Commit Graph
28150 Commits
Author SHA1 Message Date
Pádraig Brady 8ee3ca4bb9 tests: fix races in and standardize the tail tests
* tests/tail-2/F-vs-missing.sh: Use standard "fastpoll" options
(-s.1 --max-unchanged-stats=1) to speedup the non-inotify case.
Add the non-inotify case to the test. `wait` on the background
tail process to terminate which should avoid the need for the
non standard `retry_delay_ cleanup ...` on NFS.
* tests/tail-2/F-vs-rename.sh: Remove 'out' at the start of the loop,
to avoid a race in checking its contents.  Also ensure 'a' & 'b'
files are present before the tail process starts.  Use the standard
"fastpoll" options as above.
* tests/tail-2/f-vs-rename.sh: Likewise.
* tests/tail-2/append-only.sh: Use more standard variable names.
* tests/tail-2/flush-initial.sh: Use "fastpoll" options for
non-inotify platforms.  Also `wait` on the background tail to avoid
stray processes and file cleanup issues on NFS.
* tests/tail-2/inotify-hash-abuse.sh: Always run non-inotify case.
Use "fastpoll" options.  Use a more standard retry_delay_ instead
of a hardcoded sleep loop.  Add a `wait` on the background tail.
* tests/tail-2/inotify-hash-abuse2.sh: Likewise.
* tests/tail-2/inotify-rotate-resources.sh: Wait just on the
specific tail $pid needed.
* tests/tail-2/inotify-rotate.sh: Use "fastpoll" options.
* tests/tail-2/pid.sh: Use standard variable names.
Add a `wait` on the background tails.
* tests/tail-2/pipe-f2.sh: Likewise.
* tests/tail-2/tail-n0f.sh: Likewise.
* tests/tail-2/retry.sh: Use "fastpoll" options.
* tests/tail-2/symlink.sh: Likewise.
* tests/tail-2/wait.sh: Likewise.  Speedup by using sub second
parameters to timeout(1).  Improve the part ensuring that
-F never follows a renamed file.
* tests/tail-2/infloop-1.sh: Remove invalid test.  tail(1) was not
being passed the --pid=$yes_pid option, retry_delay_ wasn't used
to avoid races, and yes could write huge files before being killed.
* tests/local.mk: Remove the invalid test reference.
* tests/tail-2/assert-2.sh: Rewrite using retry_delay_().  Since
no longer hardcoding large delays, remove the VERY_EXPENSIVE tag.
* tests/tail-2/assert.sh: Likewise.
2015-05-11 14:41:17 +01:00
Paul Eggert a976d7c9ab build: rewrite is_ENOTSUP without an #if directive
* src/system.h (is_ENOTSUP): Avoid in-function #if directive.
2015-05-10 08:34:16 -07:00
Jim Meyering aa7f31fc4a build: avoid a warning form gcc's new -Wlogical-op
Without this change, very recent gcc (e.g., version 6.0.0 20150509)
would print the following when configured with --enable-gcc-warnings:

