[Mesa-dev] [PATCH v3 13/19] i965/gen6 depth surface: program 3DSTATE_DEPTH_BUFFER to top of surface

Pohjolainen, Topi topi.pohjolainen at intel.com
Mon Aug 4 01:41:31 PDT 2014


On Fri, Aug 01, 2014 at 01:27:49PM -0700, Jordan Justen wrote:
> On Fri, Aug 1, 2014 at 1:53 AM, Pohjolainen, Topi
> <topi.pohjolainen at intel.com> wrote:
> > On Fri, Aug 01, 2014 at 12:53:43AM -0700, Jordan Justen wrote:
> >> (bf25ee2 for gen6)
> >>
> >> Previously we would always find the 2D sub-surface of interest,
> >> and then program the surface to this location. Now we always
> >> program the 3DSTATE_DEPTH_BUFFER at the start of the surface.
> >> To select the lod/slice, we utilize the lod & minimum array
> >> element fields.
> >>
> >> We also must disable brw_workaround_depthstencil_alignment for
> >> gen >= 6. Now the hardware will handle alignment when rendering
> >> to additional slices/LODs.
> >>
> >> v3:
> >>  * Set depth_mt bo RELOC offset to 0, as was done in bf25ee2
> >>
> >> Signed-off-by: Jordan Justen <jordan.l.justen at intel.com>
> >> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=56127
> >> ---
> >>  src/mesa/drivers/dri/i965/brw_misc_state.c   |  4 +-
> >>  src/mesa/drivers/dri/i965/gen6_blorp.cpp     | 71 +++++++++-------------------
> >>  src/mesa/drivers/dri/i965/gen6_depth_state.c | 35 ++++++++++----
> >>  3 files changed, 51 insertions(+), 59 deletions(-)
> >>
> >> diff --git a/src/mesa/drivers/dri/i965/brw_misc_state.c b/src/mesa/drivers/dri/i965/brw_misc_state.c
> >> index 76e22bd..e3980fc 100644
> >> --- a/src/mesa/drivers/dri/i965/brw_misc_state.c
> >> +++ b/src/mesa/drivers/dri/i965/brw_misc_state.c
> >> @@ -261,10 +261,10 @@ brw_workaround_depthstencil_alignment(struct brw_context *brw,
> >>     if (stencil_irb)
> >>        brw->depthstencil.stencil_mt = get_stencil_miptree(stencil_irb);
> >>
> >> -   /* Gen7+ doesn't require the workarounds, since we always program the
> >> +   /* Gen6+ doesn't require the workarounds, since we always program the
> >>      * surface state at the start of the whole surface.
> >>      */
> >> -   if (brw->gen >= 7)
> >> +   if (brw->gen >= 6)
> >>        return;
> >>
> >>     /* Check if depth buffer is in depth/stencil format.  If so, then it's only
> >> diff --git a/src/mesa/drivers/dri/i965/gen6_blorp.cpp b/src/mesa/drivers/dri/i965/gen6_blorp.cpp
> >> index dca6bfc..5a56442 100644
> >> --- a/src/mesa/drivers/dri/i965/gen6_blorp.cpp
> >> +++ b/src/mesa/drivers/dri/i965/gen6_blorp.cpp
> >> @@ -787,10 +787,6 @@ static void
> >>  gen6_blorp_emit_depth_stencil_config(struct brw_context *brw,
> >>                                       const brw_blorp_params *params)
> >>  {
> >> -   struct gl_context *ctx = &brw->ctx;
> >> -   uint32_t draw_x = params->depth.x_offset;
> >> -   uint32_t draw_y = params->depth.y_offset;
> >> -   uint32_t tile_mask_x, tile_mask_y;
> >>     uint32_t surfwidth, surfheight;
> >>     uint32_t surftype;
> >>     unsigned int depth = MAX2(params->depth.mt->logical_depth0, 1);
> >> @@ -814,12 +810,6 @@ gen6_blorp_emit_depth_stencil_config(struct brw_context *brw,
> >>        break;
> >>     }
> >>
> >> -   brw_get_depthstencil_tile_masks(params->depth.mt,
> >> -                                   params->depth.level,
> >> -                                   params->depth.layer,
> >> -                                   NULL,
> >> -                                   &tile_mask_x, &tile_mask_y);
