This is the first step in cutting the Gordian linkage/linkage-impl knot.
This basically changes the axis along which we organize call descriptor
building logic from having platform-specific files dedicated to all call
descriptor types to having call-descriptor-type-specific files that have
The next step is to factor the JS, code stub, and runtime call descriptors
similarly, dumping them into:
compiler/js-linkage.cc
compiler/runtime-linkage.cc
compiler/code-stub-linkage.cc
or, alternatively, all of them just into compiler/js-linkage.cc.
This also anticipates a wasm-linkage.cc file in the future.
R=bmeurer@chromium.org,danno@chromium.org
BUG=
Review URL: https://codereview.chromium.org/1266603002
Cr-Commit-Position: refs/heads/master@{#29931}
Adds interpreter entry and exit trampoline builtins. Also implements the
Return bytecode handler and fixes a few bugs in InterpreterAssembler
highlighted by running on other architectures.
MIPS and MIPS64 port contributed by Paul Lind (paul.lind@imgtec.com)
BUG=v8:4280
LOG=N
Review URL: https://codereview.chromium.org/1245133002
Cr-Commit-Position: refs/heads/master@{#29929}
The idle time handler should never return DONE or DO_SCAVENGE for
background tabs. Upon receiving DONE chrome will stop sending idle notifications.
BUG=chromium:515174
LOG=NO
Review URL: https://codereview.chromium.org/1269583002
Cr-Commit-Position: refs/heads/master@{#29926}
When a Property or a VariableProxy is used as the left hand side of an
assignment statement, there is no need to allocate a LOAD_IC feedback
vector slot for it. Alter the numbering phase to support this.
BUG=
Review URL: https://codereview.chromium.org/1262803002
Cr-Commit-Position: refs/heads/master@{#29924}
All runtime function get a context anyway, which is the same as the
explicit one in case of DeclareGlobals and DeclareLookupSlot. So
we can remove the additional parameter there.
As an additional bonus, improve the runtime interface to DeclareLookupSlot.
R=yangguo@chromium.org
Review URL: https://codereview.chromium.org/1261863002
Cr-Commit-Position: refs/heads/master@{#29923}
Replaces the 64-bit div instruction with 32-bit division in DivI.
Also fixes the Ddiv implementation in the simulator.
TEST=mjsunit/asm/int32div
BUG=
Review URL: https://codereview.chromium.org/1265603002
Cr-Commit-Position: refs/heads/master@{#29920}
A Parameter operator contains a ParameterInfo payload, not an int
payload, so the OpParameter<int> cast is invalid (and only worked by
coincidence currently).
BUG=chromium:515215
LOG=n
R=yangguo@chromium.org
Review URL: https://codereview.chromium.org/1266743002
Cr-Commit-Position: refs/heads/master@{#29919}
This fixes several bugs related to float64 pool constants.
Note that float64 pool constants are still disabled by default, and are
only used if --enable-vldr-imm is provided.
BUG=
Review URL: https://codereview.chromium.org/1260953002
Cr-Commit-Position: refs/heads/master@{#29917}
Moves some uncommon type checking from ToString and
NonStringToString into DefaultString. This should
speed up string operations.
LOG=N
BUG=none
Review URL: https://codereview.chromium.org/1256323004
Cr-Commit-Position: refs/heads/master@{#29913}
script_executed and last_run are read/written by multiple threads. Also
externalized_shared_contents_ is modified by multiple threads.
BUG=4306
R=jarin@chromium.org
LOG=n
Review URL: https://codereview.chromium.org/1252623003
Cr-Commit-Position: refs/heads/master@{#29912}
When the main thread terminates, it forcibly terminates all Worker threads.
When this happens, the threads objects were only half-created; they had a
JavaScript Worker object, but not a C++ worker object.
This CL fixes that bug, as well as some other fixes:
* Signatures on Worker methods
* Use SetAlignedPointerFromInternalField instead of using an External.
* Remove state_ from Worker. Simplify to atomic bool running_.
BUG=chromium:511880
R=jarin@chromium.org
LOG=n
Review URL: https://codereview.chromium.org/1255563002
Cr-Commit-Position: refs/heads/master@{#29911}
Changes the interpreter to use a BytecodeArray pointer and an offset to avoid
having an inner pointer to a BytecodeArray object in registers during dispatch.
BUG=v8:4280
LOG=N
Review URL: https://codereview.chromium.org/1254293006
Cr-Commit-Position: refs/heads/master@{#29910}
Moves some uncommon type checking from ToNumber and
NonNumberToNumber into DefaultNumber. This should
speed up numeric operations especially as more "value types
are added.
