Port e127f58410
Original Commit Message:
Handling of class constructors was moved from CallFunction to Call
in [1].
When reducing calls with spread we forward varargs directly to
CallFunction, if we are spreading to inlined arguments or arguments of
the outermost function.
In that case we didn't check for class constructors and therefore didn't
raise an exception.
This CL adds checks for class constructors to all JSCall* nodes in
JSCallReducer that missed them before.
[1] https://crrev.com/c/3186434R=pthier@chromium.org, joransiu@ca.ibm.com, junyan@redhat.com, midawson@redhat.com
BUG=
LOG=N
Change-Id: I02cba90369354f064201daa1bf8812e17cb2dc21
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3234040
Reviewed-by: Junliang Yan <junyan@redhat.com>
Commit-Queue: Milad Fa <mfarazma@redhat.com>
Cr-Commit-Position: refs/heads/main@{#77480}
... when the v8_enable_external_code_space build flag is enabled.
Bug: v8:11880
Change-Id: I754c6229dcd25f81ef6dfbedc5885ac025c0aeff
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3164458
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77479}
The cctest enables concurrent-sparplug during startup, which creates
a job_handler_, then disables the flag before destroying the isolate,
which calls the destructor of ConcurrentBaselineCompiler.
This does not call job_handle_->Cancel and fails in a DCHECK inside
the default_job implementation.
Bug: v8:12054
Change-Id: I8975e2ad2c0aafaa86aedb13018617a08f0fded9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3234718
Commit-Queue: Victor Gomes <victorgomes@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Auto-Submit: Victor Gomes <victorgomes@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77475}
Port 8bbb44e537
Original Commit Message:
Large character classes may easily be created when unicode
properties (e.g.: /\p{L}/u and /\P{L}/u) are used - these are
expanded internally into character classes that consist of hundreds
of character ranges. Previously to this CL, we'd emit branching code
for each of these ranges, leading to very large regexp code objects.
This CL adds a new codegen mode for large character classes (where
'large' currently means > 16 ranges). Instead of emitting branching
code inline, the ranges are written into a ByteArray and we call into
the C function IsCharacterInRangeArray for the actual branching logic.
The ByteArray is smaller than emitted code and is deduplicated if the
same character class is matched repeatedly in the same pattern.
Note this mode is *not* implemented for the interpreter, since we
currently don't have a constant pool for irregexp bytecode, and thus
cannot reference ByteArrays.
R=jgruber@chromium.org, joransiu@ca.ibm.com, junyan@redhat.com, midawson@redhat.com
BUG=
LOG=N
Change-Id: I2ded01fa2767e56e72be81b949eefb5fb85b7013
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3231981
Reviewed-by: Junliang Yan <junyan@redhat.com>
Commit-Queue: Milad Fa <mfarazma@redhat.com>
Cr-Commit-Position: refs/heads/main@{#77473}
Handling of class constructors was moved from CallFunction to Call
in [1].
When reducing calls with spread we forward varargs directly to
CallFunction, if we are spreading to inlined arguments or arguments of
the outermost function.
In that case we didn't check for class constructors and therefore didn't
raise an exception.
This CL adds checks for class constructors to all JSCall* nodes in
JSCallReducer that missed them before.
[1] https://crrev.com/c/3186434
Bug: chromium:1260623
Change-Id: Id39cdfd09ff5aae804ae30d96909518e408c9613
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229369
Commit-Queue: Patrick Thier <pthier@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77472}
These two tests rely on the functions being immediately compiled with SP.
Concurrent SP might add a delay and fail the test.
Bug: v8:12054
Change-Id: I52dbdfeb5c49c23f2d20a43bb8a87aba1f16358e
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3233233
Commit-Queue: Victor Gomes <victorgomes@chromium.org>
Commit-Queue: Patrick Thier <pthier@chromium.org>
Auto-Submit: Victor Gomes <victorgomes@chromium.org>
Reviewed-by: Patrick Thier <pthier@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77471}
When --always-sparkplug, we compile on the main thread. This fixes the
CodeBuilder path when also running with --concurrent-sparkplug
Bug: v8:12054
Change-Id: Ifafcd68b635e99ca39c5eebc3e1ff839a2193eb4
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3233232
Commit-Queue: Victor Gomes <victorgomes@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Auto-Submit: Victor Gomes <victorgomes@chromium.org>
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77470}
last_index_ is already incremented in Enqueue and points to one past
the end of the array.
