| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
The gettext implementation in musl does not define some of the internal
symbols that AM_GNU_GETTEXT used to check for. This got fixed in gettext
0.19.8, but the AM_GNU_GETTEXT_VERSION macro requires a specific version
even if the system contains a higher one. We switch to use the new
AM_GNU_GETTEXT_REQUIRE_VERSION macro which requires at least a specific
version, and bump the AM_GNU_GETTEXT_VERSION to the first one that had
support for the new macro.
Signed-off-by: Guillem Jover <guillem@hadrons.org>
|
|
|
|
| |
This is read-only data, so don't include it in the read/write page.
|
|
|
|
| |
Signed-off-by: Temuri Doghonadze <temuri.doghonadze@gmail.com>
|
|
|
|
|
|
| |
This version brings some improvements & fixes over 1.11, and was
released around the same time as autoconf-2.69 that we already
require.
|
|
|
|
|
|
|
|
| |
[guillem@hadrons.org:
- Add missing Plural-Forms field.
- Fix typo. ]
Signed-off-by: Guillem Jover <guillem@hadrons.org>
|
|
|
|
|
| |
The AC_CONFIG_HEADER macro was deprecated a while ago, and the new
one behaves the same but without triggering a warning.
|
|
|
|
|
|
| |
On some platforms, libc does not provide gettext, and hence, more work
is necessary to get gettext to link. AM_GNU_GETTEXT covers for this
case and emits a $(LTLIBINTL) that we can use on those platforms.
|
|
|
|
|
|
|
| |
This special argument was not previously documented, and it is not
obvious from it might be supported.
Signed-off-by: Guillem Jover <guillem@hadrons.org>
|
|
|
|
|
|
|
| |
The man page mentions mount(8) option, but has no reference to that man
page, which makes it more difficult to know where to further look into.
Signed-off-by: Guillem Jover <guillem@hadrons.org>
|
|
|
|
|
|
|
| |
Either quote or mark terms as literal, so that both users and automated
tools, such as spell checkers do not flag these as spelling errors.
Signed-off-by: Guillem Jover <guillem@hadrons.org>
|
|
|
|
| |
Signed-off-by: Guillem Jover <guillem@hadrons.org>
|
|
|
|
|
|
|
|
| |
On GNU/kFreeBSD ENODATA is not defined, so we need to define it instead
to ENOATTR which is defined there. This is a regression from the upstream
build system modernization.
Signed-off-by: Guillem Jover <guillem@hadrons.org>
|
|
|
|
|
|
| |
The Hurd is intended to have no hardcoded limits, and POSIX makes it
possible for a system to not define PATH_MAX. Switch to the simpler
solution which is to generate the constants at compile time.
|
|
|
|
|
|
|
|
| |
This package will not build if autconf2.13 is installed. Adding AC_PREREQ
allows the automatic checks to run the correct version of autconf,
whichever is installed, or to get a clear error message about the problem.
Signed-off-by: Guillem Jover <guillem@hadrons.org>
|
|
|
|
|
|
|
| |
This was accidentally dropped in the build refactor in commit
c8f23c92177c5a82ab3699b1b0d4acbee9afb770 ("modernize build system").
Reported-by: Robert Davies <rob@robdavies.net>
|
|
|
|
| |
Typos found with codespell.
|
|
|
|
|
|
|
| |
Using 64bit versions on 32bit architectures should be enabled with
--enable-largefile, this makes it portable across musl and glibc
Signed-off-by: Khem Raj <raj.khem@gmail.com>
|
|
|
|
| |
Practically speaking, it should be the same.
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
LT_REVISION was computed from PACKAGE_VERSION by stripping the dots from
PACKAGE_VERSION. This caused LT_REVISION to change from 2253 down to
230 when the version changed from 2.2.53 to 2.3.0. This broke ldconfig
versioning and caused the library symlinks to break after a package
update.
https://bugzilla.redhat.com/1938459#c9
Reported by: Kamil Dudka <kdudka@redhat.com>
|
| |
|
|
|
|
|
| |
Add a convenience rule for updating the po files to the main Makefile. Use
that for updating the po files. Update the German translation.
|
|
|
|
|
|
| |
--access has a different indent than the rest of the options.