> >> -
> >>     min_array_element = params->depth.layer;
> >>
> >>     lod = params->depth.level - params->depth.mt->first_level;
> >> @@ -838,55 +828,42 @@ gen6_blorp_emit_depth_stencil_config(struct brw_context *brw,
> >>
> >>     /* 3DSTATE_DEPTH_BUFFER */
> >>     {
> >> -      uint32_t tile_x = draw_x & tile_mask_x;
> >> -      uint32_t tile_y = draw_y & tile_mask_y;
> >> -      uint32_t offset =
> >> -         intel_miptree_get_aligned_offset(params->depth.mt,
> >> -                                          draw_x & ~tile_mask_x,
> >> -                                          draw_y & ~tile_mask_y, false);
> >> -
> >> -      /* According to the Sandy Bridge PRM, volume 2 part 1, pp326-327
> >> -       * (3DSTATE_DEPTH_BUFFER dw5), in the documentation for "Depth
> >> -       * Coordinate Offset X/Y":
> >> -       *
> >> -       *   "The 3 LSBs of both offsets must be zero to ensure correct
> >> -       *   alignment"
> >> -       *
> >> -       * We have no guarantee that tile_x and tile_y are correctly aligned,
> >> -       * since they are determined by the mipmap layout, which is only aligned
> >> -       * to multiples of 4.
> >> -       *
> >> -       * So, to avoid hanging the GPU, just smash the low order 3 bits of
> >> -       * tile_x and tile_y to 0.  This is a temporary workaround until we come
> >> -       * up with a better solution.
> >> -       */
> >> -      WARN_ONCE((tile_x & 7) || (tile_y & 7),
> >> -                "Depth/stencil buffer needs alignment to 8-pixel boundaries.\n"
> >> -                "Truncating offset, bad rendering may occur.\n");
> >> -      tile_x &= ~7;
> >> -      tile_y &= ~7;
> >> -
> >>        intel_emit_post_sync_nonzero_flush(brw);
> >>        intel_emit_depth_stall_flushes(brw);
> >>
> >>        BEGIN_BATCH(7);
> >> +      /* 3DSTATE_DEPTH_BUFFER dw0 */
> >>        OUT_BATCH(_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
> >> +
> >> +      /* 3DSTATE_DEPTH_BUFFER dw1 */
> >>        OUT_BATCH((params->depth.mt->pitch - 1) |
> >>                  params->depth_format << 18 |
> >>                  1 << 21 | /* separate stencil enable */
> >>                  1 << 22 | /* hiz enable */
> >>                  BRW_TILEWALK_YMAJOR << 26 |
> >>                  1 << 27 | /* y-tiled */
> >> -                BRW_SURFACE_2D << 29);
> >> +                surftype << 29);
> >> +
> >> +      /* 3DSTATE_DEPTH_BUFFER dw2 */
> >>        OUT_RELOC(params->depth.mt->bo,
> >>                  I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
> >> -                offset);
> >> +                0);
> >> +
> >> +      /* 3DSTATE_DEPTH_BUFFER dw3 */
> >>        OUT_BATCH(BRW_SURFACE_MIPMAPLAYOUT_BELOW << 1 |
> >> -                (params->depth.width + tile_x - 1) << 6 |
> >> -                (params->depth.height + tile_y - 1) << 19);
> >> +                (surfwidth - 1) << 6 |
> >> +                (surfheight - 1) << 19 |
> >> +                lod << 2);
> >> +
> >> +      /* 3DSTATE_DEPTH_BUFFER dw4 */
> >> +      OUT_BATCH((depth - 1) << 21 |
> >> +                min_array_element << 10 |
> >> +                (depth - 1) << 1);
> >
> > Here we could have a comment saying that the latter "depth - 1" is render
> > target view extent, and that the spec mandates it to always match the value
> > of depth (the former "depth - 1"). The same applies for
> > gen6_emit_depth_stencil_hiz() further down. This is up to you.
