Our testing framework uses a special i18n "poisoned localization"
feature to find messages that ought to stay constant but are
incorrectly marked to be translated. This feature has been made
into a runtime option (it used to be a compile-time option).
* ab/dynamic-gettext-poison:
Makefile: ease dynamic-gettext-poison transition
i18n: make GETTEXT_POISON a runtime option
The procedure to install dependencies before testing at Travis CI
is getting revamped for both simplicity and flexibility, taking
advantage of the recent move to the vm-based environment.
* sg/travis-install-dependencies:
travis-ci: install packages in 'ci/install-dependencies.sh'
Change the GETTEXT_POISON compile-time + runtime GIT_GETTEXT_POISON
test parameter to only be a GIT_TEST_GETTEXT_POISON=<non-empty?>
runtime parameter, to be consistent with other parameters documented
in "Running tests with special setups" in t/README.
When I added GETTEXT_POISON in bb946bba76 ("i18n: add GETTEXT_POISON
to simulate unfriendly translator", 2011-02-22) I was concerned with
ensuring that the _() function would get constant folded if NO_GETTEXT
was defined, and likewise that GETTEXT_POISON would be compiled out
unless it was defined.
But as the benchmark in my [1] shows doing a one-off runtime
getenv("GIT_TEST_[...]") is trivial, and since GETTEXT_POISON was
originally added the GIT_TEST_* env variables have become the common
idiom for turning on special test setups.
So change GETTEXT_POISON to work the same way. Now the
GETTEXT_POISON=YesPlease compile-time option is gone, and running the
tests with GIT_TEST_GETTEXT_POISON=[YesPlease|] can be toggled on/off
without recompiling.
This allows for conditionally amending tests to test with/without
poison, similar to what 859fdc0c3c ("commit-graph: define
GIT_TEST_COMMIT_GRAPH", 2018-08-29) did for GIT_TEST_COMMIT_GRAPH. Do
some of that, now we e.g. always run the t0205-gettext-poison.sh test.
I did enough there to remove the GETTEXT_POISON prerequisite, but its
inverse C_LOCALE_OUTPUT is still around, and surely some tests using
it can be converted to e.g. always set GIT_TEST_GETTEXT_POISON=.
Notes on the implementation:
* We still compile a dedicated GETTEXT_POISON build in Travis
CI. Perhaps this should be revisited and integrated into the
"linux-gcc" build, see ae59a4e44f ("travis: run tests with
GIT_TEST_SPLIT_INDEX", 2018-01-07) for prior art in that area. Then
again maybe not, see [2].
* We now skip a test in t0000-basic.sh under
GIT_TEST_GETTEXT_POISON=YesPlease that wasn't skipped before. This
test relies on C locale output, but due to an edge case in how the
previous implementation of GETTEXT_POISON worked (reading it from
GIT-BUILD-OPTIONS) wasn't enabling poison correctly. Now it does,
and needs to be skipped.
* The getenv() function is not reentrant, so out of paranoia about
code of the form:
printf(_("%s"), getenv("some-env"));
call use_gettext_poison() in our early setup in git_setup_gettext()
so we populate the "poison_requested" variable in a codepath that's
won't suffer from that race condition.
* We error out in the Makefile if you're still saying
GETTEXT_POISON=YesPlease to prompt users to change their
invocation.
* We should not print out poisoned messages during the test
initialization itself to keep it more readable, so the test library
hides the variable if set in $GIT_TEST_GETTEXT_POISON_ORIG during
setup. See [3].
See also [4] for more on the motivation behind this patch, and the
history of the GETTEXT_POISON facility.
1. https://public-inbox.org/git/871s8gd32p.fsf@evledraar.gmail.com/
2. https://public-inbox.org/git/20181102163725.GY30222@szeder.dev/
3. https://public-inbox.org/git/20181022202241.18629-2-szeder.dev@gmail.com/
4. https://public-inbox.org/git/878t2pd6yu.fsf@evledraar.gmail.com/
Signed-off-by: Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Ever since we started using Travis CI, we specified the list of
packages to install in '.travis.yml' via the APT addon. While running
our builds on Travis CI's container-based infrastructure we didn't
have another choice, because that environment didn't support 'sudo',
and thus we didn't have permission to install packages ourselves. With
the switch to the VM-based infrastructure in the previous patch we do
get a working 'sudo', so we can install packages by running 'sudo
apt-get -y install ...' as well.