Bug: v8:12054
Change-Id: I63b9a315464af9366f39c863fbd40e58adb69f1b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3233230
Commit-Queue: Victor Gomes <victorgomes@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Auto-Submit: Victor Gomes <victorgomes@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77469}
Add a readonly --dict-property-const-tracking flag so that we can
correctly specify negative implications for --concurrent-inlining and
--turboprop, and correctly fail flag contradictions where needed.
Change-Id: I1b20c6a2bef42c868ccde3b8db60f1107eb29550
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3233222
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77466}
Search for free memory ranges within the +/- 2GB boundary to the
embedded builtins. So that code range can be allocated close to the
binary to enable short builtin calls when pointer compression is
disabled.
Bug: v8:12045, v8:11527
Change-Id: I4698625882c3c7c39aff73b0bc874ddcfc990881
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3212466
Commit-Queue: Hao A Xu <hao.a.xu@intel.com>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77465}
Large character classes may easily be created when unicode
properties (e.g.: /\p{L}/u and /\P{L}/u) are used - these are
expanded internally into character classes that consist of hundreds
of character ranges. Previously to this CL, we'd emit branching code
for each of these ranges, leading to very large regexp code objects.
This CL adds a new codegen mode for large character classes (where
'large' currently means > 16 ranges). Instead of emitting branching
code inline, the ranges are written into a ByteArray and we call into
the C function IsCharacterInRangeArray for the actual branching logic.
The ByteArray is smaller than emitted code and is deduplicated if the
same character class is matched repeatedly in the same pattern.
Note this mode is *not* implemented for the interpreter, since we
currently don't have a constant pool for irregexp bytecode, and thus
cannot reference ByteArrays.
Bug: v8:11069
Change-Id: I2d728e42d85114b796c637f791848731a104cd54
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229377
Reviewed-by: Patrick Thier <pthier@chromium.org>
Auto-Submit: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77463}
Huge functions can take a really long time (several minutes) in register
allocation. This is caused by a big number of virtual registers combined
with a big number of blocks.
We can avoid such long-running register allocation by falling back to
the mid-tier register allocation if we detect a huge number of virtual
registers.
Note that this is mostly relevant for bigger WebAssembly modules, but we
implement it as a general TurboFan flag.
The flag is off by default for now, since there seem to be bugs lurking
in the mid-tier register allocator. Once those are fixed, we can stage
it behind --future and then ship it.
R=mslekova@chromium.orgCC=thibaudm@chromium.org
Bug: v8:12320
Change-Id: Iff14de456c2b3a91ee2b2b12221295dd56b69463
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3231336
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77461}
mjsunit/regress/regress-crbug-9161 had two spinlocks on an atomic:
1. WaitUntil(lock == kStageRunning)
2. WaitUntil(lock == kStageDone)
But, in theory the worker updating the "lock" could progress all the way
to kStageDone before the first loop manages to check the lock value
again.
We can make this more robust by checking:
1. WaitUntil(lock != kStageInit)
2. WaitUntil(lock == kStageDone)
That way both loops check for _any_ state past the state they want to
progress past.
Bug: v8:11437
Change-Id: I5220e61070a305301c678928edb0925c04dae970
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3231339
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Shu-yu Guo <syg@chromium.org>
Reviewed-by: Shu-yu Guo <syg@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77460}
... by explicitly passing pointer compression cage base value to various
IsXXX() and map() calls in order to avoid using incorrect auto-computed
cage base value when applied to objects allocated in external code space.
This CL also introduces IsCodeObject(HeapObject) predicate which checks
the IS_EXECUTABLE bit in the page header's flags.
Bug: v8:11880
Change-Id: Ib44398c3125392e46e939044a9bd27e09d7944d5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229368
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77459}
This CL adds support for BigInt.asIntN, the necessary operations and
extensions of the compiler's type system to allow lowering of BigInts
to word64 representations that are interpreted as signed 64 bit BigInts.
