This change correctly sets the {CatchPrediction} field in exception
handler tables for bytecode and optimized code. It also adds tests
independent of promise handling for this prediction, to ensure all our
backends are in sync on their prediction.
R=rmcilroy@chromium.org,yangguo@chromium.org
TEST=mjsunit/compiler/debug-catch-prediction
BUG=v8:4674
LOG=n
Review URL: https://codereview.chromium.org/1690973002
Cr-Commit-Position: refs/heads/master@{#33906}
This adds test cases for exception handlers that require a context
switch when entering the catch-block or the finally-block, triggered
through nested contexts within the try-block.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1681933002
Cr-Commit-Position: refs/heads/master@{#33845}
By now only the default %TypedArray%.prototype.sort compare function
and the JS implementation of SameValueZero were still using the odd
%_IsMinusZero intrinsic, whose semantics both included a number check
(actually HeapNumber test) plus testing if the heap number stores the
special -0 value. In both cases we already know that we deal with
number so we can reduce it to a simple number test for -0, which can
be expressed via dividing 1 by that value and checking the sign of
the result. In case of the compare function, we can be even smarter
and work with the reciprocal values in case x and y are equal to 0
(although long term we should probably rewrite the fast case for
the typed array sorting function in C++ anyway, which will be way,
way faster than our handwritten callback-style, type-feedback
polluted JS implementation).
R=yangguo@chromium.org
Review URL: https://codereview.chromium.org/1680783002
Cr-Commit-Position: refs/heads/master@{#33833}
Avoid the hacking in JSIntrinsicLowering and provide a proper simplified
operator ObjectIsReceiver instead that is used to implement %_IsJSReceiver
which is used by our JavaScript builtins and the JSInliner.
R=jarin@chromium.org
BUG=v8:4544
LOG=n
Review URL: https://codereview.chromium.org/1657863004
Cr-Commit-Position: refs/heads/master@{#33675}
The previous versions of Math.max and Math.min made it difficult to
optimize those (that's why we already have custom code in Crankshaft),
and due to lack of ideas what to do about the variable number of
arguments, we will probably need to stick in special code in TurboFan
as well; so inlining those builtins is off the table, hence there's no
real advantage in having them around as "not quite JS" with extra work
necessary in the optimizing compilers to still make those builtins
somewhat fast in cases where we cannot inline them (also there's a
tricky deopt loop in Crankshaft related to Math.min and Math.max, but
that will be dealt with later).
So to sum up: Instead of trying to make Math.max and Math.min semi-fast
in the optimizing compilers with weird work-arounds support %_Arguments
%_ArgumentsLength, we do provide the optimal code as native builtins
instead and call it a day (which gives a nice performance boost on some
benchmarks).
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1641083003
Cr-Commit-Position: refs/heads/master@{#33582}
This patch moves the semantics of 'const' in sloppy mode to match those
in strict mode, that is, const makes lexical (let-like) bindings, must
have an initializer, and does not create properties of the global object.
R=adamk
LOG=Y
BUG=v8:3305
CQ_INCLUDE_TRYBOTS=tryserver.chromium.linux:linux_chromium_rel_ng;tryserver.blink:linux_blink_rel
Review URL: https://codereview.chromium.org/1571873004
Cr-Commit-Position: refs/heads/master@{#33218}
Adds support for the CallRuntimeForPair bytecode to the Bytecode Graph
Builder. Modifies the FrameState support to allow updating of output
registers.
Also adds Eval tests to test-run-bytecode-graph-builder since these are
enabled by CallRuntimeForPair support.
BUG=v8:4280
LOG=N
Review URL: https://codereview.chromium.org/1570623007
Cr-Commit-Position: refs/heads/master@{#33186}
Deopt support is added on two levels. On the IR level,
a new ObjectState node is added, which represenents an
object to be materialized. ObjectState nodes appear as
inputs of FrameState and StateValues nodes. On the
instruction select/code-generation level, the
FrameStateDescriptor class handles the nesting
introduced by ObjectState, and ensures that deopt code
with CAPTURED_OBJECT/DUPLICATED_OBJECT entries are
generated similarly to what crankshaft's escape
analysis does.