Let's make use of this and install necessary packages in
'ci/install-dependencies.sh', so all the dependencies (i.e. both
packages and "non-packages" (P4 and Git-LFS)) are handled in the same
file. Install gcc-8 only in the 'linux-gcc' build job; so far it has
been unnecessarily installed in the 'linux-clang' build job as well.
Print the versions of P4 and Git-LFS conditionally, i.e. only when
they have been installed; with this change even the static analysis
and documentation build jobs start using 'ci/install-dependencies.sh'
to install packages, and neither of these two build jobs depend on and
thus install those.
This change will presumably be beneficial for the upcoming Azure
Pipelines integration [1]: preliminary versions of that patch series
run a couple of 'apt-get' commands to install the necessary packages
before running 'ci/install-dependencies.sh', but with this patch it
will be sufficient to run only 'ci/install-dependencies.sh'.
[1] https://public-inbox.org/git/1a22efe849d6da79f2c639c62a1483361a130238.1539598316.git.gitgitgadget@gmail.com/
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The commit-graph and multi-pack-index features introduce optional
data structures that are not required for normal Git operations.
It is important to run the normal test suite without them enabled,
but it is helpful to also run the test suite using them.
Our continuous integration scripts include a second test stage that
runs with optional GIT_TEST_* variables enabled. Add the following
two variables to that stage:
GIT_TEST_COMMIT_GRAPH
GIT_TEST_MULTI_PACK_INDEX
This will slow down the operation, as we build a commit-graph file
after every 'git commit' operation and build a multi-pack-index
during every 'git repack' operation. However, it is important that
future changes are compatible with these features.
Signed-off-by: Derrick Stolee <dstolee@microsoft.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
In a recent update in 2.18 era, "git pack-objects" started
producing a larger than necessary packfiles by missing
opportunities to use large deltas.
* nd/pack-deltify-regression-fix:
pack-objects: fix performance issues on packing large deltas
The Travis CI scripts were taught to ship back the test data from
failed tests.
* sg/travis-retrieve-trash-upon-failure:
travis-ci: include the trash directories of failed tests in the trace log
The trash directory of a failed test might contain invaluable
information about the cause of the failure, but we have no access to
the trash directories of Travis CI build jobs. The only feedback we
get from there is the build job's trace log, so...
Modify 'ci/print-test-failures.sh' to create a tar.gz archive of the
trash directory of each failed test, encode that archive with base64,
and print the resulting block of ASCII text, so it gets embedded in
the trace log. Furthermore, run tests with '--immediate' to
faithfully preserve the failed state.
Extracting the trash directories from the trace log turned out to be a
bit of a hassle, partly because of the size of these logs (usually
resulting in several hundreds or even thousands of lines of
base64-encoded text), and partly because these logs have CRLF, CRCRLF
and occasionally even CRCRCRLF line endings, which cause 'base64 -d'
from coreutils to complain about "invalid input". For convenience add
a small script 'ci/util/extract-trash-dirs.sh', which will extract and
unpack all base64-encoded trash directories embedded in the log fed to
its standard input, and include an example command to be copy-pasted
into a terminal to do it all at the end of the failure report.
A few of our tests create sizeable trash directories, so limit the
size of each included base64-encoded block, let's say, to 1MB. And
just in case something fundamental gets broken and a lot of tests fail
at once, don't include trash directories when the combined size of the
included base64-encoded blocks would exceed 1MB.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Coccinelle's and in turn 'make coccicheck's exit code only indicates
that Coccinelle managed to finish its analysis without any errors
(e.g. no unknown --options, no missing files, no syntax errors in the
semantic patches, etc.), but it doesn't indicate whether it found any
undesired code patterns to transform or not. To find out the latter,
one has to look closer at 'make coccicheck's standard output and look
for lines like:
SPATCH result: contrib/coccinelle/<something>.cocci.patch
And this only indicates that there is something to transform, but to
see what the suggested transformations are one has to actually look
into those '*.cocci.patch' files.
This makes the automated static analysis build job on Travis CI not
particularly useful, because it neither draws our attention to
Coccinelle's findings, nor shows the actual findings. Consequently,
new topics introducing undesired code patterns graduated to master
on several occasions without anyone noticing.
