intel: faster WSI swapchain operation

Use poll instead of clock_nanosleep for WSI x11_swap_chain_wait.
That allows faster response to present events while still handling
cases when xcb will not recognize and wake on some present event packets.
1 file changed