Two unittests test correctness of the IR level implementation.
Correctness for instruction selection / code generation
is tested by mjsunit tests.
R=jarin@chromium.org,mstarzinger@chromium.org
BUG=v8:4586
LOG=n
Review URL: https://codereview.chromium.org/1485183002
Cr-Commit-Position: refs/heads/master@{#33115}
Bugfixes and improvements in escape analysis include:
* Handling of ObjectIsSmi (non-escaping)
* Handling of nested phi replacements
* Handling of phis with arity > 2
* Resilience against effectful nodes dangling from start
* Allocations escape now, if non-const load/store is performed
* Fixed a bug where non-allocated objects where tracked
* Allow fixed double arrays to be tracked
R=mstarzinger@chromium.org
BUG=v8:4586
LOG=n
Review URL: https://codereview.chromium.org/1510973006
Cr-Commit-Position: refs/heads/master@{#32833}
This patch improves escape analysis and fixes bugs
triggered by clusterfuzz. Impovements include:
* Handling of LoadElement/StoreElement if index is a
constant
* Handling of JSStoreProperty: invalidate all information,
as the store could have altered any field.
* Treat phis that use an allocation as escaping
* Improve resolution of replacements
R=mstarzinger@chromium.org
BUG=v8:4586
LOG=n
Review URL: https://codereview.chromium.org/1499143002
Cr-Commit-Position: refs/heads/master@{#32656}
Reason for revert:
Blocks the roll:
https://codereview.chromium.org/1497763004/
Original issue's description:
> [es6] Correctify and unify ArrayBuffer and SharedArrayBuffer constructors.
>
> The ArrayBuffer and SharedArrayBuffer constructors should raise an
> exception when called with no arguments or undefined length. Also
> unified the ArrayBuffer and SharedArrayBuffer implementations as C++
> builtins, and removed some (now) obsolete runtime entries.
>
> R=yangguo@chromium.org
>
> Committed: https://crrev.com/3235ccbb7826ceec2188f6ebab98fc851b54f60e
> Cr-Commit-Position: refs/heads/master@{#32590}
TBR=yangguo@chromium.org,cbruni@chromium.org,adamk@chromium.org,bmeurer@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/1501673002
Cr-Commit-Position: refs/heads/master@{#32606}
The ArrayBuffer and SharedArrayBuffer constructors should raise an
exception when called with no arguments or undefined length. Also
unified the ArrayBuffer and SharedArrayBuffer implementations as C++
builtins, and removed some (now) obsolete runtime entries.
R=yangguo@chromium.org
Review URL: https://codereview.chromium.org/1500543002
Cr-Commit-Position: refs/heads/master@{#32590}
* Add a sibling interface to InterpreterAssembler called
CodeStubAssembler which provides a wrapper around the
RawMachineAssembler and is intented to make it easy to build
efficient cross-platform code stubs. Much of the implementation
of CodeStubAssembler is shamelessly stolen from the
InterpreterAssembler, and the idea is to eventually merge the
two interfaces somehow, probably moving the
InterpreterAssembler interface over to use the
CodeStubAssembler. Short-term, however, the two interfaces
shall remain decoupled to increase our velocity developing the
two systems in parallel.
* Implement the StringLength stub in TurboFan with the new
CodeStubAssembler. Replace and remove the old Hydrogen-stub
version.
* Remove a whole slew of machinery to support JavaScript-style
code stub generation, since it ultimately proved unwieldy,
brittle and baroque. This cleanup includes removing the shared
code stub context, several example stubs and a tangle of build
file changes.
BUG=v8:4587
LOG=n
Review URL: https://codereview.chromium.org/1475953002
Cr-Commit-Position: refs/heads/master@{#32508}
This is the first part of escape analysis for turbofan.
At the moment, there is no deopt support, and support
for loops is partial (only binary Phis are handled).
The CL includes 4 unittests.
There are also 8 new mjsunit tests, some of which are
skiped as they require features not yet implemented.
BUG=v8:4586
LOG=n
Review URL: https://codereview.chromium.org/1457683003
Cr-Commit-Position: refs/heads/master@{#32498}
If the input type does not help us, we are conservative and truncate (rather than guessing signed).
