On Tue, Dec 23, 2003 at 10:54:08AM +0100, Michael Schmitz wrote: > > > No, that's aty128fb related (screen corruption when fbdev accel is on). > > > There's a known workaround; document it (maybe add a big fat warning in > > > the postinstall) and let upstream handle it (any idea, Ben?) > > > > Mmm, i have to look into this bug report again. What is the workaround ? > > Disabling the accel ? This seems like a lousy workaround. Maybe i should > > If you have a better idea - it only affects text console, and unless you > find someone to debug the accel routines, it's better than screen garbage. > Ben disabled text accel in 2.6, BTW.
Ok, so i could backport this fix, no ? > > disable the accel by default though, and it could be reenabled by a > > command line option. > > Sure, set noaccel=1 and rename the boot option to accel. But it's a rare > glitch from what I've heard, so mentioning the fbset -accel false fix in > the README might be enough. Mmm, it makes better impression when it works at once, without options though, i will look at it. BTW, the newer powerpc kernel packages have been accepted this morning :)) Friendly, Sven Luther