Under certain conditions front-end currently generates as{TypeError}
type checks in constant expressions. Support evaluating these type
checks as constant expressions instead of rejecting them.
Change-Id: Id432c4c5062d7bab89c29fef308b42b2cbd64de8
Reviewed-on: https://dart-review.googlesource.com/41761
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This reverts commit 12407c3ea0.
Reason for revert: co19/Language/Expressions/Object_Identity/double_t01 fails in -c dartkp -r dart_precompiled mode.
Original change's description:
> [kernel] Don't always treat doubles as constants
>
> Previously, when the VM parsed a doubles from the dill file it passed
> control over to the constant evaluator which would then case the
> value and canonicalize it.
> That doesn't really seem to make sense when its not used in a const
> context, nor is it done the same way for integers for instance.
>
> This CL changes that and treats doubles the same way as integers.
>
> Below timings are statistics run through "ministat" based on 5 runs
> of each configuration (now and before). The timings is for running
> the dill file with the VM, i.e. the dill file was precompiled via
> fasta.
>
> Program adding 50,000 doubles to a list and prints the last one:
>
> -56.6 +/- 28.2389
> -10.9988% +/- 5.48753%
> (~510 ms -> ~460 ms)
>
> Program adding 1,000,000 doubles to a list and prints the last one:
>
> -3177.2 +/- 263.54
> -31.4643% +/- 2.60987%
> (~10 seconds -> ~7 seconds)
>
> Hello world program:
>
> No difference proven at 95.0% confidence
>
> Change-Id: I7dd748e759b3e433bc806b2ef2505d29f0b109b6
> Reviewed-on: https://dart-review.googlesource.com/42000
> Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
> Commit-Queue: Jens Johansen <jensj@google.com>
TBR=ahe@google.com,vegorov@google.com,jensj@google.com
Change-Id: If8efa9917c009f2df68a3ef00331ab45bf8bf67b
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/41981
Reviewed-by: Jens Johansen <jensj@google.com>
Commit-Queue: Jens Johansen <jensj@google.com>
Previously, when the VM parsed a doubles from the dill file it passed
control over to the constant evaluator which would then case the
value and canonicalize it.
That doesn't really seem to make sense when its not used in a const
context, nor is it done the same way for integers for instance.
This CL changes that and treats doubles the same way as integers.
Below timings are statistics run through "ministat" based on 5 runs
of each configuration (now and before). The timings is for running
the dill file with the VM, i.e. the dill file was precompiled via
fasta.
Program adding 50,000 doubles to a list and prints the last one:
-56.6 +/- 28.2389
-10.9988% +/- 5.48753%
(~510 ms -> ~460 ms)
Program adding 1,000,000 doubles to a list and prints the last one:
-3177.2 +/- 263.54
-31.4643% +/- 2.60987%
(~10 seconds -> ~7 seconds)
Hello world program:
No difference proven at 95.0% confidence
Change-Id: I7dd748e759b3e433bc806b2ef2505d29f0b109b6
Reviewed-on: https://dart-review.googlesource.com/42000
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Commit-Queue: Jens Johansen <jensj@google.com>
The list of recognized methods in the VM includes information about
result type. If there is no known/concrete result cid, the list
contains dynamic. When StaticCallInstr for recognized method is created,
its result type was pulled from the list even if it was dynamic.
This change avoids setting result type to dynamic in such case
and fixes the assertion for conflicting result types from the list of
recognized methods and from global type inference.
Now, if list of recognized methods has dynamic, and inferred type
is available, the latter will be used.
Change-Id: I7da7313f93ca08a142226026a4645124c30cfd8f
Reviewed-on: https://dart-review.googlesource.com/41080
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
This the second part of fixing #32077 - dealing with mints.
Example:
The runtime of a program adding 100,000 different mints to
a list and prints the last one:
Before: ~12.5 seconds
Now: ~1.2 seconds.
Change-Id: Iad6da1749886bab4e4d3941748cdbab51661c2c3
Reviewed-on: https://dart-review.googlesource.com/41161
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Commit-Queue: Jens Johansen <jensj@google.com>
This the first part of fixing #32077 - dealing with doubles.
Example:
The runtime of a program adding 100,000 different doubles to
a list and prints the last one:
Before: ~12 seconds
Now: ~1.3 seconds.
Change-Id: I7f64b8763fdd37dae5d32d3275f8b9c991d4e43a
Reviewed-on: https://dart-review.googlesource.com/40300
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Slightly alter async/await transformation in such a way that
:await_jump_var/:await_ctx_var are set before _awaitHelper(...) is invoked -
just like in the AST transformer used by the legacy pipeline. This guarantees
that recursive invocations will continue from the right "jump" point (instead of
continuing from the current jump point).
Fixes https://github.com/dart-lang/sdk/issues/32098
Change-Id: I8899669ceefe3f2edba6531897834b2c275efada
Reviewed-on: https://dart-review.googlesource.com/40520
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Jens Johansen <jensj@google.com>
This CL corrects handling of overflows in range analysis to
account for wrap-around (which happens with new integer semantics).
