The setup code has been cleaned up to avoid leaks around the
repository_format structure.
* ma/clear-repository-format:
setup: fix memory leaks with `struct repository_format`
setup: free old value before setting `work_tree`
A recent update broke "is this object available to us?" check for
well-known objects like an empty tree (which should yield "yes",
even when there is no on-disk object for an empty tree), which has
been corrected.
* jk/virtual-objects-do-exist:
rev-list: allow cached objects in existence check
On platforms where "git fetch" is killed with SIGPIPE (e.g. OSX),
the upload-pack that runs on the other end that hangs up after
detecting an error could cause "git fetch" to die with a signal,
which led to a flakey test. "git fetch" now ignores SIGPIPE during
the network portion of its operation (this is not a problem as we
check the return status from our write(2)s).
* jk/no-sigpipe-during-network-transport:
fetch: ignore SIGPIPE during network operation
fetch: avoid calling write_or_die()
"git fsck --connectivity-only" omits computation necessary to sift
the objects that are not reachable from any of the refs into
unreachable and dangling. This is now enabled when dangling
objects are requested (which is done by default, but can be
overridden with the "--no-dangling" option).
* jk/fsck-doc:
fsck: always compute USED flags for unreachable objects
doc/fsck: clarify --connectivity-only behavior
"git rebase" that was reimplemented in C did not set ORIG_HEAD
correctly, which has been corrected.
* js/rebase-orig-head-fix:
built-in rebase: set ORIG_HEAD just once, before the rebase
built-in rebase: demonstrate that ORIG_HEAD is not set correctly
built-in rebase: use the correct reflog when switching branches
built-in rebase: no need to check out `onto` twice
The final report from "git bisect" used to show the suspected
culprit using a raw "diff-tree", with which there is no output for
a merge commit. This has been updated to use a more modern and
human readable output that still is concise enough.
* jk/bisect-final-output:
bisect: make diff-tree output prettier
bisect: fix internal diff-tree config loading
bisect: use string arguments to feed internal diff-tree
CFLAGS now can be tweaked when invoking Make while using
DEVELOPER=YesPlease; this did not work well before.
* ab/makefile-help-devs-more:
Makefile: allow for combining DEVELOPER=1 and CFLAGS="..."
Makefile: move the setting of *FLAGS closer to "include"
Makefile: Move *_LIBS assignment into its own section
Makefile: add/remove comments at top and tweak whitespace
Makefile: move "strip" assignment down from flags
Makefile: remove an out-of-date comment
Some compilers don't allow NULL to be passed for a va_list,
and e.g. "gcc (Raspbian 6.3.0-18+rpi1+deb9u1) 6.3.0 20170516"
errors out like this:
trace2/tr2_tgt_event.c:193:18:
error: invalid operands to binary &&
(have ‘int’ and ‘va_list {aka __va_list}’)
if (fmt && *fmt && ap) {
^^
I couldn't find any hints that va_list and pointers can be mixed,
and no hints that they can't either. Morten Welinder comments:
"C99, Section 7.15, simply says that va_list "is an object type suitable for
holding information needed by the macros va_start, va_end, and
va_copy". So clearly not guaranteed to be mixable with pointers...
The portable solution is to use "va_list" everywhere in the callchain.
As a consequence, both trace2_region_enter_fl() and trace2_region_leave_fl()
now take a variable argument list.
Signed-off-by: Torsten Bögershausen <tboegi@web.de>
Acked-by: Jeff Hostetler <jeffhost@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Remove the rebase.useBuiltin setting, which was added as an escape
hatch to disable the builtin version of rebase first released with Git
2.20.
See [1] for the initial implementation of rebase.useBuiltin, and [2]
and [3] for the documentation and corresponding
GIT_TEST_REBASE_USE_BUILTIN option.
Carrying the legacy version is a maintenance burden as seen in
7e097e27d3 ("legacy-rebase: backport -C<n> and --whitespace=<option>
checks", 2018-11-20) and 9aea5e9286 ("rebase: fix regression in
rebase.useBuiltin=false test mode", 2019-02-13). Since the built-in
version has been shown to be stable enough let's remove the legacy
version.
As noted in [3] having use_builtin_rebase() shell out to get its
config doesn't make any sense anymore, that was done for the purposes
of spawning the legacy rebase without having modified any global
state. Let's instead handle this case in rebase_config().
