Using strong mode with a mix of opted in and opted out libraries may result in a static call with bad arguments (missing required named argument) that may crash the VM with an assert fault. Ideally, we should emit code to throw a NoSuchMethodError when the code is executed. For now, we report the error as an informative exception during compilation instead of aborting.
The name of the enclosing function needs to be canonical when creating an instance of _FallThroughError when compiling a switch statement.
Change-Id: I50349fcc5b5036fbdec4697238b1e840de15ec07
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/144498
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Adds another interface target to instance calls, pointing to the torn
off method for tearoff getter calls. This is then used to direct the
call to the method extractor for the method.
Increases the fraction of dynamically dispatched instance calls in
Flutter Gallery that are handled by the dispatch table from 95.8% to
98.4%. The remaining 92 IC calls are all on static receiver type
dynamic or Function (34 function calls, 22 method calls, 21 operator
calls, 14 getter calls and 1 setter call).
Change-Id: I0bde1a8271c8563f4631534bd39d1254cd779380
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/142155
Commit-Queue: Aske Simon Christensen <askesc@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
#if defined(DART_PRECOMPILED_RUNTIME)
#error "AOT runtime should not use compiler sources (including header files)"
#endif // defined(DART_PRECOMPILED_RUNTIME)
Remove #if defined(DART_PRECOMPILED_RUNTIME) from most compiler sources.
Change-Id: Id175c83fdbea38d9d5e1371ff433e3888f2afe8a
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/143523
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
In order to have instance calls with unboxed parameters, all
the possible targets need to be able to those parameters unboxed.
Then, the methods were partitioned according to the class hierarchy,
and the inferred types from TFA are used to determine the parameters
that could be unboxed.
This is done for non-nullable integer and double parameters, and
return value.
dart-aot linux x64:
Bench2D 8.028%
DeltaBlue 4.688%
ParticleSystemPaint 27.79%
Flutter release android-armv8:
velocity_tracker_bench -2.026% (less is better)
flutter_gallery_total_size -0.1460%
flutter_gallery_instructions_size -0.2715%
flutter_gallery_readonlydata_size 0.0539%
Issue https://github.com/dart-lang/sdk/issues/33549
Change-Id: Iedd9b4321c3d78b406ea1708e15c747754433027
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/138501
Commit-Queue: Victor Agnez Lima <victoragnez@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Samir Jindel <sjindel@google.com>
Static methods and constructors may receive unboxed parameters.
This change adds a bitmap to the RawFunction object indicating
which parameters are unboxed and their types. Also, the stack frame
bitmap was updated to include new bits for those parameters.
The ArgumentsDescriptor object was updated to store not only the number
of arguments, but also their size, since 32-bit platforms would use two
words for each int64 and double value.
In addition, static methods that return non-nullable integers
and doubles may return these values unboxed.
Dart AOT x64:
TypedData.Uint8ClampedListViewVarBench +11.73%
DeltaBlue +9.32%
Meteor (Intel Xeon) +4.64%
Flutter release android-armv8:
velocity_tracker_bench -2.32% (less is better)
flutter_gallery_readonlydata_size +0.18%
flutter_gallery_total_size +0.13%
flutter_gallery_instructions_size +0.12%
Issue https://github.com/dart-lang/sdk/issues/33549
Change-Id: I97e310f8f30213920af2333fbdedf193d03db5c3
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/135313
Commit-Queue: Victor Agnez Lima <victoragnez@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
On 1/27/20, the nnbd specification changed weak and strong mode instance checks
to make them behave uniformly across legacy and opted-in libraries.
Therefore, it is not necessary anymore to propagate the library mode in
generated code to the runtime.
Change-Id: I42d3ddc6e9a921899aeac21be6374c7893a6d27c
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/138111
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Commit-Queue: Régis Crelier <regis@google.com>
Sets a threshold of five continuations determining if the old
if-else or the new igoto-based implementation will be used.
Informal benchmarking on x64 and arm_x64 point towards the overhead
of the igoto-based impl. dropping off around this point.
Benchmarks of this CL (threshold=5) show drastic improvement in
Calls.IterableManualIterablePolymorphicManyYields of about ~35-65%
across {dart,dart-aot}-{ia32,x64,armv7hf,armv8}.
Note: This is a second revision of a CL landed and reverted from:
https://dart-review.googlesource.com/c/sdk/+/133586
The fix for the previous revert was swapping the parameters to
BuildIndirectEntry(int, int) in kernel_binary_flowgraph.cc.
