>> I'm afraid I don't know enough about the Radeon driver to say whether >> this is a driver bug or not. Is the card being diddled in a way the >> driver cannot detect or work around?
>Impossible to tell without seeing the X server log. I don't believe so, I believe this is actually expected behaviour. As far as I can tell this is what is happening. 1) From reset the card tries to decided what is connected to both the ports. 2) If a display is connected to the DVI port that becomes the primary display and the VGA port the secondary display. Else the VGA port is the primary and the DVI port is not enabled. 3) If dual head is then forcibly enabled when it doesn't think that there is anything connected to the second display it comes up in clone mode. This is either undocumented behaviour or is a default in the Radeon driver if the card has failed to autodected over the the DCC connection to the monitor. In my case I believe the problem was caused by the a KVM switch between the card and the monitor. DCC seems to be very flaky :/ The workaround of putting the MonitorLayout option just forces the card into a particular mode so it ignores whatever is comming over the DCC. I've heard it is possible to damage the card with an incorrect MonitorLayout line but I can't think of a physical mechanism for this actually happening. I therefore think that this is worthy of a mention in the documentation for the radeon driver but is not actually a bug. It might warrent some more investigation in the radeon driver to see if my hypothesis is correct. -p -- Paul Gotch "What goes up must come down, ask any system administrator" --------------------------------------------------------------------
