The memory pressure notification logic wasn't correct and given the current users of
the compiler dispatcher aren't posting speculative tasks, it isn't particularly useful.
After removing this, the abort logic can also be simplified significantly by removing the
non-blocking abort logic.
BUG=v8:8041
Cq-Include-Trybots: luci.chromium.try:linux_chromium_rel_ng
Change-Id: I584533b58fb717fdca46cc620822914d6bdb28b8
Reviewed-on: https://chromium-review.googlesource.com/c/1278495
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56609}
Updated the test so that it uses assertPromiseResult, which makes sure that
a promise rejection is not swallowed. The change is reflected in the actual
async ids, checked in the test.
Bug: v8:8300
Change-Id: Ie227ca74a8cf4e0e079809b21c3abc5a5f87c11a
Reviewed-on: https://chromium-review.googlesource.com/c/1278388
Commit-Queue: Maya Lekova <mslekova@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56608}
There was a small race where an idle task could be posted after the compiler dispatcher
had aborted and the CancellableTaskRunner had been cancelled. This was causing flakyness
on the bots. This fixes this by moving the idle task posting into the same lock block as
the notification to the main thread that the background task has completed.
BUG=v8:8041
Change-Id: I43ca4cea807bfdfeb13f6d1c4a67a4d8a4f6291f
Reviewed-on: https://chromium-review.googlesource.com/c/1278494
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56607}
We generally try to use the most specific type, but sometimes we still
used LiftoffRegister even though it's statically known that it always
holds a GP register. Make these uses use Register directly. Note that
the conversion between Register and LiftoffRegister is a noop in
Release builds.
R=titzer@chromium.org
Bug: v8:6600
Change-Id: I1cf9aa727fb6fd71bbc1eed77df5e9d01e35ddee
Reviewed-on: https://chromium-review.googlesource.com/c/1278727
Reviewed-by: Ben Titzer <titzer@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56605}
This fixes a potential data race that can happen when a freshly
allocated layout descriptor is set on an existing map while the
concurrent marker is using the map to visit an object.
Change-Id: If8ff1c9368b24beb15fefe4a78a21e0f0784d2e8
Reviewed-on: https://chromium-review.googlesource.com/c/1278726
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56604}
Change-Id: Iec5e6baff16260317f693188b01230ab0c2bb86f
Reviewed-on: https://chromium-review.googlesource.com/c/1278809
Reviewed-by: Georg Neis <neis@chromium.org>
Commit-Queue: Hai Dang <dhai@google.com>
Cr-Commit-Position: refs/heads/master@{#56603}
This class was defined in function-body-decoder.cc, but it's not an
implementation of function body decoding, but rather the interface
between the decoder and the WasmGraphBuilder. Hence move it out to its
own file.
R=titzer@chromium.org, mstarzinger@chromium.org
Bug: v8:8238
Change-Id: Ib9bf47e90a3683f578b30b6de74d01da81b2be93
Reviewed-on: https://chromium-review.googlesource.com/c/1278391
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56602}
As part of the standard objects serialization, collect the identities of
the Array prototype and the Object prototype from all native contexts.
This information is then be consulted at compile time instead of having
to call Isolate::IsInAnyContext (which accesses the heap).
Bug: v8:7790
Change-Id: I26a8271afadfce243a8032e4a012f249ce3c2a60
Reviewed-on: https://chromium-review.googlesource.com/c/1275815
Commit-Queue: Georg Neis <neis@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56599}
Yesterday I added a call to memmove in CSA for pointer arrays when they
are in new space and the concurrent marker isn't running (protected by
mask kPointersFromHereAreInteresting, CL here:
https://chromium-review.googlesource.com/c/v8/v8/+/1243104/12). The bug
was that I didn't emit the check if dealing with a SMI array. However,
the GC subsystem at that point doesn't distinguish between SMI and
OBJECT FixedArrays. This fix brings the CSA code in line with that.
