Tests on branches of non-head commits are failing, because the depth of
1 doesn't allow checking them out.
Remove `depth` as was the case before fa44af5cd2,
so that Travis can determine the minimum depth to check out.
fa44af5cd2 travis: Full clone for git subtree check (MarcoFalke)
Pull request description:
Lazy workaround and fix for #12388
Tree-SHA512: ef2843c07bfe00b91a6f7ee551172b54db62a7e0120c734be2fa5e8cb2cb232e34f69239cae3cb319734e321822040ee7dcff00b8f0387a01afa6b3d26cc1ca3
Instead of calling combine_logs from test_runner when the environment variable
PYTHON_DEBUG is set, add a combinedlogslen argument. This makes it easier
to call when running locally, and allows a different length of combined logs
to be printed.
Linters should only be run on pull request builds. Once bad code style
has been merged into master, it's too late. Master and other branches
should not fail to build because of linter warnings.
487aff421 Check subtree consistency in Travis (Pieter Wuille)
e1d0cc23a Improve git-subtree-check.sh (Pieter Wuille)
Pull request description:
Apparently many of our subtrees get modified by PRs in this repository, without getting noticed.
To improve upon this:
* Make git-subtree-check.sh capable of doing a weaker consistency check (that doesn't need access to external repositories), but which should be sufficient to detect unintended changes. It can be fooled by a fake subtree merge commit, but that would hopefully be obvious to reviewers.
* Make Travis invoke this subtree check for each of our subtrees.
Note that Travis is currently expected to fail on this PR, as 2 out of 4 subtrees (`src/secp156k1` and `src/univalue` have been modified directly in master).
Tree-SHA512: 465b680392d3daf38a8c1dda77d6f74b1d1c23324c378774777fb95aa673e119a8f7e3ccc124e41d97b5ac8975f3d79f3015797d2d309666582394364917ec4e
The most recent update replaced the minimal image with a large one for the
'generic' image. Switching back to 'minimal' should reduce dependencies and
maybe speed us up some.
It should also eliminiate the need for aa2e0f09e.
This adds a new CHECK_DOC check that looks for newly introduced trailing
whitespace. Existing trailing whitespace (of which there is plenty!)
will not trigger an error.
This is written in a generic way so that new lint-*.sh scripts can be
added to contrib/devtools/, as I'd like to contribute additional lint
checks in the future.
77aa9e59e test: Check RPC argument mapping (Wladimir J. van der Laan)
Pull request description:
Parse the dispatch tables from the server implementation files, and the conversion table from the client (see #10751).
Perform the following consistency checks:
- Arguments defined in conversion table, must be present in dispatch table. If not, it was probably forgotten to add them to the dispatch table, and they will not work.
- Arguments defined in conversion table must have the same names as in the dispatch table. If not, they will not work.
- All aliases for an argument must either be present in the conversion table, or not. Anything in between means an oversight and some aliases won't work.
Any of these results in an error.
It also performs a consistency check to see if the same named argument is sometimes converted, and sometimes not. E.g. one RPC call might have a 'verbose' argument that is converted,
another RPC call might have one that is not converted. This is not necessarily wrong, but points at a possible error (as well as makes the API harder to memorize) - so it is emitted as a warning (could upgrade this to error).
This test is added to travis and run when `CHECK_DOC`. Currently fails with the following output:
```
* Checking consistency between dispatch tables and vRPCConvertParams
ERROR: createrawtransaction argument 3 (named optintorbf in vRPCConvertParams) is not defined in dispatch table
ERROR: getblock argument ['verbosity', 'verbose'] has conflicts in vRPCConvertParams conversion specifier [True, False]
WARNING: conversion mismatch for argument named verbose ([('getblock', False), ('getblockheader', True), ('getmempoolancestors', True), ('getmempooldescendants', True), ('getrawmempool', True), ('getrawtransaction', True)])
```
- ~#10698 fixes the first ERROR~
- #10747 fixes the second ERROR, as well as the WARNING
Update: #10698 was merged, leaving:
```
* Checking consistency between dispatch tables and vRPCConvertParams
ERROR: getblock argument ['verbosity', 'verbose'] has conflicts in vRPCConvertParams conversion specifier [True, False]
WARNING: conversion mismatch for argument named verbose ([('getblock', False), ('getblockheader', True), ('getmempoolancestors', True), ('getmempooldescendants', True), ('getrawmempool', True), ('getrawtransaction', True)])
```
Tree-SHA512: feabebfbeda5d4613b2b9d5265aa6bde4e1a0235297ffd48fa415ad7edc531d9ed7913fe76d191ac60d481a915a326f216bc93de3c671e45e1d14e97d07dea7a
fa4fad9d8 travis: Revert default datadir check (MarcoFalke)
Pull request description:
This still breaks builds. The default datadir check needs more work to stay in master.
Tree-SHA512: b930df4ff8cb76089030aedb2ebbbc1ecc26f9e3d1aba29c78f9f03d7611c6ecaa27afb15ee643690643439d78b98aef9c0c5b2657c7f2fe0c28a3076d6913c2
6d2aac8bb Travis: Test build against system libs (& Qt4) (Luke Dashjr)
8d82e1336 Travis: Remove bc tool from dependencies (Luke Dashjr)
Pull request description:
Also removes now-unnecessary `bc` dependency from other builds.
