- Make previous behavior non-conditional
(already default and flag uses are removed)
- Use relative paths instead of package uris as module names
under --debugging-module-names flag, false by default.
(used as server paths in the browser, need to match relative
paths in source maps)
- Pass packageConfig to Javascript bundle API
- Add tests
Closes: https://github.com/dart-lang/sdk/issues/49667
Change-Id: I57c2e2629ba5dd756cce9a8ec6ae891fb092e1d9
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/255826
Commit-Queue: Anna Gringauze <annagrin@google.com>
Reviewed-by: Nicholas Shahan <nshahan@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Update the JavaScript bundler to maintain the state necessary for
the alternative invalidation strategy. This will provide the
javascript compiler with a single library component, so this must
handle re-combining this library with any SCC as well as keeping
the old summaries around.
Revived from https://dart-review.googlesource.com/c/sdk/+/196840
Change-Id: I1dd59d813172ac68ffcb189fcf3904c844828773
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/251960
Reviewed-by: Nicholas Shahan <nshahan@google.com>
When launched as a snapshot, the initial compile is slower than the existing pub approach because of a process start time of around 500 ms. When launched as a compiled executable, initial compile times are the same as the existing pub approach. Once launched, experimental results show times to produce a kernel file of at worst 1.5-2x faster than pub's solution and at best 10x faster than pub's solution. The typical workflow of making changes and recompiling results in an average of a 5x speedup with respect to pub's implementation.
Because compiler instances use a lot of memory, there is a limit on the number of active compilers that the resident server will keep alive, and will actively bring instances down when this limit is exceeded. If the user was previously compiling a given project during the lifespan of a ResidentFrontendServer and its compiler is taken down between requests, a new compiler instance will be allocated for the request. Performance is still between 1.5x-5x faster for this case when compared to pub's compile times.
Change-Id: If9ee1ecc71d660d34faf23381c764dc11d6a5902
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/252001
Reviewed-by: Siva Annamalai <asiva@google.com>
Reviewed-by: Jake Macdonald <jakemac@google.com>
Commit-Queue: Michael Richards <msrichards@google.com>
Reviewed-by: Alexander Aprelev <aam@google.com>
Deletes the outline stubber as it's not necessary on any backend.
DDC should compile the entire sources and outline dill in one step.
dart2wasm operates similarly, and so only needs the modular transformer.
dart2js moves the erasure to a global transform.
Also, this CL reverts now unnecessary plumbing that was needed for the
outline stubber.
Change-Id: Ic085c4fad5a6bdfc7d6916f7fa575c6ef9b20110
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/253000
Reviewed-by: Joshua Litt <joshualitt@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Commit-Queue: Srujan Gaddam <srujzs@google.com>
Fixes b/235393918
@staticInterop replaces factories with a new named node, a static
method. In order to persist this transformation in modular
compilation, this needs to be done to outlines that can then be
consumed by the source library. In order to allow erasure at the
time of 'performOutlineTransformations', coreTypes is added to that
API.
Change-Id: I90d17fff8bbe143982fcd12cfb06dc3e8d58781a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/247928
Commit-Queue: Srujan Gaddam <srujzs@google.com>
Reviewed-by: Joshua Litt <joshualitt@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
This change includes kernel-related changes:
* --compact-async option is added to kernel compilers (front-end
server and gen_kernel). This option disables desugaring of async
and await on kernel AST. Note that 'await for' is still desugared.
* File offset of the 'await' is now written for AwaitExpression nodes
in the kernel binaries (will be used for async stack traces).
* Async/async*/sync* functions and AwaitExpression nodes are supported
in TFA.
Design doc: go/compact-async-await.
TEST=ci
Issue: https://github.com/dart-lang/sdk/issues/48378
Change-Id: I4233086b7434bc48347f4220645b0be5f9133456
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/241842
Reviewed-by: Slava Egorov <vegorov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Otherwise it fails internally, because the package is not null safe
in google3, and cannot be made because it has dependencies
(compiler and dev_compiler) that are not null safe. But the binary
protocol only talks to null safe code, so can be upgraded to the
newer language version.
Bug: https://buganizer.corp.google.com/issues/229124206
Change-Id: I215799de7d40a673c4c11d4d3a9e1c936b979df0
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/241143
Reviewed-by: Jake Macdonald <jakemac@google.com>
Commit-Queue: Konstantin Shcheglov <scheglov@google.com>
- removes the `loadMacro` api entirely
- adds new apis to MultiMacroExecutor
- registerExecutorFactory
- unregisterExecutorFactory
- libraryIsRegistered
- Removes the MacroClassIdentifier, instantiateMacro takes a library uri and class name
- Drop precompiledMacroUris from CompilerOptions, only have a macroExecutor now
Change-Id: Ic33933b34dbfd637da3d841ff496582f688dcd87
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/239466
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Konstantin Shcheglov <scheglov@google.com>
Commit-Queue: Jake Macdonald <jakemac@google.com>
This moves macro implementation into its own subfolder and refactors
the IdentifierImpl implementation to be split into identifier based on
TypeDeclarationBuidler, TypeBuilder, MemberBuilder and
FormalParameterBuilder.
