To be honest, your suggestion to put the check inside framebuffer_check()
sounds like a better idea: we wouldn't even hit any driver-specific
code-path when the check fails. Daniel, do you think there'd be an issue
with this approach?

Reply via email to