R=ulan@chromium.org
Bug: v8:8294
Change-Id: I9eb033c358911e8337562dbc91af8f0e6fbd2ed3
Reviewed-on: https://chromium-review.googlesource.com/c/1278386
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56598}
We were re-definining the FunctionSig typedef in several places. This
CL moves it to value-type.h, since it's a signature over ValueType.
R=titzer@chromium.org
Bug: v8:8238
Change-Id: Id5e8a55c7e0f98d61235e32a5e6cd12e04d26947
Reviewed-on: https://chromium-review.googlesource.com/c/1278387
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56597}
This also adds support in the c1 visualizer output generator to resolve
splintered live ranges before they have been merged back in.
Bug: v8:8238
Change-Id: I61e5f6da2e8b77ca9b42fae9004c77163d11960f
Reviewed-on: https://chromium-review.googlesource.com/c/1270817
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Stephan Herhut <herhut@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56595}
The register allocator has an optimization that makes it try reuse
spill-slots between phi-inputs and assign the same spill slot to a phi
output, if possible.
Furthermore, if the phi has no immediate register use, the output is
kept spilled.
With this change, no longer require that a phi input was spilled in an
immediate predecessor. Instead, it is good enough if it was spilled
at definition.
Change-Id: I317b9d5f9a9dda6a47972cdf84dc2627b996f3db
Reviewed-on: https://chromium-review.googlesource.com/c/1273053
Commit-Queue: Stephan Herhut <herhut@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56594}
Adds support for enqueuing parallel parse / compile tasks for eagerly
compiled IIFEs during parsing. If the --parallel-compile-tasks flag is
enabled, the parser will pre-parse eager top-level IIFEs and enqueue a
task on the compiler dispatcher to do the actual parsing / compilation
on a worker thread.
Currently we always enqueue the task, but we likely want to only
enqueue parallel tasks where the script has multiple IIFEs or a
substantial amount of top-level script code before the IIFE to avoid
the main thread having to immediately block on the parallel task. This
work will be done as a follow-up.
BUG=v8:8041
Change-Id: If68d7c374548cabd4ec32f1fb6752da7d6aaae6b
Reviewed-on: https://chromium-review.googlesource.com/c/1275354
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56593}
This will ensure that the Isolate tear-down sequence runs the
destructors even if the first callback already ran. The second callback
might never be executed, since it runs on a background thread which
might not be executed before tear-down of the Isolate.
R=titzer@chromium.org, mlippautz@chromium.org
Bug: v8:8208, chromium:893549
Change-Id: Ibc148e71bc06936033c5b0f0e70bbb2b5a8e8094
Reviewed-on: https://chromium-review.googlesource.com/c/1276629
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56592}
Change the way we start collecting async stack traces by storing the
current microtask as a root instead of trying to make sense of the
last frame we see. This makes it possible to use the zero cost async
stack traces in Node.js as well (where the last JavaScript frame we
see is not the actual async function, but some frame related to the
main event loop usually).
In addition to the benefit that it now works with Node.js, we can also
extend the new machinery to look through (almost arbitrary) promise
chains. For example this code snippet
```js
(async function() {
await Promise.resolve().then(() =>
console.log(new Error().stack));
})();
```
can be made to also show the async function frame, even though at the
point where the stack trace is collected we don't have any async
function on the stack. But instead there's a PromiseReactionJobTask
as "current microtask", and we can dig into the chained promise to
see where the async execution is going to continue and eventually
find the await promise in the chain.
This also removes the removes the need to allocate `.generator_object`
specially during scope resolution.
Bug: v8:7522
Ref: nodejs/node#11865
Tbr: ulan@chromium.org
Design-Document: bit.ly/v8-zero-cost-async-stack-traces
Change-Id: Ib96cb17c2f75cce083a24e5ba2bbb7914e20d203
Reviewed-on: https://chromium-review.googlesource.com/c/1277505
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56590}
If the outer classifier will end up parsing an arrow parameter list we'd only
be in an inner classifier if the outer was already known to be non-simple.
