On Mon, 22 Jan 2001, Bastien Nocera wrote:
> Quoting Michael Schmitz <[EMAIL PROTECTED]>: > > > > > It's what you suspect: the offset between the start of consecutive > > scan > > > > lines in memory. No idea why the r128 driver would pick that pitch > > though. > > > > > > Perhaps because the RAGE128 needs line lenghts that are multiples of > > 64? > > > > Does it? If that's the case, why would the X server not honor the pitch > > (as it seems, from the description of the staircase effect)? > > > > But maybe that wasn't caused by the pitch/line_len but by weird mode > > settings. hsync 28-38 and vsync 43-50 still look crummy to me though. > > Definetely, > My XF86Config for iBook users (although you'll need to change r128 to ati as > I > understood) has the correct timings for the LCD screen. > That is: > Section "Monitor" > Identifier "Monitor0" > VendorName "Apple Computer" > ModelName "LCD Screen" > HorizSync 30-40 > VertRefresh 60-70 > Mode "800x600" > DotClock 40.001 > HTimings 800 858 986 1056 > VTimings 600 601 605 628 > Flags "+HSync" "+VSync" > EndMode > EndSection No change, sorry. Still claims a pitch of 832, and the screen is still screwed up. Note that the vsync 43-50 was a hack to force the driver to use the fb settings by invalidating the built in modes. The r128 doesnt seem to require a multiple of 64 line width, as it works OK against the fb, but it may perfer it. -- Michael > I'd like both of the persons who had problems (I think Michael Hope got it to > work though) to test this with their machines. Before blaming the driver, we > must make sure that the user is not causing the error (don't take that as an > insult, it's not one). > These timings are the same for old iBooks (mach64 based) and the new ones > (r128 > based) as the LCD is the same. > > Cheers > > /Bastien Nocera, who wants a 15" LCD screen on his iBook >