There's still a bunch of references to git-legacy-rebase in po/*.po,
but those will be dealt with in time by the i18n effort.
Even though this configuration variable only existed two releases
let's not entirely delete the entry from the docs, but note its
absence. Individual versions of git tend to be around for a while due
to distro packaging timelines, so e.g. if we're "lucky" a given
version like 2.21 might be installed on say OSX for half a decade.
That'll mean some people probably setting this in config, and then
when they later wonder if it's needed they can Google search the
config option name or check it in git-config. It also allows us to
refer to the docs from the warning for details.
1. 55071ea248 ("rebase: start implementing it as a builtin",
2018-08-07)
2. d8d0a546f0 ("rebase doc: document rebase.useBuiltin", 2018-11-14)
3. 62c23938fa ("tests: add a special setup where rebase.useBuiltin is
off", 2018-11-14)
3. https://public-inbox.org/git/nycvar.QRO.7.76.6.1903141544110.41@tvgsbejvaqbjf.bet/
Acked-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Remember --allow-empty, --allow-empty-message and
--keep-redundant-commits when cherry-pick stops for a conflict
resolution.
Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When cherry-pick stops for a conflict resolution it forgets
--allow-empty --allow-empty-message and --keep-redundant-commits.
Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
reformat save_opts() to remove excessively long lines.
Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
AsciiDoc and Asciidoctor do not agree on what to write in the header and
footer of each man-page, i.e., the very first and the very last line of
*.[157]. Those differences can certainly be interesting in their own
right, but they clutter the output of `./doc-diff --from-asciidoc
--to-asciidoctor HEAD HEAD` quite a bit since the diff contains some
10-15 lines of noise per file diffed.
Teach doc-diff to cut away the first two and last two lines, i.e., the
header/footer and the empty line immediately following/preceding it.
Because Asciidoctor uses an extra empty line compared to AsciiDoc,
remove one more line at each end of the file, but only if it's empty.
An alternative approach might be to pass down `--no-header-footer`,
which both AsciiDoc and Asciidoctor understand, but it has some
drawbacks. First of all, the result doesn't build -- `xmlto` stumbles on
the resulting xml since it has multiple root elements. Second, it cuts
too much -- dropping the header loses the synopsis, which would be
interesting to diff.
Like in the previous commit, encode this option into the directory name
of the "installed" and "rendered" files. Otherwise, we wouldn't be able
to trust that what we use out of that cache actually corresponds to the
options given for this run. (We could optimize this caching a little
since this flag doesn't affect the contents of "installed" at all, but
let's punt on that.)
Signed-off-by: Martin Ågren <martin.agren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Provide `--from-asciidoctor` and `--to-asciidoctor` to select that the
"from" resp. "to" commit should be built with Asciidoctor, and provide
an `--asciidoctor` shortcut for giving both. Similarly, provide
--{from-,to-,}asciidoc for explicitly selecting AsciiDoc.
Implement this using the USE_ASCIIDOCTOR flag. Let's not enforce a
default here, but instead just let the Makefile fall back on whatever is
in config.mak, so that `./doc-diff foo bar` without any of of these new
options behaves exactly like it did before this commit.
Encode the choice into the directory names of our "installed" and
"rendered" files, so that we can run `./doc-diff --from-asciidoc
--to-asciidoctor HEAD HEAD` without our two runs stomping on each other.
Signed-off-by: Martin Ågren <martin.agren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In `render_tree()`, `$1` is documented to be the commit-ish/oid and we
use it as that with `git checkout`, but we mostly use it to form the
name of various directories. To separate these concerns, and because we
are about to construct the directory names a bit differently, take two
distinct arguments instead.
Signed-off-by: Martin Ågren <martin.agren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If you build the documentation switching between different options,
e.g., to build with both Asciidoc and Asciidoctor, you'll probably find
yourself running `make -C Documentation clean` either too often (wasting
time) or too rarely (getting mixed builds).
Track the flags we're using in the documentation build, similar to how
the main Makefile tracks CFLAGS and prefix flags. Track ASCIIDOC_COMMON
directly rather than its individual components -- that should make it
harder to forget to update the tracking if/when we modify the build
commands.
