* tests/tail-2/inotify-hash-abuse: Use kill rather than wait
to determine if the tail process is still running.
* tests/tail-2/inotify-hash-abuse2: Ditto.
* tests/ls/infloop: OpenBSD4.5's /bin/sh would mistakenly include
"set -x"-output in an application's stderr stream when stderr is
redirected before stdout. This was causing one spurious test failure.
The work-around: redirect stdout first.
Reported by Nelson Beebe.
* tests/misc/sort-version: Don't use <<- and indented here-doc contents.
s/<<-/<</ and unindent the here-document contents. Otherwise,
bash would ignore the indented delimiter and use EOF, thus silently
skipping this test. OpenBSD5.4's shell reported the failure:
$ printf 'cat<<-x\n foo\n x\n'|sh
sh: <stdin>[4]: here document `x' unclosed
[Exit 1]
by contrast, bash warns but still exits successfully:
$ printf 'cat<<-x\n foo\n x\n'|bash && echo you lose
bash: line 3: warning: here-document at line 1 delimited by \
end-of-file (wanted `x')
foo
x
you lose
* tests/check.mk: Prepend /usr/xpg4/bin to the $PATH if present.
Using the more standard utilities allows tests such as misc/printenv,
which uses the -E option to grep, to complete.
Configure is supposed to detect insufficient XATTR support.
However, if a system has the required headers, but no library,
the configure script would mistakenly enable USE_XATTR.
* m4/xattr.m4 (gl_FUNC_XATTR): If the attr_copy_file function
is not found, don't set USE_XATTR.
Nelson Beebe reported a link failure on RHEL 5.3.
Also, do not let the combination of --disable-xattr and
a stray LIB_XATTR environment setting perturb the build.
* NEWS (Build-related): Mention it.
* src/ls.c: Include <sys/capability.h> later, to avoid build
failure with a header from libcap-2.16-1 or earlier.
See http://bugzilla.redhat.com/483548 for details.
Before this change, with too long a file name, the name would
abut the date field on the left and possibly also the "Page N"
field on the right, rather than leaving a one-space separator
in each case. Fixes a regression introduced on Mar 6 2009,
by commit a4053c5291
* src/pr.c (print_header): Ensure that there is at least one
space before and after the file name part of the header line.
* NEWS: Mention it.
* tests/pr/W20l24f-ll: s/xPage/ x Page/.
* THANKS: Update.
Reported by Denis McKeon, in https://savannah.gnu.org/bugs/?28492.
* configure.ac: Require autoconf-2.62 and automake-1.11.1 or newer.
* bootstrap.conf (buildreq): Require automake-1.11.1 or newer,
to ensure people use a version with the fix for CVE-2009-4029.
Note that the coreutils-8.2 tarball included a fixed Makefile.in.
Require autoconf-2.62, per automake.
* src/ls.c (print_color_indicator): When using 'LINK target' in
dircolors, treat broken symlink as C_ORPHAN.
* tests/misc/ls-misc (sl-dangle2, sl-dangle3, sl-dangle4)
(sl-dangle5): Test for it, and add more coverage.
* NEWS: Document it.
* THANKS: Update.
Reported by Chris Jones.
* src/tail.c (usage): Reword tail -F description, so that it no
longer mentions details specific to the non-inotify implementation.
Also, join diagnostic strings (while staying under the 509-byte limit)
to ease formatting of translations. The latter was prompted by
a report from Stéphane Raimbault.
* doc/coreutils.texi (tail invocation): Update description here, too.
This fixes a bug whereby tail -F would fail to track changes
to a file that was a target of a rename, and when the source of
the rename was another tailed file.
* src/tail.c (tail_forever_inotify): Ensure the wd is not already
present in the hash table before trying to add it. When a new watch
descriptor is added to the `wd_to_name' hash table, check that it is
not already present. If it is present then remove the previous element.
* src/tail.c (tail_forever_inotify): Avoid modifying fdspec->wd while
it is in the wd_to_name hash table. Once it is removed, it can be
added using the new `wd' as key for the hash table. This fixes the
abort-inducing bug reported by Rob Wortman in
http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/19372
* src/tail.c (tail_forever_inotify): Do not use f[i] in a context
where i may be larger than the largest valid index. In the final
"if" clause in which we'd remove an inotify watch, we might have
used f[n_files]. Use fspec instead, since it is guaranteed to
be defined.
* src/tail.c (fremote): Add a comment.
Move definition "up" to precede first use, so we can
remove its prototype and the #if..#endif around each use.
(any_remote_file): Rename from any_remote_files.
* src/tail.c (struct File_spec): Add a flag to record if file is remote.
(recheck): If we're using inotify then check if the file has gone remote
and if so, drop it with a warning.
(any_remote_files): A new function to check for any open remote files.
(tailable_stdin): A new function to refactor the check for whether
a tailable file was specified through stdin.
(fremote): A new function to check if a file descriptor
refers to a remote file.
(tail_forever_inotify): Add some comments.
(tail_file): Record if a file is remote when initially opened.
(main): Disable inotify if any remote files specified.
Also document the caveat about remounted files not
being noticed by inotify.
* NEWS: Mention the fix.
* src/wc.c (main): Set stdout to line buffered mode
to ensure parallel running instances don't intersperse
their output. This adds 6.5% to the run time in the worst case
of many zero length files, but has neglible impact for
standard sized files.
* tests/misc/wc-parallel: New test for atomic output.
* tests/Makefile.am: Reference it.
* NEWS: Mention the fix
This is similar to commit 710fe413, 20-10-2009,
"md5sum, sha*sum, sum: line-buffer the printed checksums"
* src/stat.c (human_fstype): Add the following FS types:
fuseblk, rpc_pipefs. Also fix a typo of minux3 to minix3,
and mention the fs-magic-compare make target to help update the list.
* NEWS: Mention the fix.
* dist-check.mk (built_programs): Use $(bin_PROGRAMS), not $(PROGRAMS).
Otherwise, my-instcheck would fail due to non-installation of e.g.,
the noinst_PROGRAMS, setuidgid and getlimits.
(taint-distcheck): Correct the grep command that checks for libtool
traces in configure.
Regression introduced in coreutils 8.1 due to a bug in the Linux
kernel implementation of utimensat with mtime of UTIME_OMIT.
* gnulib: Update to latest, to pick up utimensat fix.
* NEWS: Mention the change.
* THANKS: Update.
Reported by John Stanley.
* bootstrap.conf (gnulib_modules): Remove the strverscmp module
which is not used since commit e505736f, on 03-10-2008,
"ls and sort: use filevercmp instead of strverscmp"
* doc/coreutils.texi (sort invocation): Reference the additional
info about filevercmp rather than the unused strverscmp.
(Details about version sort): Add some examples that are not
handled well by fileversmp.
* src/ls.c: Change a comment referencing the now unused strverscmp.
* src/remove.c (rm_fts): Fix incorrect comparison of
device and inode numbers.
* tests/rm/one-file-system2: Add a separate test so
that it can be run as a normal user (It doesn't need to mount).
* tests/Makefile.am: Reference it.
* NEWS: Mention the fix.
Reported by Jan Larres.
* dist-check.mk (null_AM_MAKEFLAGS): Remove LIBTOOL. Adding it was
erroneous, since it is required when building from a distribution
tarball of a libtool-using project. Reported by Ralf Wildenhues.
(my-distcheck): Reorganize to use a subshell and set -e, so that
failures propagate "out". Without this change, setting LIBTOOL=false
would cause a failure that would then be ignored, probably due to a
problem in $(install-transform-check).
* dist-check.mk (built_programs): More generic, but still assumes src/.
Don't set GZIP in environment when untarring.
(my-distcheck): Use $(DIST_ARCHIVES), rather than assuming that
there is always a .tar.gz file.
* dist-check.mk (null_AM_MAKEFLAGS): Define here, not in maint.mk.
(built_programs): Likewise.
(my-distcheck): Move comments to...
(coreutils-path-check): ...the code they refer to.
Remove obsolete comments.
(null_AM_MAKEFLAGS): Add gperf, even though it's not used here.
* gnulib: Update to latest, for fixed maint.mk.
* gl/lib/tempname.c.diff: Adjust patch to apply to gnulib, now that
most TABs in indentation have been converted to spaces by running
this command: f=tempname.c.diff; patch-xform $f > k && mv k $f
This bug showed up via valgrind as a "Conditional jump or move
depends on uninitialized value(s)" error.
* src/tail.c (ignore_fifo_and_pipe): New function.
(main): Use it only when tailing forever.
The code to compute n_viable and mark some F[i] as ignored would call
isapipe on an uninitialized file descriptor. But n_viable and those
.ignored marks are useful/used only when tailing forever. This bug
was introduced via commit f0ff8c73 (7.6), "tail: make the new
piped-stdin test as portable as the old one".
* NEWS (Bug fixes): Mention it.
* NEWS (Bug fixes): Mention implications of the "make distcheck" change.
This was introduced on 2008-07-22 by commit 9bb0d576, "tests: ensure
"make check" w/tainted build dir no longer impacts $HOME".
* src/sort.c (main): Reset the SIGCHLD handler to the default
as otherwise wait() could return an error.
* tests/misc/sort-compress: Set the CHLD handler in a subshell
to SIG_IGN to ensure the sort command resets it to SIG_DFL.
* NEWS: Mention the fix.
* src/timeout.c (main): Reset the SIGCHLD handler to the default
as otherwise wait() could return -1 and set errno to ECHILD.
This condition was ignored until commit 0b1dcf33, on 31-08-2009,
"timeout: defensive handling of all wait() errors"
but subsequently timeout would run the command correctly
but then fail with an error message.
* tests/misc/timeout: In a subshell set the CHLD handler to
SIG_IGN to ensure the timeout command resets it to SIG_DFL.
* NEWS: Mention the fix.
* tests/misc/stty-row-col: Linux virtual consoles at least,
issue an error if you try to increase their size, so skip the
test if we can't increase the dimensions of the tty by 1 cell.
Reported by Matthew Burgess.
* dist-check.mk (tmpdir): Rename from TMPDIR. Use ./tests/torture
unconditionally, rather than $TMPDIR-with-default-to-/tmp.
Otherwise, running "make distcheck" could leave an empty /tmp/coreutils
directory behind.
(tp): Simplify, now that it's always in the build-dir.
(taint-distcheck): Set HOME earlier, in case $(MAKE) misbehaves.
(my-instcheck, coreutils-path-check): Add diagnostics, so it's easier
to diagnose when each runs.
(coreutils-path-check): Run configure with --quiet, to reduce output.
Inspired by Ralf Wildenhues' report of /tmp/coreutils being left behind.
* tests/ls/readdir-mountpoint-inode: With some systems, stat can
succeed on a mount point and report that the inode number is 0.
Since ls displays "?" for those, that would otherwise show up as a
difference. Skip such mount points. Reported by Sergei Steshenko
in http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/19142
* tests/tail-2/inotify-race: Note the caveats of the test.
I.E. the intermittent skips and the gdb hang reported
by Alan Curry. Add extra info to the log on why the test
is skipped as it may be due to multiple reasons. Mark
the test as very expensive so that it's not normally run.
If getgroups failed with ENOSYS, mgetgroups would unnecessarily
fail, and that provoked id into freeing an uninitialized pointer.