Review URL: https://codereview.chromium.org/1455103002
Cr-Commit-Position: refs/heads/master@{#32075}
The %_CallFunction doesn't implement the call sequence properly, it
doesn't do the receiver wrapping, nor does it check for
classConstructor. Also the eager deoptimization for %_CallFunction was
seriously b0rked (we must have been lucky with TurboFan so far).
R=yangguo@chromium.org
BUG=v8:4413
LOG=n
Review URL: https://codereview.chromium.org/1419813010
Cr-Commit-Position: refs/heads/master@{#31821}
TurboFan is actually able to generate property access to all prototypes
of all primitives, except the special Oddball primitives that have no
wrapper counterparts (namely null and undefined from the ES6 point of
view).
R=jarin@chromium.org
BUG=v8:4470
LOG=n
Review URL: https://codereview.chromium.org/1409163007
Cr-Commit-Position: refs/heads/master@{#31739}
This introduces a JSConvertReceiver operator to model the implicit
conversion of receiver values for sloppy callees. It is used by the
JSInliner for now, but can also be used to model direction function
calls that bypass call stubs.
Also note that a hint is passed to said operator whenever the source
structure constrains the receiver value type. This hint allows for
optimizations in the lowering of the operator.
The underlying specification in ES6, section 9.2.1.2 is the basis for
this implementation.
R=bmeurer@chromium.org
TEST=mjsunit/compiler/receiver-conversion
BUG=v8:4493, v8:4470
LOG=n
Review URL: https://codereview.chromium.org/1412223015
Cr-Commit-Position: refs/heads/master@{#31598}
This makes sure that initializing assignments of let-declared variables
perform an environment lookup and hence keep the variable alive. This is
needed because full-codegen contains debug code verifying the variable
is still inside the TDZ at the initializing assignment.
R=jarin@chromium.org
TEST=mjsunit/compiler/regress-variable-liveness-let
BUG=v8:4493
LOG=n
Review URL: https://codereview.chromium.org/1420573002
Cr-Commit-Position: refs/heads/master@{#31437}
When calling into C++ builtins, we need to make sure that the argument
count register contains the correct number of arguments, otherwise the
CEntryStub will not be able to leave the stack in the correct state.
R=ishell@chromium.org
BUG=v8:4413
LOG=n
Review URL: https://codereview.chromium.org/1391543002
Cr-Commit-Position: refs/heads/master@{#31120}
Introduce %_ToNumber intrinsic, which just calls to the existing
ToNumberStub, and remove all uses of our custom JavaScript plus
intrinsics based ToNumber and friends.
Also replace the TO_NUMBER_INLINE macro with TO_NUMBER,
which is currently a wrapper for %_ToNumber. Newly written JS
code should use TO_NUMBER (similar to TO_STRING, TO_INT32,
and friends).
Also finally remove the DefaultString/DefaultNumber builtins, which
are basically the ES5 version of ToPrimitive. Now all code uses the
ES6 version, which is implemented in Object::ToPrimitive and
JSReceiver::ToPrimitive in C++.
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_layout_dbg,v8_linux_nosnap_dbg
R=jarin@chromium.org
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/1384443002
Cr-Commit-Position: refs/heads/master@{#31054}
Introduce a new macro TO_STRING that maps to %_ToString and use that
instead of calling into any of the ToString/NonStringToString JavaScript
builtins. Also remove the TO_STRING_INLINE macro, which is basically
obsolete with %_ToString. We still have a few uses of ToString left (via
the utils export mechanism), where we need to investigate whether we
will tank badly if we replace them with TO_STRING as well.
CQ_INCLUDE_TRYBOTS=tryserver.v8:v8_linux_layout_dbg,v8_linux_nosnap_dbg
R=yangguo@chromium.org
BUG=v8:4307
LOG=n
Review URL: https://codereview.chromium.org/1323543002
Cr-Commit-Position: refs/heads/master@{#30895}
Inner functions must be eagerly parsed for scope analysis, but the full AST is
also kept around even though it's not needed.
This CL mitigates this problem by allocating some AstNodes of the inner function
to a temporary Zone which is deallocated once the scope information has been
built. The remaining nodes (such as VariableProxy) must persist until scope
analysis actually happens, and have to be allocated to a parser-persistent Zone.