The only way to draw attention in such an automated setting is to fail
the build job. Therefore, modify the 'ci/run-static-analysis.sh'
build script to check all the resulting '*.cocci.patch' files, and
fail the build job if any of them turns out to be not empty. Include
those files' contents, i.e. Coccinelle's suggested transformations, in
the build job's trace log, so we'll know why it failed.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Currently the static analysis build job runs Coccinelle using a single
'make' job. Using two parallel jobs cuts down the build job's run
time from around 10-12mins to 6-7mins, sometimes even under 6mins
(there is quite large variation between build job runtimes). More
than two parallel jobs don't seem to bring further runtime benefits.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Let's start with some background about oe_delta_size() and
oe_set_delta_size(). If you already know, skip the next paragraph.
These two are added in 0aca34e826 (pack-objects: shrink delta_size
field in struct object_entry - 2018-04-14) to help reduce 'struct
object_entry' size. The delta size field in this struct is reduced to
only contain max 1MB. So if any new delta is produced and larger than
1MB, it's dropped because we can't really save such a large size
anywhere. Fallback is provided in case existing packfiles already have
large deltas, then we can retrieve it from the pack.
While this should help small machines repacking large repos without
large deltas (i.e. less memory pressure), dropping large deltas during
the delta selection process could end up with worse pack files. And if
existing packfiles already have >1MB delta and pack-objects is
instructed to not reuse deltas, all of them will be dropped on the
floor, and the resulting pack would be definitely bigger.
There is also a regression in terms of CPU/IO if we have large on-disk
deltas because fallback code needs to parse the pack every time the
delta size is needed and just access to the mmap'd pack data is enough
for extra page faults when memory is under pressure.
Both of these issues were reported on the mailing list. Here's some
numbers for comparison.
Version Pack (MB) MaxRSS(kB) Time (s)
------- --------- ---------- --------
2.17.0 5498 43513628 2494.85
2.18.0 10531 40449596 4168.94
This patch provides a better fallback that is
- cheaper in terms of cpu and io because we won't have to read
existing pack files as much
- better in terms of pack size because the pack heuristics is back to
2.17.0 time, we do not drop large deltas at all
If we encounter any delta (on-disk or created during try_delta phase)
that is larger than the 1MB limit, we stop using delta_size_ field for
this because it can't contain such size anyway. A new array of delta
size is dynamically allocated and can hold all the deltas that 2.17.0
can. This array only contains delta sizes that delta_size_ can't
contain.
With this, we do not have to drop deltas in try_delta() anymore. Of
course the downside is we use slightly more memory, even compared to
2.17.0. But since this is considered an uncommon case, a bit more
memory consumption should not be a problem.
Delta size limit is also raised from 1MB to 16MB to better cover
common case and avoid that extra memory consumption (99.999% deltas in
this reported repo are under 12MB; Jeff noted binary artifacts topped
out at about 3MB in some other private repos). Other fields are
shuffled around to keep this struct packed tight. We don't use more
memory in common case even with this limit update.
A note about thread synchronization. Since this code can be run in
parallel during delta searching phase, we need a mutex. The realloc
part in packlist_alloc() is not protected because it only happens
during the object counting phase, which is always single-threaded.
Access to e->delta_size_ (and by extension
pack->delta_size[e - pack->objects]) is unprotected as before, the
thread scheduler in pack-objects must make sure "e" is never updated
by two different threads.
The area under the new lock is as small as possible, avoiding locking
at all in common case, since lock contention with high thread count
could be expensive (most blobs are small enough that delta compute
time is short and we end up taking the lock very often). The previous
attempt to always hold a lock in oe_delta_size() and
oe_set_delta_size() increases execution time by 33% when repacking
linux.git with with 40 threads.
Reported-by: Elijah Newren <newren@gmail.com>
Helped-by: Elijah Newren <newren@gmail.com>
Helped-by: Jeff King <peff@peff.net>
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Developer support. Use newer GCC on one of the builds done at
TravisCI.org to get more warnings and errors diagnosed.
* nd/travis-gcc-8:
travis-ci: run gcc-8 on linux-gcc jobs
Switch from gcc-4.8 to gcc-8. Newer compilers come with more warning
checks (usually in -Wextra). Since -Wextra is enabled in developer
mode (which is also enabled in travis), this lets travis report more
warnings before other people do it.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some recent optimizations have been added to pack-objects to reduce
memory usage and some code paths are split into two: one for common
use cases and one for rare ones. Make sure the rare cases are tested
with Travis since it requires manual test configuration that is
unlikely to be done by developers.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
While at there, document about this special mode when running the test
suite.
Signed-off-by: Nguyễn Thái Ngọc Duy <pclouds@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Running test scripts under -x option of the shell is often not a
useful way to debug them, because the error messages from the
commands tests try to capture and inspect are contaminated by the
tracing output by the shell. An earlier work done to make it more
pleasant to run tests under -x with recent versions of bash is
extended to cover posix shells that do not support BASH_XTRACEFD.