Signed-off-by: Martin Ågren <martin.agren@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
d76ce4f734 ("log,diff-tree: add --combined-all-paths option",
2019-02-07) added tests for files containing tabs.
When the tests are run with bash, the lack of quoting during the file
setup causes 'ambiguous redirect' errors.
Signed-off-by: Todd Zullinger <tmz@pobox.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The "git reflog expire" command when given an unknown reference has
since 4264dc15e1 ("git reflog expire", 2006-12-19) when this command
was implemented emit an error, but this has never been tested for.
Let's test for it, also under gc.reflogExpire{Unreachable,}=never in
case a future change is tempted to take shortcuts in the presence of
such config.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change a couple of tests that weren't using the helper to use it. This
makes the trailing "--unset" unnecessary.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Change an idiom we're using to ensure that gc_before_repack() only
does work once (see 62aad1849f ("gc --auto: do not lock refs in the
background", 2014-05-25)) to be more obvious.
Nothing except this function cares about the "pack_refs" and
"prune_reflogs" variables, so let's not leave the reader wondering if
they're being zero'd out for later use somewhere else.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
There's been a lot of changing of the hardcoded "40" values to
the_hash_algo->hexsz, but we've so far missed this one where we
hardcoded 38 for the loose object file length.
This is because a SHA-1 like abcde[...] gets turned into
objects/ab/cde[...]. There's no reason to suppose the same won't be
the case for SHA-256, and reading between the lines in
hash-function-transition.txt the format is planned to be the same.
In the future we may want to further modify this code for the hash
function transition. There's a potential pathological case here where
we'll only consider the loose objects for the currently active hash,
but objects for that hash will share a directory storage with the
other hash.
Thus we could theoretically have e.g. 1k SHA-1 loose objects, and 1
million SHA-256 objects. Then not notice that we need to pack them
because we're currently using SHA-1, even though our FS may be
straining under the stress of such humongous directories.
So assuming that "gc" eventually learns to pack up both SHA-1 and
SHA-256 objects regardless of what the current the_hash_algo is,
perhaps this check should be changed to consider all files in
objects/17/ matching [0-9a-f] 38 or 62 characters in length (i.e. both
SHA-1 and SHA-256).
But none of that is something we need to worry about now, and
supporting both 38 and 62 characters depending on "the_hash_algo"
removes another case of SHA-1 hardcoding.
As noted in [1] I'm making no effort to somehow remove the hardcoding
for "2" as in "use the first two hexdigits for the directory
name". There's no indication that that'll ever change, and somehow
generalizing it here would be a drop in the ocean, so there's no point
in doing that. It also couldn't be done without coming up with some
generalized version of the magical "objects/17" directory. See [2] for
a discussion of that directory.
1. https://public-inbox.org/git/874l84ber7.fsf@evledraar.gmail.com/
2. https://public-inbox.org/git/87k1mta9x5.fsf@evledraar.gmail.com/
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Since its inception, the perf-lib.sh script has manually handled the
"--tee" option (and other options which imply it, like "--valgrind")
with a cut-and-pasted block from test-lib.sh. That block has grown stale
over the years, and has at least three problems:
1. It uses $SHELL to re-exec the script, whereas the version in
test-lib.sh learned to use $TEST_SHELL_PATH.
2. It does an ad-hoc search of the "$*" string, whereas test-lib.sh
learned to carefully parse the arguments left to right.
3. It never learned about --verbose-log (which also implies --tee),
so it would not trigger for that option.
This last one was especially annoying, because t/perf/run uses the
GIT_TEST_OPTS from your config.mak to run the perf scripts. So if you've
set, say, "-x --verbose-log" there, it will be passed as part of most
perf runs. And while this script doesn't recognize the option, the
test-lib.sh that we source _does_, and the behavior ends up being much
more annoying:
- as the comment at the top of the block says, we have to run this
tee code early, before we start munging variables (it says
GIT_BUILD_DIR, but the problematic variable is actually
GIT_TEST_INSTALLED).
- since we don't recognize --verbose-log, we don't trigger the block.
We go on to munge GIT_TEST_INSTALLED, converting it from a relative
to an absolute path.
- then we source test-lib.sh, which _does_ recognize --verbose-log. It
re-execs the script, which runs again. But this time with an
absolute version of GIT_TEST_INSTALLED.