BUG=417697
LOG=N
Review URL: https://codereview.chromium.org/1304923004
Cr-Commit-Position: refs/heads/master@{#30685}
The HUnaryMathOperation cannot be eliminated in general, because the
spec requires a ToNumber conversion on the input, which is observable
of course.
BUG=v8:4389
LOG=y
Review URL: https://codereview.chromium.org/1307413003
Cr-Commit-Position: refs/heads/master@{#30343}
Until now, TF-generated code stubs piggy-backed off of the builtin
context. Since generation of code stubs is lazy, stubs generated at
different times in different native contexts would contain embedded
pointers different builtin contexts, leading to cross-context references
and memory leaks.
After this CL, all TF-generated code stubs are generated inside a
internal thinned-out, native context that lives solely for the
purpose of hosting generated code stubs.
Review URL: https://codereview.chromium.org/1213203007
Cr-Commit-Position: refs/heads/master@{#29593}
Optimize string "length" property access based on static type
information if possible, but also optimistically optimize the access
based on type feedback from the LoadIC.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1216593003
Cr-Commit-Position: refs/heads/master@{#29543}
- Add a TurboFanIC class, derived from TurboFanCodeStub, that
automatically distinguishes between versions of the IC called from
optimized and unoptimized code.
- Add appropriate InterfaceDescriptors for both the versions of the
stub called from unoptimized and optimized code
- Change the MathFloor TF stub generator to output either the
for-optimized or for-unoptimized version based on the minor_key
parameter.
Committed: https://crrev.com/8f13b655b8a10dae2116dd18b32f09337bb2d410
Cr-Commit-Position: refs/heads/master@{#29534}
Review URL: https://codereview.chromium.org/1225943002
Cr-Commit-Position: refs/heads/master@{#29539}
Reason for revert:
Failure on compiling runtime.js on windows
Original issue's description:
> Add unoptimized/optimized variants of MathFloor TF code stub
>
> - Add a TurboFanIC class, derived from TurboFanCodeStub, that
> automatically distinguishes between versions of the IC called from
> optimized and unoptimized code.
> - Add appropriate InterfaceDescriptors for both the versions of the
> stub called from unoptimized and optimized code
> - Change the MathFloor TF stub generator to output either the
> for-optimized or for-unoptimized version based on the minor_key
> parameter.
>
> Committed: https://crrev.com/8f13b655b8a10dae2116dd18b32f09337bb2d410
> Cr-Commit-Position: refs/heads/master@{#29534}
TBR=mvstanton@chromium.org,bmeurer@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/1220783006
Cr-Commit-Position: refs/heads/master@{#29535}
- Add a TurboFanIC class, derived from TurboFanCodeStub, that
automatically distinguishes between versions of the IC called from
optimized and unoptimized code.
- Add appropriate InterfaceDescriptors for both the versions of the
stub called from unoptimized and optimized code
- Change the MathFloor TF stub generator to output either the
for-optimized or for-unoptimized version based on the minor_key
parameter.
Review URL: https://codereview.chromium.org/1225943002
Cr-Commit-Position: refs/heads/master@{#29534}
Currently we lower shifts directly to machine operators, and add an
appropriate Word32And to implement the & 0x1F operation on the right
hand side required by the specification. However for Word32And we assume
Int32 in simplified lowering, which is basically changes the right hand
side bit interpretation for the shifts from Uint32 to Int32, which is
obviously wrong. So now we represent that explicitly by proper
simplified operators for the shifts, which are lowered to machine in
simplified lowering.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1213803008
Cr-Commit-Position: refs/heads/master@{#29465}
ARM64's `fmin` and `fmax` instructions don't have the same behaviour as
TurboFan's Float(32|64)(Min|Max) functions.
BUG=4206
LOG=N
Review URL: https://codereview.chromium.org/1200123004
Cr-Commit-Position: refs/heads/master@{#29229}
This fixes the graph wiring of implicit JSToNumber nodes inserted by
JSTypedLowering, to be correctly hooked into a surrounding exceptional
continuation.