  src/copy.c:165:30: error: logical 'or' of equal expressions \
    [-Werror=logical-op]
    && (errno == EOPNOTSUPP || errno == ENOTSUP || errno == ENOSYS))
                           ^
* src/system.h (is_ENOTSUP): New function.
* src/copy.c (punch_hole): Use it.
* src/ls.c (errno_unsupported, gobble_file): Use it.
2015-05-09 22:49:07 -07:00
Pádraig Brady e981643ae3 doc: standardize messages about the '-' stdin FILE
* src/system.h (emit_stdin_note): A new function, refactoring
the usage note about the '-' FILE implying stdin.
* src/base64.c (usage): Use the new function to emit the
note in a standard location and with standard separation.
* src/cat.c (usage): Likewise.
* src/csplit.c (usage): Likewise.
* src/cut.c (usage): Likewise.
* src/expand.c (usage): Likewise.
* src/fmt.c (usage): Likewise.
* src/head.c (usage): Likewise.
* src/md5sum.c (usage): Likewise.
* src/nl.c (usage): Likewise.
* src/od.c (usage): Likewise.
* src/paste.c (usage): Likewise.
* src/pr.c (usage): Likewise.
* src/ptx.c (usage): Likewise.
* src/shred.c (usage): Likewise.
* src/shuf.c (usage): Likewise.
* src/sort.c (usage): Likewise.
* src/sum.c (usage): Likewise.
* src/tac.c (usage): Likewise.
* src/tail.c (usage): Likewise.
* src/tsort.c (usage): Likewise.
* src/unexpand.c (usage): Likewise.
* src/wc.c (usage): Likewise.
* src/join.c (usage): Adjust the separation used for
the message referring to FILE1 or FILE2 as stdin.
* src/comm.c (usage): Add a message using the same
wording (translation) as used in join.
* src/split.c (usage): Reword to using FILE rather than
INPUT, allowing use of emit_stdin_note().  Also remove
the mention of "fixed-size" pieces as this isn't now
always the case.
Fixes http://pad.lv/1450179
2015-04-30 18:25:01 +01:00
Pádraig Brady 88a0354824 tests: don't skip df tests with /proc/self/mountinfo
* tests/df/no-mtab-status.sh: getmntent is no longer called
when /proc/self/mountinfo is present, thus causing the test
to be skipped.  Therefore wrap fopen() to ignore mountinfo,
and use the test genmntent table instead.
* tests/df/skip-duplicates.sh: Likewise.
2015-04-30 18:24:47 +01:00
Yunlian Jiang c77a96ccba build: fix SINGLE_BINARY build when printf is a macro
* src/coreutils.c (usage): include coreutils.h outside
the printf call, because if it's a macro you will get the error:
embedding a #include directive within macro arguments is not supported
2015-04-28 23:48:30 +01:00
Pádraig Brady bdf6837655 maint: avoid a new -Werror=return-type warning in yes.c
* src/yes.c (main): Simplify the logic so that the
compiler can see this function always returns a value.
This was seen with GCC 5.0 in SINGLE_BINARY mode.
2015-04-28 23:07:13 +01:00
Pádraig Brady 09152f2d75 tests: run ls capability coloring test irrespective of $LS_COLORS
* tests/ls/no-cap.sh: Ensure the test isn't skipped even if
capability coloring is disabled in the current $LS_COLORS.
Also just enable/disable capability coloring to avoid the
dircolors(1) overhead.
2015-04-28 21:30:25 +01:00
Pádraig Brady 824bcc6434 maint: use gnulib styling with the online manual
The equivalent styling added in v8.23-155-g3b98ee7,
is now applied to gnulib using projects by default.
2015-04-27 11:09:48 +01:00
Pádraig Brady b817f62769 build: rely on gnulib to determine printf routines are safe
gnulib now only checks that the printf routines never crash,
which is all coreutils currrently requires, and so we revert
commit v8.23-81-gf57bfbb to let gnulib decide whether to replace
the system printf routines.
2015-04-27 11:00:07 +01:00
Pádraig Brady 0e7ac9a460 maint: fix printf format for signed integers
With GCC 5 and the newly added warnings from gnulib, ensure the
correct signed integer is passed for the printf format,
to avoid -Werror=format= failures.
2015-04-27 10:55:26 +01:00
Pádraig Brady 17fba994bd build: update gnulib submodule to latest
Fix file-has-acl build failure on RHEL/Centos 6.
Fix GCC 5 warnings with printf and in fts.c.
2015-04-27 10:27:16 +01:00
Pádraig Brady 0dcde34931 build: reduce gettext dependency to 0.19.2
* bootstrap.conf: 0.19.2 is available on openSUSE-13.2,
Debian-8.0, and Ubuntu-14.10.  Given there were issues
with earlier 0.19 gettext releases, set this as the new minimum.
* configure.ac (AM_GNU_GETTEXT_VERSION): Likewise.
Reported by Bernhard Voelker
2015-04-27 10:18:16 +01:00
Paul Eggert 997265d56e ls: on GNU/Linux, remove dependency on libacl
* src/local.mk (src_ls_LDADD): Change from LIB_ACL to LIB_HAS_ACL.
2015-04-26 22:10:46 -07:00
Paul Eggert f0b437f3cc build: update gnulib submodule to latest
* bootstrap.conf (gnulib_modules): Add file-has-acl.
(buildreq): Bump autopoint and gettext to 0.19.4.
* configure.ac (AM_GNU_GETTEXT_VERSION):
Bump to 0.19.4.
* gl/lib/tempname.c.diff, gl/lib/tempname.h.diff:
Merge recent gnulib changes.
2015-04-26 22:10:38 -07:00
Michael Witten 02f678a350 doc: fix grammar issue in truncate info
* doc/coreutils.texi (truncate invocation): The word 'their' is
incorrect; 'each file' is the antecedent, and is singular,
so 'its' is the correct pronoun.
2015-04-23 20:13:50 +01:00
Pádraig Brady 1018d6a7e8 build: fix potential factor build failure on arm and mips
* src/longlong.h: Sync with the latest longlong.h from libgmp to:
- Use __builtin_c[lt]zl on arm64.
- Fix sparc64 vis3 build failure due to missing __clz_tab.
- Avoid a clang build issue on mips.
- Support thumb2 arm 32 bit system.
* src/cfg.mk (sc_ensure_comma_after_id_est): Exclude longlong.h
to ease merges.
2015-04-22 11:11:26 +01:00
Pádraig Brady 8878c4ef1b maint: avoid -Werror=strict-overflow warnings with GCC 5
All warnings were of the form: "assuming signed overflow does not occur
when simplifying conditional to constant [-Werror=strict-overflow]"