- As a result, we copy the absolute version of GIT_TEST_INSTALLED into
perf_results_prefix. Instead of writing our results to the expected
"test-results/build_1234abcd.p1234-whatever.times", we instead write
them to "test-results/_full_path_to_repo_t_perf_build_1234...".
The aggregate.perl script doesn't expect this, and so it prints
"<missing>" for each result (even though it spent considerable time
running the tests!).
We can solve all of these in one blow by just deleting our custom
handling, and relying on the inclusion of test-lib.sh to handle --tee,
--verbose-log, etc.
There's one catch, though. We want to handle GIT_TEST_INSTALLED after
we've included test-lib.sh, since we want it un-munged in the re-exec'd
version of the script. But if we want to convert it from a relative
to an absolute path, we must do so before we load test-lib.sh, since it
will change our working directory. So we compute the absolute directory
first, store it away, then include test-lib.sh, and finally assign to
GIT_TEST_INSTALLED as appropriate.
Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The pkt-line formatted lines contained the wrong pkt-len.
Signed-off-by: Mike Hommey <mh@glandium.org>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Add a definition for what overlay means in the context of git, to
clarify the recently introduced overlay-mode in git checkout.
Helped-by: Elijah Newren <newren@gmail.com>
Helped-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Thomas Gummerer <t.gummerer@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
As far as this developer can tell, the conversion from a Perl script to
a built-in caused the regression in the difftool that it no longer runs
outside of a Git worktree (with `--no-index`, of course).
It is a bit embarrassing that it took over two years after retiring the
Perl version to discover this regression, but at least we now know, and
can do something, about it.
This fixes https://github.com/git-for-windows/git/issues/2123
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
`OPT_ARGUMENT()` is intended to keep the specified long option in `argv`
and not to do anything else.
However, it would make a lot of sense for the caller to know whether
this option was seen at all or not. For example, we want to teach `git
difftool` to work outside of any Git worktree, but only when
`--no-index` was specified.
Note: nothing in Git uses OPT_ARGUMENT(). Even worse, looking through
the commit history, one can easily see that nothing even
ever used it, apart from the regression test.
So not only do we make `OPT_ARGUMENT()` more useful, we are also about
to introduce its first real user!
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We will always spawn something from `git difftool`, so we will always
have to set `GIT_DIR` and `GIT_WORK_TREE`.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Spell out --no-rerere-autoupdate explictly to make searching
easier. This matches the other --no options in the man page.
Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This option was missing from the man pages of these commands.
Signed-off-by: Phillip Wood <phillip.wood@dunelm.org.uk>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Checking gc_auto_threshold in too_many_loose_objects() was added in
17815501a8 ("git-gc --auto: run "repack -A -d -l" as necessary.",
2007-09-17) when need_to_gc() itself was also reliant on
gc_auto_pack_limit before its early return:
gc_auto_threshold <= 0 && gc_auto_pack_limit <= 0
When that check was simplified to just checking "gc_auto_threshold <=
0" in b14d255ba8 ("builtin-gc.c: allow disabling all auto-gc'ing by
assigning 0 to gc.auto", 2008-03-19) this unreachable code should have
been removed. We only call too_many_loose_objects() from within
need_to_gc() itself, which will return if this condition holds, and in
cmd_gc() which will return before ever getting to "auto_gc &&
too_many_loose_objects()" if "auto_gc && !need_to_gc()" is true
earlier in the function.
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It is quite possible that the loose object cache gets stale when new
objects are written. In that case, get_oid() would potentially say that
it cannot find a given object, even if it should find it.
Let's blow away the loose object cache as well as the read packs and try
again in that case.
Note: this does *not* affect the code path that was introduced to help
avoid looking for the same non-existing objects (which made some
operations really expensive via NFS): that code path is handled by the
`OBJECT_INFO_QUICK` flag (which does not even apply to `get_oid()`,
which has no equivalent flag, at least at the time this patch was
written).
This incidentally fixes the problem identified earlier where an
interactive rebase wanted to re-read (and validate) the todo list after
an `exec` command modified it.
Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The interactive rebase simply complains about an "invalid line" when the
object hash of, say, a `pick` line could not be parsed.
Let's tell the user what happened in a little more detail.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
We specifically support `exec` commands in `git rebase -i`'s todo lists
to rewrite the very same todo list. Of course, we need to validate that
todo list when re-reading it.