Bug: v8:9407
Change-Id: Id4f1f45437c1caf94e01c7b4e063c2ae2386c88a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3198070
Commit-Queue: Nico Hartmann <nicohartmann@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77458}
We increment the size before enqueueing the next element.
This guarantees that size > 0 when decrementing.
Bug: v8:12325
Change-Id: Ida256d9b22a9dd5cacb21312f099ee7186e2ca53
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3231335
Commit-Queue: Victor Gomes <victorgomes@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Auto-Submit: Victor Gomes <victorgomes@chromium.org>
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77457}
A few of the changes added by https://crrev.com/c/3229379 are
causing compilation errors on platforms without sparkplug.
Change-Id: Ic6088b33ba910ae5a96881fa7609e30d35db5d8a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3226548
Reviewed-by: Victor Gomes <victorgomes@chromium.org>
Commit-Queue: Milad Fa <mfarazma@redhat.com>
Cr-Commit-Position: refs/heads/main@{#77456}
Disable --concurrent-inlining when v8_dict_property_const_tracking is
enabled, since the latter doesn't support the former. Add statusfile
variables to mark this as an incompatible flag, and otherwise disable it
via flags.
Change-Id: Ie2e7dac6f4a1ddc97e7e7f4d4d2c77e638e216a0
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229565
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77455}
Previously having a `Symbol.toStringTag` property holding a string
somewhere in the prototype chain would always take predence over trying
to determine a reasonable name from the "constructor" (in case of
subclassing). This would lead to confusing behavior when diagnosing
objects in DevTools, for example
```js
class A extends URLSearchParams {};
new A()
```
would show `URLSearchParam` as class name for the instance rather than
`A`.
With this CL, we change the lookup logic to explicitly check for
`Symbol.toStringTag` and "constructor" along each step of the prototype
chain (skipping the "constructor" for the leaf object) and pick the
first one that yields a string (that is the function debug name in case
of "constructor").
Fixed: chromium:980018
Change-Id: Ic920b4bae02f965bc383c711f8de89c0de55fcac
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3231078
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Kim-Anh Tran <kimanh@chromium.org>
Auto-Submit: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Kim-Anh Tran <kimanh@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77453}
`a || b` only evaluates b if a is false. `a | b` always evaluates
both a and b. If a and b are of type bool, `||` is usually what you
want, so clang now warns on `|` where both arguments are of type bool.
In this case the difference isn't important, but || is more
conventional to express this
Bug: chromium:1255745
Change-Id: I8fb090abc9863f7db8761bddb1440613a49bc84b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3231077
Auto-Submit: Hans Wennborg <hans@chromium.org>
Commit-Queue: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77452}
This is a reland of 929b83fb7b
This version of the CL also fixes initialization of the
marking_barrier_ in the LocalHeap constructor.
This CL also got rebased on Victor's CL in https://crrev.com/c/3229361.
It added a code_space_allocator_ in LocalHeap which needs to be
initialized a bit later on the main thread as well.
Original change's description:
> [heap] Attach to shared isolate after setting up main thread
>
> Attach to the shared isolate after the main thread was set up. Otherwise
> it could happen that a shared GC initiated from another isolate might
> see no threads are running and performs the safepoint operation in the
> middle of isolate deserialization.
>
> We use DisallowSafepoints to check that the isolate doesn't join a
> global safepoint before deserialization is complete. DisallowSafepoints
> used to prevent only invocations of Safepoint() but was updated to
> also prevent Park() and Unpark() invocations. Each state change could
> cause the thread to reach a safepoint, which would allow a shared GC
> to run.
>
> We now also DCHECK that every isolate has at least one local heap and
> that shared collections aren't started before deserialization is
> complete.
