These tests show that we handle >> and >>> similarly wrt to allowing
integer operations on double constants with integer values.
https://github.com/dart-lang/sdk/issues/45438 proposes to disallow these,
but that requires a new js evaluation strategy and the issue in 45376
is therefore just a consequence of the current approach.
Change-Id: I703b57d5a6562ff1c1410858b22a96655f59dc4c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/193406
Reviewed-by: Dmitry Stefantsov <dmitryas@google.com>
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Before this CL we could for instance get this result:
```
$ out/ReleaseX64/dart --enable-asserts pkg/front_end/test/fasta/weak_suite.dart -DsemiFuzz=false -DskipVm=true -- weak/general/error_recovery/issue_39058.crash
[...]
1 failed:
weak/general/error_recovery/issue_39058.crash: Pass
```
The reason is, that before it onle rewrote the expectations if there was
exactly one expectation.
`weak/general/error_recovery/issue_39058.crash` for instance (currently)
has two expecations: RuntimeError and SemiFuzzFailure.
All combinations of `-DsemiFuz` and `-DskipVm=true` would thus keep both
of those expectations.
For the combination `-DsemiFuzz=false -DskipVm=true` it actually passes,
though, which doesn't match so it fails saying that the expectations
should be updated.
For any of the other expectations it would be fine because it would get
one of the expected outcomes:
* `-DsemiFuzz=true -DskipVm=true`: Fuzzing is enabled so it would get a
SemiFuzzFailure which matches one of the expectations.
* `-DsemiFuzz=true -DskipVm=false`: Fuzzing is enabled so it would get a
SemiFuzzFailure which matches one of the expectations.
* `-DsemiFuzz=false -DskipVm=false`): It doesn't skip the VM so it gets
a RuntimeError which matches one of the expectations.
This CL fixes the issue by removing the "wrong" expectation(s), leaving
the correct one(s) and adding 'pass' of all other expectation(s) have
been removed.
Change-Id: I8137d16fdfeb322cb76d6af97cafc279222ae205
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/192921
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Commit-Queue: Jens Johansen <jensj@google.com>
This removes the @fields and @=fields canonical name
encodings that allowed for encoding of conflicting members
and didn't support field<->getter/setter conversion
between dills or between outline and full dill.
TEST=existing tests+add aot expectation tests
Change-Id: I119b0c95f90e456356146cdc2d9241de4c1b4fff
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/186680
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
This changes the CFE expectation test suites to use unsound null safety
as default to avoid passing --enable-experiments=no-non-nullable to opt
out old test folders.
This change removes most test folders from the 'strong' tester, since
this now only supports sound null safety, which requires all libraries
to be opted in. Instead, the 'weak' tester now runs all test folders.
Additionally the 'outline' tester is changed to use unsound null safety
so that it call be used on all test folders.
The 'fast_strong' tester is removed since it should just be run locally
and it can be run by passing an option to the other testers.
References to non-existing 'shaker' folder has been removed from
testing.json.
Change-Id: Ibcc306f7b27d06b9637c205238bc408194f1d062
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/184787
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
This CL:
* Makes the standard filesystem more synchronous by default.
This makes reading files for normal compiles faster.
* Adds a more asynchronous version of `exists` and `readAsBytes`.
This potentially allows for reading files faster when in a context
where the files can actually be read in parallel. The client have
to make the choise though.
* Skips `Uri.base.resolveUri` when the uri has a scheme. This makes it
slightly faster in my tests, but does not remove any `.` and `..`.
Having those when having a scheme seems sort of weird though.
If this turns out to be a problem we can add it back.
* Re-orders what the standard filesystem checks in `exists` to assume files
which seems more likely to happen. This - for files - means less checks
and should make checking for existance faster. (This was noticed by Siggi).
Using the same measurement procedure as in
https://dart-review.googlesource.com/c/sdk/+/85442
timing how long fasta spends on *reading* files when compiling dart2js
(when compiling via the VM) I get a good speedup.