It is also totally legitimate to extend the todo list by `pick` lines
using short names of commits that were created only after the rebase
started.
And this is where the loose object cache interferes with this feature:
if *some* loose object was read whose hash shares the same first two
digits with a commit that was not yet created when that loose object was
created, then we fail to find that new commit by its short name in
`get_oid()`, and the interactive rebase fails with an obscure error
message like:
error: invalid line 1: pick 6568fef
error: please fix this using 'git rebase --edit-todo'.
Let's first demonstrate that this is actually a bug in a new regression
test, in a separate commit so that other developers who do not believe
me can cherry-pick it to confirm the problem.
This new regression test generates two commits whose hashes share the
first two hex digits (so that their corresponding loose objects live in
the same subdirectory of .git/objects/, and are therefore supposed to be
in the same loose object cache bin).
It then picks the first, to make sure that the loose object cache is
initialized and cached that object directory, then generates the second
commit and picks it, too. Since the commit was generated in a different
process than the sequencer that wants to pick it, the loose object cache
had no chance of being updated in the meantime.
Technically, we would need only one `exec` command in this regression
test case, but for ease of implementation, it uses a pseudo-recursive
call to the same script.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In 't9811-git-p4-label-import.sh' the test 'tag that cannot be
exported' runs
!(p4 labels | grep GIT_TAG_ON_A_BRANCH)
to check that the given string is not printed by 'p4 labels'. This is
problematic, because according to POSIX [1]:
"If the pipeline begins with the reserved word ! and command1 is a
subshell command, the application shall ensure that the ( operator
at the beginning of command1 is separated from the ! by one or more
<blank> characters. The behavior of the reserved word ! immediately
followed by the ( operator is unspecified."
While most common shells still interpret this '!' as "negate the exit
code of the last command in the pipeline", 'mksh/lksh' don't and
interpret it as a negative file name pattern instead. As a result
they attempt to run a command made up of the pathnames in the current
directory (it contains a single directory called 'main'), which, of
course, fails the test.
We could fix it simply by adding a space between the '!' and '(', but
instead let's fix it by removing the unnecessary subshell.
[1] http://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#tag_18_09_02
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
According to the comments in 't/lib-git-p4.sh', sometimes 'p4d' seems
to hang, and to deal with that 'start_p4d' starts a watchdog process
to kill it after a long-enough timeout ($P4D_TIMEOUT, defaults to
300s). This watchdog process is implemented as a background subshell
loop iterating once every second until the timeout expires, producing
a few lines of trace output on each iteration when the test script is
run with '-x' tracing enabled. The watchdog loop's trace gets
intermixed with the real test output and trace, and makes that harder
to read.
Send the trace output of this loop to /dev/null to avoid polluting the
real test output.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
'lib-git-p4.sh' uses timeouts in a watchdog process to kill a
potentially stuck 'p4d' process and for certain cleanup operation
between tests. It does so by first computing when the timeout should
expire, and then repeatedly asking for the current time in seconds
until it exceeds the expiration time, and for portability reasons it
uses a one-liner Python script to ask for the current time.
Replace these timeouts with downcounters, which, though not
necessarily shorter, are much simpler, at least in the sense that they
don't execute the Python interpreter every second.
After this change the helper function with that Python one-liner has
no callers left, remove it.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Confusingly, the 'git p4' tests used two cleanup functions:
- 'kill_p4d' was run in the last test before 'test_done', and it not
only killed 'p4d', but it killed the watchdog process, and cleaned
up after 'p4d' as well by removing all directories used by the P4
daemon and client.
This cleanup is not necessary right before 'test_done', because
the whole trash directory is about to get removed anyway, but it
is necessary in 't9801-git-p4-branch.sh', which uses 'kill_p4d' to
stop 'p4d' before re-starting it in the middle of the test script.
- 'cleanup' was run in the trap on EXIT, and it killed 'p4d', but,
it didn't kill the watchdog process, and, contrarily to its name,
didn't perform any cleanup whatsoever.
Make it clearer what's going on by renaming and simplifying the
cleanup functions, so in the end we'll have:
- 'stop_p4d_and_watchdog' replaces 'cleanup' as it will try to live
up to its name and stop both the 'p4d' and the watchdog processes,
and as the sole function registered with 'test_atexit' it will be
responsible for no leaving any stray processes behind after 'git p4'
tests were finished or interrupted.