Bug: https://github.com/dart-lang/sdk/issues/37754
Change-Id: If035235ce8aa660af2ea7f8cfbc8066cde6329b7
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/136126
Commit-Queue: Clement Skau <cskau@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Sets a threshold of five continuations determining if the old
if-else or the new igoto-based implementation will be used.
Informal benchmarking on x64 and arm_x64 point towards the overhead
of the igoto-based impl. dropping off around this point.
Benchmarks of this CL (threshold=5) show drastic improvement in
Calls.IterableManualIterablePolymorphicManyYields of about ~35-65%
across {dart,dart-aot}-{ia32,x64,armv7hf,armv8}.
Bug: https://github.com/dart-lang/sdk/issues/37754
Change-Id: I6e113f1f98e9ab0f994cf93004227d616e9e4d07
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/133586
Commit-Queue: Clement Skau <cskau@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This CL fixes crashes
../../runtime/vm/compiler/frontend/kernel_binary_flowgraph.cc: 3637: error: expected: nnbd_mode == parsed_function()->function().nnbd_mode()
due to Function::nnbd_mode() returning incorrect result when function
belongs to a synthetic mixin application class.
Change-Id: I7860eb6120b7291848d655d8151af2472421920d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/134785
Reviewed-by: Régis Crelier <regis@google.com>
Commit-Queue: Alexander Markov <alexmarkov@google.com>
Introduces NativeRepresentation and NativeLocation for the compilation of FFI.
NativeRepresentations are able to express all representations (or types) of the native ABIs we bind to with FFI, this is more representations that than that are used in Dart itself.
NativeLocations are able to express all locations of the native ABIs we bind to with FFI, this is more types of locations than that are used for the Dart calling convention.
See the documentation in the respective files.
These NativeLocations and NativeRepresentations are computed by the NativeCallingConvention and consumed by the Marshaller and Assemblers.
This reenginering is required for go/dart-ffi-by-value, hardfp (Arm S and D fpu registers), and iOS 64 bit (non-word-aligned stack arguments).
In addition, by using the NativeRepresentations we also get slightly reduced code size:
* The tracking of sizes is improved, so less sign/zero-extension operations are required.
* UnboxedWidthExtenderInstr is fully removed, the size extension is done inside the native moves, coalescing moves and size extension when possible.
* BitCastInstr is only used when really needed. This reduces code-size on arm32 softfp.
This fixes the iOS arm64 calling convention, manually tested with https://github.com/flutter/flutter/pull/46078 and https://dart-review.googlesource.com/c/sdk/+/131074.
Fixes: https://github.com/dart-lang/sdk/issues/39637
Issue: https://github.com/dart-lang/sdk/issues/36309
Issue: https://github.com/dart-lang/sdk/issues/36730
Change-Id: I8878bc0f314277bab4ca22f417c6295ecc017720
Cq-Include-Trybots: luci.dart.try:vm-ffi-android-debug-arm-try,vm-ffi-android-debug-arm64-try,app-kernel-linux-debug-x64-try,vm-kernel-linux-debug-ia32-try,vm-kernel-win-debug-x64-try,vm-kernel-win-debug-ia32-try,vm-kernel-precomp-linux-debug-x64-try,vm-dartkb-linux-release-x64-abi-try,vm-kernel-precomp-android-release-arm64-try,vm-kernel-asan-linux-release-x64-try,vm-kernel-linux-release-simarm-try,vm-kernel-linux-release-simarm64-try,vm-kernel-precomp-android-release-arm_x64-try,vm-kernel-precomp-obfuscate-linux-release-x64-try,dart-sdk-linux-try,analyzer-analysis-server-linux-try,analyzer-linux-release-try,front-end-linux-release-x64-try,vm-kernel-precomp-win-release-x64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/129081
Commit-Queue: Daco Harkes <dacoharkes@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Fix the tests by disabling the optimisation that turns getter calls into
raw gets if the field is late, and by calling recording the null init
store in bytecode mode.
Change-Id: I8f12e3237cd32c890b5cef8d00c32940e937330d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/134180
Commit-Queue: Liam Appelbe <liama@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
This CL is a code cleanup landed separately from https://dart-review.googlesource.com/c/sdk/+/129081 to make the CL size smaller.
Refactorings:
* Split up runtime/vm/compiler/ffi.{cc,h} in multiple files.
* Move FrameRebase to compiler/ffi and prevent it from modifying Locations.
