT,
> + };
Are you confident that blorp never requires ISL_SURF_USAGE_CUBE_MAP_BIT?
> +
> + uint32_t offset, tile_x, tile_y;
> + offset = brw_blorp_compute_tile_offsets(surface, &tile_x, &tile_y);
> +
> + uint32_t surf_offset;
> + uint32_t *dw = brw_state_batch(brw, AUB_TRACE_SURFACE_STATE,
> + ss_info.num_dwords * 4, ss_info.ss_align,
> + &surf_offset);
> +
> + const uint32_t mocs = is_render_target ? ss_info.rb_mocs : ss_info.tex_mocs;
> +
> + isl_surf_fill_state(&brw->isl_dev, dw, .surf = &surf, .view = &view,
> + .address = surface->mt->bo->offset64 + offset,
> + .aux_surf = aux_surf, .aux_usage = aux_usage,
> + .aux_address = aux_offset,
> + .mocs = mocs, .clear_color = clear_color,
> + .x_offset_sa = tile_x, .y_offset_sa = tile_y);
Whoa! Where did x_offset_sa and y_offset_sa come from? It's not on
master, and I didn't see it in any earlier patches.
And now I skip ahead to patch 21, as you recommended.
> +
> + /* Emit relocation to surface contents */
> + drm_intel_bo_emit_reloc(brw->batch.bo,
> + surf_offset + ss_info.reloc_dw * 4,
> + surface->mt->bo,
> + dw[ss_info.reloc_dw] - surface->mt->bo->offset64,
> + read_domains, write_domain);
> +
> + if (aux_surf) {
> + /* On gen7 and prior, the bottom 12 bits of the MCS base address are
> + * used to store other information. This should be ok, however, because
> + * surface buffer addresses are always 4K page alinged.
> + */
> + assert((aux_offset & 0xfff) == 0);
> + drm_intel_bo_emit_reloc(brw->batch.bo,
> + surf_offset + ss_info.aux_reloc_dw * 4,
> + surface->mt->mcs_mt->bo,
> + dw[ss_info.aux_reloc_dw] & 0xfff,
> + read_domains, write_domain);
> + }
> +
> + return surf_offset;
> +}
> +
> /**
> * Perform a HiZ or depth resolve operation.
> *
> diff --git a/src/mesa/drivers/dri/i965/brw_blorp.h b/src/mesa/drivers/dri/i965/brw_blorp.h
> index c8f6219..beef90e 100644
> --- a/src/mesa/drivers/dri/i965/brw_blorp.h
> +++ b/src/mesa/drivers/dri/i965/brw_blorp.h
> @@ -400,6 +400,12 @@ brw_blorp_compile_nir_shader(struct brw_context *brw, struct nir_shader *nir,
> struct brw_blorp_prog_data *prog_data,
> unsigned *program_size);
>
> +uint32_t
> +brw_blorp_emit_surface_state(struct brw_context *brw,
> + const struct brw_blorp_surface_info *surface,
> + uint32_t read_domains, uint32_t write_domain,
> + bool is_render_target);
> +
> void
> gen6_blorp_init(struct brw_context *brw);
More information about the mesa-dev
mailing list