Felix wrote:
-- csc-orig.scm 2013-02-12 20:47:23.194996000 -0800
csc.scm 2013-02-12 20:47:23.194996000 -0800
@@ -274,7 274,9 @@
(else
(list (conc -L\ library-dir \
(if (and deployed (eq? (software-version) 'freebsd))
- (list -z origin)
(if
jrap...@bmedctr.com jrap...@bmedctr.com wrote:
However, I keep wondering how the version compiled without '-z origin'
managed to pass the deployment test at all, if in fact '-z origin' is really
necessary in order for the deployed application to find the necessary files.
Is the test really
Previously I wrote:
The correct thing to do here is to change '-z origin' into
'-Wl,-z,origin'. This way both GCC and Clang will do the right thing.
(I currently tested this by compiling C files; I'll be able to test csc
modified that way in a dozen of hours or so, unless any of you folks do
Felix fe...@call-with-current-continuation.org wrote:
Yup, just tested csc modified that way; -deploy works correctly with
gcc, clang and even pcc.
Did you build chicken with pcc?
Nope, and now that I tried, I can't do it either (it complains about
dynamic relocations in crt1.o).
What I did
Christian Kellermann ck...@pestilenz.org wrote:
Since it doesn't work for you on OpenBSD, I propose to test it on
NetBSD and DragonFlyBSD too, and then commit a similar patch that
would only add -Wl,-z,origin where it is know to work.
Yes, that has been my intention.
Cool.
In any case,
Hi, folks. While running gmake check the deployment test fails, saying:
/libexec/ld-elf.so.1: Shared object libchicken.so.6 not found,
required by rev-app
The full log is at [1]. The chicken in question is bootstrapped using
4.7.0, and installed into /tmp/c0 (see the log for full
Hi, folks. I'd like to bug you once again about the change to add
support for parallel build (i.e. building chicken with gmake -j). We
discussed this last month [3], but didn't reach any conclusion. This
was also discussed in ticket 526 [2], which is still open.
Anyway, the patch applicable to
So, is this a known issue? If not, how do I diagnose it further?
-deploy is known to be broken for NetBSD and OpenBSD. Does FreeBSD
support the $ORIGIN setting for RPATH?
It appears that it does, but you need to additionally pass -z origin
to the linker. For example deploying like this seems
I wrote:
Here's the test:
rm -rf /rmp/cd mkdir -p /tmp/cd cd /tmp/cd
Ugh, typo over there, it's /tmp/cd, not /rmp/cd.
You have been warned.
echo '(write 555)' | tee a.scm b.scm
csc -deploy a.scm
ldd a/a
csc -deploy -L -z origin b.scm
ldd b/b
Christian Kellermann ck...@pestilenz.org wrote:
I have also tried it with the attached patch, that should give the
correct flags, maybe we are still missing something...
Does the patch improve the situation for you?
Indeed it does for me.
Since it doesn't work for you on OpenBSD, I propose
Toby Thain wrote:
Builds for me on OS X 10.6 64-bit, OS X 10.5 dual-processor G5.
make -j8 reduces the make step from 3:39 to 0:47 on my 8-core system.
However, make check does not pass on either system:
Does it pass without the patch?
What about with the patch but using -j1?
Did you build a boot chicken for bootstrapping? Note that you will need a
chicken installed for this.
$ gmake PLATFORM=bsd boot-chicken
$ gmake PLATFORM=bsd CHICKEN=./chicken-boot
$ gmake PLATFORM=bsd install
I have put up a little tutorial on the wiki for this here
Mario Domenech Goulart mario.goul...@gmail.com wrote:
Should I try to adopt the patch from ticket 526 to the latest sources
and repeat the test? I will if this will push the things forward (this
will happen tomorrow though).
I'd suggest the patch from ticket 526, since it's been tested a
Hi, folks. I'm trying to build the latest git sources (I've got 4.7.0
installed); the build proceeds for a while and then fails like this:
$ gmake PLATFORM=bsd
...
chicken setup-api.scm [...] -output-file setup-api.c
Syntax error (import): cannot import from undefined module
Hi, folks. When I try to run a parallel build of chicken 4.7.0 via:
gmake -j4 PLATFORM=bsd
... it eventually fails saying this:
gcc -shared -Wl,-soname,libchicken.so \
-o libchicken.so library.o eval.o data-structures.o ports.o files.o
extras.o lolevel.o utils.o tcp.o srfi-1.o
Mario Domenech Goulart wrote:
This issue has been addressed by http://bugs.call-cc.org/ticket/526
OK, I see.
Any news on that bug? Will the fix be committed any time soon?
(I maintain the FreeBSD port for chicken, and some people are asking
why is it that the port doesn't build in parallel
The problem is that the build system is a critical part of CHICKEN and
it has to be extensively tested on all platforms. Testing it is a bit
hard, since even if make is finished successfully, it doesn't
necessarily means that CHICKEN was built correctly.
Very well. What is the proper
Hi, folks. I have a problem with Chicken's eval. Here's a transcript
of a csi session:
(c)2008-2010 The Chicken Team
(c)2000-2007 Felix L. Winkelmann
Version 4.4.0
freebsd-unix-gnu-x86 [ manyargs dload ptables ]
compiled 2010-03-29 on landmine.tx97.net (FreeBSD)
#;1 (eval '(case 1 ((1) 0))
18 matches
Mail list logo
