platform/upstream/coreutils.git
18 years agoSetting TIME_STYLE=long-iso in the environment would make the
Jim Meyering [Sun, 11 Jun 2006 08:12:37 +0000 (08:12 +0000)]
Setting TIME_STYLE=long-iso in the environment would make the
cp/same-file test fail.
* tests/envvar-check (vars): Add TIME_STYLE to the list.
* tests/cp/same-file: Revert last change.
Source the envvar-check script, to ensure that TIME_STYLE
settings don't affect these tests.

18 years ago* tests/cp/same-file: Execute 'ls' in the C locale, so that it
Paul Eggert [Sun, 11 Jun 2006 07:59:16 +0000 (07:59 +0000)]
* tests/cp/same-file: Execute 'ls' in the C locale, so that it
uses POSIX time stamp formats.  Problem reported by John Nixon in
<http://lists.gnu.org/archive/html/bug-coreutils/2006-06/msg00062.html>.

18 years ago(__attribute__): Don't define if already defined.
Paul Eggert [Sun, 11 Jun 2006 07:11:31 +0000 (07:11 +0000)]
(__attribute__): Don't define if already defined.

18 years agoMention the AIX-strndup-bug vs. dircolors workaround.
Jim Meyering [Sat, 10 Jun 2006 09:41:39 +0000 (09:41 +0000)]
Mention the AIX-strndup-bug vs. dircolors workaround.

18 years agoApply this change from gnulib:
Jim Meyering [Sat, 10 Jun 2006 09:10:12 +0000 (09:10 +0000)]
Apply this change from gnulib:

2006-05-30  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
    Bruno Haible  <bruno@clisp.org>

* strndup.c (strndup) [!_LIBC]: Don't undefine macro definition.

18 years agoApply this change from gnulib:
Jim Meyering [Sat, 10 Jun 2006 09:01:41 +0000 (09:01 +0000)]
Apply this change from gnulib:

2006-05-30  Ralf Wildenhues  <Ralf.Wildenhues@gmx.de>
    Bruno Haible  <bruno@clisp.org>

* strndup.m4 (gl_FUNC_STRNDUP): Replace the AC_REPLACE_FUNCS with a
check for the declaration of strnlen and a run test that exposes the
AIX 5.1 strnlen bug.  In the failure case, #define strndup to
rpl_strndup.

18 years ago*** empty log message ***
Jim Meyering [Sat, 10 Jun 2006 07:58:09 +0000 (07:58 +0000)]
*** empty log message ***

18 years agoRequire a "Version N.M" line at the top of the ChangeLog
Jim Meyering [Sat, 10 Jun 2006 07:58:05 +0000 (07:58 +0000)]
Require a "Version N.M" line at the top of the ChangeLog
file only when making the actual release, not when running
"make distcheck".

(maintainer-distcheck): Don't depend on
changelog-check.
(alpha beta major): Depend on it here, instead.

18 years agoEnsure that cat works with any of the options, -A -v -e -E -T,
Jim Meyering [Thu, 8 Jun 2006 20:13:37 +0000 (20:13 +0000)]
Ensure that cat works with any of the options, -A -v -e -E -T,
when applied to files in /proc and /sys, even when the FIONREAD
ioctl produces nonsensical results.  Before this change, cat would
produce no output (or truncated output), for some linux kernels.

* src/cat.c (write_pending): New function, factored out of cat.
(cat): Also interpret a negative ioctl/FIONREAD count as indicating
that there are bytes to read.  Some versions of linux-2.6.16 do that.
Write any pending output before returning.
Reported by Dan Jacobson in <http://bugs.debian.org/370583>.
* NEWS: Mention this bug fix.
* tests/misc/cat-proc: New file.  Test for the above.
* tests/misc/Makefile.am (TESTS): Add cat-proc.

18 years ago(eval4): Detect overflow properly when multiplying INTMAX_MIN * -1.
Paul Eggert [Thu, 8 Jun 2006 02:53:25 +0000 (02:53 +0000)]
(eval4): Detect overflow properly when multiplying INTMAX_MIN * -1.

18 years agoDon't claim x86 behavior is erroneous in a comment.
Paul Eggert [Wed, 7 Jun 2006 05:58:57 +0000 (05:58 +0000)]
Don't claim x86 behavior is erroneous in a comment.

