The `__future__` we relied on is now, where the 3 specific things are
all included [since Python 3.0](https://docs.python.org/3/library/__future__.html):
* absolute_import
* print_function
* unicode_literals
* division
These import statements are no-ops and are no longer necessary.
This will make sure that compiler developers are using the new driver when they build the compiler locally and use it.
- Adds a new build-script product category: before_build_script_impl for products we wish to build before the impl products.
- Adds a new EarlySwiftDriver product to that category, which gets built with the host toolchain.
- Adds an escape hatch: --skip-early-swift-driver
- Adjusts the swift CMake configuration with an additional step: swift_create_early_driver_symlinks which (if one was built) creates symlinks in the swift build bin directory to the EarlySwiftDriver swift-driver and swift-help executables.
- Adds a new test subset : only_early_swiftdriver, which will get built into a corresponding CMake test target: check-swift-only_early_swiftdriver-* which runs a small subset of driver-related tests against the Early SwiftDriver.
- This subset is run always when the compiler itself is tested (--test is specified)
- With an escape disable-switch: --skip-test-early-swift-driver
- All tests outside of only_early_swiftdriver are forced to run using the legacy C++ driver (to ensure it gets tested, still).
NOTE: SwiftDriver product (no 'Early') is still the main product used to build the driver for toolchain installation into and for executing the product's own tests. This change does not affect that.
Added a pass through of lit's `--filter` option to utils/run-test, to
allow running a subset of tests, for example `./utils/run-test
--filter=irgen` runs tests in `test/IRGen` and any other test with
"irgen" in its name.
For argparse, the type value of each argument must be a callable. The previous type was simply pointing to the class that implements the callable, and not to an instance of it.
This should recover the previous functionality for the --target and the --param arguments.
Seems that this is already applied in the build-script, so no changes are necessary there.
Since 1b4b7f9e17 has been committed, `update-checkout` clones "llvm-project" repo but does not clone "llvm" repo itself alone by default.
That may cause `run-test` failure because "lit.py" cannot be found.
This commit fixes the issue.
The unified LLVM builds use a slightly different structure than the
Swift standalone builds. Modify the run-test utility tool to allow using
it in both standalone and unified builds.
The manipulation of host-test and skip-android-host was a little
different than the equivalent skip-ios-host and similar variables. These
changes make them closer and allows executing only the compiler tests,
but skip the test that need an Android device to run.
- Disables the upload command of the tests if the subset is the
non-executable tests. The non-executable test do not need to be
uploaded, and in the case of Android, a device doesn't need to be
connected, so trying to connect to one will fail.
- Fix a problem where the swift_interpreter feature was removed without
first checking if it was really added.
- Only enable the host tests (the compiler tests) in the Android CI
preset (there's no device attached to that server, but currently only
the Linux tests were being executed, which doesn't make a lot of
sense).
- Move the decision about which platform support device/host tests into
the platform themselves, which allows Android to have device/host
tests. Also modify a little bit the logic around enabling/disabling
the test suite to allow running only the host tests of a platform.
- Fix the suffix name for the target of non-executable tests in a couple
of places.
Stress tests are, by definition, stressful. They intentionally burn a
lot of resources by using randomness to hopefully surface state machine
bugs. Additionally, many stress tests are multi-threaded these days and
they may attempt to use all of the available CPUs to better uncover
bugs. In isolation, this is not a problem, but the test suite as a whole
assumes that individual tests are single threaded and therefore running
multiple stress tests at once can quickly spiral out of control.
This change formalizes stress tests and then treats them like long
tests, i.e. tested via 'check-swift-all' and otherwise opt-in.
Finally, with this change, the CI build bots might need to change if
they are still only testing 'validation' instead of all of the tests.
I see three options:
1) Run all of the tests. -- There are very few long tests left these
days, and the additional costs seems small relative to the cost of
the whole validation test suite before this change.
2) Continue checking 'validation', now sans stress tests.
3) Check 'validation', *then* the stress tests. If the former doesn't
pass, then there is no point in the latter, and by running the stress
tests separately, they stand a better chance of uncovering bugs and
not overwhelming build bot resources.