LOG=N
BUG=513196, v8:4124
Review URL: https://codereview.chromium.org/1260273002
Cr-Commit-Position: refs/heads/master@{#29908}
Reason for revert:
Suspected to cause Canary crashes
Original issue's description:
> Reland^2 "Enable loads and stores to global vars through property cell shortcuts installed into parent script context".
>
> This reverts commit 362b378501.
>
> R=ishell@chromium.org
>
> Committed: https://crrev.com/440ae014e56924b52337c3747221b79283f07b81
> Cr-Commit-Position: refs/heads/master@{#29849}
TBR=ishell@chromium.org,v8-mips-ports@googlegroups.com,plind44@gmail.com,bmeurer@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
Review URL: https://codereview.chromium.org/1260423002
Cr-Commit-Position: refs/heads/master@{#29905}
Literals should stay together to simplify allocation folding. This CL takes the pretenuring information from the outermost literal and applies it to the inner liberals.
BUG=chromium:514721
LOG=n
R=mvstanton@chromium.org
Review URL: https://codereview.chromium.org/1263773002 .
Cr-Commit-Position: refs/heads/master@{#29903}
This allows variants to be named on test failures (follow
up) and then to be used in the test runner for a repro.
This also speeds up variant iteration for test262 and fixes
a bug with variants for benchmarks.
BUG=chromium:511215
NOTREECHECKS=true
LOG=n
Review URL: https://codereview.chromium.org/1245623005
Cr-Commit-Position: refs/heads/master@{#29899}
When a prototype object migrates from a slow to a fast map, where the slow map
was registered as a user of its own prototype, then the registration must be
transferred to the new map (just like MigrateToMap does for all other cases).
BUG=chromium:513602
LOG=y
NOTREECHECKS=true
Review URL: https://codereview.chromium.org/1263543004
Cr-Commit-Position: refs/heads/master@{#29898}
Scoping rules are different on recompile vis-a-vis global loads.
BUG=chromium:514526
LOG=y
NOTREECHECKS=true
Review URL: https://codereview.chromium.org/1256413005
Cr-Commit-Position: refs/heads/master@{#29896}
port 5dff4bdff0 (r29886).
original commit message:
No need to pass the name explicitly to the stubs; the runtime can
extract the name from the ScopeInfo (the extension of the
ScriptContext) on-demand easily without any performance impact.
BUG=
Review URL: https://codereview.chromium.org/1259063004
Cr-Commit-Position: refs/heads/master@{#29892}
port cac64b9f63 (r29881)
original commit message:
We don't need the hole check and slow runtime mode for read-only
properties this way.
BUG=
Review URL: https://codereview.chromium.org/1263473002
Cr-Commit-Position: refs/heads/master@{#29891}
Don't use different read/write slots for context globals, but
let them share the same slot, which reduces the number of
initial misses, and also saves some memory for large scripts.
R=yangguo@chromium.org
Review URL: https://codereview.chromium.org/1258213002
Cr-Commit-Position: refs/heads/master@{#29889}
Reason for revert:
This CL may be the reason for the spike on memory corruption. Tentatively reverting this CL.
BUG=chromium:512780
LOG=n
Original issue's description:
> Activate preserving of optimized code map accross GCs.
>
> This enables --noflush-optimized-code-cache which allows preserving
> entries in the optimized code map accross GCs. This only applies to
> values being reachable through other paths.
>
> R=hpayer@chromium.org,hablich@chromium.org
>
> Committed: https://crrev.com/1a8776db25b63c4ce718423772d1fd13f58eeab5
> Cr-Commit-Position: refs/heads/master@{#29755}
TBR=hablich@chromium.org,mstarzinger@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review URL: https://codereview.chromium.org/1255043003
Cr-Commit-Position: refs/heads/master@{#29888}
We can apply the shortcut used for data properties of the JSGlobalObject
to builtin properties as well. This mostly affects the custom properties
we use for the Math functions (i.e. rngstate for Math.random() and kMath
for sin, cos and friends).
Drive-by-fix: Also mark the internal builtin typed arrays properties
(rngstate, kMath and rempio2result) as READ_ONLY, as they should not be
touched after genesis.
R=yangguo@chromium.org
Review URL: https://codereview.chromium.org/1254143003
Cr-Commit-Position: refs/heads/master@{#29887}
No need to pass the name explicitly to the stubs; the runtime can
extract the name from the ScopeInfo (the extension of the
ScriptContext) on-demand easily without any performance impact.
Review URL: https://codereview.chromium.org/1259963002
Cr-Commit-Position: refs/heads/master@{#29886}