Note that the measurements are in microseconds:
Before this CL (run like `out/ReleaseX64/dart pkg/compiler/bin/dart2js.dart 2>&1 | grep "Read file (total)" | tail -n 1` 10 times):
Read file (total): 94678
Read file (total): 94232
Read file (total): 91726
Read file (total): 77472
Read file (total): 95582
Read file (total): 88113
Read file (total): 89245
Read file (total): 91575
Read file (total): 96291
Read file (total): 95730
With this CL (run like `out/ReleaseX64/dart pkg/compiler/bin/dart2js.dart 2>&1 | grep "Read file (total)" | tail -n 1` 10 times):
Read file (total): 68379
Read file (total): 69320
Read file (total): 72930
Read file (total): 69692
Read file (total): 68685
Read file (total): 73548
Read file (total): 64649
Read file (total): 71951
Read file (total): 73486
Read file (total): 70621
Difference at 95.0% confidence
-21138.3 +/- 4193
-23.111% +/- 4.5843%
(Student's t, pooled s = 4462.56)
Furthermore, using an internal benchmark (discussed in an email thread)
I get these numbers (note that the measurements are in milliseconds):
before:
4453
4249
4187
4190
4216
after:
2310
2379
2449
2467
2407
Difference at 95.0% confidence
-1856.6 +/- 131.443
-43.5924% +/- 3.08625%
(Student's t, pooled s = 90.1257)
Using the asynchronous versions of exists and readAsBytes and checking for
existance and reading in parallel on the same internal benchmark it gets to
1137
1185
1067
1089
1189
For completion, comparing to using "raw" `File` (again using the internal benchmark):
sequential standard
2377
2419
2485
2373
2431
sequential io_sync
2279
2182
2237
2312
2324
io_sync faster:
-150.2 +/- 76.3122
-6.21432% +/- 3.15731%
(Student's t, pooled s = 52.3245)
sequential standardasync
4865
4915
4753
5185
4760
sequential io_async
4746
4834
4989
4840
4891
No difference proven at 95.0% confidence
parallel standard
2526
2563
2652
2616
2685
parallel io_sync
2335
2184
2244
2341
2401
io_async faster
Difference at 95.0% confidence
-307.4 +/- 111.037
-11.785% +/- 4.25691%
(Student's t, pooled s = 76.1341)
parallel standardasync
1137
1185
1067
1089
1189
parallel io_async
1105
1088
1145
1048
1113
No difference proven at 95.0% confidence
TEST=Existing tests.
Change-Id: I8ba56ab0768df8672bcdb693782d3f1eec86b683
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/185101
Commit-Queue: Jens Johansen <jensj@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Gary Roumanis <grouma@google.com>
Since the flag is now enabled by default, there should be no mention of it.
There are still some uses in front_end/testcases that are not just removable
(it also uses `no-non-nullable`). There migth be more uses that are not
as easily found as grepping for `--enable-experiment
Removes two VM tests where fixing them meant they were just duplicating
the corresponding non *_2/ tests.
Fixes#44941
TEST= Large number of tests chaged.=(no-)?non-nullable`.
Change-Id: Ief755981ccde9a5482fcdf408c2929c74433a710
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/183688
Commit-Queue: Lasse R.H. Nielsen <lrn@google.com>
Reviewed-by: Nate Bosch <nbosch@google.com>
This is in preparation to migrate package:kernel to null safety.
For the visitor interfaces to support non-nullable return types, the
implementations must avoid using `null` as return value in its base case.
TEST=Refactoring
Change-Id: Ie8fa5d41b99850d9e4abb59634c72920c64128d9
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/183691
Reviewed-by: Jens Johansen <jensj@google.com>
Commit-Queue: Johnni Winther <johnniwinther@google.com>
This is in preparation to migrate package:kernel to null safety.
For the visitor interfaces to support non-nullable return types, the
implementations must avoid using `null` as return value in its base case.
TEST=Refactoring
Change-Id: Ie5e4153f8d3779d94957bb13b3d2d2a942040ff2
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/179760
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
This CL:
* Adds a "reordering fuzzing step", i.e. it tries to sort the fields,
procedures etc in both ascending order and descending order
to verify that there are on crashes and no changes in error vs
no error. This has been tested on http://dartbug.com/34803
before the fix landed.
* Introduces an attempt at "visiting the outline" of the "direct from
parser ast".
Change-Id: Ic3f5a776839030bb48a35eeee8575b37c19be453
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/182504
Commit-Queue: Jens Johansen <jensj@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Currently - i.e. without this CL - (almost) every test run by the strong
test suite verifies the platform libraries as well. That's basically a
waste of time.
This CL makes the suite skip it for the platform.
Runtimes changes (on my computer) as follows:
With -DskipVm=true:
With this CL: ~1:55
Without/before this CL: ~8:30
Without -DskipVm=true:
With this CL: ~4:35
Without/before this CL: ~11:30
Change-Id: I6c46b86935bea29911775cbc0544ea7a0b60b779
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/180562
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Commit-Queue: Jens Johansen <jensj@google.com>
This CL introduces a 'semi-fuzz' option in the strong tests, that is
disabled by default.
The current version just compiles with the incremental compiler
and invalidate one file at a time, seing that we get the same
libaries in the output and that we either get errors in both cases
or no errors in both cases.
Future versions should for instance split libraries into more
libraries or parts etc., and the semiFuzz option should be enabled
by default.
Change-Id: I47f7d0f10798b0f2b19c9f2a02287f9431f67608
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/180145
Commit-Queue: Jens Johansen <jensj@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Emphasize that the operation is going away,
and mark constructor as deprecated.
TEST= Refactoring+deprecation only, covered by existing tests.
Change-Id: I82aa044cd2cf7bf347b624371399f44bda8f4a07
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/173261
Commit-Queue: Lasse R.H. Nielsen <lrn@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Reviewed-by: Nate Bosch <nbosch@google.com>
This CL records which const constructors were evaluated
(i.e. contributes coverage) but doesn't actually
save the coverage into the dill file.
runtimes for constant evaluation:
dart2js: No difference proven at 95.0% confidence
flutter gallery: No difference proven at 95.0% confidence
big internal app: No difference proven at 95.0% confidence
sdk sizes:
vm: no change
dart2js: no change
flutter: no change
compile sizes:
dart2js: no change
flutter gallery: no change
big internal app: no change
Change-Id: I2567446c1b78180d73fd7613bc9e780e1a067f5b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/170435
Commit-Queue: Jens Johansen <jensj@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
This is in preparation to migrate package:kernel to null safety.