Change-Id: I9fdfb5539b3a7989869f7190f0919699fe8e5c90
Reviewed-on: https://chromium-review.googlesource.com/c/1276630
Reviewed-by: Adam Klein <adamk@chromium.org>
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56589}
This is a reland of c285380ca8
Original change's description:
> Create a fast path to get migration target when updating map
>
> During map updating, store the pointer to new map in the
> raw_transitions slot of the old map that is deprecated from map
> transition tree. Thus, we can get the migration target directly
> instead of TryReplayPropertyTransitions when updating map.
>
> This can improve Speedometer2.0 Elm-TodoMVC case by ~5% on ATOM
> Chromebook and ~9% on big-core Ubuntu.
>
> Change-Id: I56f9ce5183bbdd567b964890f623ef0ceed9b7db
> Reviewed-on: https://chromium-review.googlesource.com/1233433
> Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
> Reviewed-by: Igor Sheludko <ishell@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#56303}
Change-Id: Idf0b7716b92a6a15bfe58721c2c34dbd02b31137
Reviewed-on: https://chromium-review.googlesource.com/c/1270261
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Commit-Queue: Shiyu Zhang <shiyu.zhang@intel.com>
Cr-Commit-Position: refs/heads/master@{#56588}
On ia32, we were pinning too many registers, resulting in no unpinned
byte registers left (we only have three byte registers since {ebx}
is reserved for the root register).
It turns out that on most paths, we don't actually need to pin any
registers, since {Store} is often the last call for an operation (like
any store or set_global). If registers need to be pinned, only pass
those that must be kept alive across the {Store}. This allows to
compute a more narrow set of pinned registers on demand inside {Store}.
Plus minor drive-by changes.
R=titzer@chromium.org
Bug: chromium:894374, chromium:894307, v8:6600
Change-Id: Ic4d7131784c193dc7a2abf0e504d9973f6d5c5f1
Reviewed-on: https://chromium-review.googlesource.com/c/1275819
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56587}
We don't need to store the native context explicitly anymore, the
broker already has it.
Bug: v8:7790
Change-Id: I1096953e3c56bed9d3a8d7d37b108888ef4ac7ec
Reviewed-on: https://chromium-review.googlesource.com/c/1270594
Commit-Queue: Georg Neis <neis@chromium.org>
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56586}
The primary purpose of this is to untangle a circular dependency
objects.h -> handles.h -> objects.h. Most compilation units only
need message-template.h, without the rest of messages.h.
Bonus: change the enum to an enum class for improved type safety.
Bug: v8:3770
Cq-Include-Trybots: luci.v8.try:v8_linux_noi18n_rel_ng
Change-Id: I8102c55197a450811de2588a68a08e7f99ea6b9e
Reviewed-on: https://chromium-review.googlesource.com/c/1272193
Commit-Queue: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Adam Klein <adamk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56583}
Port a63987a41a
Original Commit Message:
This JSAsyncFunctionObject represents the implicit generator object
inside of async functions, and also holds the outer promise for the
async functions. This in turn allows us to get rid of the .promise
in the Parser / BytecodeGenerator completely, and will make it
possible to build zero-cost async stack traces independent of the
concrete synchronous part of the stack frame (which currently breaks
in Node.js).
In the bytecode all the async function operations now take this new
JSAsyncFunctionObject instead of passing both the .generator_object
and the .promise, which further simplifies and shrinks the bytecode.
It also reduces the size of async function frames, potentially making
the suspend/resume cheaper.
This also changes `await` to use intrinsics instead of calling to
special JSFunctions on the native context, and thus reduces the size of
the native contexts.
to TurboFan.
R=bmeurer@chromium.org, joransiu@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=
LOG=N
Change-Id: Ife0aa45b11580f316e657942485907cf78336e4b
Reviewed-on: https://chromium-review.googlesource.com/c/1276867
Commit-Queue: Junliang Yan <jyan@ca.ibm.com>
Reviewed-by: Joran Siu <joransiu@ca.ibm.com>
Cr-Commit-Position: refs/heads/master@{#56581}
We only need to report once that we're in an invalid path for binding patterns
and arrow formals.