Meanwhile, we were not using xalloc_die properly. Both issues
are better solved in gnulib, by introducing xgetgroups; this
patch uses the new interface.
Regression introduced by commit 6a31fd8d7.
* gnulib: Update, for mgetgroups improvments.
* src/id.c (print_full_info): Adjust caller to die on allocation
failure, and no longer worry about ENOSYS.
* src/group-list.c (print_group_list): Likewise.
* src/setuidgid.c (main): Likewise.
* NEWS: Mention the fix.
* THANKS: Update.
Reported by Scott Harrison.
While "rm ''" would properly fail, "rm F1 '' F2" would fail
to remove F1 and F2, due to the empty string argument.
This bug was introduced on 2009-07-12, via commit 4f73ecaf,
"rm: rewrite to use fts".
* gnulib: Update to latest, for fixed fts.c.
* NEWS (Bug fixes): Describe it.
* tests/rm/empty-name: Adjust for changed diagnostic.
(mk_file): Define, copied from misc/ls-misc.
(empty-name-2): New test, for today's fix.
* lib/xfts.c (xfts_open): Reflect the change in fts_open, now that
it no longer fails immediately when one argument is the empty string.
Assert that the bit flags were not the cause of failure.
* po/POTFILES.in: Remove xfts.c.
* THANKS: Update.
Reported by Ladislav Hagara.
* bootstrap (get_version): Don't use perl's $] special
variable, as that requires updating all bootstrap.conf files to
use perl's x.yyyzzz version format. Instead make the regular
expression more general to support version formats from older
perl-5.005_002 (5.5.2) and perl-5.11 which has other numbers
in the version line.
Counterpart to commit 8fe40b84bd, since test-link.c uses rename,
and we override gnulib with a rename() replacement that can xalloc_die.
* gl/modules/link-tests.diff: New file.
Cygwin 1.5 has a broken sleep, and the gnulib tests dragged in
rpl_sleep which then caused a link failure because it wasn't in
libcoreutils.a. We could solve it by using the gnulib sleep module.
However, sleep and usleep may interact poorly with SIGALRM, and they
have less granularity; so it is better to adopt a policy that if we
must sleep, prefer xnanosleep.
* src/sort.c (pipe_fork): Use xnanosleep, to avoid the need for
rpl_sleep on cygwin, and to reduce granularity.
(MAX_FORK_TRIES_COMPRESS, MAX_FORK_TRIES_DECOMPRESS): Increase,
to account for reduction in granularity.
* src/tail.c (tail_file): Use xnanosleep in debug code.
* cfg.mk (sc_prohibit_sleep): New rule.
* cfg.mk (gnu_ftp_host-alpha, gnu_ftp_host-beta gnu_ftp_host-stable):
(gnu_rel_host, url_dir_list): Remove definitions. The defaults,
now provided by maint.mk, are the same.
* gnulib: Update for latest, including those maint.mk additions.
These tests perform no PATH search, and used to simply delete PATH from
the environment. However, that is not portable, as seen on Cygwin,
where cygwin.dll must be resolvable via PATH when starting a sub-shell.
With commit 0cc04241, we took the alternate approach of untainting the
incoming $ENV{PATH}, but that fails when it contains an other-writable
directory. Instead, now we hard code it to '/bin:/usr/bin'.
* tests/misc/pwd-long: Hard code $ENV{PATH} to a safe value.
* tests/rm/fail-eperm: Likewise.
Reported by Gilles Espinasse, Andreas Schwab, and Bauke Jan Douma.
* src/true.c (main): There is no reason to examine argv[0],
call atexit, etc., in the usual case in which we're about to exit.
This has the side effect of making it so that these programs
no longer segfault when subjected to execve abuse.
Before this change, these commands would make "true" segfault:
printf '%s\n' '#include <unistd.h>' 'int main(int c, char**v)' \
'{ execve (v[1], 0, 0); }' > k.c && gcc k.c && ./a.out $PWD/true
Now it succeeds. Reported by Tetsuo Handa and Bart Van Assche
via Ondřej Vašík in http://bugzilla.redhat.com/537684.
Tailing forever and by-name (--follow=name, -F), tail would
sometimes fail to follow a file that had been removed via rename.
If you can't apply this patch and have tail 7.6 or newer, you can
work around the bug via the undocumented --disable-inotify option.
* src/tail.c (tail_forever_inotify): When tailing by name (-F),
do not un-watch a file upon receipt of the IN_MOVE_SELF event.
Reported by Arjan Opmeer in http://bugs.debian.org/548439.
* NEWS (Bug fixes): Mention it.
Also see http://marc.info/?l=coreutils-bug&m=125829031916515
* tests/Makefile.am (TESTS): Add tail-2/inotify-rotate.
* tests/tail-2/inotify-rotate: New test.
* tests/misc/timeout-parameters: This test could fail due to
the 1-second timeout expiring before a command of "no_such"
could be exec'd and fail. Increase to 10 seconds.
* configure.ac: Use AM_GNU_GETTEXT([external], [need-ngettext]),
rather than AM_GNU_GETTEXT([external], [need-formatstring-macros]).
Reported by Martin Jacobs in
http://thread.gmane.org/gmane.comp.parsers.bison.bugs/3181
* THANKS: Add his name.
* tests/misc/timeout-parameters: This test would fail due to
the 1-second timeout expiring before a command of "." could
be exec'd and fail. Increase to 10 seconds.
A replacement getgroups is now guaranteed to exist, but it may
fail with ENOSYS. mgetgroups is moved to gnulib, and now takes
gid_t instead of GETGROUPS_T (but setgroups still needs GETGROUPS_T).
* gnulib: Update to latest.
* gl/modules/mgetgroups: Delete, moved to gnulib.
* gl/m4/mgetgroups.m4: Likewise.
* gl/lib/mgetgroups.h: Likewise.
* gl/lib/mgetgroups.c: Likewise.
* src/group-list.c (print_group_list): Adjust callers.
* src/id.c (print_full_info): Likewise.
Capability checking was incorrectly done on just the base name
rather than on the whole path. Consequently there could be both
false positives and negatives when coloring files with capabilities.
Also capability checking was not done at all in certain cases for
non executable files. Note passing absolute rather than relative
names to cap_get_file() reduces the has_capability() overhead
from around 33% to 30%. I.E. ls --color is now around 3% faster.
* src/ls.c (struct fileinfo): Add a has_capability member.
(print_color_indicator): Refactor to pass just a fileinfo pointer
and a flag to say if we're dealing with a symlink target.
(print_name_with_quoting): Likewise.
(gobble_file): Set has_capability in the fileinfo struct. Also do
a capability check even if executable coloring is disabled.
Ditto for SETUID and SETUID coloring.
Comment on how expensive has_capability() is.
(print_long_format): Adjust to refactored print_name_with_quoting.
(quote_name): Likewise.
(print_file_name_and_frills): Likewise.
* tests/ls/capability: Test the various false positive and negatives.
* THANKS: Add reporter (Ivan Labath).
* NEWS: Mention the fix.
cat, head, ptx, shuf, tac, tail, tee, tr, and uniq used freopen
on stdout, and were potentially vulnerable. dircolors, du, and
tsort only used it on stdin, which is unaffected by freopen_safer,
but this covers all uses for consistency.
* cfg.mk (sc_require_stdio_safer): New rule.
* gl/modules/xfreopen (Depends-on): Add freopen-safer.
* gl/lib/xfreopen.c (includes): Use stdio--.h.
* src/ptx.c (includes): Likewise.
* src/shuf.c (includes): Likewise.
* src/uniq.c (includes): Likewise.
* src/dircolors.c (includes): Likewise.
* src/du.c (includes): Likewise.
* src/tsort.c (includes): Likewise.
If stdin or stdout is closed, then freopen(,stderr) can violate
the premise that STDERR_FILENO==fileno(stderr), which in turn
breaks mktemp -q.
* bootstrap.conf (gnulib_modules): Add freopen-safer.
* src/mktemp.c (includes): Use stdio--.h.
* tests/misc/close-stdout: Enhance test to catch bug.
* bootstrap.conf (gnulib_modules): Add do-release-commit-and-tag.
* build-aux/do-release-commit-and-tag: Remove file. Now it's in gnulib.
* gnulib: Update submodule to the latest, to get the just-moved script.
Reverts earlier patch - fflush() can succeed but fclose() fail for
some cases of write failures, and we want to catch those.
* src/mktemp.c (stdout_closed): New variable.
(maybe_close_stdout): New function, borrowed from dd.c.
(main): Track whether stdout has been closed.
* src/mktemp.c (main): Rather than calling close_stream (which would
make atexit-called close_stdout try to close it a second time),
check for write failure via ferror and fflush.
Now that mkstemps is supported, we might as well use it.
* src/mktemp.c (TMPDIR_OPTION): New enum value.
(longopts): Add new option.
(usage): Document it.
(count_trailing_X_s): Rename...
(count_consecutive_X_s): ...to this, and add parameter.
(mkstemp_len, mkdtemp_len): Add parameter.
(main): Implement new option.
(AUTHORS): Add myself.
* AUTHORS (mktemp): Likewise.
* tests/misc/mktemp: Test new option.
* doc/coreutils.texi (mktemp invocation): Document it.
* NEWS: Likewise.
Fixes http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=548316.
In glibc 2.11 and gnulib, gen_tempname added a parameter
suffixlen (unfortunately, it is typed as int rather than
size_t, for historical compatibility to a poor choice by BSD).
* gnulib: Import latest changes.
* gl/lib/tempname.h.diff: Accommodate new suffixlen parameter.
* gl/lib/tempname.c.diff (check_x_suffix): Allow for X in suffix
beyond x_suffix_len.
(gen_tempname_len): Add suffixlen parameter.
(__gen_tempname): Update caller.
* src/mktemp.c (mkstemp_len, mkdtemp_len): Update callers.
Diffs are more robust than wholesale replacement, because bootstrap
will inform us of any incompatible changes made in upstream gnulib.
* gl/lib/tempname.h: Change...
* gl/lib/tempname.h.diff: ...to diff.
* gl/lib/tempname.c: Change...
* gl/lib/tempname.c.diff: ...to diff.
* src/mktemp.c (main): Remove just-created file if stdout had
problems.
* bootstrap.conf (gnulib_modules): Add remove.
* tests/misc/close-stdout: Test it.
* NEWS: Document it.
* src/du.c (symlink_deref_bits): New global, decl moved from ...
(main): ...here.
(process_file): When fts detects a directory cycle that can't
be due to symlinks, report it and arrange to exit nonzero.
* NEWS (Bug fixes): Mention it.
rm -f must not print a diagnostic for a nonexistent file. However,
most linux-based kernel unlinkat functions set errno to EROFS when
the named file (regardless of whether it exists) would lie on a
read-only file system. remove.c now performs an extra fstatat call
in that case, to determine whether the file exists.
* src/remove.c (excise): Map EROFS to ENOENT, if a file is nonexistent.
Reported by Steven Drake in <http://savannah.gnu.org/bugs/?27923>.
* NEWS (Changes in behavior): Mention it.
* gl/lib/mbsalign.c (mbsalign): Mark unused parameter.
* bootstrap.conf (gnulib_modules): Remove obsolete
rename-dest-slash.
* gnulib-tests/Makefile.am (AM_CFLAGS): Reduce set of warnings for
gnulib tests.
