i965/rbc: Allocate mcs directly

such as we do for compressed msaa. In case of non-compressed simgle
sampled buffers the allocation of mcs is deferred until there is
actually a clear operation that needs the mcs.
In case of render buffer compression the mcs buffer always needed
and there is no real reason to defer the allocation. By doing it
directly allows to drop quite a bit unnecessary complexity.

Patch leaves brw_predraw_set_aux_buffers() a no-op. Subsequent
patches will re-use it and it seemed cleaner to leave it instead
of removing and re-introducing.

Signed-off-by: Topi Pohjolainen <[email protected]>
Reviewed-by: Jason Ekstrand <[email protected]>

4 files changed, 26 insertions, 63 deletions

diff --git a/src/mesa/drivers/dri/i965/brw_blorp.c b/src/mesa/drivers/dri/i965/brw_blorp.c
index b0fbb64ce38..fdaf4293c19 100644
--- a/src/mesa/drivers/dri/i965/brw_blorp.c
+++ b/src/mesa/drivers/dri/i965/brw_blorp.c
@@ -287,8 +287,6 @@ brw_blorp_blit_miptrees(struct brw_context *brw,
    intel_miptree_slice_resolve_depth(brw, src_mt, src_level, src_layer);
    intel_miptree_slice_resolve_depth(brw, dst_mt, dst_level, dst_layer);
 
-   intel_miptree_prepare_mcs(brw, dst_mt);
-
    DBG("%s from %dx %s mt %p %d %d (%f,%f) (%f,%f)"
        "to %dx %s mt %p %d %d (%f,%f) (%f,%f) (flip %d,%d)\n",
        __func__,
@@ -689,6 +687,9 @@ do_single_blorp_clear(struct brw_context *brw, struct gl_framebuffer *fb,
        !brw_is_color_fast_clear_compatible(brw, irb->mt, &ctx->Color.ClearColor))
       can_fast_clear = false;
 
+   const bool is_lossless_compressed = intel_miptree_is_lossless_compressed(
+                                          brw, irb->mt);
+
    if (can_fast_clear) {
       /* Record the clear color in the miptree so that it will be
        * programmed in SURFACE_STATE by later rendering and resolve
@@ -708,7 +709,8 @@ do_single_blorp_clear(struct brw_context *brw, struct gl_framebuffer *fb,
        * it now.
        */
       if (!irb->mt->mcs_mt) {
-         if (!intel_miptree_alloc_non_msrt_mcs(brw, irb->mt)) {
+         assert(!is_lossless_compressed);
+         if (!intel_miptree_alloc_non_msrt_mcs(brw, irb->mt, false)) {
             /* MCS allocation failed--probably this will only happen in
              * out-of-memory conditions.  But in any case, try to recover
              * by falling back to a non-blorp clear technique.
@@ -757,7 +759,7 @@ do_single_blorp_clear(struct brw_context *brw, struct gl_framebuffer *fb,
                   clear_color, color_write_disable);
       blorp_batch_finish(&batch);
 
-      if (intel_miptree_is_lossless_compressed(brw, irb->mt)) {
+      if (is_lossless_compressed) {
          /* Compressed buffers can be cleared also using normal rep-clear. In
           * such case they behave such as if they were drawn using normal 3D
           * render pipeline, and we simply mark the mcs as dirty.
diff --git a/src/mesa/drivers/dri/i965/brw_draw.c b/src/mesa/drivers/dri/i965/brw_draw.c
index 9b1e18c51cf..cab67c94e1c 100644
--- a/src/mesa/drivers/dri/i965/brw_draw.c
+++ b/src/mesa/drivers/dri/i965/brw_draw.c
@@ -409,8 +409,8 @@ brw_predraw_set_aux_buffers(struct brw_context *brw)
       struct intel_renderbuffer *irb =
          intel_renderbuffer(fb->_ColorDrawBuffers[i]);
 
-      if (irb) {
-         intel_miptree_prepare_mcs(brw, irb->mt);
+      if (!irb) {
+         continue;
       }
    }
 }
diff --git a/src/mesa/drivers/dri/i965/intel_mipmap_tree.c b/src/mesa/drivers/dri/i965/intel_mipmap_tree.c
index 7b971839a0a..427657c8771 100644
--- a/src/mesa/drivers/dri/i965/intel_mipmap_tree.c
+++ b/src/mesa/drivers/dri/i965/intel_mipmap_tree.c
@@ -789,6 +789,20 @@ intel_miptree_create(struct brw_context *brw,
        intel_miptree_supports_non_msrt_fast_clear(brw, mt)) {
       mt->fast_clear_state = INTEL_FAST_CLEAR_STATE_RESOLVED;
       assert(brw->gen < 8 || mt->halign == 16 || num_samples <= 1);
+
+      /* On Gen9+ clients are not currently capable of consuming compressed
+       * single-sampled buffers. Disabling compression allows us to skip
+       * resolves.
+       */
+      const bool lossless_compression_disabled = INTEL_DEBUG & DEBUG_NO_RBC;
+      const bool is_lossless_compressed =
+         unlikely(!lossless_compression_disabled) &&
+         brw->gen >= 9 && !mt->is_scanout &&
+         intel_miptree_supports_lossless_compressed(brw, mt);
+
+      if (is_lossless_compressed) {
+         intel_miptree_alloc_non_msrt_mcs(brw, mt, is_lossless_compressed);
+      }
    }
 