18 years agoThe 'expr' command now detects and reports integer overflow.
Paul Eggert [Wed, 7 Jun 2006 05:52:41 +0000 (05:52 +0000)]
The 'expr' command now detects and reports integer overflow.

18 years ago(integer_overflow): New function.
Paul Eggert [Wed, 7 Jun 2006 05:52:27 +0000 (05:52 +0000)]
(integer_overflow): New function.
(eval4, eval3): Check for integer overflow.

18 years agoThe 'expr' command now detects and reports integer overflow.
Paul Eggert [Wed, 7 Jun 2006 05:52:07 +0000 (05:52 +0000)]
The 'expr' command now detects and reports integer overflow.

18 years agoFix problems when building with Solaris/SVR4/etc. make, which uses a
Paul Eggert [Mon, 5 Jun 2006 20:05:26 +0000 (20:05 +0000)]
Fix problems when building with Solaris/SVR4/etc. make, which uses a
different and somewhat bogus implementation of VPATH.

18 years agoRename file to avoid bugs with VPATH builds and Solaris 'make'.
Paul Eggert [Mon, 5 Jun 2006 20:05:00 +0000 (20:05 +0000)]
Rename file to avoid bugs with VPATH builds and Solaris 'make'.

18 years agoDon't assign to PROG; no longer needed
Paul Eggert [Mon, 5 Jun 2006 20:03:52 +0000 (20:03 +0000)]
Don't assign to PROG; no longer needed
now that Makefile.am sets PROG to the basename.

18 years ago(PROG): Take the basename of $$tst,
Paul Eggert [Mon, 5 Jun 2006 20:03:27 +0000 (20:03 +0000)]
(PROG): Take the basename of $$tst,
in case Solaris make has prepended the directory.
(TESTS): Adjust to test script renamings.

18 years ago* regexec.c (group_nodes_into_DFAstates): Fix a buffer overrun
Paul Eggert [Mon, 5 Jun 2006 05:21:46 +0000 (05:21 +0000)]
* regexec.c (group_nodes_into_DFAstates): Fix a buffer overrun
reported by Andreas Schwab.

18 years ago*** empty log message ***
Jim Meyering [Sat, 3 Jun 2006 09:04:26 +0000 (09:04 +0000)]
*** empty log message ***

18 years agoMake `cp --link --no-dereference' work also on systems where the
Jim Meyering [Sat, 3 Jun 2006 09:04:22 +0000 (09:04 +0000)]
Make `cp --link --no-dereference' work also on systems where the
link system call cannot create a hard link to a symbolic link.

(copy_internal) [LINK_FOLLOWS_SYMLINKS]: Don't use
the link syscall on a symlink when it would do the wrong thing.
Based on the patch by Aurelien Jarno: <http://bugs.debian.org/329451>

18 years ago`cp --link --no-dereference' now works also on systems where the
Jim Meyering [Sat, 3 Jun 2006 09:04:05 +0000 (09:04 +0000)]
`cp --link --no-dereference' now works also on systems where the
link system call cannot create a hard link to a symbolic link.

18 years agoTest for this:
Jim Meyering [Sat, 3 Jun 2006 09:03:19 +0000 (09:03 +0000)]
Test for this:

  Make `cp --link --no-dereference' work also on systems where the
  link system call cannot create a hard link to a symbolic link.
  * src/copy.c (copy_internal) [LINK_FOLLOWS_SYMLINKS]: Don't use
  the link syscall on a symlink when it would do the wrong thing.
  Based on the patch by Aurelien Jarno: <http://bugs.debian.org/329451>

18 years ago(TESTS): Add link-no-deref.
Jim Meyering [Sat, 3 Jun 2006 09:02:53 +0000 (09:02 +0000)]
(TESTS): Add link-no-deref.

18 years ago*** empty log message ***
Jim Meyering [Sat, 3 Jun 2006 08:29:28 +0000 (08:29 +0000)]
*** empty log message ***

18 years agoFix some HP-UX 11.23 porting bugs.
Paul Eggert [Thu, 1 Jun 2006 07:10:48 +0000 (07:10 +0000)]
Fix some HP-UX 11.23 porting bugs.