For the visitor interfaces to support non-nullable return types, the
implementations must avoid using `null` as return value in its base case.
TEST=Refactoring
Change-Id: I9f9b318982148d844be9826a5f8c88374a9fc402
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/172180
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
Even when non-nullable is enabled by default, enabling the experiment
explicitly should result in the experiment release version (and not
the experiment enabled version) to be used for opting in.
For this change, the semantics of parseExperimentalFlags was change
to _not_ normalize the flags to a full mapping including default values.
For this reason all uses of such maps are renamed to
'explicitExperimentalFlags'.
Closes#43879
Change-Id: I0d0262e68ec1403549abcfd305ae3a4404fe93e3
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/168654
Reviewed-by: Jake Macdonald <jakemac@google.com>
Commit-Queue: Johnni Winther <johnniwinther@google.com>
This changes the reported failure from Crash to something more specific
to the encounter error, like ExpectationFileMismatch. Validation errors
for the .yaml file still produce Crash, and so does a couple of the
checks on Component integrity that were also used outside
incremental_load_from_dill_suite.dart
Change-Id: I48aa8cb25198d24e986b0f354db2b370dc9b8ac8
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/167565
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
This increases our "local" coverage of the constant evaluator.
Run with something like
`pkg/front_end/test/fasta/strong_tester.dart -DupdateExpectations=true -DstressConstantEvaluator=true -DskipVm=true`
there is now almost coverage of everything (or a comment in the code
suggests that the path is probably unreachable).
Note that
`out/ReleaseX64/dart pkg/front_end/test/vm_service_coverage_constant_evaluator.dart pkg/front_end/test/fasta/strong_tester.dart -DupdateExpectations=true -DstressConstantEvaluator=true -DskipVm=true`
has erroneous misses, for instance (but not limited to) asserts.
These are caused by errors in the VM.
A follow-up CL will be created to turn on "stressConstantEvaluator" by
default and print out any sucesses in the expect files.
Change-Id: I837837be4f0487fdbe57c5107e4b3415de189177
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/163060
Commit-Queue: Jens Johansen <jensj@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
This adds support lowering the encoding of top-level/static fields
with initializers as if they were marked as late fields. This ensures
that LateInitialization is thrown if final fields are written to during
initialization.
Closes#42956
Change-Id: I488fdddd87ebd935a0cdaf82a724e9b87d5f91ba
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/160724
Reviewed-by: Dmitry Stefantsov <dmitryas@google.com>
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Rename test.options files and TestOptions to folder.options and FolderOptions,
respectively. Rename link.options files and LinkDependencyOptions to
test.options and TestOptions.
This is to better reflect the current usage of these options.
Change-Id: Ibd3c49fa5bd1c9ce8a9eb5910417c12cd45845d2
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/160723
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
This CL does three things:
* It adds a test that shows how a stub is added if it compiles all
"user libraries" from source, but not when it loads some libraries
from a dill. This is bad and can lead to leaks (and possibly other
things).
* It adds the transformation phase to the tests using the link option
for the linked component. This is needed to avoid a potential crash
(assert error) in the mixin transformation triggered by the new test
added above.
* It adds the possibility to create a non-linked compilation in a
testcase directory with a link.options file by naming the file
"<whatever>.no_link.dart". This way one can have two expect files
close together that shows differences between using linked
dependencies and not without too much duplicated test code.
Change-Id: I170c66fb3e0d4d1263217d64234dc0e88388747d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/156912
Commit-Queue: Jens Johansen <jensj@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
This CL embeds the sdk_nnbd/lib/_internal/allowed_experiments.json
into the CFE and uses this to allow experiments enabled on a
per library basis. The file is embedded through generated code to
avoid reliance on access to the file itself.
A presubmit check is also added to ensure that the json file and
the generated code are in sync.
This work is in preparation for https://github.com/dart-lang/sdk/issues/41538
Change-Id: Ic03c0fc69684d124c685f2fe6e80423a7e751530
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/149064
Commit-Queue: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
Add a kernel transformation which removes redundant casts and reduces
some casts to null checks. This transformation assumes that type
system is sound and subtyping relation is transitive. So in NNBD weak
mode this transformation is always correct. In NNBD strong mode
this transformation is correct only if opted-out libraries are not
allowed.
The following patterns are handled:
If S <: T (this includes S <: T? in weak mode)
S x; x as T => x
If S <: T? in strong mode
S x; x as T => (x == null) ? x as T : x
This kernel transformation is not specific to the VM and can be reused
by other back-ends such as dart2js and DDC.
Fixes https://github.com/dart-lang/sdk/issues/40892
Issue https://github.com/dart-lang/sdk/issues/40893
Change-Id: I466ba96d763d5358011c768400c5225fa7bf5e67
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/139314
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Johnni Winther