    return mt;
@@ -1563,7 +1577,8 @@ intel_miptree_alloc_mcs(struct brw_context *brw,
 
 bool
 intel_miptree_alloc_non_msrt_mcs(struct brw_context *brw,
-                                 struct intel_mipmap_tree *mt)
+                                 struct intel_mipmap_tree *mt,
+                                 bool is_lossless_compressed)
 {
    assert(mt->mcs_mt == NULL);
    assert(!mt->disable_aux_buffers);
@@ -1605,16 +1620,6 @@ intel_miptree_alloc_non_msrt_mcs(struct brw_context *brw,
       layout_flags |= MIPTREE_LAYOUT_FORCE_HALIGN16;
    }
 
-   /* On Gen9+ clients are not currently capable of consuming compressed
-    * single-sampled buffers. Disabling compression allows us to skip
-    * resolves.
-    */
-   const bool lossless_compression_disabled = INTEL_DEBUG & DEBUG_NO_RBC;
-   const bool is_lossless_compressed =
-      unlikely(!lossless_compression_disabled) &&
-      brw->gen >= 9 && !mt->is_scanout &&
-      intel_miptree_supports_lossless_compressed(brw, mt);
-
    /* In case of compression mcs buffer needs to be initialised requiring the
     * buffer to be immediately mapped to cpu space for writing. Therefore do
     * not use the gpu access flag which can cause an unnecessary delay if the
@@ -1656,47 +1661,6 @@ intel_miptree_alloc_non_msrt_mcs(struct brw_context *brw,
    return mt->mcs_mt;
 }
 
-void
-intel_miptree_prepare_mcs(struct brw_context *brw,
-                          struct intel_mipmap_tree *mt)
-{
-   if (mt->mcs_mt)
-      return;
-
-   if (brw->gen < 9)
-      return;
-
-   /* Single sample compression is represented re-using msaa compression
-    * layout type: "Compressed Multisampled Surfaces".
-    */
-   if (mt->msaa_layout != INTEL_MSAA_LAYOUT_CMS || mt->num_samples > 1)
-      return;
-
-   /* Clients are not currently capable of consuming compressed
-    * single-sampled buffers.
-    */
-   if (mt->is_scanout)
-      return;
-
-   assert(intel_tiling_supports_non_msrt_mcs(brw, mt->tiling) ||
-          intel_miptree_supports_lossless_compressed(brw, mt));
-
-   /* Consider if lossless compression is supported but the needed
-    * auxiliary buffer doesn't exist yet.
-    *
-    * Failing to allocate the auxiliary buffer means running out of
-    * memory. The pointer to the aux miptree is left NULL which should
-    * signal non-compressed behavior.
-    */
-   if (!intel_miptree_alloc_non_msrt_mcs(brw, mt)) {
-      _mesa_warning(NULL,
-                    "Failed to allocated aux buffer for lossless"
-                    " compressed %p %u:%u %s\n",
-                    mt, mt->logical_width0, mt->logical_height0,
-                    _mesa_get_format_name(mt->format));
-   }
-}
-
 /**
  * Helper for intel_miptree_alloc_hiz() that sets
  * \c mt->level[level].has_hiz. Return true if and only if
diff --git a/src/mesa/drivers/dri/i965/intel_mipmap_tree.h b/src/mesa/drivers/dri/i965/intel_mipmap_tree.h
index 07db51b305e..2f1b8ebf0e1 100644
--- a/src/mesa/drivers/dri/i965/intel_mipmap_tree.h
+++ b/src/mesa/drivers/dri/i965/intel_mipmap_tree.h
@@ -709,11 +709,8 @@ intel_miptree_supports_lossless_compressed(struct brw_context *brw,
 
 bool
 intel_miptree_alloc_non_msrt_mcs(struct brw_context *brw,
-                                 struct intel_mipmap_tree *mt);
-
-void
-intel_miptree_prepare_mcs(struct brw_context *brw,
-                          struct intel_mipmap_tree *mt);
+                                 struct intel_mipmap_tree *mt,
+                                 bool is_lossless_compressed);
 
 enum {
    MIPTREE_LAYOUT_ACCELERATED_UPLOAD       = 1 << 0,