* Remove use of <algorithm>.
Change-Id: Ice2a96470a3242a731255383b9647fabab107ac5
Cq-Include-Trybots: luci.dart.try:vm-ffi-android-debug-arm-try,vm-ffi-android-debug-arm64-try,app-kernel-linux-debug-x64-try,vm-kernel-linux-debug-ia32-try,vm-kernel-win-debug-x64-try,vm-kernel-win-debug-ia32-try,vm-kernel-precomp-linux-debug-x64-try,vm-dartkb-linux-release-x64-abi-try,vm-kernel-precomp-android-release-arm64-try,vm-kernel-asan-linux-release-x64-try,vm-kernel-linux-release-simarm-try,vm-kernel-linux-release-simarm64-try,vm-kernel-precomp-android-release-arm_x64-try,vm-kernel-precomp-obfuscate-linux-release-x64-try,dart-sdk-linux-try,analyzer-analysis-server-linux-try,analyzer-linux-release-try,front-end-linux-release-x64-try,vm-kernel-precomp-win-release-x64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/133224
Commit-Queue: Daco Harkes <dacoharkes@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This reverts commit 5f81efb4a3.
Reason for revert: 3xH bot is green again.
Original change's description:
> Revert "[vm/aot] Only patch call-sites to monomorphic entry-points when the receiver is proven to be a heap object."
>
> This reverts commit ab2026af45.
>
> Reason for revert: Flutter 3xH is broken on Golem.
>
> Original change's description:
> > [vm/aot] Only patch call-sites to monomorphic entry-points when the receiver is proven to be a heap object.
> >
> > Flutter Gallery ARMv8: -0.8% total, -1.3% instructions
> > Flutter Gallery ARMv7: -0.7% total, -1.2% instructions
> >
> > There are some performance regressions, in four categories:
> >
> > 1. Arithmetic on num types (e.g. MinLib, MaxLib):
> > We believe arithmetic on num types is rare in practice, and can be optimized later
> > by specializing the call-sites.
> >
> > 2. Lack of types (e.g. MeshDecompression, ImagingGaussianBlur, Crypto*):
> > These (and similar benchmarks) are written in a very untyped style, which is not
> > representative of Dart 2 style. We've confirmed that the regressions disappear if
> > the benchmarks are annotated with appropriate types.
> >
> > 3. Megamorphic calls which are specialized in the benchmark (e.g. DeltaBlue):
> > DeltaBlue uses string interpolation in one place, so the toString() method in the
> > string interpolation helper is monomorphic for _Smis. However, in a realistic programs,
> > string interpolation is used much more frequently and with different types, so it's very
> > unlikely that this call would be monomorphic. When string interpolation is removed, there's
> > a remaining 1.3% regression on DeltaBlue which is not yet accounted for.
> >
> > 4. Noisy benchmarks
> > The remaining regressions are on benchmarks which show frequent jitter historically.
> >
> > There are also 1-4% improvements on many benchmarks (Golem's "noise" analysis is hiding
> > many of them, look at the graphs).
> >
> > For the reasons above, we believe the regressions justified by the code size improvement.
> >
> > Change-Id: Ic9b281d4383a6111de9d6f44347976ffa61a6ca6
> > Cq-Include-Trybots:luci.dart.try:vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-release-simarm-try,vm-kernel-precomp-linux-release-simarm64-try,vm-kernel-precomp-bare-linux-release-simarm64-try
> > Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/128667
> > Commit-Queue: Samir Jindel <sjindel@google.com>
> > Reviewed-by: Ryan Macnak <rmacnak@google.com>
> > Reviewed-by: Martin Kustermann <kustermann@google.com>
>
> TBR=kustermann@google.com,rmacnak@google.com,alexmarkov@google.com,sjindel@google.com
>
> Change-Id: I411aae4b230f2a08146ad3bf3e7a10aa6592db0e
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-debug-x64-try, vm-kernel-precomp-linux-release-simarm-try, vm-kernel-precomp-linux-release-simarm64-try, vm-kernel-precomp-bare-linux-release-simarm64-try
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/131660
> Reviewed-by: Samir Jindel <sjindel@google.com>
> Commit-Queue: Samir Jindel <sjindel@google.com>
TBR=kustermann@google.com,rmacnak@google.com,alexmarkov@google.com,sjindel@google.com
# Not skipping CQ checks because original CL landed > 1 day ago.
Change-Id: I36e4e9d598033fcc934c2cdf5b061aa11255d3b2
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-debug-x64-try, vm-kernel-precomp-linux-release-simarm-try, vm-kernel-precomp-linux-release-simarm64-try, vm-kernel-precomp-bare-linux-release-simarm64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/131837
Reviewed-by: Samir Jindel <sjindel@google.com>
Commit-Queue: Samir Jindel <sjindel@google.com>
Running all existing tests with/without lazy async stacks ensures that
the existing functionality for --lazy-async-stacks does not regress.