- 'stop_and_cleanup_p4d' replaces 'kill_p4d' as it will stop 'p4d'
(and the watchdog) and remove all directories used by the P4
daemon and cliean, so it can be used mid-script to stop and then
re-start 'p4d'.
Note that while 'cleanup' sent a single SIGKILL to 'p4d', 'kill_p4d'
was quite brutal, as it first sent SIGTERM to the daemon repeatedly,
either until its pid disappeared or until a given timeout was up, and
then it sent SIGKILL repeatedly, for good measure. This is overkill
(pardon the pun): a single SIGKILL should be able to take down any
process in a sensible state, and if a process were to somehow end up
stuck in the dreaded uninterruptible sleep state then even repeated
SIGKILLs won't bring immediate help. So ditch all the repeated
SIGTERM/SIGKILL parts, and use a single SIGKILL to stop 'p4d', and
make sure that there are no races between asynchron signal delivery
and subsequent restart of 'p4d' by waiting for it to die.
With this change the 'retry_until_fail' helper has no callers left,
remove it.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use 'test_atexit' to run cleanup commands to stop 'p4d' at the end of
the test script or upon interrupt or failure, as it is shorter,
simpler, and more robust than registering such cleanup commands in the
trap on EXIT in the test scripts.
Note that one of the test scripts, 't9801-git-p4-branch.sh', stops and
then re-starts 'p4d' twice in the middle of the script; take care that
the cleanup functions to stop 'p4d' are only registered once.
Note also that 'git p4' tests invoke different functions in the trap
on EXIT ('cleanup') and in the last test before 'test_done'
('kill_p4d'). Register both of these functions with 'test_atexit' for
now, and a a later patch in this series will then clean up the
redundancy.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use 'test_atexit' to run cleanup commands to stop the credentials
helper at the end of the test script or upon interrupt or failure, as
it is shorter, simpler, and more robust than registering such cleanup
commands in the trap on EXIT in the test scripts.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use 'test_atexit' to run cleanup commands to stop httpd at the end of
the test script or upon interrupt or failure, as it is shorter,
simpler, and more robust than registering such cleanup commands in the
trap on EXIT in the test scripts.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Use 'test_atexit' to run cleanup commands to stop 'git-daemon' at the
end of the test script or upon interrupt or failure, as it is shorter,
simpler, and more robust than registering such cleanup commands in the
trap on EXIT in the test scripts.
Note that in 't5570-git-daemon.sh' the daemon is stopped and then
re-started in the middle of the test script; take care that the
cleanup functions to stop the daemon are only registered once.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When running Apache, 'git daemon', or p4d, we want to kill them at the
end of the test script, otherwise a leftover daemon process will keep
its port open indefinitely, and thus will interfere with subsequent
executions of the same test script.
So far, we stop these daemon processes "manually", i.e.:
- by registering functions or commands in the trap on EXIT to stop
the daemon while preserving the last seen exit code before the
trap (to deal with a failure when run with '--immediate' or with
interrupts by ctrl-C),
- and by invoking these functions/commands last thing before
'test_done' (and sometimes restoring the test framework's default
trap on EXIT, to prevent the daemons from being killed twice).
On one hand, we do this inconsistently, e.g. 'git p4' tests invoke
different functions in the trap on EXIT and in the last test before
'test_done', and they neither restore the test framework's default trap
on EXIT nor preserve the last seen exit code. On the other hand, this
is error prone, because, as shown in a previous patch in this series,
any output from the cleanup commands in the trap on EXIT can prevent a
proper cleanup when a test script run with '--verbose-log' and certain
shells, notably 'dash', is interrupted.
Let's introduce 'test_atexit', which is loosely modeled after
'test_when_finished', but has a broader scope: rather than running the
commands after the current test case, run them when the test script
finishes, and also run them when the test is interrupted, or exits
early in case of a failure while the '--immediate' option is in
effect.
When running the cleanup commands at the end of a successful test,
then they will be run in 'test_done' before it removes the trash
directory, i.e. the cleanup commands will still be able to access any
pidfiles or socket files in there. When running the cleanup commands
after an interrupt or failure with '--immediate', then they will be
run in the trap on EXIT. In both cases they will be run in
'test_eval_', i.e. both standard error and output of all cleanup
commands will go where they should according to the '-v' or
'--verbose-log' options, and thus won't cause any troubles when
interrupting a test script run with '--verbose-log'.
