On Mon, 2002-09-09 at 11:11, Juergen Keil wrote: > > I've had a look at revision 1.17 of radeon_accel.c on DRI CVS, and it > seems you've changed (among many other things) the > Mono8x8PatternFillFlags to use BIT_ORDER_IN_BYTE_MSBFIRST instead of > BIT_ORDER_IN_BYTE_LSBFIRST; and removed the RADEON_GMC_BYTE_LSB_TO_MSB > flag in RADEONSetupForMono8x8PatternFill. This was the change that has > fixed the problem for me as well, using xfree86 CVS. > > > The other change you made related to "8x8 mono pattern filled > rectangles" - the "Take care of endianness" stuff in function > RADEONSetupForMono8x8PatternFill() - does not make a difference on a > little endian architecture like x86 but I guess it's fixing some > corruption with swapped vertical lines that should occur on big endian > architectures.
Exactly. > Have you been able to reproduce the issue with the horizontal flipped > pixels and the 2d accelerated "8x8 mono pattern filled rectangles" op? Yes. > I'm asking because I posted the same question on usenet and from the 3-4 > people that responded, apparently noone was able to reproduce this problem. > What's the explanation for this? Maybe they were running with DRI enabled, where there was no 8x8 mono acceleration before these changes of mine. -- Earthling Michel D�nzer (MrCooper)/ Debian GNU/Linux (powerpc) developer XFree86 and DRI project member / CS student, Free Software enthusiast _______________________________________________ Xpert mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xpert
