In a poorly coded app, the framebuffer can be partially drawn, an FBO
switched, switch back to the framebuffer and keep drawing, etc.
Reordering would fix this, but for now we need to just be careful about
flushing scanout too.
Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
{
struct panfrost_context *ctx = pan_context(pctx);
- /* Flush when switching away from an FBO, but not if the framebuffer
+ /* Flush when switching framebuffers, but not if the framebuffer
* state is being restored by u_blitter
*/
- if (!panfrost_is_scanout(ctx) && !ctx->blitter->running) {
- panfrost_flush(pctx, NULL, 0);
+ bool is_scanout = panfrost_is_scanout(ctx);
+ bool has_draws = ctx->draw_count > 0;
+
+ if (!ctx->blitter->running && (!is_scanout || has_draws)) {
+ panfrost_flush(pctx, NULL, PIPE_FLUSH_END_OF_FRAME);
}
ctx->pipe_framebuffer.nr_cbufs = fb->nr_cbufs;