R=bmeurer@chromium.org
TEST=mjsunit/compiler/try-binop,test262
Review URL: https://codereview.chromium.org/1178153004
Cr-Commit-Position: refs/heads/master@{#28975}
This adds some basic tests of the interaction between try-catch and
try-finally statements and OSR in TurboFan. The try-osr test suite
follows the structure of try-deopt closely.
R=titzer@chromium.org
TEST=mjsunit/compiler/try-osr
Review URL: https://codereview.chromium.org/1165103003
Cr-Commit-Position: refs/heads/master@{#28855}
This allows try-catch and try-finally constructs to be separately
enabled and disabled. We plan to stage try-catch support soon.
R=titzer@chromium.org
Review URL: https://codereview.chromium.org/1157863015
Cr-Commit-Position: refs/heads/master@{#28848}
First steps only, the TurboFan compilation is still triggered from C++ land.
Includes some simplifications/cleanups, too.
Review URL: https://codereview.chromium.org/1150263002
Cr-Commit-Position: refs/heads/master@{#28581}
This was already done for other binary operations, so it's basically
copying the existing functionality to shift left and shift right
logical/arithmetic.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1140883003
Cr-Commit-Position: refs/heads/master@{#28389}
Use these check points to optimize comparisons where we already know
that one side cannot be a String (or turn into a string via
ToPrimitive).
Also remove bunch of useless DoNotCrash tests for the scheduler that are
painful to maintain and add almost no value.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1140583004
Cr-Commit-Position: refs/heads/master@{#28383}
This stub will be used as the basis of a Math.floor-specific CallIC to
detect and track calls to floor that return -0.
Along the way:
- Create a TurboFanCodeStub super class from which the StringLength and
MathRound TF stubs derive.
- Fix the ugly hack that passes the first stub parameter as the "this"
pointer in the the TF-compiled JS function.
- Fix bugs in the ia32/x64 disassembler.
Review URL: https://codereview.chromium.org/1137703002
Cr-Commit-Position: refs/heads/master@{#28339}
The goal is to port all of error stack trace formatting to C++.
We will do this bottom up, by first porting helper functions.
Eventually, CallSite methods will only be used when a custom
error stack trace formatter is defined via Error.prepareStackTrace.
R=jkummerow@chromium.org
Review URL: https://codereview.chromium.org/1060583008
Cr-Commit-Position: refs/heads/master@{#28095}
Deoptimization infrastructure already handles it correctly.
This change fixes repetitive deoptimizations in the code like this:
var u32 = new Uint32Array(1);
u32[0] = -1;
function tr(x) { return x|0; }
function ld() { return tr(u32[0]); }
while (true) ld();
Currently inlined tr will contain HArgumentsObject that is considered uint32-unsafe use and prevents u32[0] from becoming uint32 load - instead a speculative int32 load is generated which just deopts.
BUG=
Review URL: https://codereview.chromium.org/1077113002
Cr-Commit-Position: refs/heads/master@{#27781}
With this change, we remember the types of frame state inputs (in a new
operator, called TypedStateValues). Instead of inferring the value types
when building translations, we used the recorded types.
The original approach was not reliable because the passes after
simplified lowering can change node types, and this in turn confuses
the translation builder.
BUG=chromium:468727
LOG=n
R=bmeurer@chromium.org
Review URL: https://codereview.chromium.org/1015423002
Cr-Commit-Position: refs/heads/master@{#27310}
The change introduces a second frame state (for the state before
the operation) for the StoreProperty nodes. If the store writes
into a typed array, the frame state is used for lazy deopt from
the to-number conversion that is performed by the store.
BUG=v8:3963
LOG=n
R=bmeurer@chromium.org
Review URL: https://codereview.chromium.org/997983004
Cr-Commit-Position: refs/heads/master@{#27285}
In constructing the transfer between loop copies, we need to merge the backedges from all the previous copies of the given loop. The control reduction will work out which ones are really reachable.
R=mstarzinger@chromium.org
BUG=
Review URL: https://codereview.chromium.org/1004993004
Cr-Commit-Position: refs/heads/master@{#27246}
This makes sure only the %_DeoptimizeNow intrinsic is inlined, and
not the %DeoptimizeNow one. It hence re-establishes the invariant
that JSIntrinsicLowering only deals with inline intrinsics.