18 years agoDon't assume that >&- works.
Paul Eggert [Thu, 1 Jun 2006 07:10:34 +0000 (07:10 +0000)]
Don't assume that >&- works.

18 years agoDon't assume that >&- works. Add a /dev/full test.
Paul Eggert [Thu, 1 Jun 2006 07:10:15 +0000 (07:10 +0000)]
Don't assume that >&- works.  Add a /dev/full test.

18 years agoDon't assume that \< \> works in sed.
Paul Eggert [Thu, 1 Jun 2006 07:09:51 +0000 (07:09 +0000)]
Don't assume that \< \> works in sed.

18 years ago*** empty log message ***
Jim Meyering [Tue, 30 May 2006 06:22:20 +0000 (06:22 +0000)]
*** empty log message ***

18 years ago(usage): Add `v' to the list of sorting-related options.
Jim Meyering [Tue, 30 May 2006 06:22:16 +0000 (06:22 +0000)]
(usage): Add `v' to the list of sorting-related options.
From Justin Pryzby.

18 years agoUpdate from gnulib.
Jim Meyering [Mon, 29 May 2006 06:55:36 +0000 (06:55 +0000)]
Update from gnulib.

18 years ago*** empty log message ***
Jim Meyering [Sun, 28 May 2006 12:12:47 +0000 (12:12 +0000)]
*** empty log message ***

18 years agoSource lang-default. From Ralf Wildenhues.
Jim Meyering [Sun, 28 May 2006 12:11:35 +0000 (12:11 +0000)]
Source lang-default.  From Ralf Wildenhues.

18 years ago*** empty log message ***
Jim Meyering [Sun, 28 May 2006 09:33:00 +0000 (09:33 +0000)]
*** empty log message ***

18 years agoAIX 4.3.3 gives a different diagnostic.
Jim Meyering [Sun, 28 May 2006 09:32:54 +0000 (09:32 +0000)]
AIX 4.3.3 gives a different diagnostic.
Recognize it, too.  Reported by Ralf Wildenhues, in
http://lists.gnu.org/archive/html/bug-coreutils/2006-05/msg00192.html

18 years ago(gl_C99_STRTOLD): Use a link test rather than a
Paul Eggert [Sun, 28 May 2006 09:16:57 +0000 (09:16 +0000)]
(gl_C99_STRTOLD): Use a link test rather than a
compile test, for Tru64 4.0D.

18 years ago*** empty log message ***
Jim Meyering [Sun, 28 May 2006 09:03:12 +0000 (09:03 +0000)]
*** empty log message ***

18 years ago*** empty log message ***
Jim Meyering [Sun, 28 May 2006 09:02:42 +0000 (09:02 +0000)]
*** empty log message ***

18 years ago*** empty log message ***
Jim Meyering [Sat, 27 May 2006 15:51:23 +0000 (15:51 +0000)]
*** empty log message ***

18 years agoUse `AM_MAKEINFOFLAGS' rather than
Jim Meyering [Sat, 27 May 2006 15:51:19 +0000 (15:51 +0000)]
Use `AM_MAKEINFOFLAGS' rather than
overwriting `MAKEINFO', so that `missing' can do its job.

18 years ago*** empty log message ***
Jim Meyering [Sat, 27 May 2006 15:13:35 +0000 (15:13 +0000)]
*** empty log message ***

18 years ago*** empty log message ***
Jim Meyering [Sat, 27 May 2006 15:12:05 +0000 (15:12 +0000)]
*** empty log message ***

18 years agochgrp now supports --preserve-root, --no-preserve-root
Jim Meyering [Sat, 27 May 2006 15:01:36 +0000 (15:01 +0000)]
chgrp now supports --preserve-root, --no-preserve-root

18 years ago*** empty log message ***
Jim Meyering [Sat, 27 May 2006 15:00:22 +0000 (15:00 +0000)]
*** empty log message ***

18 years ago*** empty log message ***
Jim Meyering [Sat, 27 May 2006 14:44:44 +0000 (14:44 +0000)]
*** empty log message ***

18 years agoSupport new options: --preserve-root and --no-preserve-root.
Jim Meyering [Sat, 27 May 2006 14:44:41 +0000 (14:44 +0000)]
Support new options: --preserve-root and --no-preserve-root.
Somehow this program was skipped when those options were added to chown,
chmod, and rm.  Reported by vaqflabuopac@spammotel.com in
<http://bugs.debian.org/365656>.