Change-Id: I8c7edc1c2a9f431c98e09725d0534e661db76634
Reviewed-on: https://chromium-review.googlesource.com/c/1276626
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56579}
If the instance template is not marked as undetectable, we can
allow it to be called as a constructor. This broke previously with
commit ff05633408.
R=verwaest@chromium.org
Bug: v8:7670
Change-Id: I6ecde33bd7532bea4786b2282efce9060bb76276
Reviewed-on: https://chromium-review.googlesource.com/c/1272579
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56577}
Now that the preparser doesn't track errors anymore, it also doesn't make sense
to prepare the message.
Change-Id: Ifc69e67f9220be69812257b0fc18b55097236dbd
Reviewed-on: https://chromium-review.googlesource.com/c/1275818
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56574}
It's rarely used, so just using {compiler::WasmGraphBuilder} directly
is easier to understand and does not add much to the code length.
R=titzer@chromium.org
Bug: v8:8238
Change-Id: I8eafaac59295047a8540ebe256f00dff61ddae82
Reviewed-on: https://chromium-review.googlesource.com/c/1276426
Reviewed-by: Ben Titzer <titzer@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56571}
These test mainly the data structures of JSWeakFactory / JSWeakCell.
BUG=v8:8179
Change-Id: I20ffd07c18bbb2e21c69d11aa65d1e245203cc82
Reviewed-on: https://chromium-review.googlesource.com/c/1267939
Commit-Queue: Marja Hölttä <marja@chromium.org>
Reviewed-by: Sathya Gunasekaran <gsathya@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56569}
Simpliy track arrow-formal-parameters through errors and copy it from binding
pattern errors in AccumulateImpl rather than using a separate flag.
Change-Id: I8146718f490f3c9a5de3c30dec2aa3218831523f
Reviewed-on: https://chromium-review.googlesource.com/c/1275816
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56568}
A new CSA function, MoveElements() does an efficient memmove
operation when the ElementsKind or new-space status allows it.
A few other TNode cleanups applied in the file, for example,
preferring the StoreFixedDoubleArrayHole() function.
Change-Id: Ia0848c066eebbbbe321f81afe0cfa7df7567cbb7
Reviewed-on: https://chromium-review.googlesource.com/c/1268235
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56567}
The tests are scratching at the overall timeout.
TBR=mslekova@chromium.org
NOTRY=true
Change-Id: I817337eff76dcf36306d7f5ceaaef34eea7886d3
Reviewed-on: https://chromium-review.googlesource.com/c/1276385
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56565}
This will allow us to use ts-node to execute tests written
in TypeScript.
Bug: v8:7327
Change-Id: I0804db1f112448350c5e91135242e6ec6706d231
Reviewed-on: https://chromium-review.googlesource.com/c/1274086
Commit-Queue: Sigurd Schneider <sigurds@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56564}
In the preparser we don't need to worry about adding duplicate error objects to
the list, so we can simply unconditionally set the error flag. Restructure the
accumulator so we can also guarantee that we're checking the same flag as we're
setting.
Change-Id: I6a22cae468e77e5c6283c6fe937ca655f73991ac
Reviewed-on: https://chromium-review.googlesource.com/c/1275813
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56561}
The {MaskOut} operation is faster than using {operator~}, since
{operator~} needs to ensure to return a valid register list, so it
contains an additional AND operation with the mask of all registers.
Hence use {MaskOut} more consistently, and remove the now unused
{operator~} on {LiftoffRegList}.
R=ahaas@chromium.org
Bug: v8:6600
Change-Id: Icf072e564c4e391cafa2abeeba53cf275cee75f8
Reviewed-on: https://chromium-review.googlesource.com/c/1275810
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#56560}