* gl/modules/rename-tests.diff (Makefile.am): New file, to add
LIBINTL to LDADD, since we avoid canonicalize-lgpl module.
* gl/lib/regcomp.c.diff (regerror, calc_next)
(build_collating_symbol, parse_bracket_element, build_equiv_class)
(free_tree): Mark unused parameters.
* gl/lib/regex_internal.h.diff (re_string_elem_size_at): New file,
to mark unused parameters.
* gl/lib/printf-args.c.diff (PRINTF_FETCHARGS): New file, to avoid
type mismatch.
* gl/lib/vasnprintf.c (VASNPRINTF): New file, to avoid shadowing
local variable name.
* .gitignore: Ignore temporary build artifacts.
This script automates the process of updating NEWS, performs
the resulting final commit (thus with a consistent log message),
and applies a signed tag (v$VERSION) to the result.
* build-aux/do-release-commit-and-tag: New script.
* README-release: Document it.
Run this command to remove the factored-out "fail=0" lines.
perl -ni -e '/^fail=0$/ or print' $(g grep -l '^fail=0$')
* tests/test-lib.sh: Initialize fail=0 here, not in 300+ scripts.
* tests/...: nearly all bourne shell scripts
Suggested by Eric Blake.
* configure.ac (GNULIB_WARN_CFLAGS): Define.
* lib/Makefile.am (AM_CFLAGS): Use $(GNULIB_WARN_CFLAGS)
rather than $(WARN_CFLAGS) and add $(WERROR_CFLAGS).
* gl/lib/regcomp.c.diff: New file.
* gl/lib/regex_internal.c.diff: New file.
* gl/lib/regexec.c.diff: New file.
These programs can print non-fatal diagnostics to stderr prior to
exec'ing a subsidiary program. However, if we thought the situation
warranted a diagnostic, we insist that the diagnostic be printed
without error, rather than blindly exec, as it may be a security risk.
For an example, try 'nice -n -1 nice 2>/dev/full'. Failure to raise
priority (by lowering niceness) is not fatal, but failure to inform
the user about failure to change priority is dangerous.
* src/nice.c (main): Declare failure if writing advisory message
to stderr fails.
* src/nohup.c (main): Likewise.
* src/su.c (main): Likewise.
* tests/misc/nice: Test this.
* tests/misc/nohup: Likewise.
* NEWS: Document this.
* doc/Makefile.am (check-texinfo): Begin moving each individual test
into its own rules.
(sc-avoid-builtin, sc-avoid-path): New rules.
Extracted from check-texinfo.
(syntax_checks): Add them.
* src/printf.c (usage): Merge strings with echo.c to
aid translators. Move the description for \NNN beside
the other numeric escape codes. Don't mention
"character" as that suggests character conversion.
* src/echo.c (usage): Likewise.
Also mention the \xHH escape sequence.
Allows for unambiguous processing when environment values (or even
non-portable names!) contain newline.
* src/env.c (longopts): Add new option.
(usage): Document it.
(main): Implement it.
* src/printenv.c (longopts): New variable.
(usage): Document new option.
(main): Implement it.
* doc/coreutils.texi (Common options): New macro optNull.
(du invocation, env invocation, printenv invocation): Use it.
* NEWS: Mention this.
* tests/misc/env-null: New test.
* tests/Makefile.am (TESTS): Run it.
* tests/misc/sort-compress: Remove non-portable over-restriction
of PATH; besides, commit 3ea177e changed sort to no longer default
to gzip.
* tests/rm/fail-eperm: Untaint, rather than clear, PATH.
* tests/misc/pwd-long: Likewise. Also skip test if long path
cannot be created.
(normalize_to_cwd_relative): Use eq rather than ==, since cygwin
perl doesn't properly handle 64-bit ino_t numerically.
The comment in env.c about -- handling has not matched the behavior
in the code since the initial commit back in 1992.
* src/env.c: Fix bogus comment.
* tests/misc/env: Further tweaks, avoiding PATH problems inherent
in testing -i, and testing program name containing =.
* doc/coreutils.texi (env invocation): Mention that intermediate
program is needed to invoke program with name containing =.
* src/env.c (main): Use unsetenv rather than putenv to remove
items from environ, and check for failure.
* bootstrap.conf (gnulib_modules): Add unsetenv.
* tests/misc/env: Test this.
* NEWS: Document it.
* tests/rm/one-file-system (cleanup_): Unmount a/b, rather than
"$other_partition_tmpdir", to accommodate those who link /etc/mtab
to /proc/mounts. Reported by Gilles Espinasse in
http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/18508
* tests/misc/nice (tests): Accommodate a nice program for which
"nice -n -1 nice" prints nothing. It should print -1 or (usually) 0.
Otherwise, we'd get syntax errors.
* src/timeout.c (install_signal_handlers): Handle any user
specified signal, so that if it does not cause the child
to exit then we don't exit and orphan the child. Previously this
for example, would leave an orphan dd process running:
timeout -sUSR1 1s dd if=/dev/zero of=/dev/null
* NEWS: Mention the fix.
* gnulib-tests/Makefile.am (AM_CFLAGS): Add WARN_CFLAGS.
* configure.ac (enable-gcc-warnings): Also use -funit-at-a-time,
to silence gcc 4.3.4 -Wdisabled-optimization.
* .gitignore: Ignore some more files.
If new data becomes available between the initial read and when tail
registers the inotify watch descriptors, ensure that it is read
before a new event happens on the file.
* tests/Makefile.am (TESTS): Add tail-2/inotify-race.
* tests/tail-2/inotify-race: New file.
* NEWS (Bug fixes): Mention it.
* src/tail.c (check_fspec): New function.
(tail_forever_inotify): Ensure there is no new data before entering the
inotify events wait loop.
* src/md5sum.c (main): Set stdout to line buffered mode
to ensure parallel running instances don't intersperse
their output. This adds 5% to the run time in the worst case
of many zero length files, or 2% with standard file sizes.
* src/sum.c (main): Likewise.
* tests/misc/md5sum-parallel: New test for atomic output.
* tests/Makefile.am: Reference it.
* NEWS: Mention the fix
* README-hacking: Suggest to use ./configure --quiet so that
any warnings are easily noticed.
* m4/gmp.m4 (cu_GMP): Warn if libgmp is not available.
* m4/jm-macros.m4 (coreutils_MACROS): Normalize the libcap warning.
* m4/xattr.m4 (gl_FUNC_XATTR): Warn if libattr is not available.
* src/touch.c (no_dereference): New flag variable.
(longopts): Add -h/--no-dereference.
(touch): Add symlink handling.
(usage): Document new option.
(main): Accept new option.
* NEWS: Document it.
* doc/coreutils.texi (touch invocation): Likewise. Also mention
birthtime.
* tests/touch/no-dereference: New test.
* tests/Makefile.am (TESTS): Run it.
* tests/ls/abmon-align: Don't remove (1d;) the first line of output.
That was making the test consider only 11 of 12 month names.
Rewrite not to use \(.*\), as that provoked a malfunction in GNU sed
on powerpc Mac OS X (though we don't know yet whether this is due to a
sed bug, or to miscompilation). Nelson Beebe reported the test failure.
Setting the envvars, LIB_FDATASYNC, LIB_XATTR or LIB_CRYPT
could cause a configure-time and/or build-time malfunction.
Typically, a configure-time function-in-library test is performed
via code like this:
LIB_VAR=
AC_SUBST([LIB_VAR])
prefix_saved_LIBS=$LIBS
AC_SEARCH_LIBS([FUNC], [LIB_NAME],
[test "$ac_cv_search_FUNC" = "none required" ||
LIB_VAR=$ac_cv_search_FUNC])
LIBS=$prefix_saved_LIBS
However, in each of the files affected by this change, the LIB_VAR=
initialization was omitted. Thus, when set in the environment, its
value would propagate into generated Makefiles when FUNC is not found
in LIB_NAME.
* m4/jm-macros.m4 (coreutils_MACROS): Initialize AC_SUBST'd var
* m4/lib-check.m4 (cu_LIB_CHECK): Likewise.
* m4/xattr.m4 (gl_FUNC_XATTR): Likewise.
* src/copy.c (utimens_symlink): Simplify by using lutimens.
* m4/jm-macros.m4 (coreutils_MACROS): Drop utimensat; gnulib does
this for us.
* tests/cp/preserve-slink-time: Recognize lutimes support.
* tests/tail-2/pid: Run tail -f --pid=... on an actual file, not on
/dev/null, to avoid this failure on FreeBSD 6.1: tail: /dev/null:
cannot change nonblocking mode: Inappropriate ioctl for device
Before, on a system without the uname function, the build
system would detect that and not build/install a uname program.
Now that gnulib guarantees a uname function, ...
* configure.ac: Don't check for the uname function.
* src/Makefile.am (build_if_possible__progs): Move uname...
(EXTRA_PROGRAMS): ...to this list.
* src/chcon.c (main): Now that gnulib provides getfilecon wrappers,
we can revert most of the 2009-10-05 commit 3a97d664, "chcon: exit
immediately if SELinux is disabled", since chcon is still useful as
long as the file system provides handlers for the security.*
name space. gnulib's getfilecon wrappers ensure that an offending
context now evokes a return value of -1.
Prompted by comments from Stephen Smalley in
http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/18378/focus=18394
* NEWS (Bug fixes): Mention it.
* src/stat.c (human_fstype) [CIFS, HFS]: Add new file system types.
Prompted by a report from Stuart Kemp.
Normalize the form of a few hexadecimal magic numbers.
Alphabetize on S_MAGIC_ case names.
* src/Makefile.am (fs-magic-compare, fs-def, fs-magic): New rules, to
automate comparison of our list with that in the Linux statfs man page.
* NEWS (Bug fixes): Mention it.
* src/md5sum.c (split_3): Accept openssl checksum syntax, which
differs only by two spaces from that of the bsd checksum tools:
openssl: MD5(f)= d41d8cd98f00b204e9800998ecf8427e
bsd: MD5 (f) = d41d8cd98f00b204e9800998ecf8427e
This change happens to avoid an abort in chcon when SELinux is
disabled while operating on a file with an "unlabeled" context from
back in 2006. However, that same abort can still be triggered by the
same file when running chcon with SELinux enabled. This bug in chcon
will be fixed in a subsequent commit via a getfilecon wrapper. See
http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/18378/focus=18384
for how to correct your disk attributes to avoid triggering this bug.
* src/chcon.c (main): Exit immediately if SELinux is disabled.
Reported in http://bugzilla.redhat.com/527142 by Yanko Kaneti.
* src/runcon.c (main): Do not hardcode program name in error message.
* THANKS: Update.
* src/tail.c (tail_forever, tail_forever_inotify): Close a race in
tail_forever_inotify where new data written after the file check by
a now dead process, but before the pid check, is not output. We use
the POSIX guarantee that read() and write() are serialized wrt each
other even in separate processes, to assume full file consistency
after exit() and so poll for new data _after_ the writer has exited.
This also allows us to not redundantly _wait_ for new data if the
process is dead.
* tests/tail-2/pid: Remove the now partially invalid sub second sleep
check as we now don't unconditionally wait, and replace it with a check
for the redundant sleep. Also clarify some of the existing comments.
* NEWS: Mention the fix.
* bootstrap.conf (gnulib_modules): Add gnu-web-doc-update.
Remove gendocs, since gnu-web-doc-update depends on it.
