Description of problem: Build problem with libsafe on x86_64. end of the errors received. intercept.c:1550: warning: int format, different type arg (arg 3) intercept.c: In function `realpath': intercept.c:1584: warning: int format, different type arg (arg 3) intercept.c:1599: warning: unsigned int format, different type arg (arg 2) intercept.c:1599: warning: unsigned int format, different type arg (arg 3) make[2]: *** [intercept.o] Error 1 make[2]: Leaving directory `/rpmbuild/extras/cvs/rpms/libsafe/devel/libsafe-2.0-16/src' make[1]: *** [libsafe] Error 2 make[1]: Leaving directory `/rpmbuild/extras/cvs/rpms/libsafe/devel/libsafe-2.0-16' error: Bad exit status from /var/tmp/rpm-tmp.13915 (%build) RPM build errors: Bad exit status from /var/tmp/rpm-tmp.13915 (%build) make: *** [x86_64] Error 1
http://people.redhat.com/wtogami/temp/libsafe-2.0-16c.src.rpm
Michael Schwendt suggested that libsafe could be/should be pulled from extras entirely. At least he told me to exclude it from the build for now.
I'm interested to hear the reasons. Does it screw package dependencies when libsafe is installed in buildroot during rpmbuild? If not then it should be fine. Software behavior should be the same. Steve would you concur?
I don't agree with pulling from the build. I'd like to hear the reason. It solves a different problem than some of the other security tools. Some people are not able to use SE Linux because of strange local configurations. Exec Shield may not solve all problems. There are programs that need exec shield turned off. Also, for people still running and maintaining RH 7.3, or other legacy distributions..there is no hope of using exec shield. That said, I don't know if it will actually *run* for x86_64. Its stack layout has to be identical to IA32 family or I'll need to do some investigation and re-work some code.
Steve, should we ExclusiveArch: i386 for now?
Did the old version without my "c" patch work for x86_64? If not, I'd say we should ExclusiveArch it for now. I'll fix the warnings so that maybe we can try it on an x86_64 to find out. BTW, I don't have access to a x86_64 machine right now.
Re: comment 2 Only rationale for that suggestion was that libsafe had not been updated or rebuilt since Red Hat Linux 8.0/9 and fedora.us' "patches" repository, http://download.fedora.us/patches/redhat/9/i386/RPMS.stable/ and hence belonged to the potentially unmaintained packages which should be checked before a rebuild is published. The FC3Status page http://fedoraproject.org/wiki/Extras_2fFC3Status contained a comment about that.
Well "lack of maintainer" is no longer a problem. And I agree that libsafe is a useful optional add-on. I say we ExclusiveArch: i386 for now and keep this bug open. Maybe it can be fixed for x86_64 later.
I agree with ExclusiveArch. But adding x86_64 support is more of an RFE than a bug report though.
ExclusiveArch: i386 is set. Closing since x86_64 is not an intended feature.
*** Bug 158451 has been marked as a duplicate of this bug. ***