mckenney@RPI.EDU (Bruce McKenney) (10/13/89)
Chris Hanson came up with a (quick!) pointer to the source of the trouble. For the benefit of any R6.1.x users, the file "mul.c" has a declaration: fast long Hi_A, Hi_B, Lo_A, Lo_B, Lo_C, Middle_C; which should be modified to: fast long Hi_A, Hi_B, Lo_A, Lo_B, Middle_C; fast unsigned long Lo_C; This certainly fixed this on our 370, and I suspect, from what I know of the 68000 multiply (which also produces a result in which the low-half is effectively "unsigned") should fix the Sun-3 as well. I also understand that this has been "fixed in Release 7". Many thanks to all who responded. Bruce McKenney mckenney@itsgw.rpi.edu Bruce_McKenney@mts.rpi.edu userarkc@rpitsmts.bitnet PS: It Definitely fixes the problem on the Sun-3 -- I just saw it.