* sg/test-x:
travis-ci: run tests with '-x' tracing
t/README: add a note about don't saving stderr of compound commands
t1510-repo-setup: mark as untraceable with '-x'
t9903-bash-prompt: don't check the stderr of __git_ps1()
t5570-git-daemon: don't check the stderr of a subshell
t5526: use $TRASH_DIRECTORY to specify the path of GIT_TRACE log file
t5500-fetch-pack: don't check the stderr of a subshell
t3030-merge-recursive: don't check the stderr of a subshell
t1507-rev-parse-upstream: don't check the stderr of a shell function
t: add means to disable '-x' tracing for individual test scripts
t: prevent '-x' tracing from interfering with test helpers' stderr
Build the executable in 'script' phase in Travis CI integration, to
follow the established practice, rather than during 'before_script'
phase. This allows the CI categorize the failures better ('failed'
is project's fault, 'errored' is build environment's).
* sg/travis-build-during-script-phase:
travis-ci: build Git during the 'script' phase
Now that the test suite runs successfully with '-x' tracing even with
/bin/sh, enable it on Travis CI in order to
- get more information about test failures, and
- catch constructs breaking '-x' with /bin/sh sneaking into our test
suite.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Travis updates.
* sg/travis-linux32-sanity:
travis-ci: don't fail if user already exists on 32 bit Linux build job
travis-ci: don't run the test suite as root in the 32 bit Linux build
travis-ci: don't repeat the path of the cache directory
travis-ci: use 'set -e' in the 32 bit Linux build job
travis-ci: use 'set -x' for the commands under 'su' in the 32 bit Linux build
The split-index mode had a few corner case bugs fixed.
* tg/split-index-fixes:
travis: run tests with GIT_TEST_SPLIT_INDEX
split-index: don't write cache tree with null oid entries
read-cache: fix reading the shared index for other repos
The 32 bit Linux build job runs in a Docker container, which lends
itself to running and debugging locally, too. Especially during
debugging one usually doesn't want to start with a fresh container
every time, to save time spent on installing a bunch of dependencies.
However, that doesn't work quite smootly, because the script running
in the container always creates a new user, which then must be removed
every time before subsequent executions, or the build script fails.
Make this process more convenient and don't try to create that user if
it already exists and has the right user ID in the container, so
developers don't have to bother with running a 'userdel' each time
before they run the build script.
The build job on Travis CI always starts with a fresh Docker
container, so this change doesn't make a difference there.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Travis CI runs the 32 bit Linux build job in a Docker container, where
all commands are executed as root by default. Therefore, ever since
we added this build job in 88dedd5e7 (Travis: also test on 32-bit
Linux, 2017-03-05), we have a bit of code to create a user in the
container matching the ID of the host user and then to run the test
suite as this user. Matching the host user ID is important, because
otherwise the host user would have no access to any files written by
processes running in the container, notably the logs of failed tests
couldn't be included in the build job's trace log.
Alas, this piece of code never worked, because it sets the variable
holding the user name ($CI_USER) in a subshell, meaning it doesn't
have any effect by the time we get to the point to actually use the
variable to switch users with 'su'. So all this time we were running
the test suite as root.
Reorganize that piece of code in 'ci/run-linux32-build.sh' a bit to
avoid that problematic subshell and to ensure that we switch to the
right user. Furthermore, make the script's optional host user ID
option mandatory, so running the build accidentally as root will
become harder when debugging locally. If someone really wants to run
the test suite as root, whatever the reasons might be, it'll still be
possible to do so by explicitly passing '0' as host user ID.
Finally, one last catch: since commit 7e72cfcee (travis-ci: save prove
state for the 32 bit Linux build, 2017-12-27) the 'prove' test harness
has been writing its state to the Travis CI cache directory from
within the Docker container while running as root. After this patch
'prove' will run as a regular user, so in future build jobs it won't
be able overwrite a previously written, still root-owned state file,
resulting in build job failures. To resolve this we should manually
delete caches containing such root-owned files, but that would be a
hassle. Instead, work this around by changing the owner of the whole
contents of the cache directory to the host user ID.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Some of our 'ci/*' scripts repeat the name or full path of the Travis
CI cache directory, and the following patches will add new places
using that path.
Use a variable to refer to the path of the cache directory instead, so
it's hard-coded only in a single place.