18 years ago*** empty log message ***
Jim Meyering [Sat, 27 May 2006 13:43:12 +0000 (13:43 +0000)]
*** empty log message ***

18 years ago(check-texinfo): Use literal `{' only in brackets, i.e., [{] or [}],
Jim Meyering [Sat, 27 May 2006 13:43:07 +0000 (13:43 +0000)]
(check-texinfo): Use literal `{' only in brackets, i.e., [{] or [}],
to avoid triggering an error from Solaris 2.6's grep.

18 years ago*** empty log message ***
Jim Meyering [Sat, 27 May 2006 13:32:10 +0000 (13:32 +0000)]
*** empty log message ***

18 years agoUse `$(EGREP)' instead of `grep -E'.
Jim Meyering [Sat, 27 May 2006 13:32:04 +0000 (13:32 +0000)]
Use `$(EGREP)' instead of `grep -E'.

18 years ago(small_open, large_open): New macros.
Paul Eggert [Thu, 25 May 2006 21:57:22 +0000 (21:57 +0000)]
(small_open, large_open): New macros.
(__open, __open64) [!_LIBC]: Remove.
(__gen_tempname): Use small_open and large_open instead of __open
and __open64.  This fixes a portability bug on HP-UX 11.11i
reported by Simon Wing-Tang in
<http://lists.gnu.org/archive/html/bug-coreutils/2006-05/msg00114.html>.

18 years agoRemove mention of --seed.
Paul Eggert [Thu, 25 May 2006 16:48:05 +0000 (16:48 +0000)]
Remove mention of --seed.

18 years ago(usage): Remove mention of --seed.
Paul Eggert [Thu, 25 May 2006 16:47:43 +0000 (16:47 +0000)]
(usage): Remove mention of --seed.

18 years ago*** empty log message ***
Jim Meyering [Wed, 24 May 2006 07:35:32 +0000 (07:35 +0000)]
*** empty log message ***

18 years agoRequire autoconf-2.59d, for its fixed version
Jim Meyering [Wed, 24 May 2006 07:35:28 +0000 (07:35 +0000)]
Require autoconf-2.59d, for its fixed version
of AC_FUNC_LSTAT_FOLLOWS_SLASHED_SYMLINK.

18 years agoFix misspelling.
Paul Eggert [Tue, 23 May 2006 07:48:31 +0000 (07:48 +0000)]
Fix misspelling.

18 years agoNote problem with string.h.
Paul Eggert [Tue, 23 May 2006 07:48:12 +0000 (07:48 +0000)]
Note problem with string.h.

18 years agofilemode.c: Don't include <string.h>.
Paul Eggert [Tue, 23 May 2006 07:47:51 +0000 (07:47 +0000)]
filemode.c: Don't include <string.h>.

18 years agoDon't include <string.h>; this include was
Paul Eggert [Tue, 23 May 2006 07:47:32 +0000 (07:47 +0000)]
Don't include <string.h>; this include was
inadvertantly put into the previous patch.

18 years ago* filemode.c (setst): Remove.
Paul Eggert [Mon, 22 May 2006 20:11:45 +0000 (20:11 +0000)]
* filemode.c (setst): Remove.
(strmode): Rewrite to avoid setst.  This makes the code shorter,
(arguably) clearer, and the generated code is a bit smaller on my
Debian GNU/Linux stable x86 host.

18 years agoDocument the internals better. Most of this change
Paul Eggert [Mon, 22 May 2006 18:01:48 +0000 (18:01 +0000)]
Document the internals better.  Most of this change
was written by Bruno Haible.

18 years ago*** empty log message ***
Jim Meyering [Sun, 21 May 2006 22:04:34 +0000 (22:04 +0000)]
*** empty log message ***

18 years ago*** empty log message ***
Jim Meyering [Sat, 20 May 2006 22:18:20 +0000 (22:18 +0000)]
*** empty log message ***

18 years ago(fts_open): Fail with EINVAL if a caller violates this rule:
Jim Meyering [Sat, 20 May 2006 22:18:16 +0000 (22:18 +0000)]
(fts_open): Fail with EINVAL if a caller violates this rule:
Either FTS_LOGICAL or FTS_PHYSICAL must be provided to the
fts_open() function.