> 
> Actually, I think we might be programming the field wrong for 3D textures.
> "For 3D Surfaces:
> This field indicates the extent of the accessible ???R??? coordinates
> minus 1 on the LOD
> currently being rendered to.
> For 1D and 2D Surfaces:
> This field must be set to the same value as the Depth field."
> 
> So, I think we need a new variable:
> const unsigned view_extent =
>    gl_target == GL_TEXTURE_3D ?
>       minify(depth, lod) : depth;
> 
> Then programming this state packet should be clearer. (The distinction
> between the depth and view extent fields.)
> 
> Can I look into fixing this separately, since it probably impacts all gen >= 6?

Sounds good to me - nice finding. I guess we are also missing tests.

> 
> -Jordan
> 
> >> +
> >> +      /* 3DSTATE_DEPTH_BUFFER dw5 */
> >>        OUT_BATCH(0);
> >> -      OUT_BATCH(tile_x |
> >> -                tile_y << 16);
> >> +
> >> +      /* 3DSTATE_DEPTH_BUFFER dw6 */
> >>        OUT_BATCH(0);
> >>        ADVANCE_BATCH();
> >>     }
> >> @@ -894,17 +871,13 @@ gen6_blorp_emit_depth_stencil_config(struct brw_context *brw,
> >>     /* 3DSTATE_HIER_DEPTH_BUFFER */
> >>     {
> >>        struct intel_mipmap_tree *hiz_mt = params->depth.mt->hiz_mt;
> >> -      uint32_t hiz_offset =
> >> -         intel_miptree_get_aligned_offset(hiz_mt,
> >> -                                          draw_x & ~tile_mask_x,
> >> -                                          (draw_y & ~tile_mask_y) / 2, false);
> >>
> >>        BEGIN_BATCH(3);
> >>        OUT_BATCH((_3DSTATE_HIER_DEPTH_BUFFER << 16) | (3 - 2));
> >>        OUT_BATCH(hiz_mt->pitch - 1);
> >>        OUT_RELOC(hiz_mt->bo,
> >>                  I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
> >> -                hiz_offset);
> >> +                0);
> >>        ADVANCE_BATCH();
> >>     }
> >>
> >> diff --git a/src/mesa/drivers/dri/i965/gen6_depth_state.c b/src/mesa/drivers/dri/i965/gen6_depth_state.c
> >> index 5e3981c..860233c 100644
> >> --- a/src/mesa/drivers/dri/i965/gen6_depth_state.c
> >> +++ b/src/mesa/drivers/dri/i965/gen6_depth_state.c
> >> @@ -51,6 +51,7 @@ gen6_emit_depth_stencil_hiz(struct brw_context *brw,
> >>     unsigned int min_array_element;
> >>     GLenum gl_target = GL_TEXTURE_2D;
> >>     unsigned int lod;
> >> +   const struct intel_mipmap_tree *mt = depth_mt ? depth_mt : stencil_mt;
> >>     const struct intel_renderbuffer *irb = NULL;
> >>     const struct gl_renderbuffer *rb = NULL;
> >>
> >> @@ -112,8 +113,16 @@ gen6_emit_depth_stencil_hiz(struct brw_context *brw,
> >>
> >>     lod = irb ? irb->mt_level - irb->mt->first_level : 0;
> >>
> >> +   if (mt) {
> >> +      width = mt->logical_width0;
> >> +      height = mt->logical_height0;
> >> +   }
> >> +
> >>     BEGIN_BATCH(7);
> >> +   /* 3DSTATE_DEPTH_BUFFER dw0 */
> >>     OUT_BATCH(_3DSTATE_DEPTH_BUFFER << 16 | (7 - 2));
> >> +
> >> +   /* 3DSTATE_DEPTH_BUFFER dw1 */
> >>     OUT_BATCH((depth_mt ? depth_mt->pitch - 1 : 0) |
> >>               (depthbuffer_format << 18) |