Pay extra attention to the 32 bit Linux build: it runs in a Docker
container, so pass the path of the cache directory from the host to
the container in an environment variable. Note that an environment
variable passed this way is exported inside the container, therefore
its value is directly available in the 'su' snippet even though that
snippet is single quoted. Furthermore, use the variable in the
container only if it's been assigned a non-empty value, to prevent
errors when someone is running or debugging the Docker build locally,
because in that case the variable won't be set as there won't be any
Travis CI cache.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The script 'ci/run-linux32-build.sh' running inside the Docker
container of the 32 bit Linux build job uses an && chain to break the
build if one of the commands fails. This is problematic for two
reasons:
- The && chain is broken, because there is this in the middle:
test -z $HOST_UID || (CI_USER="ci" && useradd -u $HOST_UID $CI_USER) &&
Luckily it is broken in a way that it didn't lead to false
successes. If installing dependencies fails, then the rest of the
first && chain is skipped and execution resumes after the ||
operator. At that point $HOST_UID is still unset, causing
'useradd' to error out with "invalid user ID 'ci'", which in turn
causes the second && chain to abort the script and thus break the
build.
- All other 'ci/*' scripts use 'set -e' to break the build if one of
the commands fails. This inconsistency among these scripts is
asking for trouble: I forgot about the && chain more than once
while working on this patch series.
Enable 'set -e' for the whole script and for the commands executed
under 'su' as well.
While touching every line in the 'su' command block anyway, change
their indentation to use a tab instead of spaces.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Split index mode only has a few dedicated tests, but as the index is
involved in nearly every git operation, this doesn't quite cover all the
ways repositories with split index can break. To use split index mode
throughout the test suite a GIT_TEST_SPLIT_INDEX environment variable
can be set, which makes git split the index at random and thus
excercises the functionality much more thoroughly.
As this is not turned on by default, it is not executed nearly as often
as the test suite is run, so occationally breakages slip through. Try
to counteract that by running the test suite with GIT_TEST_SPLIT_INDEX
mode turned on on travis.
To avoid using too many cycles on travis only run split index mode in
the linux-gcc target only. The Linux build was chosen over the Mac OS
builds because it tends to be much faster to complete.
The linux gcc build was chosen over the linux clang build because the
linux clang build is the fastest build, so it can serve as an early
indicator if something is broken and we want to avoid spending the extra
cycles of running the test suite twice for that.
Helped-by: Lars Schneider <larsxschneider@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>
Ever since we started building and testing Git on Travis CI (522354d70
(Add Travis CI support, 2015-11-27)), we build Git in the
'before_script' phase and run the test suite in the 'script' phase
(except in the later introduced 32 bit Linux and Windows build jobs,
where we build in the 'script' phase').
Contrarily, the Travis CI practice is to build and test in the
'script' phase; indeed Travis CI's default build command for the
'script' phase of C/C++ projects is:
./configure && make && make test
The reason why Travis CI does it this way and why it's a better
approach than ours lies in how unsuccessful build jobs are
categorized. After something went wrong in a build job, its state can
be:
- 'failed', if a command in the 'script' phase returned an error.
This is indicated by a red 'X' on the Travis CI web interface.
- 'errored', if a command in the 'before_install', 'install', or
'before_script' phase returned an error, or the build job exceeded
the time limit. This is shown as a red '!' on the web interface.
This makes it easier, both for humans looking at the Travis CI web
interface and for automated tools querying the Travis CI API, to
decide when an unsuccessful build is our responsibility requiring
human attention, i.e. when a build job 'failed' because of a compiler
error or a test failure, and when it's caused by something beyond our
control and might be fixed by restarting the build job, e.g. when a
build job 'errored' because a dependency couldn't be installed due to
a temporary network error or because the OSX build job exceeded its
time limit.
The drawback of building Git in the 'before_script' phase is that one
has to check the trace log of all 'errored' build jobs, too, to see
what caused the error, as it might have been caused by a compiler
error. This requires additional clicks and page loads on the web
interface and additional complexity and API requests in automated
tools.
Therefore, move building Git from the 'before_script' phase to the
'script' phase, updating the script's name accordingly as well.
'ci/run-builds.sh' now becomes basically empty, remove it. Several of
our build job configurations override our default 'before_script' to
do nothing; with this change our default 'before_script' won't do
anything, either, so remove those overriding directives as well.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Every once in a while our explicit .gitignore files get out of sync
when our build process learns to create new artifacts, like test
helper executables, but the .gitignore files are not updated
accordingly.
