Re: [patch] voluntary-preempt-2.6.9-rc1-bk4-Q9

[Alan Cox <alan@lxorguk.ukuu.org.uk>]

On Sad, 2004-09-04 at 01:04, Lee Revell wrote:
> This is looking more and more like a video driver problem:

Not really. The delay is too small and X is smarter than this. (except a
VIA case that only recently got squished).

> The video cards have a command FIFO that is written to via the PCI bus.
> They also have a status register, read via the PCI bus, which says
> whether the command FIFO is full or not. The hack is to not check
> whether the command FIFO is full before attempting to write to it, thus
> saving a PCI bus read.

On problem cards X defaults to polling the status FIFO. You can tell it
to be rude but you have to actively do so. Newer PCI 2.x specs also have
a thing or two to say on the subject