* If there is an overflow while doing computations in range
analysis, the resulting range is approximated as full int64 range.
* For symbolic range boundaries 'symbol + offset', offset is checked
to stay within [kMinInt64 - kSmiMin, kMaxInt64 - kSmiMax] in order
to guarantee that overflow doesn't occur.
Issue: https://github.com/dart-lang/sdk/issues/31920
Change-Id: I2c16adbe3597e9b718ed2f6ce7210426fcc9e6a6
Reviewed-on: https://dart-review.googlesource.com/39423
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
These two values are not necessarily in sync: FLAG_use_field_guards can be set to
false while kernel-service is running from an app-jit snapshot with field
guards.
This caused major flakiness on kernel opt-counter-threshold bot.
This was a bug introduced by refactoring in d6eb8c6c27.
Change-Id: Icba874d319c1e96df90cae2d81e0bce3e752808b
Reviewed-on: https://dart-review.googlesource.com/38620
Reviewed-by: Erik Corry <erikcorry@google.com>
This relands 75a9579ea0.
The approach works as follows:
Step 1: Kernel transform. Under the closed world assumption compute the
set of selectors dispatched dynamically, then mark all procedures that don't
match any of those selectors as 'not-dispatched-dynamically'.
Step 2: VM backend. When building IR for a function if this function was
marked as not-dispatched-dynamically then omit type checks for any parameter
that is not marked as generic-covariant-impl, as such arguments are guaranteed
to be checked on the caller side (by front-end).
+------------------------+------------+----------+-----------+--------------+
| benchmark | baseline | current | with opt | improved by |
+------------------------+------------+----------+-----------+--------------+
| stock_layout_iteration | 2366.3786 | 2724.3 | 2562.75 | -5.93% |
| stock_build_iteration | 3824.3 | 4914.8 | 4681 | -4.76% |
+------------------------+------------+----------+-----------+--------------+
* Flutter gallery Instructions size is reduced by 11% (8748720 bytes to 7846368 bytes).
Baseline is at 6196496 bytes.
Alternatively to annotating individual procedures, I considered annotating Program node
with a set of dynamically dispatched selectors. Decoding and passing this information
around proved to be quite cumbersome in the "streaming" world, so I opted for a simpler
approach where all individual procedures are annotated.
Bug: https://github.com/dart-lang/sdk/issues/3179
Change-Id: I2f32a609e3872c74d5ae7bbd97555453aaedf15f
Reviewed-on: https://dart-review.googlesource.com/38125
Commit-Queue: Samir Jindel <sjindel@google.com>
Reviewed-by: Samir Jindel <sjindel@google.com>
This reverts commit 75a9579ea0.
Reason for revert: function_subtype_bound_closure7_test fails in darkp
Original change's description:
> [vm/kernel/aot] Skip unnecessary type checks on parameters of instance methods.
>
> The approach works as follows:
>
> Step 1: Kernel transform. Under the closed world assumption compute the
> set of selectors dispatched dynamically, then mark all procedures that don't
> match any of those selectors as 'not-dispatched-dynamically'.
>
> Step 2: VM backend. When building IR for a function if this function was
> marked as not-dispatched-dynamically then omit type checks for any parameter
> that is not marked as generic-covariant-impl, as such arguments are guaranteed
> to be checked on the caller side (by front-end).
>
> # Performance Impact
>
> +------------------------+------------+----------+-----------+--------------+
> | benchmark | baseline | current | with opt | improved by |
> +------------------------+------------+----------+-----------+--------------+
> | stock_layout_iteration | 2366.3786 | 2724.3 | 2562.75 | -5.93% |
> | stock_build_iteration | 3824.3 | 4914.8 | 4681 | -4.76% |
> +------------------------+------------+----------+-----------+--------------+
>
> * Flutter gallery Instructions size is reduced by 11% (8748720 bytes to 7846368 bytes).
> Baseline is at 6196496 bytes.
>
> # Alternative Implementations
>
> Alternatively to annotating individual procedures, I considered annotating Program node
> with a set of dynamically dispatched selectors. Decoding and passing this information
> around proved to be quite cumbersome in the "streaming" world, so I opted for a simpler
> approach where all individual procedures are annotated.
>
> Change-Id: I363db6d5dd1138fe25917646313c16d0b213c3b4
> Reviewed-on: https://dart-review.googlesource.com/37822
> Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
> Reviewed-by: Alexander Markov <alexmarkov@google.com>
> Reviewed-by: Samir Jindel <sjindel@google.com>
TBR=vegorov@google.com,alexmarkov@google.com,sjindel@google.com
Change-Id: I0fd6bf3e6edfc3e605da2b996f9a0da6c409d01c
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/37802
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
The approach works as follows:
Step 1: Kernel transform. Under the closed world assumption compute the
set of selectors dispatched dynamically, then mark all procedures that don't
match any of those selectors as 'not-dispatched-dynamically'.