* src/dd.c (cache_round): Use an appropriately sized unsigned type,
to avoid possibility of undefined signed overflow.
* src/mknod.c (main): Likewise.
* src/pr.c (pad_down): Likewise.
* src/wc.c (main): Likewise.
* src/tail.c (main): Assert that argc >= 0 thus allowing the
compiler to assume without implication that argc - optind
is positive.
2015-04-22 11:11:19 +01:00
Pádraig Brady 01f4065b12 dircolors: add 'MISSING' to the default database
* src/dircolors.hin: Add the MISSING entry, to indicate
this as a possibility in new templates output from dircolors,
and also to ease comparison with existing databases that
generally do define a MISSING entry.
2015-04-17 23:51:32 +01:00
Andreas Gruenbacher 88e491044b doc: clarify how cp behaves with default ACLs
* doc/coreutils.texi (cp invocation): Mention that when copying files
without preserving permissions, the umask or a default ACL affect
the mode of new files.
* THANKS.in: Remove committer.
Related to http://bugs.gnu.org/8527
2015-04-13 09:39:37 +01:00
Pádraig Brady 1b1c40e1d6 df: fix --local hanging with inaccessible remote mounts
* src/df.c (filter_mount_list): With -l, avoid stating remote mounts.
* init.cfg: Avoid test hangs with inaccessible remote mounts.
* tests/df/no-mtab-status.sh: Skip with inaccessible remote mounts.
* tests/df/skip-rootfs.sh: Likewise.
* tests/df/total-verify.sh: Likewise.
* NEWS: Mention the bug fix.
Reported at http://bugzilla.redhat.com/1199679
2015-04-13 09:20:55 +01:00
Mitchel Humpherys 74bf6670eb doc: correct pluralization for mkfifo and mknod
* doc/coreutils.texi: `mkfifo' and `mknod' use the optContext macro
which adds a description for the SELinux security context in addition to
the single option already described in each case.  The result in both
cases is two options being introduced as `option' (singular).  Fix this
by introducing them as `options' (plural).
2015-04-09 08:49:34 +02:00
Pádraig Brady bf180f8f5a df: fix use of uninitialized variable reported by valgrind
Conditional jump or move depends on uninitialised value(s)
    at 0x40380C: get_field_values (df.c:840)
    by 0x403E16: get_dev (df.c:994)
    by 0x404D65: get_all_entries (df.c:1364)
    by 0x405926: main (df.c:1714)