R=jarin@chromium.org
TEST=mjsunit/compiler/eager-deopt-simple
Review URL: https://codereview.chromium.org/988333003
Cr-Commit-Position: refs/heads/master@{#27070}
This just contains test, no fixes. Note that some of the tests are
still disabled because they either fail or we don't want ClusterFuzz
to pick up the flag yet.
R=jarin@chromium.org
TEST=cctest/test-run-jsexceptions/Deopt,mjsunit/compiler/try-deopt
Review URL: https://codereview.chromium.org/972943004
Cr-Commit-Position: refs/heads/master@{#26968}
This prevents eliminating effectful statements before the loop.
BUG=
Review URL: https://codereview.chromium.org/830923002
Cr-Commit-Position: refs/heads/master@{#25953}
We cannot just clear the result register optimistically, because the
register allocator might assign the same register to result and buffer.
TEST=mjsunit/compiler/regress-445858
BUG=chromium:445858
LOG=y
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/828303002
Cr-Commit-Position: refs/heads/master@{#25950}
Set a valid reason for disabling optimization when using
%NeverOptimizeFunction.
TEST=mjsunit/compiler/regress-445732
BUG=chromium:445732
LOG=y
TBR=machenbach@chromium.org
Review URL: https://codereview.chromium.org/832003002
Cr-Commit-Position: refs/heads/master@{#25949}
Change InstructionOperand to use a 64-bit field for encoding the operand
information instead of the 32-bit field that was used before. Ideally we
wouldn't use the Zone-allocated bit field at all, and use an integer
instead of the pointer; but that requires fixing the register allocator
first, which will take some time.
TEST=mjsunit/compiler/regress-3786
BUG=v8:3786
LOG=y
Review URL: https://codereview.chromium.org/826673002
Cr-Commit-Position: refs/heads/master@{#25941}
The `right == 0` checks only worked for `0 <= right < 32`. This patch
replaces the checks with simple tests for negative results.
The attached test can detect this error, but the test relies on a broken
flag (--noopt-safe-uint32-operations), so it is skipped for now. See
issue 3487 for details.
BUG=
R=ulan@chromium.org
Review URL: https://codereview.chromium.org/487913005
git-svn-id: https://v8.googlecode.com/svn/branches/bleeding_edge@23243 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
version is passing all the existing test + a bunch of new tests
(packaged in the change list, too).
The patch extends the SlotRef object to describe captured and duplicated
objects. Since the SlotRefs are not independent of each other anymore,
there is a new SlotRefValueBuilder class that stores the SlotRefs and
later materializes the objects from the SlotRefs.
Note that unlike the previous implementation of SlotRefs, we now build
the SlotRef entries for the entire frame, not just the particular
function. This is because duplicate objects might refer to previous
captured objects (that might live inside other inlined function's part
of the frame).
We also need to store the materialized objects between other potential
invocations of the same arguments object so that we materialize each
captured object at most once. The materialized objects of frames live
in the new MaterielizedObjectStore object (contained in Isolate),
indexed by the frame's FP address. Each argument materialization (and
deoptimization) tries to lookup its captured objects in the store before
building new ones. Deoptimization also removes the materialized objects
from the store. We also schedule a lazy deopt to be sure that we always
get rid of the materialized objects and that the optmized function
adopts the materialized objects (instead of happily computing with its
captured representations).
Concerns:
- Is the FP address the right key for a frame? (Note that deoptimizer's
representation of frame is different from the argument object
materializer's one - it is not easy to find common ground.)
- Performance is suboptimal in several places, but a quick local run of
benchmarks does not seem to show a perf hit. Examples of possible
improvements: smarter generation of SlotRefs (build other functions'
SlotRefs only for captured objects and only if necessary), smarter
lookup of stored materialized objects.
- Ideally, we would like to share the code for argument materialization
with deoptimizer's materializer. However, the supporting data structures
(mainly the frame descriptor) are quite different in each case, so it
looks more like a separate project.
Thanks for any feedback.