Step 2: VM backend. When building IR for a function if this function was
marked as not-dispatched-dynamically then omit type checks for any parameter
that is not marked as generic-covariant-impl, as such arguments are guaranteed
to be checked on the caller side (by front-end).
# Performance Impact
+------------------------+------------+----------+-----------+--------------+
| benchmark | baseline | current | with opt | improved by |
+------------------------+------------+----------+-----------+--------------+
| stock_layout_iteration | 2366.3786 | 2724.3 | 2562.75 | -5.93% |
| stock_build_iteration | 3824.3 | 4914.8 | 4681 | -4.76% |
+------------------------+------------+----------+-----------+--------------+
* Flutter gallery Instructions size is reduced by 11% (8748720 bytes to 7846368 bytes).
Baseline is at 6196496 bytes.
# Alternative Implementations
Alternatively to annotating individual procedures, I considered annotating Program node
with a set of dynamically dispatched selectors. Decoding and passing this information
around proved to be quite cumbersome in the "streaming" world, so I opted for a simpler
approach where all individual procedures are annotated.
Change-Id: I363db6d5dd1138fe25917646313c16d0b213c3b4
Reviewed-on: https://dart-review.googlesource.com/37822
Commit-Queue: Vyacheslav Egorov <vegorov@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Samir Jindel <sjindel@google.com>
When the type to test against is instantiated and has no type arguments
there is a high probability that we receive instances of that class or
subclasses at runtime.
This CL therefore extends the fast-path of AssertAssignable/InstanceOf
by checking whether the instance class id is within the cid ranges that
directly/indirectly implement/extend the type to test against.
Currently we have an almost depth-first preorder numbering of class ids
in AOT, but there are exceptions. So each class can have a number of
cid-ranges as subclasses / classes which implement it's interface.
This seems to improve performance of dart-aot-v2
* flutter stock build by 15+%
* DeltaBlueClosures by 10+%
and reduces code size on
* flutter gallery by -3%
Issue https://github.com/dart-lang/sdk/issues/31798
Change-Id: I07dd91589cc3fcd8c5952bdba339e2e2a459e08e
Reviewed-on: https://dart-review.googlesource.com/35620
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Régis Crelier <regis@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
This reverts commit 12759870a4.
Reason for revert: Broke many bots.
Original change's description:
> [VM] Use IR for code in [CatchEntryInstr]s to populate captured exception/stacktrace variables
>
> This fixes an issue when a program got loaded via dill, a function
> with a try-catch got optimized and the exception/stacktrace variables
> got captured.
>
> Change-Id: Ia6b62f2a0986c78b90afe7fae25025ca4e5b09db
> Reviewed-on: https://dart-review.googlesource.com/35182
> Commit-Queue: Martin Kustermann <kustermann@google.com>
> Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
TBR=vegorov@google.com,kustermann@google.com
Change-Id: I2ec7b4bea401e6d5665f6e789eb8e241b99db7b3
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://dart-review.googlesource.com/36200
Reviewed-by: Martin Kustermann <kustermann@google.com>
Commit-Queue: Martin Kustermann <kustermann@google.com>
This fixes an issue when a program got loaded via dill, a function
with a try-catch got optimized and the exception/stacktrace variables
got captured.
Change-Id: Ia6b62f2a0986c78b90afe7fae25025ca4e5b09db
Reviewed-on: https://dart-review.googlesource.com/35182
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
This CL extends support of mixed operations in strong-mode AOT
from double/smi to double/int (if underlying code generator supports
conversions from int64 to double).
* MintToDouble instruction is renamed to Int64ToDouble.
* Code generation of Int64ToDouble is implemented on x64.
Issue: https://github.com/dart-lang/sdk/issues/30480
Change-Id: I3142302e80c2785be71d7a8baff42d40f0b5b1b8
Reviewed-on: https://dart-review.googlesource.com/35382
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
The original change is available in patchset 1.
The changes since then are just status file updates.
Change-Id: I0c45c4b09a187bdd60c6a38bf23f6b7a0cf80c97
Reviewed-on: https://dart-review.googlesource.com/34400
Reviewed-by: Jens Johansen <jensj@google.com>
Update status files.
This cl includes implementing these two features on all platforms:
1) Support calling generic functions via DartEntry::InvokeFunction().
2) Support native generic functions. These are currently allowed, but type
arguments are ignored, and therefore not accessible from the C++ side.
Change-Id: Id39e8ca46c2ba1ba3d46946c16712a8572ff64ea
Reviewed-on: https://dart-review.googlesource.com/34023
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Siva Annamalai <asiva@google.com>
A new Expression was added in bc2560069a
but C++ wasn't updated to handle it.
The crash reported in 31868 is caused by trying to skip past such an
expression, which is now possible.
Actually using that expression is still unhandled though, although thats
a separate issue.
Fixes#31868.
Bug: 31868
Change-Id: Ife0aadb97cee398ee1925fb52f91127225253768
Reviewed-on: https://dart-review.googlesource.com/34361
Reviewed-by: Vyacheslav Egorov <vegorov@google.com>
Commit-Queue: Jens Johansen <jensj@google.com>
Ryan Macnak