* src/df.c (get_dev): Initialize the fsu.fsu_bavail_top_bit_set
member, when adding placeholder entries.
(main): Avoid a "definitely lost" memory leak warning from valgrind,
reported by Bernhard Voelker.
2015-04-03 14:27:27 +01:00
Pádraig Brady ccf1cd87f8 doc: clarify that ls --sort=time is newest first
* src/ls.c (usage): Add punctuation to avoid ambiguity in the
description of the --time option.  Mention that both the -u
and --sort=time default order is newest first.
2015-04-01 02:04:19 +01:00
Stephane Chazelas d313a0b242 tail: fix -f to follow changes after a rename
* src/tail.c (tail_forever_inotify): Only monitor write()s and
truncate()s to files in --follow=descriptor mode, thus avoiding
the bug where we removed the watch on renamed files.
Also adjust the inotify event processing code that is
now significant only in --follow=name mode.
* tests/tail-2/F-vs-rename.sh: Improve this existing test by running
in both polling and inotify modes.
* tests/tail-2/f-vs-rename.sh: A new test based on the existing one.
* tests/local.mk: Reference the new test.
* NEWS: Mention the bug.
Fixes http://bugs.gnu.org/19760
2015-03-31 11:55:32 +01:00
Pádraig Brady 52209ee547 doc: move numfmt info to the 'Numeric operations' section
* doc/coreutils.texi: Move numfmt info to this section,
as numfmt functionality aligns more with seq and factor,
than fmt and pr etc.
2015-03-29 02:44:57 +01:00
Paul Eggert 9b2782ca6d nohup: clarify stdin redirection
Problem reported by Isaac Schwabacher in:
http://bugs.gnu.org/20214
* doc/coreutils.texi (nohup invocation): Clarify that when nohup's
stdin gets redirected, it's unreadable.
* doc/coreutils.texi (nohup invocation):
* src/nohup.c (usage): Don't promise /dev/null.
2015-03-27 15:02:35 -07:00
Assaf Gordon cbed4b3376 tests: fix false test failure with df on Debian/kFreeBSD
* tests/fs/skip-duplicates.sh: On this platform .mnt_opts is significant
so define to empty to avoid a NULL deref in read_file_system_list().
Fixes http://bugs.gnu.org/20210
2015-03-27 09:43:42 +00:00
Christoph Anton Mitterer 41d0fc1c37 doc: disambiguate the ls --color description
* src/ls.c (usage): Avoid the implication that the
default ls behavior is to --color=always.
Reported in http://bugs.debian.org/781208
2015-03-26 09:51:50 +00:00
Pádraig Brady bf11fd17a9 doc: clarify the date standard output formats
* src/date.c (usage): Use FMT rather than TIMESPEC as the parameter,
since it's simpler to understand and can be better aligned.
Give an example for the --iso-8601 output format.
Adjust the example used for the 3 standard formats to be unambiguous
with respect to day/mon ordering and use of leading zeros in the time.
Reorder the options descriptions slightly, so that the
3 standards options are together.
Indent the multi-line descriptions so that grouping is obvious.
Remove a redundant description of the --rfc-3339 format,
which is obvious in the existing example.
Separate these 3 standards options to their own translatable string
to simplify translation.
Change 'date and time' to 'date/time' in the --iso-8601 description
to be consistent with --rfc-3339 and to help avoid the implication
that the time is always output or even output by default.
Fixes http://bugs.gnu.org/20203
2015-03-26 08:40:31 +00:00
Pádraig Brady 3b98ee7e99 maint: apply basic styling to online manual
* README-release: Reference http://www.gnu.org/s/coreutils/manual.css
to apply basic styling to the online coreutils manual, consistent
with the Emacs documentation.
2015-03-24 10:25:54 +00:00
Pádraig Brady 88f170e0c0 doc: clarify the uniq -D man page description
* src/uniq.c (usage): The description was very confusing in the man page
due to the stripped newlines.  Add punctuation for clarification.
2015-03-24 10:25:48 +00:00
Pádraig Brady 56ff7a6760 wc: use a more adaptive wc -l implementation
* src/wc.c (wc): Allow any block to select the count implementation,
rather than just using the first 10 lines.  This also simplifies
the code from 3 loops to 2.
2015-03-24 10:23:29 +00:00
Dan Jacobson 01fb984887 doc: clarify default order for ls --sort=size
* src/ls.c (usage): Mention that default order is largest first.
Fixes http://bugs.gnu.org/20172
2015-03-23 09:03:58 +00:00
Pádraig Brady 9042ff441c maint: really fix wildcard quoting in sc_tests_executable
* cfg.mk (sc_tests_executable): The previous commit avoided
the globbing, but also passed on the quoted wildcards to find(1).
We could use eval to handle the quoting, though that's a bit
awkward and dangerous, so instead explicitly disable globbing
for the whole make target subshell.  Note noglob is not available
on solaris, where we fall back to set -f.  Note also that zsh
uses set -F for this, but that's moot here.  Also correct the
find(1) expression to include the -o between each wildcard.
2015-03-20 14:34:31 +00:00
Pádraig Brady 79e27fd67e maint: fix wildcard quoting in sc_tests_executable
* cfg.mk (sc_tests_executable): If there are files with
$TEST_EXTENSIONS in the current directory, then the
lack of quoting of the $test_extensions_rx contents
could result in globbing and an inconsequential run.
find(1) produces warnings only with more than one expansion.
2015-03-20 01:16:06 +00:00
Kristoffer Brånemyr 1025243b6a wc: speedup counting of short lines
Using a test file generated with:
  yes | head -n100M > 2x100M.txt

