On 08/21/2013 01:28 AM, Claudio Bley wrote: > Only compile securityselinuxhelper.c if xattr support was detected to > avoid this error: > > securityselinuxhelper.c:34:24: fatal error: attr/xattr.h: No such file > or directory compilation terminated. > > Since all SELinux tests depend upon the securityselinuxhelper library, > these test programs are now only build when xattr support is > available. > > Signed-off-by: Claudio Bley <cbley@xxxxxxxxxx> > --- > I'm by no means an automake/SELinux expert and thus I am unsure whether this > is the right thing to do. > > Especially since in commit 95577af442e5 Eric commented that the > "Makefile already guaranteed xattr support." which is obviously not the > the case, but maybe I'm just missing something... Ah, there were TWO programs that use the file, and only one of them was guarded in the Makefile. Thanks for beating me to this fix. > if WITH_SECDRIVER_SELINUX > +if WITH_ATTR > if WITH_TESTS > noinst_LTLIBRARIES += libsecurityselinuxhelper.la ... > > if WITH_QEMU > -if WITH_ATTR > securityselinuxlabeltest_SOURCES = \ > securityselinuxlabeltest.c testutils.h testutils.c \ > testutilsqemu.h testutilsqemu.c > securityselinuxlabeltest_LDADD = $(qemu_LDADDS) $(SELINUX_LIBS) > securityselinuxlabeltest_DEPENDENCIES = libsecurityselinuxhelper.la \ > ../src/libvirt.la > endif > endif > endif As a followup, it would be nice to take advantage of automake's nesting matching - if you write: endif WITH_QEMU endif WITH_TESTS endif WITH_ATTR in reverse order, then Automake will validate that you correctly nested things. We ought to do that everywhere in our Makefile.am; and that means I ought to write a syntax-check rule to enforce it... guess I know what I'm tackling next... -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org
Attachment:
signature.asc
Description: OpenPGP digital signature
-- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list