This is the mail archive of the
insight@sourceware.org
mailing list for the Insight project.
Compiling on amd64 (ubuntu 7.10)
- From: David Haworth <dave at fen-net dot de>
- To: insight at sourceware dot org
- Date: Mon, 26 Nov 2007 19:26:57 +0100
- Subject: Compiling on amd64 (ubuntu 7.10)
Hi,
I'm having trouble compiling insight 6.7.1 for an amd64 host (ubuntu 7.10).
Target is arm-elf, but I don't think this is anything to do with the
problem.
Configure works fine (or appears to):
./configure --prefix=/usr/local/xgcc/broken --target=arm-elf
but then make ends with an error (after a lot of warnings). The warnings are
mostly of the form:
warning: passing argument 2 of 'foo' with different width due to prototype
and the error is:
cc1: warnings being treated as errors
.././gdb/gdbtk/generic/gdbtk-register.c:348: warning: cast from pointer to integer of different size
After "fixing" this one using typecasts the compilation then proceeds to the
next error:
.././gdb/gdbtk/generic/gdbtk-stack.c:298: warning: cast from pointer to integer of different size
and (after fixing this one):
.././gdb/gdbtk/generic/gdbtk-wrapper.c:158: warning: cast to pointer from integer of different size
(and lots more of a similar nature).
After fixing these, the compilation completes without further errors,
but the resulting binary doesn't work poperly. The pure gdb binary appears tos
be OK, but the GUI version (arm-elf-insight or arm-elf-gdb -w) doesn't. It starts
up, but doesn't show a source window - instead, a window with title "top" is shown.
Other things fail too. The target settings window doesn't show at all, and
Run->Connct to target gives an error message "Error: can't read "_status": no
such variable.
The fixes that I made, although hacks I admit, should work OK since they are mostly
casting from into to pointer via long, and then back again from pointer to int via
long. However, the warnings look a lot more suspicious.
Has anyone got any experience in getting insight to compile for a 32-bit target
on a 64-bit host?
Any help gratefully accepted.
Dave
--
David Haworth, Germany.
PS. I can supply the complete set of warnings & errors if anyone wants them -
they seem a bit to spost to the list. Please mail me off -list if
they will help.