Use Travis CI to help catch such issues earlier: check that there are
no untracked files at the end of any build jobs building Git (i.e. the
64 bit Clang and GCC Linux and OSX build jobs, plus the GETTEXT_POISON
and 32 bit Linux build jobs) or its documentation, and fail the build
job if there are any present.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The Clang and GCC 64 bit Linux build jobs download and store the P4
and Git LFS executables under the current directory, which is the
working tree that we are about to build and test. This means that Git
commands like 'status' or 'ls-files' would list these files as
untracked. The next commit is about to make sure that there are no
untracked files present after the build, and the downloaded
executables in the working tree are interfering with those upcoming
checks.
Therefore, let's download P4 and Git LFS in the home directory,
outside of the working tree.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Travis CI dutifully builds and tests each new branch tip, even if its
tree has previously been successfully built and tested. This happens
often enough in contributors' workflows, when a work-in-progress
branch is rebased changing e.g. only commit messages or the order or
number of commits while leaving the resulting code intact, and is then
pushed to a Travis CI-enabled GitHub fork.
This is wasting Travis CI's resources and is sometimes scary-annoying
when the new tip commit with a tree identical to the previous,
successfully tested one is suddenly reported in red, because one of
the OSX build jobs happened to exceed the time limit yet again.
So extend our Travis CI build scripts to skip building commits whose
trees have previously been successfully built and tested. Use the
Travis CI cache feature to keep a record of the object names of trees
that tested successfully, in a plain and simple flat text file, one
line per tree object name. Append the current tree's object name at
the end of every successful build job to this file, along with a bit
of additional info about the build job (commit object name, Travis CI
job number and id). Limit the size of this file to 1000 records, to
prevent it from growing too large for git/git's forever living
integration branches. Check, using a simple grep invocation, in each
build job whether the current commit's tree is already in there, and
skip the build if it is. Include a message in the skipped build job's
trace log, containing the URL to the build job successfully testing
that tree for the first time and instructions on how to force a
re-build. Catch the case when a build job, which successfully built
and tested a particular tree for the first time, is restarted and omit
the URL of the previous build job's trace log, as in this case it's
the same build job and the trace log has just been overwritten.
Note: this won't kick in if two identical trees are on two different
branches, because Travis CI caches are not shared between build jobs
of different branches.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Reviewed-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
It seems that Travis CI creates the cache directory for us anyway,
even when a previous cache doesn't exist for the current build job.
Alas, this behavior is not explicitly documented, therefore we don't
rely on it and create the cache directory ourselves in those build
jobs that read/write cached data (currently only the prove state).
In the following commit we'll start to cache additional data in every
build job, and will access the cache much earlier in the build
process.
Therefore move creating the cache directory to 'ci/lib-travisci.sh' to
make sure that it exists at the very beginning of every build job.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Reviewed-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
To make this info message stand out from the regular build job trace
output.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Reviewed-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
When a build job running the test suite fails, our
'ci/print-test-failures.sh' script scans all 't/test-results/*.exit'
files to find failed tests and prints their verbose output. However,
if a build job were to fail before it ever gets to run the test suite,
then there will be no files to match the above pattern and the shell
will take the pattern literally, resulting in errors like this in the
trace log:
cat: t/test-results/*.exit: No such file or directory
------------------------------------------------------------------------
t/test-results/*.out...
------------------------------------------------------------------------
cat: t/test-results/*.out: No such file or directory
Check upfront and proceed only if there are any such files present.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Reviewed-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
This change follows suit of 6272ed319 (travis-ci: run previously
failed tests first, then slowest to fastest, 2016-01-26), which did
this for the Linux and OSX build jobs. Travis CI build jobs run the
tests parallel, which is sligtly faster when tests are run in slowest
to fastest order, shortening the overall runtime of this build job by
about a minute / 10%.
Note, that the 32 bit Linux build job runs the tests suite in a Docker
container and we have to share the Travis CI cache directory with the
container as a second volume. Otherwise we couldn't use a symlink
pointing to the prove state file in the cache directory, because
that's outside of the directory hierarchy accessible from within the
container.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Reviewed-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
The change in commit 4f2636667 (travis-ci: use 'set -x' in 'ci/*'
scripts for extra tracing output, 2017-12-12) left a couple of rough
edges:
- 'ci/run-linux32-build.sh' is executed in a Docker container and
therefore doesn't source 'ci/lib-travisci.sh', which would enable
tracing executed commands. Enable 'set -x' in this script, too.
