http://bugs.winehq.com/show_bug.cgi?id=740
*** shadow/740 Fri May 31 07:36:05 2002 --- shadow/740.tmp.28580 Fri May 31 10:51:49 2002 *************** *** 216,218 **** --- 216,242 ---- "/ fgBits" and "/ bgBits". It would be interesting to research where fgBits and bgBits come from. One of these could be 0 for some reasons. I don't have time right now to research it, but this guess may help you. + + ------- Additional Comments From ronis@onsager.chem.mcgill.ca 2002-05-31 10:51 ------- + I've added some trace calls and found the problem; it turns out that fgBits=0, and + that's the divisior in mous.c:270. The origial TRACE call shows: + + trace:cursor:X11DRV_GetCursor Bitmap 32x32 planes=1 bpp=24 bytesperline=96 + bBitsPerPixel=24. + + I also put in some traces for the loops at line 233. xmax=ymax=32 and + threshold=64. It turns out that the entire theImage array is zero; hence, + the first block of the if at line 253 is never entered, and fgBits stays zero. + + If I understand what's going on correctly, a zero mask is allowed. + + I tried protecting the divisions by bgBits and fgBits by testing to see that + they weren't zero--setting, bg.red=bg.geeen=bg.blue=0 and similarly for the fg + struct if the division would fail; this get's me farther, but the application + still fails. + + + + + +