On Thu, 8 Sep 2005, Petr Salinger wrote:
This can be also fixed with following in debian/rules.
--- debian/rules~ 2005-08-31 07:18:37.000000000 +0000
+++ debian/rules 2005-09-08 08:05:21.000000000 +0000
@@ -28,11 +28,13 @@
export ARBHOME=`pwd` ; \
export LC_ALL=C ; \
export PATH=$${ARBHOME}/bin:$${PATH} ; \
+ export LD_LIBRARY_PATH=$${ARBHOME}/lib:$${LD_LIBRARY_PATH}; \
$(MAKE) all
# export GCC=gcc-2.95 ; \ ####
export ARBHOME=`pwd` ; \
export LC_ALL=C ; \
export PATH=$${ARBHOME}/bin:$${PATH} ; \
+ export LD_LIBRARY_PATH=$${ARBHOME}/lib:$${LD_LIBRARY_PATH}; \
$(MAKE) arbxtras
touch build-stamp
The missing -fpic/-fPIC can be fixed by:
--- WINDOW/Makefile~ 2005-09-08 08:37:53.000000000 +0000
+++ WINDOW/Makefile 2005-09-08 08:37:53.000000000 +0000
@@ -21,7 +21,7 @@
$(CCPLIB) $(cflags) -c $< $(CPPINCLUDES)
.c.o:
- $(ACC) $(cflags) -c $< $(CPPINCLUDES)
+ $(CCLIB) $(cflags) -c $< $(CPPINCLUDES)
DEPENDS = $(OBJECTS:.o=.depend)
depends: $(DEPENDS)
Thanks for the patch.
But arb depends on xview which is not 64 bit clean and is unavailable
on all 64 bit archs, including amd64.
So the bug submitter will have have bad luck anyway?
How was he able to even start the build process if xview-dev is in the
Build-Dependencies?
Even with patches above and xview patch from
http://debian-ppc64.alioth.debian.org/gcc4/patches/xview_3.2p1.4-19.0.0.1.gcc4.patch
Ahh, this was the trick to get Build-Dependencies solved???
arb cannot be built on amd64 - "error: cast from 'void*' to 'int' loses
precision".
This can be "fixed" by changing cast (int) to (long),
(as in attached patch) but according to warnings in build log,
the arb is definitely not 64 bit clean.
Yes - as the author said - this is a problem which has lower priority than
getting arb fixed for i386 first which has also issues to solve.
Therefore, I suggest to apply only patch for debian/rules and
WINDOW/Makefile and restrict architectures similarly as xview.
I would set
Architecture: alpha arm hppa i386 m68k mips mipsel powerpc s390 sparc
which matches xviewg-dev architectures.
Andreas J., I am just curious, are these patches also sufficient for ppc64 ?
Another question to Andreas: If I exclude amd86 from this architecture list
the bug you reported is not really solved but it will not occure any more
because it is circumvented. What would you prefer:
[ ] Close the bug
[ ] Reduce severity to wishlist
[ ] Keep as is
Kind regards and thanks for the help
Andreas.
--
http://fam-tille.de
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