* gnu-web-doc-update: Remove file, now that we get it from gnulib.
* tests/tail-2/pid: When using the timeout program to ensuring that
tail -s.1 --pid=$PID_T_MAX does not wait forever, use a timeout longer
than 1 second. A 1-second timeout could be too short on a very busy
system, and result in a timeout, and hence false-positive failure.
2009-09-30 Jim Meyering <meyering@redhat.com>
Previously for `ls -Ls` (but not `ls -Lsl`), we referenced
the st_blocks returned from the previous failed stat() call.
This undefined value was seen to be 0 for dangling symlinks at least.
* src/ls.c (print_file_name_and_frills, length_of_file_name_and_frills):
Don't use st_blocks if the previous stat() failed
* tests/ls/dangle: Add a test case
* NEWS: Mention the fix, and roll up related items into a single entry.
* bootstrap.conf (gnulib_modules): Add freopen, strsignal, fsync.
Exposed via make CFLAGS=-DGNULIB_POSIXCHECK 2>&1 \
|perl -lne '/.* use gnulib module (\S+).*/ and print $1' \
|sort |uniq -c|sort -nr
(avoided_gnulib_modules): Don't avoid the "lock" module.
Now it's required, as a dependency of the strsignal module.
* src/stat.c (do_stat): Interpret a command line argument of "-"
to mean "standard input", like many other tools do.
(do_statfs): Fail upon any attempt to use "-".
* NEWS (Changes in behavior): Mention it.
* tests/misc/stat-hyphen: New test, to exercise the above.
* tests/Makefile.am (TESTS): Add misc/stat-hyphen.
ls prints inode numbers two ways: for long (-l) listings,
and for short ones, e.g., ls -li and ls -i. The code to print
long listings properly printed "?" when the inode was unknown,
but the code for handling short listings would print 0 instead.
Factor out the formatting code into a new function so ls prints
the right string ("?") from both places:
* NEWS (Bug fixes): Mention it.
* src/ls.c (format_inode): New function.
(print_long_format): Use it here.
(print_file_name_and_frills): Use it here, too.
* tests/ls/dangle: Exercise this fix.
Reported by Yang Ren in http://bugzilla.redhat.com/525400
* src/ls.c (print_dir): Diagnosing the cycle is not enough.
Also set exit status to 2. This is what Solaris' /bin/ls does, too.
* tests/ls/infloop: Rework test: match both expected stdout and stderr.
Require an exit status of 2 in this case.
* doc/coreutils.texi (ls invocation): Mention that a loop provokes
in an exit status of 2.
* NEWS (Bug fixes): Mention it.
Reported by Yang Ren in http://bugzilla.redhat.com/525402.
* THANKS: Correct ordering of Yang Ren's names.
* bootstrap (bootstrap_epilogue): Define a default, empty function.
Remove coreutils-specific code, and instead,
invoke this new function at the end of this script.
* bootstrap.conf (bootstrap_epilogue): Define, to override the default.
This is because bitwise operators are:
- confusing and inconsistent in a boolean context
- non short circuiting
- brittle in C89 where bool can be an int (so > 1)
* bootstrap.conf (obsolete_gnulib_modules): Move rename...
(gnulib_modules): ...here. Add symlink.
* NEWS: Document the change in readlink.
* doc/coreutils.texi (readlink invocation): Likewise.
* tests/readlink/can-f: Update test to new semantics, and add test
of loop.
* src/df.c: Don't include "canonicalize.h". No longer needed,
since canonicalize_file_name is now guaranteed to be declared
in <stdlib.h>, thanks to gnulib.
* src/ls.c (print_color_indicator): This reinstates commit f3f1ccfd,
21-10-2008, "ls: make it possible to disable file capabilities checking"
which was inadvertently reverted with commit 3a169f4c, 14-09-2009,
"ls: handle disabling of colors consistently ...".
* src/ls.c (print_color_indicator): Use consistent syntax for
all file and directory subtypes, and fall back to the color
of the base type if there is no enabled color for the subtype.
This allows turning off specific colors for o+w dirs for example.
* tests/ls/color-dtype-dir: Add a case to test that turning off
coloring for o+w directories, falls back to standard dir color.
* NEWS: Mention the fix
Introduced by commit ac467814, 2005-09-05,
"Colorize set-user-ID ... files and sticky ... directories."
* src/ls.c (usage): Shorten the --color ancillary info by
two lines, while replacing --color=none with --color=never.
Mention "always" is the default parameter of the --color option,
along with the primary help for that option.
Mention the ancillary --color info in the --color primary help.
* doc/coreutils.texi: Add the missing arch, base64, link, readlink,
and unlink entries. Also remove extraneous '.' from some entries
and try to align all entries on the same column.
* THANKS: Add Benno Schulenberg.
* src/system.h: Rename emit_bug_reporting_address() to
emit_ancillary_info() and update it to not print the translation
project address in en_* locales, and _do_ print it in the 'C'
(and other) locales so that it's included in the default man page.
Also mention how to invoke the texinfo documentation for each command.
Also move the "hard-locale.h" include to the 8 files that now use it.
* man/help2man: Strip the newly added texinfo reference from the
--help output as a more verbose version is already added by help2man.
Suggestion from C de-Avillez
* src/mktemp.c (mkstemp_len, mkdtemp_len): Update callers of
gen_tempname_len.
* gl/lib/tempname.c, gl/lib/tempname.h: Rebase against recently
API-modified copy of tempname module in gnulib.
Reported by Lluís Batlle.
* bootstrap.conf (gnulib_modules): Drop rmdir-errno.
* src/rmdir.c (errno_rmdir_non_empty): Check both cases allowed by
POSIX, rather than relying on configure-time check that might
fail during cross-compilation. Reverts commit 9b6eb98d41.
* src/copy.c (copy_reg): Fix initial value of access_changed variable.
This was introduced by Pádraig Brady in commit cca83faf, 2009-09-14,
"cp,mv: preserve extended attributes even for read-only files"
A valid command like "touch -t 197101010000.60 F" would fail due
to the suffix of ".60". This bug is fixed via the latest change
to gnulib's posixtm module.
* tests/touch/60-seconds: New test.
* tests/Makefile.am (TESTS): Add it.
* NEWS (Bug fixes): Mention it.
* tests/misc/sort-continue: Change first line to standard #!/bin/sh,
not #!/bin/bash (though it doesn't matter, since each is invoked
via $(SHELL) dir/test-name.
* tests/dd/skip-seek-past-file: Require sparse support
to ensure that when we're checking if we can create an
$OFF_T_MAX length file, that we don't actually allocate
any space. This was an issue on ecryptfs and was reported
by Bert Wesarg.
* src/copy.c (copy_reg): Temporarily set u+rw on the destination file
to allow GNU/Linux to set xattrs.
* tests/misc/xattr: Test that change.
* NEWS (Bug fixes): Mention it.
Reported by Ernest N. Mamikonyan.
* NEWS (rm -r, without -f): Mention that the N in "O(N)" represents
hierarchy depth. Suggested by Ralf Wildenhues.
(rm -r, standards conformance): Make wording more accurate.
* src/id.c (print_full_info) [POSIXLY_CORRECT]: Don't print context.
Reported by Ulrich Drepper.
* NEWS (Changes in behavior): Mention it.
* doc/coreutils.texi (id invocation): Document that id also prints the
security context, when possible, and when POSIXLY_CORRECT is not set.
* tests/id/no-context: New file. Test for this.
* tests/Makefile.am (TESTS): Add it.
* bootstrap.conf (gnulib_modules): Add faccessat. Replace strdup
with strdup-posix.
* m4/jm-macros.m4 (coreutils_MACROS): Revert previous change, now
that gnulib does it for us.
* src/remove.c (write_protected_non_symlink): Use faccessat in
more situations.
where N is the depth of the deepest hierarchy rm is processing.
* src/remove.c (write_protected_non_symlink): Use faccessat to
avoid O(N)-per-entry cost of calling euidaccess.
* m4/jm-macros.m4 (coreutils_MACROS): Check for faccessat.
* NEWS (Improvements): Mention it.
* remove.c (rm_fts): Put braces around each of the two offending blocks.
* configure.ac: Don't turn off -Wjump-misses-init.
With the rewrite of remove.c, it is no longer needed.
Before this change, :>f; ln -T f no-such/ would succeed on Solaris 10.
After it, ln fails, as it should: ln: accessing `z/': Not a directory
The command, link f no-such/, had the same problem on that system.
* bootstrap.conf (gnulib_modules): Add "link".
* tests/ln/slash-decorated-nonexistent-dest: New test.
* tests/Makefile.am (TESTS): Add it.
* NEWS (Portability): Mention the improvement.
* doc/coreutils.texi (multiplierSuffixes): Mention that
the suffix can be specified without a leading number
* src/split.c (usage): Refactor SIZE help to within a function
* src/truncate.c (usage): Likewise
* src/ls.c (usage): Likewise
* src/df.c (usage): Likewise. Also add a function with BLOCKSIZE help
* src/du.c (usage): Likewise.
* src/system.h: Define 2 functions to emit common help text
This was prompted by https://bugzilla.redhat.com/show_bug.cgi?id=511188
* src/dd.c (dd_copy) [C_UNBLOCK]: Always print the final newline for
non-empty output, not just when output size is a multiple of cbs.
* doc/coreutils.texi (dd invocation) [conv=unblock]: Mention that dd
prints a newline after each output record, not just when replacing
trailing spaces.
Reported by Ulrich Drepper.
* tests/dd/unblock: New file. Test for this.
* tests/Makefile.am (TESTS): Add it.
* NEWS (Bug fixes): Mention it.
* src/tail.c (main): Adapt piped-stdin test to use the same isapipe,
test as was used in the preceding POSIXLY_CORRECT condition.
Remove the now-subsumed POSIXLY_CORRECT test.
Reported by Pádraig Brady.
* doc/coreutils.texi (tail invocation): Document this change.
* NEWS (Changes in behavior): Reclassify, clarify.
* tests/tail2/assert: This reverts commit be853120, 25-08-2009,
"tests: tail-2/assert: avoid risk of race condition"
kill -0 doesn't send a signal and so will only confirm that the
background process was forked, which we know already because
we have its pid.
* tests/misc/cat-buf: Increase the delay between writes
to decrease the chance that dd will read both at once.
Since the test is inherently racy, print a warning via
skip_test_ rather than failing outright.
Reported by Jim Meyering.
* src/tail.c (main): Tailing a pipe "forever" is not useful,
and POSIX specifies that tail ignore the -f when there is no
file argument and stdin is a FIFO or pipe. So we do that.
In addition, GNU tail excludes "-" arguments from the list of files
to tail forever, when the associated file descriptor is connected
to a FIFO or pipe. Before this change, ":|tail -f" would hang.
Reported by Ren Yang and Ulrich Drepper.
* tests/tail-2/pipe-f: Test for this.
* tests/tail-2/pipe-f2: Ensure tail doesn't exit early for a fifo.
* tests/Makefile.am (TESTS): Add these tests.
* NEWS (POSIX conformance): Mention it.
* tests/tail-2/flush-initial: stdbuf is not built on all systems.
In any case it's redundant since stdout will automatically be buffered
since we're redirecting to file. So just call tail without using stdbuf.
* tests/ls/color-clear-to-eol: Some vendor sed programs fail
to operate on lines that are not NL-terminated.