Signed-off-by: Johannes Schindelin <johannes.schindelin@gmx.de>
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
After 'start_git_daemon' starts 'git daemon' (note the space in the
middle) in the background, it saves the background process' PID, so
the daemon can be stopped at the end of the test script. However,
'git-daemon' is not a builtin but a dashed external command, which
means that the dashless 'git daemon' executes the dashed 'git-daemon'
command, and, consequently, the PID recorded is not the PID of the
"real" daemon process, but that of the main 'git' wrapper. Now, if a
test script involving 'git daemon' is interrupted by ctrl-C, then only
the main 'git' process is stopped, but the real daemon process tends
to survive somehow, and keeps on running in the background
indefinitely, keeping the daemon's port to itself, and thus preventing
subsequent runs of the same test script.
Work this around by running 'git daemon' with the '--pidfile=...'
option to save the PID of the real daemon process, and kill that
process in 'stop_git_daemon' as well.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When a test script run with 'dash' and '--verbose-log -x' is
interrupted by ctrl-C, SIGTERM, or closing the terminal window, then
most of the time the registered EXIT trap actions are not executed.
This is an annoying issue with tests involving daemons, because they
should run cleanup commands to kill those daemon processes in the trap
on EXIT, but since these cleanup commands are not executed, the
daemons are left alive and keep their port open, thus interfering with
subsequent execution of the same test script.
The cause of this issue is the subtle combination of several factors
(bear with me, or skip over the indented part):
- Even when the test script is interrupted, the cleanup commands are
not run in the trap on INT, TERM, or HUP, but in the trap on EXIT
after the trap on the signals invokes 'exit' [1].
- According to POSIX [2]:
"The environment in which the shell executes a trap on EXIT
shall be identical to the environment immediately after the last
command executed before the trap on EXIT was taken."
Pertinent to the issue at hand is that all open file descriptors
and the state of '-x' tracing should be preserved. All shells
I've tried [3] preserve '-x'. Unfortunately, however:
- 'dash' doesn't conform to this when it comes to open file
descriptors: even when standard output and/or error are
redirected somewhere when 'exit' is invoked, anything written
to them in the trap on EXIT goes to the script's original
stdout and stderr [4].
We can't dismiss this with a simple "it doesn't conform to
POSIX, so we don't care", because 'dash' is the default
/bin/sh in some of the more popular Linux distros.
- As far as I can tell, POSIX doesn't explicitly say anything
about the environment of trap actions for various signals.
In practice it seems that most shells behave sensibly and
preserve both open file descriptors and the state of '-x'
tracing for the traps on INT, TERM, and HUP, including even
'dash'. The exceptions are 'mksh' and 'lksh': they do
preserve '-x', but not the open file descriptors.
- When a test script run with '-x' tracing enabled is interrupted,
then it's very likely that the signal arrives mid-test, i.e.:
- while '-x' tracing is enabled, and, consequently, our trap
actions on INT, TERM, HUP, and EXIT will produce trace output
as well.
- while standard output and error are redirected to a log file,
to the test script's original standard output and error, or to
/dev/null, depending on whether the test script was run with
'--verbose-log', '-v', or neither. According to the above, we
can't rely on these redirections still be in effect when
running the traps on INT, TERM, HUP, and/or EXIT.
- When a test script is run with '--verbose-log', then the test
script is re-executed with its standard output and error piped
into 'tee', in order to send the "regular" non-verbose test's
output both to the terminal and to the log file. When the test is
interrupted, then the signal interrupts the downstream 'tee' as
well.
Putting these together, when a test script run with 'dash' and
'--verbose-log -x' is interrupted, then 'dash' tries to write the
trace output from the EXIT trap to the script's original standard
error, but it very likely can't, because the 'tee' downstream of the
pipe is interrupted as well. This causes the shell running the test
script to die because of SIGPIPE, without running any of the commands
in the EXIT trap.
Disable '-x' tracing in the trap on INT, TERM, and HUP to avoid this
issue, as it disables tracing in the chained trap on EXIT as well.
