Re: another 64bit bug

[Igor Schein <igor@txc.com>]

On Tue, Nov 09, 1999 at 04:15:21PM -0500, Igor Schein wrote:
> Hi,
> 
> I built a debugging version ( generic kernel ) of latest CVS code for
> sparcv9 target ( Solaris 64 bit ) using WorkShop 5.0 compiler.
> A pretty bizarre bug showed up:
> 
> ? 60!
>   ***   segmentation fault: bug in GP (please report).
> 
> This didn't happen in 2.0.17.  Unfortunately I don't have
> dbx, so debugging is not plausible for me.

Here's the stack trace:

mulii(c47f28,c47f10,0,0,0,0) + 10
divide_conquer_prod(100c47f40,100036618,0,0,0,0) + 168
mpfact(3c,0,0,0,0,0) + 238
facteur(0,0,0,0,0,0) + 3dc
expr(0,0,0,0,0,0) + 90
seq(0,0,0,0,0,0) + 1bc
lisseq0(100ed60f0,1002530a0,0,0,0,0) + 1b0
lisseq(100ed60f0,1,0,0,0,0) + 28
readseq(100ed60f0,1,36302100,7efefeff,81010100,ff00) + 54
gp_main_loop(1e,7a120,10040af00,0,0,0) + 7c0
main(3,ffffffff7ffff768,ffffffff7ffff788,ffffffff7f4ae600,0,100000000) + 49c

Have I ever mentioned that adb is a bitch?

Igor