This affects at least Solaris 10's /bin/sed.
Reported by Pádraig Brady.
* tests/tail-2/infloop-1: Sleep 3 seconds, not 1, but in increments
of 0.1 second. Before, this test would fail ~1 time in 20 via
"make -j9 check" on a quad-core system.
Correct comment.
* src/tail.c (main) [HAVE_INOTIFY]: When stdin (i.e., "-", or no args,
but not /dev/stdin) is specified on the command line, don't use inotify.
Reported by Bill Brelsford in <http://bugs.debian.org/545422>.
* tests/tail-2/follow-stdin: New file. Test for this.
* tests/Makefile.am (TESTS): Add the test.
* NEWS (Bug fixes): Mention it.
This bug was introduced in coreutils-7.5 via commit ae494d4b,
2009-06-02, "tail: use inotify if it is available".
* tests/misc/ls-misc: Set umask to 022. A umask setting permitting
world-write access, e.g., umask o+w, would cause this test to fail.
Report by Mathias Brodala and analysis by Tom Fitzhenry in
<http://bugs.debian.org/544965>.
* src/tail.c (main): Flush any output from tail_file,
before calling tail_forever_inotify, which can block.
* tests/tail-2/flush-initial: New file. Test for the bug.
* tests/Makefile.am (TESTS): Add tail-2/flush-initial.
* NEWS (Bug fixes): Mention it.
This bug was introduced in coreutils-7.5 via commit ae494d4b,
2009-06-02, "tail: use inotify if it is available".
* src/tail.c (main): Add an undocumented ---disable-inotify option
to allow disabling inotify.
* tests/tail-2/pid: Run test in both normal and "disable_inotify" modes.
* tests/tail-2/tail-n0f: Likewise.
* tests/tail-2/wait: Likewise.
* tests/tail-2/append-only: Likewise.
* src/stty.c (mode_info) [TAB0, TAB1, TAB2]: Guard each
entry with #ifdef. Required for GNU/kFreeBSD.
Reported by Petr Salinger in http://bugs.debian.org/520368.
* tests/tail-2/wait: Increase the file name recheck frequency to
fix a failure on systems without inotify and a file timestamp precision
of 1 second (like GNU/kFreeBSD).
* tests/tail-2/wait: Silently skip a portion of the test
when running as root, rather than failing the whole test.
This regression was introduced with commit 84b5844d, 2009-09-03,
"tests: simplify and fix a race in 2 tail --follow tests".
* m4/jm-macros.m4 (coreutils_MACROS): The code to handle configure-time
enabling or disabling of libcap support was broken. It would treat any
libcap configure option as --disable-libcap because it doesn't check
$enableval at all. This change makes sure we do the sane thing:
--disable-libcap -> disable and don't run any tests
--enable-libcap -> run tests and fail if not found
default -> run tests and warn if not found
* src/df.c (main): If open or fstat fails when we're trying to ensure
that all arg-partitions are automounted, fall back on using stat.
Inspired by the report and patch from Olivier Fourdan in
http://bugzilla.redhat.com/520630.
* NEWS (Bug fixes): Mention it.
* tests/df/unreadable: New test for the above.
* tests/Makefile.am (TESTS): Add df/unreadable.
The bug was introduced in coreutils-7.3 via commit dbd17157,
2009-04-28, "df: use open(2), not stat, to trigger automounting".
* tests/tail-2/pid: Use the timeout command to determine process
longevity, rather than querying /proc/$pid/status.
The latter was racy in any case when inotify is used, as then
tail wakes up periodically even for unchanging files therefore
causing the check for "S (sleeping)" state to fail intermittently.
* tests/tail-2/wait: Likewise.
* src/copy.c (copy_reg): Don't return from the function after an
unsuccessful and required preservation of extended attributes.
This resulted in leaking the copy buffer and file descriptors.
* NEWS (Bug fixes): Mention the fix.
The bug was introduced in coreutils-7.1 via commit 0889381c, 2009-01-23,
"cp/mv: add xattr support".
* src/chown-core.c: Include "ignore-value.h".
(change_file_owner): Don't set "ent" only to ignore it.
* src/chcon.c (process_file): Likewise.
* src/chmod.c: Include "ignore-value.h".
(process_file): Don't set "ent" only to ignore it.
After diagnosing root-dev/ino failure, return false immediately:
Now that we don't set "ent" we must be sure not to use it uninitialized,
and there's no point in issuing --verbose-related output in this case.
* src/timeout.c (main): While keeping argc and argv in
sync may be marginally useful, it is redundant to update argc,
so just remove that to suppress the clang warning.
* src/timeout.c (main): Handle all possible cases of unexpected
failures from wait(). This was prompted by the clang tool reporting
the possible non-initialization of the status variable.
On most unix- and linux-based kernels, ls -i DIR_CONTAINING_MOUNT_POINT
would print the wrong inode number for any entry that is a mount point.
It would do that by relying on readdir's dirent.d_ino values, while
most readdir implementations return the inode number of the underlying,
inaccessible directory. Thus, it is not consistent with what you'd
get when applying stat to the same entry. This bug led to surprising
results like "ls -i" and "ls -i --color" printing different numbers (ls
must usually "stat" a file to colorize its name). This change makes it
so that on offending systems, ls must stat non-command-line-arguments
for which otherwise it would be able to use "for free" dirent.d_ino
values. Regardless of this change, ls is already required to stat every
command-line argument. Note: versions of GNU ls prior to coreutils-6.0
did not perform the invalid optimization, and hence always printed
correct inode numbers. Thus, for the sake of correctness, ls -i is
forgoing the readdir optimization, for any kernel (including linux!)
with POSIX-nonconforming readdir. Note that currently, only Cygwin has
been agile enough to conform.
* src/ls.c (RELIABLE_D_INO): Define.
(print_dir): Use it.
For plenty of discussion, see this long thread:
http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/14020
This bug was introduced by the 2006-02-26 commit, 33eb3efe:
"In ls, avoid calling stat for --inode (-i), when possible."
* tests/ls/readdir-mountpoint-inode: New test.
* tests/Makefile.am (TESTS): Add it.
* tests/ls/stat-vs-dirent: Don't suppress failure of this test,
now that ls -i is fixed. Though note that it doesn't test well,
since it compares only the always-stat'd command-line arguments.
* NEWS (Bug fixes): Mention it.
* src/copy.c (copy_reg): When cloning only skip the data copying
* tests/cp/reflink-perm: New test to check times and modes copied
* tests/Makefile.am: Reference the new test
* NEWS: Mention the fix
* doc/coreutils.texi (cp invocation): Document the new
"auto" and "always" options to --reflink.
* src/copy.c (copy_reg): Fall back to a standard copy
when reflink() is not supported and --reflink=auto specified.
* src/copy.h [struct cp_options] (reflink): Change type s/bool/enum/.
* src/cp.c (usage): Describe the --reflink={always,auto} options
and expand a little on what --reflink does.
(main): parse the new parameters to --reflink and allow all
--sparse options with --reflink=auto.
* src/install.c (cp_option_init): Init the enum instead of bool.
* src/mv.c (cp_option_init): Likewise.
* tests/cp/reflink-auto: A new test for falling back to normal copy.
* tests/Makefile.am: Reference the new test.
* NEWS: Mention the new feature.
* Makefile.am (.version, dist-hook, gen-ChangeLog): Use $(AM_V_GEN)
and $(AM_V_at), so that automake's silent-rules option (make V=1/V=0)
now controls whether the commands are printed at build time.
(THANKS-to-translators, check-ls-dircolors): Likewise.
* tests/test-lib.sh (require_selinux_enforcing_): New function.
* tests/mkdir/selinux: Use it.
Otherwise, this test would fail on Rawhide with SELinux disabled.
* tests/tail-2/assert: Avoid spurious failure due to race condition.
Rather than sleeping for 1 second and crossing fingers,
wait explicitly for backgrounded tail process to start.
Otherwise, this test would fail under heavy load.
Now that we prohibit indentation via TABs, there's no need for
Emacs indent-tabs-mode setting lines, so prohibit those, too.
* cfg.mk (sc_prohibit_emacs__indent_tabs_mode__setting): New rule.
...when run on a kernel older than what was implied by headers and
libraries tested at configure time.
* src/copy.c (utimens_symlink): Ignore failure when errno == ENOSYS.
* NEWS (Bug fixes): Mention it.
Reported by Todd Zullinger and Kamil Dudka.
Details in this thread:
http://thread.gmane.org/gmane.linux.redhat.fedora.devel/119834
* tests/cp/cp-mv-enotsup-xattr: Upon a set-up failiure, rather than
failing the test with a "framework failure" diagnostic, just skip it.
Russell Whitaker reported that this test failed on slackware.
* src/install.c (extra_mode): Be careful to return only a 0 or 1
value, since this is a "bool" function, and there are still some
compilers for which this is necessary. Without this change,
Bernhard Voelker reported that the Forte Developer 7 C 5.4 2002/03/09
compiler would produce a malfunctioning "install" binary. Details in
http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/17710/focus=17760
Now that "make check" no longer builds programs like arch, ...
* src/Makefile.am (dist-hook): Depend on $(all_programs), so that the
subsequent build in man/ doesn't fail due to lack of a program like
arch that is not going to be installed.
* NEWS: Mention the change.
* doc/coreutils.texi: Document the new --line-increment option.
* src/nl.c (struct option): Add --line-increment,
(usage): Describe it,
(main): Use it.
* src/Makefile.am (check): Don't depend on check-AUTHORS.
That check fails on systems for which a program like stdbuf is
not built. Instead, move this check to "make syntax-check".
(sc_check-AUTHORS): Rename rule from check-AUTHORS.
* cfg.mk (sc_check-AUTHORS): New rule.
Reported by Berhnard Voelker.
* src/sort.c (find_unit_order): Spell out 256-element static
initializer, rather than relying on C99 syntax.
Required for Forte Developer 7 C 5.4 2002/03/09 on Solaris 10.
Reported by Bernhard Voelker.
* src/copy.c (copy_internal): On systems without utimensat don't
use utimens on a symlink, as that would dereference the symlink.
* tests/cp/abuse: To work around possible attribute preservation
failures breaking the test, use cp -dR rather than cp -a.
* NEWS: Remove the description associated with the removed
experimental code which unconditionally tried to reflink() on copy.
Also clarify where --reflink works exactly.
* src/tail.c (tail_forever_inotify): Remove the redundant and
incorrect error check of the return from inotify_add_watch().
Also initialize the wd member of each File_spec to an invalid value.
Reported by C de-Avillez.
* src/dd.c (main): Install the signal handlers at startup
rather than just before the copy starts. In this way signals
received before the copy (like during a slow truncate for e.g.)
will be deferred and handled consistently.
* THANKS: Add Bernhard's email address.
* NEWS: Mention the fix.
Reported by Bernhard Voelker.
* tests/tail-2/pid: Speed up the test by specifying a
timeout of 100ms rather than the default 1s. Also
skip the test instead of failing in the unlikely case
were the pid required to be missing pid is present.
* NEWS: Mention it.
* doc/coreutils.texi (cp invocation): Describe it.
* src/copy.h (struct cp_options) [reflink]: New member.
* src/copy.c (usage): Describe it.
(copy_reg): If reflink is true try to clone the file.
(main): Check for --reflink.