18 years ago*** empty log message ***
Jim Meyering [Sat, 20 May 2006 19:00:46 +0000 (19:00 +0000)]
*** empty log message ***

18 years ago(main): Use FTS_PHYSICAL here, too.
Jim Meyering [Sat, 20 May 2006 17:27:07 +0000 (17:27 +0000)]
(main): Use FTS_PHYSICAL here, too.

18 years ago*** empty log message ***
Jim Meyering [Sat, 20 May 2006 07:20:03 +0000 (07:20 +0000)]
*** empty log message ***

18 years ago(main): Rename local, s/symlink_deref_bit/symlink_deref_bits/
Jim Meyering [Sat, 20 May 2006 07:19:53 +0000 (07:19 +0000)]
(main): Rename local, s/symlink_deref_bit/symlink_deref_bits/
and arrange for -D to set fts' FTS_PHYSICAL bit as well as
FTS_COMFOLLOW.  Spotted by Justin Pryzby.

18 years ago*** empty log message ***
Jim Meyering [Fri, 19 May 2006 22:30:55 +0000 (22:30 +0000)]
*** empty log message ***

18 years agoMerge changes from automake, retaining the ""--to...
Jim Meyering [Fri, 19 May 2006 22:30:27 +0000 (22:30 +0000)]
Merge changes from automake, retaining the ""--to...
kludge to placate overzealous `make distcheck' check.

18 years agoSync nanosleep from gnulib.
Paul Eggert [Fri, 19 May 2006 18:03:25 +0000 (18:03 +0000)]
Sync nanosleep from gnulib.

18 years ago*** empty log message ***
Jim Meyering [Fri, 19 May 2006 12:36:36 +0000 (12:36 +0000)]
*** empty log message ***

18 years ago(main): Don't let -D, -L, or -P turn off the internal
Jim Meyering [Fri, 19 May 2006 12:36:18 +0000 (12:36 +0000)]
(main): Don't let -D, -L, or -P turn off the internal
FTS_TIGHT_CYCLE_CHECK directory traversal option.
Reported by Justin Pryzby in http://bugs.debian.org/367691

18 years agomention cp -p vs. no-/proc fix
Jim Meyering [Tue, 16 May 2006 17:27:04 +0000 (17:27 +0000)]
mention cp -p vs. no-/proc fix

18 years ago(usage): Correct description of -a: s/-dpR/-dpPR/. From Tomas Pospisek.
Jim Meyering [Mon, 15 May 2006 20:19:02 +0000 (20:19 +0000)]
(usage): Correct description of -a: s/-dpR/-dpPR/. From Tomas Pospisek.

18 years ago*** empty log message ***
Jim Meyering [Mon, 15 May 2006 20:18:49 +0000 (20:18 +0000)]
*** empty log message ***

18 years ago*** empty log message ***
Jim Meyering [Mon, 15 May 2006 13:28:28 +0000 (13:28 +0000)]
*** empty log message ***

18 years ago(xfts_open): Always use FTS_CWDFD.
Jim Meyering [Mon, 15 May 2006 13:28:24 +0000 (13:28 +0000)]
(xfts_open): Always use FTS_CWDFD.

18 years ago*** empty log message ***
Jim Meyering [Mon, 15 May 2006 10:59:15 +0000 (10:59 +0000)]
*** empty log message ***

18 years agoAvoid the expense of an fstat, when possible.
Jim Meyering [Mon, 15 May 2006 10:59:10 +0000 (10:59 +0000)]
Avoid the expense of an fstat, when possible.

(O_NOFOLLOW, STREQ): Define.
(diropen_fd): Remove function.  Merge it into sole caller...
(diropen): ...here.  Use O_NOFOLLOW when appropriate.
(fts_safe_changedir): Call fstat for dev/inode check, only if the
previous open/openat call may have opened the wrong directory.