- 'ci/print-test-failures.sh' iterates over all the files containing
the exit codes of all the executed test scripts. Since there are
over 800 such files, the loop produces way too much noise with
tracing executed commands enabled, so disable 'set -x' for this
script.
- 'ci/run-windows-build.sh' busily waits in a loop for the result of
the Windows build, producing too much noise with tracing executed
commands enabled as well. Disable 'set -x' for the duration of
that loop.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
While the build logic was embedded in our '.travis.yml', Travis CI
used to produce a nice trace log including all commands executed in
those embedded scriptlets. Since 657343a60 (travis-ci: move Travis CI
code into dedicated scripts, 2017-09-10), however, we only see the
name of the dedicated scripts, but not what those scripts are actually
doing, resulting in a less useful trace log. A patch later in this
series will move setting environment variables from '.travis.yml' to
the 'ci/*' scripts, so not even those will be included in the trace
log.
Use 'set -x' in 'ci/lib-travisci.sh', which is sourced in most other
'ci/*' scripts, so we get trace log about the commands executed in all
of those scripts.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Commit 657343a60 (travis-ci: move Travis CI code into dedicated
scripts, 2017-09-10) converted '.travis.yml's default 'before_install'
scriptlet to the 'ci/install-dependencies.sh' script, and while doing
so moved setting GIT_TEST_HTTPD=YesPlease for the 64-bit GCC and Clang
Linux build jobs to that script. This is wrong for two reasons:
- The purpose of that script is, as its name suggests, to install
dependencies, not to set any environment variables influencing
which tests should be run (though, arguably, this was already an
issue with the original 'before_install' scriptlet).
- Setting the variable has no effect anymore, because that script is
run in a separate shell process, and the variable won't be visible
in any of the other scripts, notably in 'ci/run-tests.sh'
responsible for, well, running the tests.
Luckily, this didn't have a negative effect on our Travis CI build
jobs, because GIT_TEST_HTTPD is a tri-state variable defaulting to
"auto" and a functioning web server was installed in those Linux build
jobs, so the httpd tests were run anyway.
Apparently the httpd tests run just fine without GIT_TEST_HTTPD being
set, therefore we could simply remove this environment variable.
However, if a bug were to creep in to change the Travis CI build
environment to run the tests as root or to not install Apache, then
the httpd tests would be skipped and the build job would still
succeed. We would only notice if someone actually were to look
through the build job's trace log; but who would look at the trace log
of a successful build job?!
Since httpd tests are important, we do want to run them and we want to
be loudly reminded if they can't be run. Therefore, move setting
GIT_TEST_HTTPD=YesPlease for the 64-bit GCC and Clang Linux build jobs
to 'ci/lib-travisci.sh' to ensure that the build job fails when the
httpd tests can't be run. (We could set it in 'ci/run-tests.sh' just
as well, but it's better to keep all environment variables in one
place in 'ci/lib-travisci.sh'.)
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Our '.travis.yml's 'env.global' section sets a bunch of environment
variables for all build jobs, though none of them actually affects all
build jobs. It's convenient for us, and in most cases it works just
fine, because irrelevant environment variables are simply ignored.
However, $GIT_SKIP_TESTS is an exception: it tells the test harness to
skip the two test scripts that are prone to occasional failures on
OSX, but as it's set for all build jobs those tests are not run in any
of the build jobs that are capable to run them reliably, either.
Therefore $GIT_SKIP_TESTS should only be set in the OSX build jobs,
but those build jobs are included in the build matrix implicitly (i.e.
by combining the matrix keys 'os' and 'compiler'), and there is no way
to set an environment variable only for a subset of those implicit
build jobs. (Unless we were to add new scriptlets to '.travis.yml',
which is exactly the opposite direction that we took with commit
657343a60 (travis-ci: move Travis CI code into dedicated scripts,
2017-09-10)).
So move setting $GIT_SKIP_TESTS to 'ci/lib-travisci.sh', where it can
trivially be set only for the OSX build jobs.
Furthermore, move setting all other environment variables from
'.travis.yml' to 'ci/lib-travisci.sh', too, because a couple of
environment variables are already set there, and this way all
environment variables will be set in the same place. All the logic
controlling our builds is already in the 'ci/*' scripts anyway, so
there is really no good reason to keep the environment variables
separately.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
A couple of 'ci/*' scripts are shared between different build jobs:
'ci/lib-travisci.sh', being a common library, is sourced from almost
every script, while 'ci/install-dependencies.sh', 'ci/run-build.sh'
and 'ci/run-tests.sh' are shared between the "regular" GCC and Clang
Linux and OSX build jobs, and the latter two scripts are used in the
GETTEXT_POISON Linux build job as well.