before> time wc -l 2x100M.txt
  real 0.842s
  user 0.810s
  sys  0.033s

after> time wc -l 2x100M.txt
  real 0.142s
  user 0.111s
  sys  0.031s

* src/wc.c (wc): Split the loop that deals with -l into 3.
The first is used at the start of the input to determine if
the average line length is < 15, and if so the second loop is
used to look for '\n' internally to wc.  For longer lines,
memchr is used as before to take advantage of system specific
optimizations which any outweigh function call overhead.
Note the first 2 loops could be combined, though in testing,
GCC 4.9.2 at least, wasn't sophisticated enough to separate
the loops based on the "check_len" invariant.
Note also __builtin_memchr() isn't significant here as
GCC currently only applies constant folding with that.
* NEWS: Mention the improvement.
2015-03-20 00:48:52 +00:00
Giuseppe Scrivano e2e11119e0 yes: improve efficiency when all args aren't buffered
* src/yes.c (main): Even when the internal buffer isn't large enough,
output what we've buffered already, and interate over the rest.
This improves the performance in the edge case where there are
many small arguments that overflow the buffer.
* tests/misc/yes.sh: Add a test case for the many small arguments case.
2015-03-10 12:30:33 +00:00
Pádraig Brady 35217221c2 yes: output data more efficiently
yes(1) may be used to generate repeating patterns of text
for test inputs etc., so adjust to be more efficient.

Profiling the case where yes(1) is outputting small items
through stdio (which was the default case), shows the overhead
of continuously processing small items in main() and in stdio:

    $ yes >/dev/null & perf top -p $!
    31.02%  yes           [.] main
    27.36%  libc-2.20.so  [.] _IO_file_xsputn@@GLIBC_2.2.5
    14.51%  libc-2.20.so  [.] fputs_unlocked
    13.50%  libc-2.20.so  [.] strlen
    10.66%  libc-2.20.so  [.] __GI___mempcpy
     1.98%  yes           [.] fputs_unlocked@plta

Sending more data per stdio call improves the situation,
but still, there is significant stdio overhead due to memory copies,
and the repeated string length checking:

    $ yes "`echo {1..1000}`" >/dev/null & perf top -p $!
    42.26%  libc-2.20.so  [.] __GI___mempcpy
    17.38%  libc-2.20.so  [.] strlen
     5.21%  [kernel]      [k] __srcu_read_lock
     4.58%  [kernel]      [k] __srcu_read_unlock
     4.27%  libc-2.20.so  [.] _IO_file_xsputn@@GLIBC_2.2.5
     2.50%  libc-2.20.so  [.] __GI___libc_write
     2.45%  [kernel]      [k] system_call
     2.40%  [kernel]      [k] system_call_after_swapgs
     2.27%  [kernel]      [k] vfs_write
     2.09%  libc-2.20.so  [.] _IO_do_write@@GLIBC_2.2.5
     2.01%  [kernel]      [k] fsnotify
     1.95%  libc-2.20.so  [.] _IO_file_write@@GLIBC_2.2.5
     1.44%  yes           [.] main