18 years ago*** empty log message ***
Jim Meyering [Mon, 15 May 2006 10:00:30 +0000 (10:00 +0000)]
*** empty log message ***

18 years agoRestore the parts of fts that were removed on 2006-01-17, so that
Jim Meyering [Mon, 15 May 2006 10:00:16 +0000 (10:00 +0000)]
Restore the parts of fts that were removed on 2006-01-17, so that
it's easier for legacy applications designed for the version
of fts in glibc or BSD to convert to this more robust version.
Add a new mode, FTS_CWDFD, by which to enable the improved
(openat- based -- aka no-chdir) semantics.

* fts_.h (FTS_CWDFD): Define.  Callers must use this fts_open
option to enable the more robust behavior.
(FTS_OPTIONMASK): Widen accordingly.
* fts.c: Restore removed code, reverting the default behavior.

18 years ago(AC_HEADER_STDBOOL): Fix overly-picky
Paul Eggert [Mon, 15 May 2006 03:17:08 +0000 (03:17 +0000)]
(AC_HEADER_STDBOOL): Fix overly-picky
test for C99 conformance; (bool) 0.5 is an integer constant
expression, but (bool) -0.5 is not.  Problem reported by Fedor
Sergeev in <http://forum.sun.com/jive/thread.jspa?threadID=96202>.

18 years ago*** empty log message ***
Jim Meyering [Sat, 13 May 2006 08:41:37 +0000 (08:41 +0000)]
*** empty log message ***

18 years ago*** empty log message ***
Jim Meyering [Sat, 13 May 2006 08:41:06 +0000 (08:41 +0000)]
*** empty log message ***

18 years agoTest two more cases.
Jim Meyering [Sat, 13 May 2006 08:41:02 +0000 (08:41 +0000)]
Test two more cases.

18 years agoupdate from gnulib
Jim Meyering [Fri, 12 May 2006 07:18:08 +0000 (07:18 +0000)]
update from gnulib

18 years ago*** empty log message ***
Jim Meyering [Thu, 11 May 2006 19:44:47 +0000 (19:44 +0000)]
*** empty log message ***

18 years ago(sha1_buffer): Correct comment: s/MD5/SHA1/. From James Lemley.
Jim Meyering [Thu, 11 May 2006 19:44:42 +0000 (19:44 +0000)]
(sha1_buffer): Correct comment: s/MD5/SHA1/.  From James Lemley.

18 years ago.
Jim Meyering [Thu, 11 May 2006 08:58:23 +0000 (08:58 +0000)]
.

18 years agoMention that:
Jim Meyering [Thu, 11 May 2006 08:57:45 +0000 (08:57 +0000)]
Mention that:
  mv -T DIR EMPTY_DIR no longer fails unconditionally

18 years ago*** empty log message ***
Jim Meyering [Thu, 11 May 2006 08:57:32 +0000 (08:57 +0000)]
*** empty log message ***

18 years ago*** empty log message ***
Jim Meyering [Thu, 11 May 2006 08:55:08 +0000 (08:55 +0000)]
*** empty log message ***

18 years agomv -T DIR EMPTY_DIR no longer fails
Jim Meyering [Thu, 11 May 2006 08:55:04 +0000 (08:55 +0000)]
mv -T DIR EMPTY_DIR no longer fails

(copy_internal): Don't manually prohibit a move where
the destination is an existing directory.  Sometimes doing that is
valid.  Let the rename system call enforce the rules.  That is
allowed only when the source is a directory and the destination
directory (to be replaced) is empty.  Reported by Eric Blake.

18 years ago*** empty log message ***
Jim Meyering [Thu, 11 May 2006 08:54:30 +0000 (08:54 +0000)]
*** empty log message ***

18 years ago(TESTS): Add no-target-dir.
Jim Meyering [Thu, 11 May 2006 08:54:23 +0000 (08:54 +0000)]
(TESTS): Add no-target-dir.

18 years agoensure that mv did what it's supposed to
Jim Meyering [Thu, 11 May 2006 08:18:34 +0000 (08:18 +0000)]
ensure that mv did what it's supposed to

18 years ago.
Jim Meyering [Thu, 11 May 2006 07:53:25 +0000 (07:53 +0000)]
.

18 years ago*** empty log message ***
Jim Meyering [Thu, 11 May 2006 07:51:17 +0000 (07:51 +0000)]
*** empty log message ***