Wrap it in a '{ ... } 2>/dev/null' block, so the trace of the command
disabling the tracing doesn't go to standard error either [5].
Note that it's not only '-x' tracing that can be problematic, but any
shell builtin, e.g. 'echo', that writes to standard output or error in
the trap on EXIT, while a test running with 'dash' and '--verbose-log'
(even without '-x') is interrupted. As far as I can tell, this is not
an issue at the moment:
- The cleanup commands to stop the credential-helper, Apache, or
'p4d' don't use any such shell builtins.
- stop_git_daemon() does use 'say' and 'error', both wrappers around
'echo', but it redirects 'say' to fd 3, i.e. to the log file, and
while 'error' does write to standard output, it comes only after
the daemon was killed.
- The non-builtin commands that actually stop the daemons ('kill',
'apache2 -k stop', 'git credential-cache exit') are silent, so they
won't get SIGPIPE before finishing their job.
[1] The trap on EXIT must run cleanup commands, because we want to
stop any daemons when a test script run with '--immediate' fails
and exits early with error. By chaining up the trap on signals to
the trap on EXIT we can deal with cleanup commands a bit simpler,
because the tests involving daemons only have to set a single
trap.
[2] http://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#trap
[3] The shells I tried: dash, Bash, ksh, ksh93, mksh, lksh, yash,
BusyBox sh, FreeBSD /bin/sh, NetBSD /bin/sh.
[4] $ cat trap-output.sh
#!/bin/sh
trap "echo output; echo error >&2" EXIT
{ exit; } >OUT 2>ERR
$ dash ./trap-output.sh
output
error
$ wc -c OUT ERR
0 OUT
0 ERR
On a related note, 'ksh', 'ksh93', and BusyBox sh don't conform to
the specs in this respect, either.
[5] This '{ set +x; } 2>/dev/null' trick won't help those shells that
show trace output for any redirections and don't preserve open
file descriptors for the trap on INT, TERM and HUP. The only such
shells I'm aware of are 'mksh' and 'lksh'.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When cloning with --recurse-submodules a superproject with at least one
submodule with HEAD pointing to an unborn branch, the clone goes
something like this:
Cloning into 'test'...
<messages about cloning of superproject>
Submodule '<name>' (<uri>) registered for path '<submodule path>'
Cloning into '<submodule path>'...
fatal: Couldn't find remote ref HEAD
Unable to fetch in submodule path '<submodule path>'
<messages about fetching with SHA-1>
From <uri>
* branch <hash> -> FETCH_HEAD
Submodule path '<submodule path>': checked out '<hash>'
In other words, first, a fetch is done with no hash arguments (that is,
a fetch of HEAD) resulting in a "Couldn't find remote ref HEAD" error;
then, a fetch is done given a hash, which succeeds.
The fetch given a hash was added in fb43e31f2b ("submodule: try harder
to fetch needed sha1 by direct fetching sha1", 2016-02-24), and the
"Unable to fetch..." message was downgraded from a fatal error to a
notice in e30d833671 ("git-submodule.sh: try harder to fetch a
submodule", 2018-05-16).
This commit improves the notice to be clearer that we are retrying the
fetch, and that the previous messages (in particular, the fatal errors
from fetch) do not necessarily indicate that the whole command fails. In
other words:
- If the HEAD-fetch succeeds and we then have the commit we want,
git-submodule prints no explanation.
- If the HEAD-fetch succeeds and we do not have the commit we want, but
the hash-fetch succeeds, git-submodule prints no explanation.
- If the HEAD-fetch succeeds and we do not have the commit we want, but
the hash-fetch fails, git-submodule prints a fatal error.
- If the HEAD-fetch fails, fetch prints a fatal error, and
git-submodule informs the user that it will retry by fetching
specific commits by hash.
- If the hash-fetch then succeeds, git-submodule prints no
explanation (besides the ones already printed).
- If the HEAD-fetch then fails, git-submodule prints a fatal error.
It could be said that we should just eliminate the HEAD-fetch
altogether, but that changes some behavior (in particular, some refs
that were opportunistically updated would no longer be), so I have left
that alone for now.
There is an analogous situation with the fetching code in fetch_finish()
and surrounding functions. For now, I have added a NEEDSWORK.
Signed-off-by: Jonathan Tan <jonathantanmy@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