There are still a few smaller things to be done in the debugger for lazy
async stacks.
This CL also ensures:
* The inner closure, i.e. AsyncClosure/AsyncGenClosure, is not
inlined if FLAG_lazy_async_stacks is on.
* Fixes a crash in dartkb mode, when the function of the Bytecode
object is null.
* Does a simple integration of the lazy async stacks in debugger.cc -
to ensure a stack is returned via vm-service.
Issue https://github.com/dart-lang/sdk/issues/37668
Change-Id: Ibc1e887a457e2c456ae65d9ed5fa92434f122a32
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/131825
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Ben Konyi <bkonyi@google.com>
This reverts commit ab2026af45.
Reason for revert: Flutter 3xH is broken on Golem.
Original change's description:
> [vm/aot] Only patch call-sites to monomorphic entry-points when the receiver is proven to be a heap object.
>
> Flutter Gallery ARMv8: -0.8% total, -1.3% instructions
> Flutter Gallery ARMv7: -0.7% total, -1.2% instructions
>
> There are some performance regressions, in four categories:
>
> 1. Arithmetic on num types (e.g. MinLib, MaxLib):
> We believe arithmetic on num types is rare in practice, and can be optimized later
> by specializing the call-sites.
>
> 2. Lack of types (e.g. MeshDecompression, ImagingGaussianBlur, Crypto*):
> These (and similar benchmarks) are written in a very untyped style, which is not
> representative of Dart 2 style. We've confirmed that the regressions disappear if
> the benchmarks are annotated with appropriate types.
>
> 3. Megamorphic calls which are specialized in the benchmark (e.g. DeltaBlue):
> DeltaBlue uses string interpolation in one place, so the toString() method in the
> string interpolation helper is monomorphic for _Smis. However, in a realistic programs,
> string interpolation is used much more frequently and with different types, so it's very
> unlikely that this call would be monomorphic. When string interpolation is removed, there's
> a remaining 1.3% regression on DeltaBlue which is not yet accounted for.
>
> 4. Noisy benchmarks
> The remaining regressions are on benchmarks which show frequent jitter historically.
>
> There are also 1-4% improvements on many benchmarks (Golem's "noise" analysis is hiding
> many of them, look at the graphs).
>
> For the reasons above, we believe the regressions justified by the code size improvement.
>
> Change-Id: Ic9b281d4383a6111de9d6f44347976ffa61a6ca6
> Cq-Include-Trybots:luci.dart.try:vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-release-simarm-try,vm-kernel-precomp-linux-release-simarm64-try,vm-kernel-precomp-bare-linux-release-simarm64-try
> Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/128667
> Commit-Queue: Samir Jindel <sjindel@google.com>
> Reviewed-by: Ryan Macnak <rmacnak@google.com>
> Reviewed-by: Martin Kustermann <kustermann@google.com>
TBR=kustermann@google.com,rmacnak@google.com,alexmarkov@google.com,sjindel@google.com
Change-Id: I411aae4b230f2a08146ad3bf3e7a10aa6592db0e
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Cq-Include-Trybots: luci.dart.try:vm-kernel-precomp-linux-debug-x64-try, vm-kernel-precomp-linux-release-simarm-try, vm-kernel-precomp-linux-release-simarm64-try, vm-kernel-precomp-bare-linux-release-simarm64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/131660
Reviewed-by: Samir Jindel <sjindel@google.com>
Commit-Queue: Samir Jindel <sjindel@google.com>
Flutter Gallery ARMv8: -0.8% total, -1.3% instructions
Flutter Gallery ARMv7: -0.7% total, -1.2% instructions
There are some performance regressions, in four categories:
1. Arithmetic on num types (e.g. MinLib, MaxLib):
We believe arithmetic on num types is rare in practice, and can be optimized later
by specializing the call-sites.
