against libapparmor::change_hat would be linking against
change_hat@IMMUNIX_1.0 rather than change_hat@APPARMOR_1.0 (the
preferred version).
Remove -module to prevent linking portability warning when building
mod_apparmor.
Use RPM_OPT_FLAGS for CFLAGS when building rpms.
to aa_log_record.
- Added the integer that follows the colon in the
audit ID ('audit_sub_id') to aa_log_record.
- Modified the testsuite to deal with the two new
numbers.
<mathiaz@ubuntu.com> [Message-ID: <20070813201254.GD11381@mathias.mathiaz.net>]
Added comments to both file-skipping locations referencing the other
location that needs to be modified.
(The ideal solution would be for this information to be stored in one
commonly referenced location, configurable by distributors and
administratrors.)
[Message-ID: <20070813195328.GB11381@mathias.mathiaz.net>]
Ubuntu installs apxs, the apache module building binary in /usr/bin
rather than /usr/sbin, so search there as well.
<20070813195328.GB11381@mathias.mathiaz.net>].
This fixes the make install target of pam_apparmor so that it depends on
the library already being built.
/var/run/resolvconf/resolv.conf:
The Nameservice abstraction configuration file
(/etc/apparmor.d/abstractions/nameservice) permits reads access to
(amongst other paths) /etc/resolv.conf.
However, on systems using resolvconf, this is a symbolic link to
/etc/resolvconf/run/resolv.conf -- where /etc/resolvconf/run itself
is a symlink to /var/run/resolvconf.
icons" and its duplicate #297243 "Missing YaST icon: All app armor
icons".
The problem was that (a) the .desktop spec doesn't really allow having
path elements in the "Icon=" entry, and (b) that y2controlcenter-gnome
was looking in /usr/share/icons/hicolor/[size]/apps/ for the apparmor
icons, which weren't there.
Today's example is my prepending APPARMOR_REJECT to a message, rather
than APPARMOR_DENIED. This caused problems and was a source of much
hair tearing all weekend.
Turkish has 4 letter "I"s. English has only two, a lowercase dotted i and an
uppercase dotless I. Turkish has lowercase and uppercase forms of both dotted
and dotless I. So tolower(I) != i and toupper(i) != I for us.
Althought this situation is not a apparmor bug (sed can't convert these
properly) its directly affected. But these must locale independent. So
please apply following simple patch to solve this issue.
Acked-By: Steve Beattie <sbeattie@suse.de>