As part of the refactoring, the resolveTypeAnnotation implementation was
rewritten to use DartType instead of TypeBuilder.
Change-Id: Ie59160f4252627849606169b0210aaa83a27d60c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/236880
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
Commit-Queue: Johnni Winther <johnniwinther@google.com>
This is used by the modular compilers, so it unblocks more "real" blaze integration, and allows the language repo to create examples that you can actually run.
Note that this also actually enables macros when the experiment flag is passed in to this entrypoint.
Change-Id: Ia5ce82621a63363a17d16599fead5124ddf65faf
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/236340
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Commit-Queue: Jake Macdonald <jakemac@google.com>
Use `hasScheme` in place of comparing against the empty string, and
`isScheme` to compare against all other schemes.
TEST=No behavior changes.
Change-Id: Ifc9fd13c6cf37933ebd4a754c4b500dedbcb291b
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231185
Reviewed-by: Kevin Moore <kevmoo@google.com>
Commit-Queue: Nate Bosch <nbosch@google.com>
This change adds --source option to kernel compilers (front-end server
and gen_kernel). This option instructs compiler to include extra source
files into compilation even if they are not referenced from main
library.
TEST=Added test cases to pkg/frontend_server/test/frontend_server_test.dart
and pkg/vm/test/kernel_front_end_test.dart
Fixes https://github.com/dart-lang/sdk/issues/48144
Change-Id: If3d71538751e9ccfa8c82a5685d810cf811e021c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/231334
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
This CL changes the way dart: libraries are considered supported when
used in conditional imports or bool.fromEnvironment constant using
the "dart.library.*" values.
Library nodes now has an isUnsupported flag which is set according to
the "supported" property in the libraries specification. Furthermore
the Target now supplies a DartLibrarySupport interface that allows
targets to override whether dart: libraries are unsupported.
This allows the JIT/AOT to use the same platform file but still
consider dart:mirrors unsupported in AOT mode, and dart2js to consider
the internal library `dart:_dart2js_runtime_metrics` supported.
Furthermore, the internal handling is changed so that condition imports
and bool.fromEnvironments constants are computed through the same logic
for "dart.library.*" values, avoiding the need for passing these values
through the environment.
TEST=pkg/front_end/testcases/general/supported_libraries/main
Closes https://github.com/dart-lang/sdk/issues/48057
Closes https://github.com/dart-lang/sdk/issues/47814
Closes https://github.com/dart-lang/sdk/issues/47243
Closes https://github.com/dart-lang/sdk/issues/32657
Closes https://github.com/dart-lang/sdk/issues/36460
Change-Id: Ie8f8dff99167de64ced51b71d89918bf0f3bbd13
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/227020
Reviewed-by: Sigmund Cherem <sigmund@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Chloe Stefantsova <cstefantsova@google.com>
This change introduces Rapid Type Analysis (RTA) and uses it to
calculate initial set of allocated classes for Type Flow Analysis (TFA).
As a result, TFA converges much faster and AOT compilation time
improves.
RTA is less precise than TFA, so the set of allocated classes is
larger compared to the one calculated by TFA. However, it has only
marginal effect on the size of resulting AOT snapshots.
Time of AOT compilation step 2 on a large Flutter application
118.652s -> 59.907s (-49.5% / improved by a factor of 1.98x)
Snapshot size on armv8: -0.13%
Flutter gallery snapshot size in release and release-sizeopt modes
armv7 +0.19%, armv8 +0.2%
Just in case, RTA can be disabled using --no-rta option.
TEST=ci
Issue: https://github.com/dart-lang/sdk/issues/42442
Issue: b/154155290
Change-Id: Iffbdabe7d486cad2e138f7592bffcb70474ddc34
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/222500
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Slava Egorov <vegorov@google.com>
This CL adds a flag, `assume-initialize-from-dill-up-to-date`, to the
`frontend_server` which will skip the check and potential invalidation
of files from the dill we're initializing from.
Normally the sources included in the dill is used to figure out what
files that was used to produce the dill has changed and thus needs to
be invalidated after we've initialized from it. If we didn't do that
we'd be in a bad state where some changes would not be discovered.
In some circumstances, though, we might know that the sources haven't
changed and the check is thus not needed. This could for instance be
if we're initializing from a dill that was just build by another build
system.
Change-Id: Id8244125fd79465be89ac81d603cd624339fbc0e
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/219040
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Commit-Queue: Jens Johansen <jensj@google.com>
This CL removes async markers where await wasn't used.
In a few cases this also meant the ability to make the procedure not
return a future at all.
Also - at least on the VM with my benchmark run on my PC - returning
new Future.value(whatever) in a non-async procedure is faster than
returning whatever in an async procedure.
TEST=CI and existing tests.
Change-Id: I4ba888313c87de76bb0ca02c19eb1ab45f168a85
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/213480
Commit-Queue: Jens Johansen <jensj@google.com>
Reviewed-by: Johnni Winther <johnniwinther@google.com>
Johnni Winther