Our builds could benefit from these shared scripts being able to
easily tell which build job they are taking part in. Now, it's
already quite easy to tell apart Linux vs OSX and GCC vs Clang build
jobs, but it gets trickier with all the additional Linux-based build
jobs included explicitly in the build matrix.
Unfortunately, Travis CI doesn't provide much help in this regard.
The closest we've got is the $TRAVIS_JOB_NUMBER variable, the value of
which is two dot-separated integers, where the second integer
indicates a particular build job. While it would be possible to use
that second number to identify the build job in our shared scripts, it
doesn't seem like a good idea to rely on that:
- Though the build job numbering sequence seems to be stable so far,
Travis CI's documentation doesn't explicitly states that it is
indeed stable and will remain so in the future. And even if it
were stable,
- if we were to remove or insert a build job in the middle, then the
job numbers of all subsequent build jobs would change accordingly.
So roll our own means of simple build job identification and introduce
the $jobname environment variable in our builds, setting it in the
environments of the explicitly included jobs in '.travis.yml', while
constructing one in 'ci/lib-travisci.sh' as the combination of the OS
and compiler name for the GCC and Clang Linux and OSX build jobs. Use
$jobname instead of $TRAVIS_OS_NAME in scripts taking different
actions based on the OS and build job (when installing P4 and Git LFS
dependencies and including them in $PATH). The following two patches
will also rely on $jobname.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Linux build jobs on Travis CI skip the P4 and Git LFS tests since
commit 657343a60 (travis-ci: move Travis CI code into dedicated
scripts, 2017-09-10), claiming there are no P4 or Git LFS installed.
The reason is that P4 and Git LFS binaries are not installed to a
directory in the default $PATH, but their directories are prepended to
$PATH. This worked just fine before said commit, because $PATH was
set in a scriptlet embedded in our '.travis.yml', thus its new value
was visible during the rest of the build job. However, after these
embedded scriptlets were moved into dedicated scripts executed in
separate shell processes, any variable set in one of those scripts is
only visible in that single script but not in any of the others. In
this case, 'ci/install-dependencies.sh' downloads P4 and Git LFS and
modifies $PATH, but to no effect, because 'ci/run-tests.sh' only sees
Travis CI's default $PATH.
Move adjusting $PATH to 'ci/lib-travisci.sh', which is sourced in all
other 'ci/' scripts, so all those scripts will see the updated $PATH
value.
Signed-off-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
09f5e97 ("travis-ci: skip a branch build if equal tag is present",
2017-09-17) introduced the "skip_branch_tip_with_tag" function with
a broken string comparison. Fix it!
Reported-by: SZEDER Gábor <szeder.dev@gmail.com>
Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
If we push a branch and a tag pointing to the HEAD of this branch,
then Travis CI would run the build twice. This wastes resources and
slows the testing.
Add a function to detect this situation and skip the build the branch
if appropriate. Invoke this function on every build.
Helped-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Most of the Travis CI commands are in the '.travis.yml'. The yml format
does not support functions and therefore code duplication is necessary
to run commands across all builds.
To fix this, add a library for common CI functions. Move all Travis CI
code into dedicated scripts and make them call the library first.
Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
* ls/travis-relays-for-windows-ci:
travis-ci: retry if Git for Windows CI returns HTTP error 502 or 503
travis-ci: handle Git for Windows CI status "failed" explicitly
Travis CI gained a task to format the documentation with both
AsciiDoc and AsciiDoctor.
* ls/travis-doc-asciidoctor:
travis-ci: check AsciiDoc/AsciiDoctor stderr output
travis-ci: unset compiler for jobs that do not need one
travis-ci: parallelize documentation build
travis-ci: build documentation with AsciiDoc and Asciidoctor
The Git for Windows CI web app sometimes returns HTTP errors of
"502 bad gateway" or "503 service unavailable" [1]. We also need to
check the HTTP content because the GfW web app seems to pass through
(error) results from other Azure calls with HTTP code 200.
Wait a little and retry the request if this happens.
[1] https://docs.microsoft.com/en-in/azure/app-service-web/app-service-web-troubleshoot-http-502-http-503
Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
Git for Windows CI returns "completed: failed" if a build or test
failure happened. This case was processed as "Unhandled status".
Handle the case explicitly.
Signed-off-by: Lars Schneider <larsxschneider@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
