Revert "minigbm: rockchip: align NV12 planes to 64 bytes"

This reverts commit 33cca93af658650b655b4b5dcfb18a3235a100cf.

Reason for revert:

Many CTS tests are failed by the change, and video playback is corrupted on Youtube/Netflix app.

Original change's description:
> minigbm: rockchip: align NV12 planes to 64 bytes
>
> When running the mapped_texture_test with the Mali EGL implementation
> compiled with debug symbols, Mali complains that the NV12 planes are
> not sufficiently aligned (specifically, mali_cobj_surface.c line 1806).
> Fix this.
>
> BUG=b:63910596
> TEST=run Youtube App video on Kevin, at fullscreen and at normal size
>
> Change-Id: Ia204360c3aebe2bbd545ac7be0f6d095b0412f9d
> Reviewed-on: https://chromium-review.googlesource.com/556540
> Commit-Ready: Gurchetan Singh <gurchetansingh@chromium.org>
> Tested-by: Gurchetan Singh <gurchetansingh@chromium.org>
> Reviewed-by: Stphane Marchesin <marcheu@chromium.org>

Bug: b:63999659
Test: 1) run CTS on minnie: CtsVideoTestCases android.video.cts.VideoEncoderDecoderTest#testAvcGoog0Qual0720x0480, which fails on ToT and passes with the revert.
2) Play a video (size:426x240) on Youtube app. Video is corrupted on ToT and good with the revert.

Change-Id: I3418d0065733b4e0f99d1ab8c3cd416f7f454bc0
Reviewed-on: https://chromium-review.googlesource.com/583696
Commit-Ready: Pin-chih Lin <johnylin@chromium.org>
Tested-by: Pin-chih Lin <johnylin@chromium.org>
Reviewed-by: Tomasz Figa <tfiga@chromium.org>

diff --git a/rockchip.c b/rockchip.c
index d8f8cd3..66f1ea0 100644 (file)
--- a/rockchip.c
+++ b/rockchip.c
@@ -162,7 +162,16 @@ static int rockchip_bo_create_with_modifiers(struct bo *bo, uint32_t width, uint
        size_t plane;
        struct drm_rockchip_gem_create gem_create;
 
-       if (width <= 2560 &&
+       if (format == DRM_FORMAT_NV12) {
+               uint32_t w_mbs = DIV_ROUND_UP(ALIGN(width, 16), 16);
+               uint32_t h_mbs = DIV_ROUND_UP(ALIGN(height, 16), 16);
+
+               uint32_t aligned_width = w_mbs * 16;
+               uint32_t aligned_height = DIV_ROUND_UP(h_mbs * 16 * 3, 2);
+
+               drv_bo_from_format(bo, aligned_width, height, format);
+               bo->total_size = bo->strides[0] * aligned_height + w_mbs * h_mbs * 128;
+       } else if (width <= 2560 &&
                   has_modifier(modifiers, count, DRM_FORMAT_MOD_CHROMEOS_ROCKCHIP_AFBC)) {
                /* If the caller has decided they can use AFBC, always
                 * pick that */
@@ -174,34 +183,20 @@ static int rockchip_bo_create_with_modifiers(struct bo *bo, uint32_t width, uint
                        return -1;
                }
 
-               uint32_t stride, extra_size;
+               uint32_t stride;
                /*
-                * Since the ARM L1 cache line size is 64 bytes, align to that as a performance
-                * optimization. For YV12, the Mali cmem allocator requires that chroma planes are
-                * aligned to 64-bytes, so align the luma plane to 128 bytes.
+                * Since the ARM L1 cache line size is 64 bytes, align to that
+                * as a performance optimization. For YV12, the Mali cmem allocator
+                * requires that chroma planes are aligned to 64-bytes, so align the
+                * luma plane to 128 bytes.
                 */
-               extra_size = 0;
                stride = drv_stride_from_format(format, width, 0);
                if (format == DRM_FORMAT_YVU420 || format == DRM_FORMAT_YVU420_ANDROID)
                        stride = ALIGN(stride, 128);
                else
                        stride = ALIGN(stride, 64);
 
-               if (format == DRM_FORMAT_NV12) {
-                       /* RK VPU needs the buffers macroblock-aligned. */
-                       stride = ALIGN(stride, 16);
-                       height = ALIGN(height, 16);
-                       /*
-                        * The RK VPU driver requires extra space for motion vectors. The amount
-                        * of extra space required is specified in 16x16 macroblocks.
-                        */
-                       uint32_t w_mbs = DIV_ROUND_UP(stride, 16);
-                       uint32_t h_mbs = DIV_ROUND_UP(height, 16);
-                       extra_size = w_mbs * h_mbs * 128;
-               }
-
                drv_bo_from_format(bo, stride, height, format);
-               bo->total_size += extra_size;
        }
 
        memset(&gem_create, 0, sizeof(gem_create));