Window uses shared stride for UV planes and tegra_dc_window struct
defines array of 2 strides per window. That's not taken in account
during setting up of the window addresses and strides, resulting in
out-of-bounds write of the 3-rd (non-existent) V plane stride that
overwrites Y plane base address.
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
[treding@nvidia.com: explain why the V-plane stride is ignored]
Signed-off-by: Thierry Reding <treding@nvidia.com>
struct tegra_bo *bo = tegra_fb_get_plane(fb, i);
window.base[i] = bo->paddr + fb->offsets[i];
- window.stride[i] = fb->pitches[i];
+
+ /*
+ * Tegra uses a shared stride for UV planes. Framebuffers are
+ * already checked for this in the tegra_plane_atomic_check()
+ * function, so it's safe to ignore the V-plane pitch here.
+ */
+ if (i < 2)
+ window.stride[i] = fb->pitches[i];
}
tegra_dc_setup_window(dc, p->index, &window);