Sorry, I should have debugged a bit more before writing: Following up to my last message, using "fpsetmask" to ignore all exceptions except division-by-zero, the program works. Program received signal SIGFPE, Arithmetic exception. 0x281824d8 in gl_x86_transform_points3_3d_no_rot_raw () from /t/video/libGL.so.3 (gdb) info float u: status 0xb8a8: exceptions: OVERF LOS; flags: 0000; top 7 e: status 0x3800: flags: 0000; top 7 control 0x1272: compute to 53 bits; round NEAREST; mask: DENORM UNDERF LOS; last instruction: opcode 0x15f; pc 0x1f:0x28182503; operand 0x27:0x8077058 regno tag msb lsb value %st(0) => valid 40859be8f01674000000 2.65267e+40 %st(7) empty 4038c93de20000002000 2.26578e+17 %st(6) empty 403d9e87160000000000 5.71156e+18 %st(5) empty 40859be8f01674000000 2.65267e+40 %st(4) empty bffeced0010000000000 -0.807861 %st(3) empty bffed2fd950000000000 -0.824182 %st(2) empty c0008974890000000000 -2.14774 %st(1) empty 400081226e0000000000 2.01773 Should Mesa set it's own floating point masks, or just not do this? :) Kevin _______________________________________________ Mesa-bug maillist - [EMAIL PROTECTED] http://lists.mesa3d.org/mailman/listinfo/mesa-bug _______________________________________________ Mesa-dev maillist - [EMAIL PROTECTED] http://lists.mesa3d.org/mailman/listinfo/mesa-dev
