This: http://gcc.gnu.org/ml/gcc-regression/2010-05/msg00227.html is a bootstrap fail for me, ..will try to add some analysis later (looked like two or three bytes different yesterday - so maybe no an obvious prob).
Stage3: 000054d0 7f e3 fb 78 4b ff ac 65 7c 7c 1b 78 93 81 00 38 |.„˚xKˇ¨e||.x...8| 000054e0 4b ff fc b8 1c 42 00 14 38 02 f8 6c 3c 40 00 01 |Kˇ¸∏.B..8.¯l<@..| 000054f0 38 42 83 ec 7c 42 02 15 41 82 00 18 81 63 00 14 |8B.Ï|B..A....c..| 00005500 80 03 00 18 7f 8b 00 00 41 9c 05 58 38 40 00 00 |........A..X8@..| Stage2: 000054d0 7f e3 fb 78 4b ff ac 65 7c 7c 1b 78 93 81 00 38 |.„˚xKˇ¨e||.x...8| 000054e0 4b ff fc b8 1c 42 00 14 3d 60 00 01 38 0b 83 ec |Kˇ¸∏.B..=`..8..Ï| 000054f0 38 42 f8 6c 7c 42 02 15 41 82 00 18 81 63 00 14 |8B¯l|B..A....c..| 00005500 80 03 00 18 7f 8b 00 00 41 9c 05 58 38 40 00 00 |........A..X8@..| 00005510 stage3: 000053a8 b 0x5060 000053ac mulli r2,r2,0x14 000053b0 addi r0,r2,0xf86c 000053b4 lis r2,0x1 000053b8 addi r2,r2,0x83ec 000053bc add. r2,r2,r0 Stage2: 000053a8 b 0x5060 000053ac mulli r2,r2,0x14 000053b0 lis r11,0x1 000053b4 addi r0,r11,0x83ec 000053b8 addi r2,r2,0xf86c 000053bc add. r2,r2,r0 === also... Stage3: 0000a1b0 00 00 01 c3 00 00 54 18 00 00 01 c3 00 00 54 04 |...√..T....√..T.| 0000a1c0 00 00 01 c3 00 00 53 b8 00 00 03 45 00 00 00 00 |...√..S∏...E....| 0000a1d0 00 00 00 41 00 00 53 b4 00 00 03 46 00 00 83 ec |...A..S¥...F...Ï| 0000a1e0 00 00 00 41 00 00 53 9c 00 00 01 c3 00 00 53 88 |...A..S....√..S.| Stage2: 0000a1b0 00 00 01 c3 00 00 54 18 00 00 01 c3 00 00 54 04 |...√..T....√..T.| 0000a1c0 00 00 01 c3 00 00 53 b4 00 00 03 45 00 00 00 00 |...√..S¥...E....| 0000a1d0 00 00 00 41 00 00 53 b0 00 00 03 46 00 00 83 ec |...A..S∞...F...Ï| 0000a1e0 00 00 00 41 00 00 53 9c 00 00 01 c3 00 00 53 88 |...A..S....√..S.|
Confirmed and it is not fixed by the patch in http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44199#c6 .
powerpc64-apple-darwin9 does *not* exhibit this ...
fixed by r159952