From: <gb...@di...> - 2002-03-29 23:10:42
|
Hi, > It seemed to happen exactly when Basilisk wanted to show the window, thus > when it tried to write and discovered it isn't a DirectColor visual at > all. (Below the "// Show window" comment in driver_xf86dga) On my system, default visual is TrueColor (0x21) but DirectColor is available. B2 first tries to get a DirectColor visual for gamma correction. It then uses that visual, if available. Get rid of Gtk, and you will get a more precise error from the X server. > What I was thinking is, AFAIK, the new DGA 2.0 extension, included in XF86 > 4.0 and up, allows (unlike the old 1.0 version) depth and visual > switching. I believe the VidMode extension allows depth and resolution switching even for DGA 1. > Maybe the X server returned that the DirectColor visual is available and > can be switched to, unlike the code, which seems to assume the current > visual is DirectColor. Where does the code make that assumption? The code actually does try a DirectColor visual first. XMatchVisual() returned non-zero so that of visual is available and will be used. Probably, some other bits mismatch. Bye, Gwenole. |