2. Lack of types (e.g. MeshDecompression, ImagingGaussianBlur, Crypto*):
These (and similar benchmarks) are written in a very untyped style, which is not
representative of Dart 2 style. We've confirmed that the regressions disappear if
the benchmarks are annotated with appropriate types.
3. Megamorphic calls which are specialized in the benchmark (e.g. DeltaBlue):
DeltaBlue uses string interpolation in one place, so the toString() method in the
string interpolation helper is monomorphic for _Smis. However, in a realistic programs,
string interpolation is used much more frequently and with different types, so it's very
unlikely that this call would be monomorphic. When string interpolation is removed, there's
a remaining 1.3% regression on DeltaBlue which is not yet accounted for.
4. Noisy benchmarks
The remaining regressions are on benchmarks which show frequent jitter historically.
There are also 1-4% improvements on many benchmarks (Golem's "noise" analysis is hiding
many of them, look at the graphs).
For the reasons above, we believe the regressions justified by the code size improvement.
Change-Id: Ic9b281d4383a6111de9d6f44347976ffa61a6ca6
Cq-Include-Trybots:luci.dart.try:vm-kernel-precomp-linux-debug-x64-try,vm-kernel-precomp-linux-release-simarm-try,vm-kernel-precomp-linux-release-simarm64-try,vm-kernel-precomp-bare-linux-release-simarm64-try
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/128667
Commit-Queue: Samir Jindel <sjindel@google.com>
Reviewed-by: Ryan Macnak <rmacnak@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
This also fixes the implementation of 'null is T' in legacy vs opted-in library
by making it independent of strong/weak mode.
Includes updates to type propagation and call specializer.
Change-Id: I2d347ef277e08cef779faf51a972e863244ef095
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/130922
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Reviewed-by: Liam Appelbe <liama@google.com>
NNBDMode does not reflect the semantics (legacy or nnbd) of type tests anymore.
Instead, the semantics are derived from the value of the strong mode flag.
This required one function to be specialized, namely the implementation of
'null is Type' in weak mode in an opted-in library, which still requires
nnbd semantics although run in weak mode.
Relevant changes are in object.h and object.cc:
- methods NNBD_NullIsInstanceOf and NNBD_IsTopType are new.
- methods IsNullType, IsTopType, and IsNeverType are modified.
Change-Id: I36cd43d93d2cfabd110cbcc6b26487a583bb089d
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/130444
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
This does not yet include checking of required named parameters.
Also, inlined type testing code, type testing stubs, and type propagation are
not yet updated.
Change-Id: If6d8996635474e349d28b7fd207d724838d08881
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/129922
Commit-Queue: Régis Crelier <regis@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
This is done by collecting all frames on the current stack until an async/async* frame is hit
which has yielded before (i.e. is not in sync-async case).
From there on it finds the closure of the async/async* frame and starts traversing the listeners:
while (closure != null) {
yield_index = closure.context[Context::kAsyncJumpVarIndex]
pc = closure.function.code.pc_descriptors.LookupPcFromYieldIndex(yield_index);
<emit pc in frame>
closure = closure.context[Context::kAsyncCompleterVarIndex]._future._resultOrListeners.callback;
}
Bug: https://github.com/dart-lang/sdk/issues/37668
Change-Id: I97030d22e529bf8a74ecd30a8a2589cfe57ad330
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/122644
Commit-Queue: Clement Skau <cskau@google.com>
Reviewed-by: Martin Kustermann <kustermann@google.com>
Right now `Script::yield_positions()` is an array mapping function start
token positions to arrays. Those arrays contain token positions and are
indexed by yield index.
=> This `Script::yield_positions()` is not available in AOT mode.
The fast async stack implementation will need to be able to find out
where an async closure was suspended. It does so by looking at the
":await_jump_var", which contains the yield index. It then needs to
associate the yield index to token position of the yield.
This CL adds an entry into the PcDescriptors for every yield in a
async/async* function and removes `Script::yield_positions()`.
The entry will associate the yield index with the token position.
Flutter gallery total size impact for flutter-release:
- armv7: +0.016%
- armv8: +0.045%
Issue https://github.com/dart-lang/sdk/issues/37668
Change-Id: I0b2ce41e85d8f5d590201bf2fb091578d7379890
Reviewed-on: https://dart-review.googlesource.com/c/sdk/+/125408
Commit-Queue: Martin Kustermann <kustermann@google.com>
Reviewed-by: Clement Skau <cskau@google.com>
Reviewed-by: Alexander Markov <alexmarkov@google.com>
Ryan Macnak