(cp_option_init): Initialize the new member.
* src/install.c (cp_option_init): Initialize the new member.
* src/mv.c (cp_option_init): Likewise.
* tests/cp/sparse: Add a new test case.
* src/dd.c: Include "ignore-value.h"
(iwrite): When disabling O_DIRECT, try to compensate
via POSIX_FADV_DONTNEED and fsync.
Suggested by Eric Sandeen.
* gl/lib/selinux-at.c: Remove file.
* gl/lib/selinux-at.h: Likewise.
* gl/modules/selinux-at: Likewise.
* gnulib: update to latest, to get the new module.
dd oflag=direct would fail to copy a file with size that is
not a multiple of 512 (destination file system specific)
* NEWS (Bug fixes): Mention it.
* src/dd.c (iwrite): Turn off O_DIRECT for any
smaller-than-obs-sized write. Don't bother to restore it.
* tests/dd/direct: New test for the above.
* tests/Makefile.am (TESTS): Add dd/direct.
* doc/coreutils.texi (dd invocation): Mention oflag=direct
buffer size restriction.
Details in http://thread.gmane.org/gmane.comp.gnu.coreutils.bugs/17586
Reported by Eric Sandeen.
* src/copy.c (utimensat_if_possible): New function.
(copy_internal): Remove variable, "preserve_metadata".
Replace with "dest_is_symlink". That covers all cases but one:
the one in which cp --link has created hard links to non-directories.
In that case, there is no need to update attributes of the links.
Use utimensat_if_possible, to preserve timestamps of symlinks.
* NEWS (New features): Mention this.
* tests/Makefile.am (TESTS): Add cp/preserve-slink-time.
* tests/cp/preserve-slink-time: New file.
* m4/jm-macros.m4 (coreutils_MACROS): Test for utimensat.
Reported in http://bugzilla.redhat.com/230866
* m4/jm-macros.m4: Pull in SELinux libraries while checking for
matchpathcon_init_prefix (). Emit configure warning when not found
with SELinux enabled.
* NEWS (Bug fixes): Mention it.
The bug was introduced in coreutils-7.0 via commit 0647f3eb, 2008-06-02,
"accommodate older SELinux which lacks matchpathcon_init_prefix".
* gnulib: Update submodule to get latest maint.mk.
* Makefile.am (update-copyright): Remove rule; now it's in maint.mk.
* m4/check-decl.m4: Update sole remaining copyright year list.
* src/tail.c (tail_forever_inotify): The fractional
part of the delay was 1000 times too large.
* tests/tail-2/pid: Add a test to ensure the
timeout happens for this case.
* src/tail.c (tail_forever_inotify): If a PID is specified and the
watched process dies, exit with status EXIT_SUCCESS, rather than
falling through to an EXIT_FAILURE.
* build-aux/update-copyright: Handle the case in which "\n#"
appears between the final year number and the copyright holder name.
* m4/lib-check.m4: Update copyright year list.
Reported by Joel E. Denny.
* src/copy.c [HAVE_SYS_IOCTL_H]: Include <sys/ioctl.h>.
(BTRFS_IOCTL_MAGIC, BTRFS_IOC_CLONE): Define.
(clone_file): New function.
(copy_reg): Use the btrfs clone operation if possible.
* Makefile.am (update-copyright): Relax the selection rule
to match any file containing the word "Copyright".
Correct the exclusion rule so that it also excludes ChangeLog
and COPYING files that are not in the top level directory.
* build-aux/update-copyright: Handle cases in which the final
year number and copyright holder are on separate lines.
Prompted by a report from Joel E. Denny.
Also, do not invoke localtime for each line we process.
* doc/coreutils.texi: Move the "SELinux context" section
down below the "System context" nodes so that the HTML
index is generated correctly. Also some extraneous and missing
'.' characters were corrected in the index.
Reported by Benno Schulenberg.
* src/tail.c (tail_forever_inotify): When a PID is specified, use
select to block for no more than sleep_interval seconds at a time,
and check for process death upon timeout.
(main): Adapt to new tail_forever_inotify interface.
* doc/coreutils.texi (readlink invocation): Add realpath to the index,
and also mention it in the readlink description so people
searching for that functionality can easily make the connection.
* man/Makefile.am (check-x-vs-1): Using a fixed-name temporary file
is better than one ending in -$$, because many -$$-suffixed files
may accumulate over time.
* src/sort.c: Store the si_present state per key rather than globally
* tests/misc/sort: Add a check that would have previously failed.
Also add a test to demonstrate that invalid IEC/SI mixtures are not
always noticed when they're not significant to the sort.
* doc/coreutils.texi (sort invocation): Mention in the description
of -b, that the locale can also affect whether blanks are significant.
Update the list of ordering options for a --key that cause it
to not inherit any global ordering options.
Update the list of ordering options that implicitly skip whitespace.
* README-hacking: Add a little more high level info and
merge notes on specific developer prerequisites into README-prereq
* README-prereq: Remove the Fedora 8 specific information
and generalize the information to be applicable to any system.
* COPYING: Don't modify Copyright date list in imported file.
This file is one of very few exceptions in that we version-control it,
even though its primary source is another package.
Spotted by Eric Blake.
* AUTHORS: Register as the author.
* NEWS: Mention this change.
* README: Add stdbuf command to list.
* configure.ac: Only enable on ELF systems with GCC.
* cfg.mk (sc_system_h_headers): Use VC_LIST_EXCEPT rather than
VC_LIST, so we can add an exception, if needed.
* .x-sc_system_h_headers: New file. Exempt libstdbuf.c.
* Makefile.am (syntax_check_exceptions): Add .x-sc_system_h_headers.
* doc/coreutils.texi (stdbuf invocation): Add stdbuf info.
* man/.gitignore: Ignore generated manpage.
* src/.gitignore: Ignore stdbuf and libstdbuf.so binaries.
* man/Makefile.am (stdbuf.1): Add dependency.
* man/stdbuf.x: New file with example usage.
* po/POTFILES.in: Reference new command and shared library sources.
* src/Makefile.am (build_if_possible__progs): Add stdbuf and libstdbuf,
(pkglib_PROGRAMS): Reference optional shared lib,
(libstdbuf_so_LDADD): Ensure we don't link with non PIC libcoreutils.a.
(libstdbuf_so_LDFLAGS): Add -shared GCC option,
(libstdbuf_so_CFLAGS): Add -fPIC GCC option.
(check-README): Exclude libstbuf.
(check-AUTHORS): ditto.
(sc_tight_scope): Exclude functions starting with __.
* src/libstdbuf.c: The LD_PRELOAD shared library to control buffering.
* src/stdbuf.c: New file to setup env variables before execing command.
* tests/Makefile.am: Reference new test file.
* tests/misc/help-version: Set expected exit codes.
* tests/misc/invalid-opt: ditto.
* tests/misc/stdbuf: Add 9 tests.
* doc/coreutils.texi (cp invocation) [-L]: Elaborate.
[--preserve=links]: Remove comments saying that we need documentation
for just this situation. Provide more explanation and examples.
Reported by Brian M. Carlson in http://bugs.debian.org/525048.
* NEWS: Document the new feature.
* m4/jm-macros.m4: Check if inotify is present.
* src/tail.c (tail_forever_inotify): New function.
(main): Use the inotify-based function, if possible.
* tests/Makefile.am: Add new tests for tail.
* tests/test-lib.sh (require_proc_pid_status_, get_process_status_):
New functions.
* tests/tail-2/pid: New file.
* tests/tail-2/wait: New file.
* tests/tail-2/tail-n0f: Refactor code into the test-lib.sh
require_proc_pid_status_ function.
* src/ls.c: Rename hl->mh, do not colorize files with multiple
hard links by default.
* src/dircolors.c: Rename HARDLINK -> MULTIHARDLINK, hl -> mh.
* src/dircolors.hin: Do not colorize files with multiple hard links by
default.
* tests/Makefile.am: Rename the test case accordingly.
* tests/ls/multihardlink: Additionally test ls' default behavior
and factor out some duplication.
* NEWS: Mention the change in behavior.
* Makefile.am: Add a couple of missing entries to
syntax_check_exceptions for distribution.
* cfg.mk: Add a rule to ensure the syntax_check_exceptions list
stays in sync with the .x-sc_* files in the repository.
* HACKING (Amending...): Remove spurious "-e" used with commit --amend.
(log message policy): Mention the 72-column limit.
"build:" and "maint:" are common prefixes, these days.
* README-prereq: Mention new requirement on automake-1.11,
with its new location and updated dependency on autoconf-2.62.
Remove the reference to coreutils-7.0 so that this file
implicitly refers to the latest or release version of coreutils.
Fix and clarify the instructions for updating the $PATH.
Signed-off-by: Pádraig Brady <P@draigBrady.com>
* bootstrap.conf (obsolete_gnulib_modules): Remove memchr from
the list, now that it fixes a problem in some modern C libraries.
(gnulib_modules): Add it here.
This should help people notice that they are generated.
Note: each file already has a comment to that effect from help2man.
* man/Makefile.am (.x.1): Make generated files read-only.
Also, don't redirect directly to the target, $@.
* man/Makefile.am (.x.1, check-x-vs-1, check-programs-vs-x):
Mark with $(AM_V_GEN), so that automake-1.11 prints
"GEN $@" by default (stick with "GEN" in spite of the latter two
rules not officially generating anything -- they're just tests).
* src/chroot.c (set_additional_groups): Add comments.
Given an empty or all-comma group list, diagnose it and return nonzero.
When more than one group is invalid, diagnose all of them,
not just the first.
(main): Honor --groups= also when --userspec= is not specified.
Now that set_additional_groups consistently diagnoses its failures,
don't diagnose it separately here.
* tests/chroot/credentials: Do not invoke with an empty group list.
* src/chroot.c (main): Initialize both "uid" and "gid". To -1.
This also allows one to set the user-ID or primary group-ID to 0,
in case it's not that already.
* tests/chroot/credentials: Test for the above.
* dist-check.mk (warn_cflags): Remove definition, now that it's
been subsumed by --enable-gcc-warnings.
(my-distcheck): Use configure with --enable-gcc-warnings and remove
use of $(warn_cflags).
* NEWS: Document the new option
* doc/coreutils.texi (sort invocation): ditto
* src/sort.c (main): handle the new --human-numeric-sort option (-h).
(human_numcompare): A new function to compare SI and IEC suffixes
before falling back to the standard --numeric comparison.
(find_unit_order): A new helper function to find the order
of magnitude of a number string as determined by its suffix.
(check_mixed_SI_IEC): A new helper function to exit with error
if both SI and IEC suffixes are presented.
* tests/misc/sort: Add 8 tests to test the new functionality.
* THANKS: Update
* src/comm.c (usage): give more information on the -123 parameters,
with examples to show that they can be combined.
Addresses <http://savannah.gnu.org/bugs/?24974>.
* doc/coreutils.texi (comm invocation): Mention that the
column separators are suppressed along with the column.
Suggestion from Dan Jacobson.
* man/comm.x: Reference other commands that match adjacent lines.
* man/join.x: ditto.
* man/uniq.x: ditto.
* doc/coreutils.texi (uniq invocation): Simplify the
text to remove the inconsequential mentioning of order,
while implying that LC_COLLATE can alter equality comparisons.
* src/comm.c (usage): Mention LC_COLLATE is significant.