>
> Bug: v8:11708
> Change-Id: Iba3fb59dd951d5ee4fc9934158062287302fc279
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3221157
> Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Reviewed-by: Shu-yu Guo <syg@chromium.org>
> Reviewed-by: Jakob Gruber <jgruber@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#77424}
Bug: v8:11708
Change-Id: I7d44e4a5f76cc09092c2444cede10e9331222c1d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229361
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Shu-yu Guo <syg@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77448}
This is a reland of dca83ff7e7
Relanding without changes, looks like the bot that failed succeeded
again before the revert landed (https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux64%20TSAN/38915/overview)
Original change's description:
> [snapshot] Fix two stress_snapshot failures
>
> 1. Stress snapshot discards code on SFIs, which breaks flushing
> invariants. Add --stress-snapshot to those invariants.
> 2. Another test failing on IsScript, added it to the disabled list.
>
> Change-Id: Ic415923a9cc96d34b74f7450b29be99c1b53d077
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229375
> Commit-Queue: Leszek Swirski <leszeks@chromium.org>
> Commit-Queue: Jakob Gruber <jgruber@chromium.org>
> Auto-Submit: Leszek Swirski <leszeks@chromium.org>
> Reviewed-by: Jakob Gruber <jgruber@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#77430}
Change-Id: I18fd290c3794d90addc2a94765d57a3ee3ceab86
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229563
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77447}
Young generation collection requires that full GCs unmark before
starting marking.
Bug: v8:12324
Change-Id: Id6cc218057252cbf0664326126f34b07ac8ea247
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229570
Auto-Submit: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Anton Bikineev <bikineev@chromium.org>
Reviewed-by: Anton Bikineev <bikineev@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77445}
[1] fixes the behavior of StaNamedOwnProperty to no longer do prototype
lookups. This lets us revert [2] and go back to using the fast path in
the clone spread object literal bytecode.
The test case from [2] is kept.
[1] https://chromium-review.googlesource.com/c/v8/v8/+/2795831
[2] https://chromium-review.googlesource.com/c/v8/v8/+/3178969
Bug: v8:9888, chromium:1251366
Change-Id: I9d2cb69b803c403f63365f55d27c4de20ff7dafb
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3224666
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Patrick Thier <pthier@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Shu-yu Guo <syg@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77444}
No functionality change expected.
Most scopes are renamed new_scope.
Bug: v8:12244,v8:12245
Change-Id: I85d23b0ba6971f51b9bbfc1f3afeb89fb70f035e
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3227268
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77443}
No functionality change expected.
Most scopes are renamed new_scope or new_handle_scope. For some test
cases the outer scope is renamed to outer_scope since there are
multiple inner scopes.
Bug: v8:12244,v8:12245
Change-Id: I85953617e54f2140fa88c593eb7c186b570fdd04
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3227266
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77442}
Avoid generating switch statements with only a default case. Instead,
when there are no instructions that can trap, simply have
HasMemoryAccessMode() return false. This avoids a MSVC warning when
doing a 32-bit build.
To do this, remove empty TARGET_ARCH_OPCODE_WITH_MEMORY_ACCESS_MODE_LIST
definitions from instruction-codes-$arch.h files.
Change-Id: Ifed76eb9cbca169f30c188c1999e1e9be0b2c6aa
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3224807
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Commit-Queue: Lei Zhang <thestig@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77441}
Change the logic behind GenerateOneOf so that not only
ref_null would appear almost everytime. Increase the probability
of other alternatives to appear as well.
Change the name of GenerateOptRef to GenerateRef.
Bug: v8:11954
Change-Id: I8bff1661c3d3a91105bca1ffa68e1e9b0b662240
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229367
Commit-Queue: Maria Tîmbur <mtimbur@google.com>
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Reviewed-by: Manos Koukoutos <manoskouk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77438}
The gdb DCHECK frame skipping automatically skips over all the abort
etc. frames to get you to that DCHECK callsite you're actually looking
for.
However, this is annoying if you tried to call a function with a
breakpoint from the gdb prompt; the frame skipping wold skip over your
breakpoint back up to the failing DCHECK.
Now, we abort the frame walk on dummy frames inserted by gdb execution.
Change-Id: I2cf89ea9219374ad7c562c6eb13afe471038b033
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3229376
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77436}
... by updating the checks for not_mapped_symbol root and using proper
cage base values depending on the host object type.
Bug: v8:11880
Change-Id: I28908cbb5b1023addaee248028661d480734e29c
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3222760
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77435}
