Re: [Chicken-users] Chicken on Solaris 9

[Felix Winkelmann]

Bruce Hoult wrote:
> On Thu, 2003-07-03 at 15:21, Bruce Hoult wrote:
>
> > I'm attempting to build from the Chicken 1.12 tarball on the following
> > system:
>
>
> To reply to myself, there's now a far worse problem:
>
> ./chicken csi.scm -quiet -debug-level 0 -optimize-level 2 -include-path
> .  -output-file csi.c -prologue ./build.scm -uses posix -uses
> chicken-setup -uses tcp
> gmake[1]: *** [csi.c] Segmentation Fault (core dumped)
> gmake[1]: Leaving directory `/volA/home/bhoult/chicken-1.12'
>
> (gdb) bt
> #0  0xff216f58 in C_reclaim () from
> /home/bhoult/chicken-1.12/./.libs/libchicken.so.0
> #1  0xff216e14 in C_save_and_reclaim () from
> /home/bhoult/chicken-1.12/./.libs/libchicken.so.0
> #2  0xff2896e4 in f6013 () from
> /home/bhoult/chicken-1.12/./.libs/libchicken.so.0
>
>
> Any ideas what to do?
>
> If it helps, "./chicken csi.scm" gives no errors and just prints
> "compiling `csi.scm' ..."

Urgh. I fixed a couple of compiler bugs recently, but this looks
more like a different thing. Does nsample run? (that
nursery benchmarking program)

Can you try some more options to chicken (-optimize-level 2,
 etc.)? Perhaps we can pinpoint the exact spot. It might also help
if you rebuild with debugging info. f6013 is in eval.c (macroexpander),
perhaps something goes wrong there. Most likely some invalid argument,
that get subsequently passed to GC, which then dereferences garbage
data.


cheers,
felix