* src/join.c (usage): Ditto.
* src/uniq.c (usage): Ditto. Also improve the summary.
Suggestion from Andries Brouwer
Use gnulib's new priv-set module and updated write-any-file.
With them, the remove-called can_write_any_file function no
longer tries to drop the unlink-directory privilege, so now
each caller of remove must do that separately, calling
priv_set_remove_linkdir.
* bootstrap.conf (gnulib_modules): Add priv-set.
* src/rm.c: Include "priv-set.h".
(main): Call priv_set_remove_linkdir.
* src/mv.c (main): Likewise.
* gnulib: Update submodule to latest.
* src/Makefile.am (check-AUTHORS): Revert back to using
en_US.UTF-8, to ease parsing (English-only) text around the
list of names, even when .po files are not installed.
Reported by Andreas Schwab.
Along the way, use $(AM_V_GEN), not "@".
(sc_tight_scope): Use $(AM_V_GEN) here, too.
* gnulib: Update submodule to latest, for the proper_name_utf8 fix
that makes --version output print the UTF-8 rendering of author names
in more cases.
* tests/misc/date-next-dow: New file.
* tests/Makefile.am (TESTS): Add misc/date-next-dow.
* gnulib: Update submodule to latest, for getdate.y that
handles "next Monday" properly when run on a Monday.
* README-release: Specify -j1 on the command to run all tests.
Otherwise, running some in parallel would cause failures, e.g.,
for rm/ext3-perf and tail-2/assert-2.
* doc/Makefile.am (sc-lower-case-var): Don't redirect stderr of
$(PERL) (which is sometimes build-aux/missing) to /dev/null,
so that a failing "make check" explains that Perl is not installed.
Reported by James Youngman.
* src/df.c (main): When iterating over command-line arguments,
attempting to ensure each backing file system is mounted, use
open, not stat. stat is no longer sufficient to trigger
automounting, in some cases. Based on a suggestion from Ian Kent.
More details in http://bugzilla.redhat.com/497830
* src/copy.c (copy_attr_by_fs): Always print diagnostics when preserving
xattrs is required.
(copy_attr_by_name): Likewise.
(copy_reg): Always print diagnostics when preserving SELinux
context is required.
(copy_internal): Likewise. Also, do not ignore ENOTSUP and ENODATA
errors when preserving SELinux context is required.
* NEWS (Bug fixes): Mention it.
*src/copy.c: Do not warn about xattr-preservation failure when xattrs
are not supported and preservation of xattrs is not explicitly required.
Reported by Eric Sandeen in http://bugzilla.redhat.com/496142
* src/Makefile.am (dircolors.h, wheel-size.h, wheel.h, fs.h):
(version.c, version.h): Use $(AM_V_GEN) only as a prefix of an existing
command. Otherwise, it provokes warnings from some "make" programs.
Suggestion from Ralf Wildenhues.
* dist-check.mk: New file: coreutils-specific rules extracted
from maint.mk, now that maint.mk has migrated to gnulib.
* Makefile.am (EXTRA_DIST): Add dist-check.mk.
* coreutils/cfg.mk: Include $(srcdir)/dist-check.mk.
* gnulib: Update submodule to latest.
* src/Makefile.am (dircolors.h, wheel-size.h, wheel.h, fs.h):
(version.c, version.h): Mark with $(AM_V_GEN) and $(AM_V_at),
so that the latest automake prints "GEN $@" by default.
* configure.ac (AM_INIT_AUTOMAKE): Remove silent-rules. Instead,...
(AM_SILENT_RULES): Use this, with it's undocumented [yes] argument.
Those who want verbose build output may configure with
--disable-silent-rules or use "make V=1".
* src/sort.c (avoid_trashing_input): Fix an off-by-one error and
guard the use of memmove.
* NEWS (Bug fixes): Mention it.
* tests/misc/sort: Add tests to exercise the offending code.
* THANKS: Update.
Reported by Otavio Salvador in http://bugs.debian.org/525048.
run as part of "make check".
* src/Makefile.am (check-AUTHORS): Don't depend on en_US.UTF-8.
Instead, use the French UTF8 locale, if configure found one.
If not found, just skip the test.
Move the check-AUTHORS rule from "syntax-check" to "check".
* maint.mk (local-checks-available): Remove check-AUTHORS.
(check-AUTHORS): Remove rule.
* src/Makefile.am (CLEANFILES): Add $(no_install__progs),
since they too are built sometimes.
(check-AUTHORS): Split a longer-than-80 sed command.
(../AUTHORS): Remove obsolete rule.
Remove the optimization that avoided up to 50% of cp's read syscalls.
Do not assume that a short read on a regular file indicates EOF.
When reading from a file in /proc on linux [at least 2.6.9 - 2.6.29]
into a 4k-byte buffer or larger, a short read does not
always indicate EOF. For example, "cp /proc/slabinfo /tmp"
copies only 4068 of the total 7493 bytes. This optimization
(25719a3315, Improve performance a bit
by optimizing away; 2005-11-24) appears to have been worth less than
a 2% speed-up (and usually much less), so the impact of removing it
is negligible.
* src/copy.c (copy_reg): Don't exit the loop early.
* tests/cp/proc-short-read: New test, lightly based on a suggestion
from Mike Frysinger, to exercise this fix.
* tests/Makefile.am (TESTS): Add cp/proc-short-read.
* NEWS (Improve robustness): Mention this change.
With newer perl (e.g., Fedora 10's 4:5.10.0-68.fc10), tests/CuTmpdir.pm
stopped removing its temporary directories, with diagnostics like this:
cannot remove path when cwd is /c/coreutils/tests/misc/seq.tmp-e2up \
for /c/coreutils/tests/misc/seq.tmp-e2up: at \
/usr/lib/perl5/5.10.0/File/Temp.pm line 902
Chdir out of the target directory before that code runs:
* tests/CuTmpdir.pm (END): chdir '..'.
(chmod_tree): Remove explicit "chdir $dir".
* README: (Reporting bugs): List the bug-reporting address here, too,
not just in the following more test-oriented paragraph.
Reported by Tim Mooney.
All changes are no longer listed in version-controlled ChangeLog
files, so note that contributions are attributed in the commit logs.
Mention bootstrap.conf, now that it's the authoritative source of
minimal prerequisite program/version# pairs.
* tests/install/install-C: Use 2755 (set-gid), not 1755 (sticky),
to test install -C with non-permission mode bits set. At least on
FreeBSD with a UFS file system, a non-root user may not set the
sticky bit on a non-directory.
Steven Parkes reported that `id -G $USER` went into an infinite loop
on Darwin systems for users in more than 10 groups:
http://bugs.gentoo.org/show_bug.cgi?id=264007
* gl/lib/mgetgroups.c (mgetgroups): Work around buggy getgrouplist
implementations that don't update the required size correctly,
by doubling the result buffer and retrying. Also return the
parameter updated by getgrouplist rather than its return value,
as the documentation doesn't actually state the number of groups
stored is returned by getgrouplist.
* tests/misc/id-groups: Add test to exercise this logic
* tests/Makefile.am: Reference new test
* NEWS: Mention the fix
* THANKS: Update
Suggestion from Steven Schveighoffer at:
http://savannah.gnu.org/patch/?6797
to greatly speed up the random passes done by shred.
* gl/lib/randread.c: Default to using the internal
pseudorandom generator, rather than reading /dev/urandom
* src/shred.c (usage): remove mention of /dev/urandom
* src/shuf.c (usage); ditto
* src/sort.c (usage): ditto
* doc/coreutils.text: Document the new behaviour
for aquiring random data.
* maint.mk (sc_changelog): Use $(VC_LIST_EXCEPT), not find.
Don't hang when there are no ChangeLog files.
(sc_require_config_h): For the file-name-filter, use grep -l
to be slightly more efficient.
(sc_require_config_h_first, _header_without_use, sc_program_name):
Likewise.
* maint.mk (_prohibit_regexp): Add ';' at end of ':'-introduced
comment so that the comment doesn't gobble up the following test
for a missing regexp definition.
(_header_without_use): Escape all '.'s in a header file name,
not just the first one.
* NEWS: Reword an entry from 7.2 and change "linux" to "GNU/Linux".
The latter was requested by Richard Stallman.
* cfg.mk (old_NEWS_hash): Regenerate.
Reported by Samuel Thibault and Stéphane Raimbault, as the glibc fr_FR
locale has recently changed to use the official but variable width
abbreviated month names. Other glibc locales also have variable widths.
http://sourceware.org/ml/libc-locales/2008-q1/msg00035.htmlhttp://sourceware.org/bugzilla/show_bug.cgi?id=9859
* NEWS: Mention the fix
* gl/lib/mbsalign.c: A new module to align and truncate a
string in a specified number of screen cells, while handling
multi-byte characters appropriately.
* gl/lib/mbsalign.h: Ditto
* gl/modules/mbsalign: Ditto
* bootstrap.conf: Reference the new module
* src/ls.c (abmon_init): New function, precompute the abbreviated
months aligned left in a minimum width column <= 5 screen cells.
(align_nstrftime): New function, replace the first %b in the
format specification to strftime with the precomputed month string.
Note using the cached month strings speeds up `ls -lU` by around 17%
on glibc-2.7-2 on linux at least. Also if we implement this function
using heap storage rather than automatic storage, and use snprintf
instead of strcpy, ls will slow down by 2% and 1% respectively
(i.e. a net gain of 14% rather than 17%).
* tests/ls/abmon-align: A new test to test ls alignment for
various formats and locales
* tests/Makefile.am: Reference the new test
* README-prereq: Reverse order of -b/--track options.
Pádraig Brady mentioned that is required with at least git-1.5.3.6.
Invoke automake's own bootstrap script.
* bootstrap: Revert commit 9f39fa855, 2009-03-28,
"build: use automake's --silent-rules option when possible".
* configure.ac (AM_INIT_AUTOMAKE): Instead, set it here.
Require the just-released version of automake: 1.10b.
* build-aux/check.mk: Remove, not needed any more.
* tests/Makefile.am (AUTOMAKE_OPTIONS): Add parallel-tests and
color-tests.
* tests/check.mk: Do not include build-aux/check.mk any more.
(SUFFIXES, TEST_LOGS): Remove.
(TESTS_ENVIRONMENT): Use $$f rather than $$tst in $PERL invocation.
* gnulib-tests/Makefile.am: Do not include build-aux/check.mk.
(AUTOMAKE_OPTIONS): New macro, add parallel-test and color-tests.
(TEST_LOGS): Remove.
* tests/sample-test: Don't recommend using lang-default here.
It is now run for each test automatically, via TESTS_ENVIRONMENT.
* tests/dd/reblock: Don't source lang-default here.
* tests/misc/truncate-fail-diag: Likewise.
* tests/other-fs-tmpdir (CANDIDATE_TMP_DIRS): Add /dev/shm to the list.
Don't use a line of asterisks as the first line of the "why skipped"
diagnostic, now that that line is printed on 'make check's stderr.
* gl/tests/test-argv-iter.c: Move declaration "up", so as not
to ruffle feathers of crufty compilers. We accommodate such
compilers here solely because this module may migrate to gnulib.
Tiny patch by Matthew Woehlke.
* tests/test-lib.sh (skip_if_setgid_): New function.
* tests/setgid-check: Remove file.
* tests/Makefile.am (EXTRA_DIST): Remove setgid-check.