Tree-SHA512: 815215994eeba0acf27774f57cf3a0bf77bbc22834d3242a227e0d90b5948a05f8b5ef846eb384e3ee575bec60880ae215ccc3882f13b60004a62549d3b3a28f
Parse the dispatch tables from the server implementation files,
and the conversion table from the client.
Perform the following consistency checks:
- Arguments defined in conversion table, must be present in dispatch
table. If not, it was probably forgotten to add them to the
dispatch table, and they will not work.
- Arguments defined in conversion table must have the same names as
in the dispatch table. If not, they will not work.
- All aliases for an argument must either be present in the
conversion table, or not. Anything in between means an oversight
and some aliases won't work.
Any of these results in an error.
It also performs a consistency check to see if the same
named argument is sometimes converted, and sometimes not. E.g.
one RPC call might have a 'verbose' argument that is converted,
another RPC call might have one that is not converted. This is not
necessarily wrong, but points at a possible error (as well as
makes the API harder to memorize) - so it is emitted as a warning
(could upgrade this to error).
4f92b5f Run Qt wallet tests on travis (Russell Yanofsky)
Pull request description:
Currently these test failures are not caught by travis leading to bugs like:
https://github.com/bitcoin/bitcoin/pull/10506
Tree-SHA512: db1e4ff5b17bcd6fd000a3d21aa74f6b7e4c194e0663c1896a100612671910a7cdadd896b59642420ea016598895b54a8468914847ebefef105a3c47c311d4b2
9576b01 Enable xvfb in travis to allow running test_bitcoin-qt (Russell Yanofsky)
9e6817e Add new test_bitcoin-qt static library dependencies (Russell Yanofsky)
2754ef1 Add simple qt wallet test sending a transaction (Russell Yanofsky)
b61b34c Add braces to if statements in Qt test_main (Russell Yanofsky)
cc9503c Make qt test compatible with TestChain100Setup framework (Russell Yanofsky)
91e3035 Make test_bitcoin.cpp compatible with Qt Test framework (Russell Yanofsky)
Tree-SHA512: da491181848b8c39138e997ae5ff2df0b16eef2d9cdd0a965229b1a28d4fa862d5f1ef314a1736e5050e88858f329124d15c689659fc6e50fefde769ba24e523
Now that caches are distinct (https://github.com/travis-ci/travis-ci/issues/4393),
we can use the Travis minimal image.
The minimal image should take less time to setup and lead to quicker builds.
Also addressed while I'm in here:
- No need to delete the broken google-chrome repo in the minimal image
- Set the hostname to work-around an openjdk bug
- Remove the non-functional apt-cache option
- Remove useless message at completion
- Install jre where the java tests are run
This was doing more harm than good. The original intention was to speed up
builds, since a PR's ccache results will be thrown away anyway.
However, each PR maintains its own cache, so disabling writes means that
subsequent pushes don't benefit from the fresh cache. This is significant when
(for example) many headers are touched in a PR, then the PR is updated. With
this change, the updated PR will take advantage of the cache generated during
the PR's previous build.
149641e Travis: Use Blue Box VMs for IPv6 loopback support (Luke Dashjr)
c01f08d Bugfix: depends/Travis: Use --location (follow redirects) and --fail [on HTTP error response] with curl (Luke Dashjr)
5d1148c Travis: Use curl rather than wget for Mac SDK (Luke Dashjr)
1ecbb3b depends: Use curl for fetching on Linux (Luke Dashjr)
1) created rpc-tests.py
2) deleted rpc-tests.sh
3) travis.yml points to rpc-tests.py
4) Modified Makefile.am
5) Updated README.md
6) Added tests_config.py and deleted tests-config.sh
7) Modified configure.ac with script to set correct path in tests_config.py
For Gitian releases:
- Windows builds remain unchanged. libstdc++ was already linked statically.
- OSX builds remain unchanged. libstdc++ is tied to the SDK and not worth
messing with.
- Linux builds now statically link libstdc++.
For Travis:
- Match the previous behavior by adding --enable-reduce-exports as
necessary.
- Use static libstdc++ for the full Linux build.
This is a long chain of errors, and there are likely other changes that could
be made to cope in other places along that chain.
If depends don't build successfully, don't bother trying again for the sake of
better logging. That's likely to hurt more than help. In this case, qt build
failed, and on the second attempt, it appeared to be successful. However, due
to a bad object from an internal gcc error on the first build, the resulting
lib was unusable. This caused bitcoin-qt to not be built, and tests and
packaging which expected bitcoin-qt to be there failed.
The root cause:
Mingw is especially crashy when using -jX, likely compounded by low-memory
environments. I've seen multiple problems with this combo in Gitian as well.
In this case:
i686-w64-mingw32-g++: internal compiler error: Killed (program cc1plus)
...
make[3]: *** [.obj/release/qdrawhelper.o] Error 4
The workaround:
Bump Travis down to using -j2 by default. Additionaly, enable --with-gui for
the windows builds. This will cause configure to fail if qt is not working
while also testing the config flag.
Other failures which may be worth revisiting separately:
- If a depends package fails, maybe remove the workdir so that it doesn't taint
subsequent runs
- See if there's anything repeatable about the ICE when building qt
For the all-off build, enable the wallet and debug. This ensures that debug
options will catch wallet problems as well.
In order to make sure the no-wallet path is still tested, disable the wallet
in the other x86_64 build.
