Dennis Ruffer wrote:
> 
> On Nov 15, 2007, at 7:11 AM, Anton Ertl wrote:
> > What does "./gforth --no-dynamic --debug" say (in particular, about
> > the image file used)?  My suspicion is that it does not use
> > ./gforth.fi, but something else.
> 
> FPGA-1 ./gforth --no-dynamic --debug
> --no-dynamic conflicts with --ss-min-codesize, reverting to --ss-min- 
> nexts
> Opened image file: ./gforth.fi

Ok, my suspicion was wrong, so this becomes even more mysterious.

> Compiled with gcc-4.0.1 (Apple Computer, Inc. build 5370)

I have now tried this with gcc-4.0.3 20051201 (prerelease) (Debian
4.0.2-5) on a Debian system, and could not reproduce this problem.

> caught signal, throwing exception -23, ip=0x0 rp=0x55c00

-23 is an alignment exception, which is interesting, because Intel
processors normally handle unaligned addresses fine.  The only
exception I know is some SSE instructions, but I would not expect them
to be used in Gforth startup, only in FP code.

Moreover, ip has a wrong value.  I would now compile with debugging
(rm engine/*.o; make XCFLAGS=-g) and then look use gdb to see where
the problem happens, but running gdb through mail is not the best way
to debug:-).

Another possible source of the problem would be if the image was
corrupted for some reason.  Here's the start of the image I got:

[c1:~/gforth:11716] od -tx4  gforth.fi|head    
0000000 2f202123 2f727375 61636f6c 69622f6c
0000020 66672f6e 6874726f 362e302d 322d392e
0000040 30373030 20343036 6d692d2d 2d656761
0000060 656c6966 2020200a 726f6647 05336874
0000100 00000000 00000000 00033848 00400000
0000120 00004000 00004000 00003c00 00003a00
0000140 00005ff8 0000df04 00004000 00000000
0000160 4001b000 40020000 40025000 4002a000
0000200 00000000 00000004 706f6f6e 20202020
0000220 fffffff5 00000040 00000004 6c6c6163

- anton


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to