R=danno@chromium.org, mstarzinger@chromium.org
LOG=N
BUG=
Committed: https://code.google.com/p/v8/source/detail?r=18918
Review URL: https://codereview.chromium.org/103243005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18936 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
mostly to make sure that it is going in the right direction. The current
version is passing all the existing test + a bunch of new tests
(packaged in the change list, too).
The patch extends the SlotRef object to describe captured and duplicated
objects. Since the SlotRefs are not independent of each other anymore,
there is a new SlotRefValueBuilder class that stores the SlotRefs and
later materializes the objects from the SlotRefs.
Note that unlike the previous implementation of SlotRefs, we now build
the SlotRef entries for the entire frame, not just the particular
function. This is because duplicate objects might refer to previous
captured objects (that might live inside other inlined function's part
of the frame).
We also need to store the materialized objects between other potential
invocations of the same arguments object so that we materialize each
captured object at most once. The materialized objects of frames live
in the new MaterielizedObjectStore object (contained in Isolate),
indexed by the frame's FP address. Each argument materialization (and
deoptimization) tries to lookup its captured objects in the store before
building new ones. Deoptimization also removes the materialized objects
from the store. We also schedule a lazy deopt to be sure that we always
get rid of the materialized objects and that the optmized function
adopts the materialized objects (instead of happily computing with its
captured representations).
Concerns:
- Is there a simpler/more correct way to store the already-materialized
objects? (At the moment there is a custom root reference to JSArray
containing frames' FixedArrays with their captured objects.)
- Is the FP address the right key for a frame? (Note that deoptimizer's
representation of frame is different from the argument object
materializer's one - it is not easy to find common ground.)
- Performance is suboptimal in several places, but a quick local run of
benchmarks does not seem to show a perf hit. Examples of possible
improvements: smarter generation of SlotRefs (build other functions'
SlotRefs only for captured objects and only if necessary), smarter
lookup of stored materialized objects.
- Ideally, we would like to share the code for argument materialization
with deoptimizer's materializer. However, the supporting data structures
(mainly the frame descriptor) are quite different in each case, so it
looks more like a separate project.
Thanks for any feedback.
R=mstarzinger@chromium.org, danno@chromium.org
LOG=N
BUG=
Review URL: https://codereview.chromium.org/103243005
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18918 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
by escape analysis). Added several tests that expose the bug.
Summary:
LCodegen::AddToTranslation assumes that Lithium environments are
generated by depth-first traversal, but LChunkBuilder::CreateEnvironment
was generating them in breadth-first fashion. This fixes the
CreateEnvironment to traverse the captured objects depth-first.
Note:
It might be worth considering representing LEnvironment by a list
with the same order as the serialized translation representation
rather than having two lists with a subtle relationship between
them (and then serialize in a slightly different order again).
R=titzer@chromium.org, mstarzinger@chromium.org
LOG=N
BUG=
Review URL: https://codereview.chromium.org/93803003
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@18470 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This change implements a simple data-flow analysis pass over captured
objects to the existing escape analysis. It tracks the state of values
in the Hydrogen graph through CapturedObject marker instructions that
are used to construct an appropriate translation for the deoptimizer to
be able to materialize these objects again.
This can be considered a combination of scalar replacement of loads and
stores on captured objects and sinking of unused allocations.
R=titzer@chromium.org
TEST=mjsunit/compiler/escape-analysis
Review URL: https://codereview.chromium.org/21055011
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@16098 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
The current usage of this runtime function is broken as it does not
prevent inlining of the affected function but rather bails out from the
whole unit of compilation after trying to inline affected functions.
This simplifies said runtime function to avoid accidental misuse.
R=titzer@chromium.org
TEST=mjsunit/never-optimize
Review URL: https://codereview.chromium.org/19776006
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@15762 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
This allows the deoptimizer to materialize objects (e.g. the arguments
object) while deopting without having a consective stack area holding
the object values. The LEnvironment explicitly tracks locations for
these values and preserves them in the translation.
R=svenpanne@chromium.org
TEST=mjsunit/compiler/inline-arguments
Review URL: https://codereview.chromium.org/16779004
git-svn-id: http://v8.googlecode.com/svn/branches/bleeding_edge@15087 ce2b1a6d-e550-0410-aec6-3dcde31c8c00