> >>               ((enable_hiz_ss ? 1 : 0) << 21) | /* separate stencil enable */
> >> @@ -121,22 +130,32 @@ gen6_emit_depth_stencil_hiz(struct brw_context *brw,
> >>               (BRW_TILEWALK_YMAJOR << 26) |
> >>               ((depth_mt ? depth_mt->tiling != I915_TILING_NONE : 1)
> >>                << 27) |
> >> -             (depth_surface_type << 29));
> >> +             (surftype << 29));
> >>
> >> +   /* 3DSTATE_DEPTH_BUFFER dw2 */
> >>     if (depth_mt) {
> >>        OUT_RELOC(depth_mt->bo,
> >>               I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
> >> -             depth_offset);
> >> +             0);
> >>     } else {
> >>        OUT_BATCH(0);
> >>     }
> >>
> >> -   OUT_BATCH(((width + tile_x - 1) << 6) |
> >> -             ((height + tile_y - 1) << 19));
> >> -   OUT_BATCH(0);
> >> +   /* 3DSTATE_DEPTH_BUFFER dw3 */
> >> +   OUT_BATCH(((width - 1) << 6) |
> >> +             ((height - 1) << 19) |
> >> +             lod << 2);
> >> +
> >> +   /* 3DSTATE_DEPTH_BUFFER dw4 */
> >> +   OUT_BATCH((depth - 1) << 21 |
> >> +             min_array_element << 10 |
> >> +             (depth - 1) << 1);
> >>
> >> -   OUT_BATCH(tile_x | (tile_y << 16));
> >> +   /* 3DSTATE_DEPTH_BUFFER dw5 */
> >> +   OUT_BATCH(0);
> >> +   assert(tile_x == 0 && tile_y == 0);
> >
> > I noticed that the vtable function for gen < 6 is now the only one actually
> > using the tile offsets - gen7 and gen8 have their own functions also and don't
> > even assert.
> >
> > I don't expect any change in this patch on this regard, I'm just making a
> > mental note for us to address this later on. Hence:
> >
> > Reviewed-by: Topi Pohjolainen <topi.pohjolainen at intel.com>
> >
> >>
> >> +   /* 3DSTATE_DEPTH_BUFFER dw6 */
> >>     OUT_BATCH(0);
> >>
> >>     ADVANCE_BATCH();
> >> @@ -158,7 +177,7 @@ gen6_emit_depth_stencil_hiz(struct brw_context *brw,
> >>        OUT_BATCH(hiz_mt->pitch - 1);
> >>        OUT_RELOC(hiz_mt->bo,
> >>                  I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
> >> -                brw->depthstencil.hiz_offset);
> >> +                0);
> >>        ADVANCE_BATCH();
> >>        } else {
> >>        BEGIN_BATCH(3);
> >> @@ -180,7 +199,7 @@ gen6_emit_depth_stencil_hiz(struct brw_context *brw,
> >>        OUT_BATCH(2 * stencil_mt->pitch - 1);
> >>        OUT_RELOC(stencil_mt->bo,
> >>                  I915_GEM_DOMAIN_RENDER, I915_GEM_DOMAIN_RENDER,
> >> -                brw->depthstencil.stencil_offset);
> >> +                0);
> >>        ADVANCE_BATCH();
> >>        } else {
> >>        BEGIN_BATCH(3);
> >> --
> >> 2.0.1
> >>
> >> _______________________________________________
> >> mesa-dev mailing list
> >> mesa-dev at lists.freedesktop.org
> >> http://lists.freedesktop.org/mailman/listinfo/mesa-dev
> > _______________________________________________
> > mesa-dev mailing list
> > mesa-dev at lists.freedesktop.org
> > http://lists.freedesktop.org/mailman/listinfo/mesa-dev


More information about the mesa-dev mailing list