Age | Commit message (Collapse) | Author |
|
* lib/sha256.c (set_uint32): New function.
(sha256_read_ctx, sha224_read_ctx): Use it.
* lib/sha512.c (set_uint64): New function.
(sha512_read_ctx, sha384_read_ctx): Use it.
* lib/sha256.h: Remove warning about alignment constraint.
* lib/sha512.h: Likewise.
Prompted by similar changes in gnulib's sha1 and md[45] modules.
|
|
* tests/check.mk (vc_exe_in_TESTS): Adapt to new constraint
that vc-list-files be run only from $(top_srcdir).
|
|
* bootstrap.conf (gnulib_modules): Add vc-list-files.
* build-aux/vc-list-files: Remove file.
|
|
* tests/cp/parent-perm: Also check that perms of existing dest
dirs are changed to match those of corresponding src dir
|
|
* src/dd.c: Include quotearg.h.
(operand_matches): New function.
(parse_symbols, operand_is): Use it.
(parse_symbols): 1st arg is now const pointer. Don't modify it.
msgid arg is now just the message, not a format.
(scanargs): Add some 'const's to check for problems like the above.
|
|
|
|
* src/system.h: Include "openat.h" here, ...
* src/chcon.c: ... not here.
* src/chmod.c: Likewise.
* src/chown-core.c: Likewise.
* src/remove.c: Likewise.
Signed-off-by: Jim Meyering <meyering@redhat.com>
|
|
|
|
* src/rmdir.c (remove_parents, main): With --ignore-fail-on-non-empty,
suppress a diagnostic also for other errno values, which can arise
with read-only media or when the parent directory has the immutable
attribute (set via chattr +i).
(errno_may_be_empty, ignorable_failure): New functions.
* src/remove.c (is_empty_dir): Move function to ...
* src/system.h (is_empty_dir): ...here, and make it inline.
Suggested by Josselin Mouette in <http://bugs.debian.org/363011>
via Bob Proulx.
* NEWS: Mention the improvement.
|
|
* src/dd.c (operand_is): New function.
(scanargs): Use it so that we don't need to modify argv, as a nicety.
Problem reported by Adam Goode in
<http://lists.gnu.org/archive/html/bug-coreutils/2008-01/msg00264.html>.
|
|
Reported by Bob Proulx.
|
|
* man/date.x: Add a compact description of the --date=STRING.
* man/touch.x: Likewise.
Reported by A. Costa in http://bugs.debian.org/363011
|
|
* doc/coreutils.texi (dd invocation): Say that dd prints stats
upon normal termination and upon SIGINT.
Signed-off-by: Jim Meyering <meyering@redhat.com>
|
|
* Makefile.am (EXTRA_DIST): Add .version.
(.version): New rule.
(dist-hook): Don't create $(distdir)/.version here, now that it's
being distributed.
* man/Makefile.am (common_dep): Use ../.version, not ../VERSION.
(../VERSION): Remove rule.
* GNUmakefile (dummy): Create .version, not VERSION.
Add an extra "...:= $(shell..." statement to ensure that
.version exists even when the preceding code is not run.
* .gitignore: Remove both .version and VERSION.
Signed-off-by: Jim Meyering <meyering@redhat.com>
|
|
* src/ls.c (gobble_file): Work around a bug in libselinux1-2.0.15
whereby getfilecon returns 0 yet sets the context to NULL.
Reported by Jan Moringen via Michael Stone in
http://bugs.debian.org/463043
* tests/ls/Makefile.am (TESTS): Add proc-selinux-segfault.
* tests/ls/proc-selinux-segfault: Test for the above fix.
* NEWS: Mention the fix.
|
|
This avoids another spurious Mac OS 10.5.1 (Darwin 9.1) test failure.
* tests/cp/preserve-gid: Accept a group ID of 0, as well.
Reported by Elias Pipping.
|
|
rather than just "info PROG". The latter would often fail
or simply display the man page.
* man/help2man: Change the template.
Prompted by http://bugs.debian.org/399684
|
|
* src/chcon.c (usage): Remove invalid reference to lchown.
Reported by Göran Uddeborg.
|
|
* man/Makefile.am (common_dep): Don't depend on configure.ac for
version changes. Instead, depend on ../VERSION.
(../VERSION): New rule.
* Makefile.am (DISTCLEANFILES): Define.
* GNUmakefile: Update ./VERSION.
* .gitignore: List VERSION.
Reported by Sven Joachim.
|
|
* tests/cp/preserve-gid: Set group as well as owner on ".".
Reported by Elias Pipping.
|
|
* doc/coreutils.texi (mv invocation): Adjust the warning: moving a
dir-symlink-specified-with-a-trailing-slash works in a surprising
manner only on some systems. Reported by Tomas Pospisek in
http://bugs.debian.org/343652.
|
|
|
|
* Makefile.cfg (gnulib_dir): Update comment.
* Makefile.maint (announcement) [cl_date, utc_date]: Don't set
now-unused variables.
|
|
|
|
* src/Makefile.am (install_su): Add a semicolon.
|
|
* src/setuidgid.c (main): Don't misinterpret as size_t an error
return from mgetgroups. Reported by Theodoros V. Kalamatianos.
|
|
|
|
* configure.ac (AM_INIT_AUTOMAKE): Require 1.10.1.
|
|
|
|
|
|
|
|
|
|
* README: Remove a note about failing tests on SunOS 4.
On Mac OS 10.5.1 (Darwin 9.1), you'll need --disable-acl.
(Running tests as root): Recommend using "check-root", not "check".
|
|
* src/system.h (__attribute__): Remove the __STRICT_ANSI__ disjunct.
It has been unnecessary since approximately gcc-2.6, and now, leaving
it would cause gcc -Werror -ansi to fail to compile csplit.c.
* gl/lib/randread.c (__attribute__): Likewise.
|
|
|
|
Remove a few entries. Update a few others.
|
|
* tests/check.mk (vc_exe_in_TESTS): Remove t1 and t2 _first_,
in case they exist beforehand and are not writable.
* build-aux/check.mk (am__check_pre): Likewise, remove $@-t.
|
|
|
|
|
|
* Makefile.maint (announcement): Remove use of the
--gnulib-snapshot-time-stamp option.
Use new --gnulib-version option instead.
(gnulib-version): New variable.
|
|
|
|
|
|
Suggestions from James Youngman and Pádraig Brady in
http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/12218/focus=12227
|
|
|
|
* src/cp.c (make_dir_parents_private): If stat fails on the parent
directory, do not add it to the list of directories whose modes
might need fixing later. Also, do not bother invoking 'stat'
unless the stat results might be needed later.
|
|
* tests/cp/parent-perm: Accommodate the situation in which
chmod ("dir", 02755) returns 0 yet fails to set the S_ISGID bit.
|
|
* src/Makefile.am (dir_LDADD, ls_LDADD, vdir_LDADD): Remove uses
of no-longer-defined variable, $(LIB_ACL_TRIVIAL).
|
|
* src/cp.c (do_copy) [--parents]: Free the attribute list.
|
|
* tests/misc/tty-eof: Uncomment a debug "warn".
This change happens to make this test far less likely to fail.
With that statement commented out, this test would fail about
20% of the time on my desktop. Now, it's gone 100 iterations
in a row with no failure.
|
|
|