We can avoid all stdio overhead by building up the buffer
_once_ and outputting that, and the profile below shows
the bottleneck moved to the kernel:

    $ src/yes >/dev/null & perf top -p $!
    15.42%  [kernel]      [k] __srcu_read_lock
    12.98%  [kernel]      [k] __srcu_read_unlock
     9.41%  libc-2.20.so  [.] __GI___libc_write
     9.11%  [kernel]      [k] vfs_write
     8.35%  [kernel]      [k] fsnotify
     8.02%  [kernel]      [k] system_call
     5.84%  [kernel]      [k] system_call_after_swapgs
     4.54%  [kernel]      [k] __fget_light
     3.98%  [kernel]      [k] sys_write
     3.65%  [kernel]      [k] selinux_file_permission
     3.44%  [kernel]      [k] rw_verify_area
     2.94%  [kernel]      [k] __fsnotify_parent
     2.76%  [kernel]      [k] security_file_permission
     2.39%  yes           [.] main
     2.17%  [kernel]      [k] __fdget_pos
     2.13%  [kernel]      [k] sysret_check
     0.81%  [kernel]      [k] write_null
     0.36%  yes           [.] write@plt

Note this change also ensures that yes(1) will only write
complete lines for lines shorter than BUFSIZ.

* src/yes.c (main): Build up a BUFSIZ buffer of lines,
and output that, rather than having stdio process each item.
* tests/misc/yes.sh: Add a new test for various buffer sizes.
* tests/local.mk: Reference the new test.
Fixes http://bugs.gnu.org/20029
2015-03-10 00:12:30 +00:00
Pádraig Brady 3c3b760c34 build: fix make dependencies for test.1
In certain parallel build situations this would give the error:
  help2man: can't get `--help' info from man/test.td/[
  Makefile:14189: recipe for target 'man/test.1' failed

* man/local.mk (test.1): Depend on `[` rather than `test`,
as `test --help` outputs nothing.  Also move dir.1 and vdir.1
back to the main list, as they're no more exceptions than
sha1sum etc.
2015-03-07 02:09:37 +00:00
Bernhard Voelker c18b296e83 maint: be less strict about executable permissions of tests
With "umask 0027" or even "umask 0077", the git clone of coreutils
does not have the executable bit set for 'other' (or 'group).
Therefore, "make syntax-check" would fail.

* cfg.mk (sc_tests_executable): Change the -perm argument of find(1)
to only print the names of the files which are not executable by the
user, rather than insisting on ugo+x (octal 111).
2015-03-05 16:54:52 +01:00
Stéphane Aulery e005719211 doc: mention persistence caveats in sync man page
* man/sync.x ([BUGS]): Mention the caveats and reference
the system call man pages for more details.
Fixes http://bugs.gnu.org/19995

Debian bug #507085 reported by jidanni@jidanni.org
2015-03-05 12:21:50 +00:00
Pádraig Brady 5e9c277f99 tee: generalize the --write-error option to --output-error
Adjust commit v8.23-140-gfdd6ebf to add the --output-error option
instead of --write-error, and treat open() errors like write() errors.

* doc/coreutils.texi (tee invocation): s/write-error/output-error/.
* src/tee.c (main): Exit on open() error if appropriate.
* tests/misc/tee.sh: Add a case to test open() errors.
* NEWS: Adjust for the more general output error behavior.

Suggested by Bernhard Voelker.
2015-03-04 13:24:21 +00:00
Jarosław Gruca 69410690b8 maint: update stale comment about ls color sequences
* src/ls.c (color_indicator[C_END]): Comment with the correct
sequence, which was used since commit v6.10-61-g483297d
Fixes http://bugs.gnu.org/19992
2015-03-04 11:47:59 +00:00
Shane M Seymour e60547db04 tail,stat: improve support for the IBRIX file system
Note that IBRIX used to have a different magic number 0x013111A7
instead of the current 0x013111A8.  However, the former is no longer
used and the version of IBRIX it was used in is really ancient, so
it's extremely unlikely anyone is still using it.  Therefore, just
add the newer magic number.
Mark IBRIX as a 'remote' file system type as inotify support had
never been officially tested with it.