Signed-off-by: Valentin Vidic <vvidic@valentin-vidic.from.hr>
|
|
|
|
|
|
|
| |
Add a --one-filesystem option to getfacl. With this option, getfacl will not
cross mount points, similar to "tar --one-file-system".
(Patch modified by Andreas Gruenbacher.)
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
__acl_free_acl_obj() is unprepared for NULL, so just return immediately
instead of trying to free a NULL obj.
Signed-off-by: Tavian Barnes <tavianator@tavianator.com>
[Fix the same bug in acl_copy_int.]
Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
|
|
|
|
|
|
| |
... to avoid SIGSEGV with link-time optimisation enabled
Bug: https://bugzilla.redhat.com/1873975
|
|
|
|
| |
Signed-off-by: Anthony Sottile <asottile@umich.edu>
|
| |
|
|
|
|
|
| |
The run script is in $srcdir, not $PWD (aka $builddir).
This fixes `make distcheck`.
|
| |
|
| |
|
|
|
|
| |
This is for the recent "Fix acl_from_text() returning NULL on all input".
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Invoke AC_SYS_LARGEFILE to enable large-file support.
This fixes acl_get_fd and acl_get_file that invoke *stat functions,
as well their users, perm_copy_fd and perm_copy_file.
The tools are the most spectacularly affected because without the fix
they just cannot process large files, e.g.
$ truncate -s2G large-file && getfacl large-file
getfacl: large-file: Value too large for defined data type
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Commit 5dac69ce914202c587f00da638123403111ca673 introduced a regression:
apparently, gcc treats
#pragma GCC diagnostic pop
as a kind of statement, so this causes parse_acl_entry()
to return -1 unconditionally, and therefore causes acl_from_text()
to return NULL unconditionally, too.
Fix this by moving the pragma after the "if" statement.
Reported-by: Valery Inozemtsev <shrek@altlinux.org>
|
|
|
|
|
|
|
|
| |
Valgrind complains that setfacl --restore triggers uninitialized memory
accesses to cmd->c_tag and cmd->c_perm of CMD_REMOVE_ACL commands in
do_set (http://savannah.nongnu.org/bugs/?50566). In this case, the
uninitialized memory accesses have no effect. Silence valgrind by
initializing cmd->c_tag and cmd->c_perm anyway.
|
|
|
|
|
|
|
|
|
| |
When the ACL to be set is equivalent to a file mode, on filesystems
without POSIX ACL support, setfacl falls back to chmod(1) for setting
the file mode to the equivalent of the ACL.
Unfortunately it did not preserve the set-user-ID, set-group-ID, and
sticky bits in that case; fix that.
|
|
|
|
| |
[agruen: Description further clarified.]
|
|
|
|
| |
Allow setfacl commands like "setfacl -m u::rwx- file" to succeed.
|
|
|
|
|
|
|
|
| |
The root/permissions test can fail if run by a root user with umask 077,
because the "d" temporary directory ends up unreadable/unexecutable by
bin or daemon.
Reported anonymously on Savannah: http://savannah.nongnu.org/bugs/?47949
|
|
|
|
|
|
| |
The root-specific tests use two additional groups. Add them to
test/test.group and don't create / delete the non-standard group with
groupadd / groupdel.
|
| |
|
| |
|
|
|
|
|
| |
These permissions will never be added to POSIX ACLs; remove the
(already commented out) definitions.
|
|
|
|
|
|
|
| |
ls adds a '.' at the end of the permission field list on SELinux
machines, filter this out so root tests work on SELinux machines.
Signed-off-by: He Zhe <zhe.he@windriver.com>
|
|
|
|
|
|
|
|
|
|
|
| |
In this function, we call int2ext() on a temporary object on the stack.
When int2ext() checks for a NULL pointer, gcc's -Waddress warning points
out that the address of the temporary object can never be NULL, which is
correct but not helpful:
warning: the address of 'entry_obj' will always evaluate as 'true' [-Waddress]
Suppress this warning.
|