* tests/chmod/c-option: Use the new function rather than sourcing
the separate file.
* tests/cp/cp-parents: Likewise.
* tests/install/install-C: Likewise.
* tests/mkdir/parents: Likewise.
* tests/mkdir/perm: Likewise.
* tests/rm/ext3-perf: Relegate this test to the very_expensive
category, since it is failing too often for me, when run in
parallel (-j4) with other disk-hammering tests. This test fails
when it takes too long, so the fact that it fails under pressure
is an inconvenience more than a real problem.
* src/df.c (known_value): New function, which also works on AIX
/proc file systems.
(df_readable, show_dev): Use it instead of hardcoding comparison to
UINTMAX_MAX. Suggested by Jim Meyering and Matthew Woehlke.
* src/df.c (show_dev): Don't add UINTMAX_MAX to grand totals, as that
value indicates that the true value is unknown; adding it effectively
subtracts 1 from the total, whereas we want to leave the total alone.
* src/pwd.c (longopts): New variable.
(logical_getcwd): New function.
(main): Use it.
(usage): Document new options.
* doc/coreutils.texi (pwd invocation): Likewise.
* NEWS: Likewise.
* TODO (pwd): Mark it done.
* tests/misc/pwd-option: New file.
* tests/Makefile.am (TESTS): Add test.
* THANKS: Update.
Reported by Paul D. Smith, in savannah bug 24949.
* src/ls.c (print_color_indicator): Don't color hard links if disabled
(when there is an empty hl= in the LS_COLORS environment variable).
* tests/ls/hardlink: Add test case for disabled hard link highlighting.
* NEWS: Mention the change.
* maint.mk (sc_error_exit_success): Search only files
that are under version control.
(sc_makefile_path_separator_check): Likewise.
Check *.mk files, too, not just Makefile.am files.
* src/dd.c (O_FULLBLOCK): Compute its value without using a 180KB
macro. This avoids triggering a compilation failure with HP-UX's cc.
Reported by Matthew Woehlke.
* tests/Makefile.am (TESTS): Add sort-merge-fdlimit.
* tests/misc/sort-merge-fdlimit: New file.
* doc/coreutils.texi (sort invocation): Document that we now silently
lower nmerge if necessary.
Patch by Paul Eggert, Nima Nikzad, Max Chang, Alexander Nguyen,
Sahil Amoli, and Nick Graham.
This is an alternative to my 9 March patch labeled "Silently lower
nmerge; don't (sometimes incorrectly) range-check"
<http://lists.gnu.org/archive/html/bug-coreutils/2009-03/msg00070.html>.
It differs by not using 'dup' to probe for extra file descriptors;
instead, it simply calls 'open' (and 'pipe') to open files and pipes,
until one of these calls fails due to file descriptor exhaustion; it
then backs off by 1, does a merge with the files that it has opened,
and then retries with the (now-smaller) number of files.
This patch requires quite a few more changes to the source code than
the earlier patch, but it is in some sense "better" because it doesn't
need to call "dup" ahead of time in order to decide whether "open" or
"pipe" will fail. Also, it's more robust in the case where "open" or
"pipe" fails with errno==EMFILE because some system-wide limit is
exhausted.
* src/sort.c (create_temp_file): New arg SURVIVE_FD_EXHAUSTION.
(stream_open): New function, containing guts of xfopen.
(xfopen): Use it.
(pipe_fork): Set errno on failure.
(maybe_create_temp): New function, containing guts of create_temp.
(create_temp): Use it.
(open_temp): Distinguish failures due to file descriptor exhaustion
from other failures, and on fd exhaustion return a notice to caller
rather than dying. Don't test execlp's return value; when it returns,
it *always* returns -1.
(open_input_files): New function.
(mergefps): New arg FPS. It's now the caller's responsibility to open
the input and output files. All callers changed.
(mergefiles): New function.
(avoid_trashing_input, merge): Handle the case where a single merge
can't merge as much as we wanted due to file descriptor exhaustion, by
merging as much as we can and then retrying.
* tests/Makefile.am (TESTS): Add misc/sort-continue.
* tests/misc/sort-continue: New file.
* THANKS: Add Glen Lenker and Matt Pham who coauthored this patch.
* maint.mk (sc_makefile_TAB_only_indentation): New rule.
Replace each TAB+8-space sequence with two TABs.
* man/Makefile.am: Likewise.
* build-aux/check.mk: Likewise.
I used this command (run it more than once, if needed):
t=$'\t'; git grep -l -E "$t {8}"|grep -E 'Makefile|\.mk$' \
| xargs perl -pi -e 's/\t {8}/\t\t/'
* README-prereq: coreutils currently requires an unreleased
version of Automake, so mention the minimum version of Autoconf
required to build that. Suggestion from James Youngman.
* copy.c (copy_attr_by_fd): Reduce xattr diagnostics for 'cp -a'.
(copy_attr_by_name): Likewise.
* cp.c (main): Preserve xattrs with -a option, when possible.
* doc/coreutils.texi: Document that xattrs are preserved with
cp -a, with no added diagnostics.
* NEWS: Mention the change.
* tests/misc/xattr: Add tests for 'cp --preserve=all' and 'cp -a'.
Introduced by commit 790892db, 2006-06-08 "Ensure that cat works ...".
* NEWS: Mention the bugfix.
* src/cat.c (cat): Fix the typo which stopped the writing
of processed data before a blocking read() is done.
* tests/misc/cat-buf: Add to ensure processed data is not buffered.
* tests/Makefile.am: Reference the new test.
This is following on from this change:
[02c3dc9d 2008-03-06 cat: use larger buffer sizes ...]
which increased the IO block size used by cat by 8 times,
but also capped it at 32KiB.
* NEWS: Mention the change in behavior.
* src/system.h: Add a new io_blksize() function that
returns the max of ST_BLKSIZE or 32KiB, as this was
seen as a good value for a minimum block size to use
to get good performance while minimizing system call overhead.
* src/cat.c: Use it.
* src/copy.c: ditto
* src/split.c: ditto
* src/pr.c (print_header): Honor chars_per_margin also for the header.
Inspired by a patch from C de-Avillez.
This change also adjusts two other %*s directives to be %*.*s instead,
(for the padding on either side of "file_text") to correct a bug that
could make pr print one space of padding instead of none.
* NEWS (Bug fixes): Mention it.
* tests/pr/o3Jml24f-lm-lo: Adjust for new expected output.
* tests/pr/o3a3Sl24f-tn: Likewise.
* tests/pr/o3a3Snl24f-tn: Likewise.
* tests/pr/o3a3l24f-tn: Likewise.
* tests/pr/o3b3Sl24f-tn: Likewise.
* tests/pr/o3b3Snl24f-tn: Likewise.
* tests/pr/o3b3l24f-tn: Likewise.
* tests/pr/o3mSl24f-bl-tn: Likewise.
* tests/pr/o3mSnl24fbltn: Likewise.
* tests/pr/o3ml24f-bl-tn: Likewise.
* src/cat.c (max): Remove definition. Use MAX from system.h instead.
(compute_buffer_size): New function to compute the input and output
buffer sizes, which are now set at 8 times st_blksize with a maximum
of 32KiB. Previously the typical block sizes used were 1KiB for pipes
and 4KiB for files, and now will be 8KiB and 32KiB respectively.
(main): Use it.
This change can double throughput on modern systems. For timings,
see http://article.gmane.org/gmane.comp.gnu.core-utils.bugs/16040
Suggestion from Tzvi Rotshtein.
* README-release: Fix typo. it's "maintainer-clean".
Refer to HACKING for full list of dependencies.
Use cvs add's -ko option to inhibit $Id$-style keyword expansion.
* maint.mk (web-manual): Factor out definition of $(manual_title).
(vc-dist): Set XZ_OPT=-9ev for the final "make dist".
* cfg.mk (manual_title): Define it here.
Suggestions from Eric Blake.
* tests/rm/ext3-perf: If set-up takes too long, use the set-up time
as the removal time limit. Otherwise, a parallel "make check" would
usually fail when run with RUN_VERY_EXPENSIVE_TESTS=yes.
* tests/mkdir/perm: Now that this test is no longer always skipped,
make it work independently of ls: use stat, not ls, to get perm string.
* test-lib.sh (rwx_to_mode_): Use stat, not ls to get perm string
This test was run only when RUN_VERY_EXPENSIVE_TESTS=yes was set.
* bootstrap.conf (gnulib_modules): Add manywarnings.
* configure.ac: Use gl_MANYWARN_ALL_GCC, and exclude options
I don't want or that provoke too many warnings.
(WARN_CFLAGS, WERROR_CFLAGS): Define.
(lint, GNULIB_PORTCHECK): Define.
(_FORTIFY_SOURCE): Define to 2.
Running "mkdir dir; cp -rl dir dir" would create dir/dir/dir/...
rather than diagnosing the "copy-into-self" failure.
The easy fix would have been to revert this part of the change
[3ece0355 2008-11-09 cp: use far less memory in some cases]
that introduced the bug:
- remember_copied (dst_name, dst_sb.st_ino, dst_sb.st_dev);
+ if (!x->hard_link)
+ remember_copied (dst_name, dst_sb.st_ino, dst_sb.st_dev);
However, that would have induced the failure of the new cp/link-heap
test, due to the added memory pressure of recording 10k dev/ino pairs.
And besides, I liked that improvement and wanted to keep it.
Now that it's obvious recording the just-created-directory dev/ino
needn't depend on the setting of hard_link, I realized it is necessary
to record the pair only for the first directory created for each
source command-line argument.
I made that change, then noticed the new test, cp -rl a d d, would pass
when run once, yet output the into-self diagnostic twice. Also note
the side effect: it creates d/a and d/d. However, running that same
command a second time, now with the modified directory, would fail.
That turned out to be due to the fact that although the first into-self
failure was detected in copy_dir, that function would continue copying
other entries regardless -- and that would make it fail (eventually)
with the unwanted recursion.
* src/copy.c (copy_internal): This function needed an indicator of
whether, for a give command line argument, it had already created its
first directory. If so, no more need to record dev/ino pairs. If this
is the first, then do record its pair. Hence, the new parameter.
(copy_dir, copy): Update callers.
(copy_dir): Upon any into-self failure, break out of the loop.
* tests/cp/into-self: Test for the above.
Reported by Mikael Magnusson.
* src/sort.c: When no specific number of chars to skip
is specified for the end field, always skip the whole field.
Also never include leading spaces from next field.
* tests/misc/sort: Add 2 new tests for these cases.
* NEWS: Mention this bug fix.
* THANKS: Add bug reporter.
Reported by Davide Canova.
* gl/lib/argv-iter.h (_ATTRIBUTE_NONNULL_): Don't assume C99 variadic
macros are portable yet; needed for at least IRIX/MIPSpro.
* THANKS: Update.
Reported by Stuart Shelton.
* tests/dd/skip-seek-past-file: Just check for
the non system specific part of the error strings.
This was causing an erroneous failure on NetBSD 1.6 at least.
dnl Copyright (C) 2005 Free Software Foundation, Inc.
dnl Copyright (C) 2005, 2009-2010 Free Software Foundation, Inc.
dnl This file is free software; the Free Software Foundation
dnl gives unlimited permission to copy and/or distribute it,
dnl with or without modifications, as long as this notice is preserved.
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