* src/stat.c (human_fstype): Add file system ID definition.
* NEWS: Mention the improvement.

Fixes http://bugs.gnu.org/19951
2015-02-27 18:54:01 +01:00
Pádraig Brady fdd6ebf0ae tee: add --write-error to control handling of closed pipes
tee is very often used with pipes and this gives better control
when writing to them.  There are 3 classes of file descriptors
that tee can write to: files(1), pipes(2), and early close pipes(3).
Handling write errors to 1 & 2 is supported at present with the caveat
that failure writing to any pipe will terminate tee immediately.
Handling write errors to type 3 is not currently supported.
To improve the supported combinations we add these options:

 --write-error=warn
   Warn if error writing any output including pipes.
   Allows continued writing to still open files/pipes.
   Exit status is failure if any output had error.
 --write-error=warn-nopipe, -p
   Warn if error writing any output except pipes.
   Allows continued writing to still open files/pipes.
   Exit status is failure if any non pipe output had error.
 --write-error=exit
   Exit if error writing any output including pipes.
 --write-error=exit-nopipe
   Exit if error writing any output except pipes.

Use the "nopipe" variants when files are of types 1 and 3, otherwise
use the standard variants with types 1 and 2.  A caveat with the above
scheme is that a combination of pipe types (2 & 3) is not supported
robustly.  I.e. if you use the "nopipe" variants when using both type
2 and 3 pipes, then any "real" errors on type 2 pipes will not be
diagnosed.
  Note also a general issue with type 3 pipes that are not on tee's
stdout, is that shell constructs don't allow to distinguish early
close from real failures.  For example `tee >(head -n1) | grep -m1 ..`
can't distinguish between an error or an early close in "head" pipe,
while the fail on the grep part of the pipe is distinguished
independently from the resulting pipe errors.  This is a general
issue with the >() construct, rather than with tee itself.

* NEWS: Mention the new feature.
* doc/coreutils.texi (tee invocation): Describe the new option.
* src/tee.c (usage): Likewise.
(main): With --write-error ignore SIGPIPE, and handle
the various exit, diagnostics combinations.
* tests/misc/tee.sh: Tess all the new options.
Fixes http://bugs.gnu.org/11540
2015-02-24 18:12:22 +00:00
Bernhard Voelker 19d34d244b tee: close "-" file
This is a cleanup to the previous commit v8.23-138-g7ceaf1d.

* src/tee.c (tee_files): Do not exempt the "-" file from being closed,
as this is no longer stdout but a normal file.
2015-02-23 11:39:08 +01:00
Bernhard Voelker 7ceaf1d975 tee: treat '-' operand as file name as mandated by POSIX
Since v5.2.1-1247-g8dafbe5, tee(1) treated '-' as stdout while POSIX
explicitly requires to treat this as a file name.  Revert this change,
as the interleaved output - due to sending another copy of input to
stdout - is not considered to be useful.  Discussed in
http://lists.gnu.org/archive/html/coreutils/2015-02/msg00085.html

* src/tee.c (tee_files): Remove the special handling for "-" operands.
(usage): Remove the corresponding sentence.
* doc/coreutils.texi (common options): Remove the "tee -" example.
(tee invocation): Document that tee(1) now treats "-" as a file name.
* tests/misc/tee.sh: Add a test case for "tee -".
While at it, re-indent the above multi-argument processing case and
extend that to 13 operands, as POSIX mandates that, too.
* tests/misc/tee-dash.sh: Remove now-obsolete test.
* tests/local.mk (all_tests): Remove the above test.
* NEWS (Changes in behavior): Mention the change.
2015-02-20 14:00:41 +01:00
Pádraig Brady 3f2f05f067 tee: exit early if no more writable outputs
* src/tee.c (main): Don't continue reading if we can't
output anywhere.
* tests/misc/tee.sh: Ensure we exit when no more outputs.
* NEWS: Mention the change in behavior.
2015-02-18 23:41:27 +00:00
Pádraig Brady ced120406f tests: support stderr verification with returns_()
* tests/init.sh (returns_): Disable tracing for this wrapper
function, so that stderr of the wrapped command is unchanged,
allowing for verification of the contents.
2015-02-18 23:36:34 +00:00