This reverts commit 80f5dfda01.
Reason for revert: Fails CSA verification: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux%20-%20verify%20csa/21766/overview
Original change's description:
> [no-wasm] Exclude src/wasm from compilation
>
> This is the biggest chunk, including
> - all of src/wasm,
> - torque file for wasm objects,
> - torque file for wasm builtins,
> - wasm builtins,
> - wasm runtime functions,
> - int64 lowering,
> - simd scala lowering,
> - WasmGraphBuilder (TF graph construction for wasm),
> - wasm frame types,
> - wasm interrupts,
> - the JSWasmCall opcode,
> - wasm backing store allocation.
>
> Those components are all recursively entangled, so I found no way to
> split this change up further.
>
> Some includes that were recursively included by wasm headers needed to
> be added explicitly now.
>
> backing-store-unittest.cc is renamed to wasm-backing-store-unittest.cc
> because it only tests wasm backing stores. This file is excluded from
> no-wasm builds then.
>
> R=jkummerow@chromium.org, jgruber@chromium.org, mlippautz@chromium.org, petermarshall@chromium.org
>
> Bug: v8:11238
> Change-Id: I7558f2d12d2dd6c65128c4de7b79173668c80b2b
> Cq-Include-Trybots: luci.v8.try:v8_linux64_no_wasm_compile_rel
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2742955
> Commit-Queue: Clemens Backes <clemensb@chromium.org>
> Reviewed-by: Peter Marshall <petermarshall@chromium.org>
> Reviewed-by: Toon Verwaest <verwaest@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
> Reviewed-by: Jakob Gruber <jgruber@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#73344}
Bug: v8:11238
Change-Id: I93672002c1faa36bb0bb5b4a9cc2032ee2ccd814
Cq-Include-Trybots: luci.v8.try:v8_linux64_no_wasm_compile_rel
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2752866
Auto-Submit: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#73346}
This will make it easier to generate builtin calls that require the
context to be passed in that register, because this can be represented
as a {LiftoffRegister} then.
R=thibaudm@chromium.org
Bug: v8:11453
Change-Id: I00995203cc936f23d1c01fdbf8f13eb929f228ae
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2752151
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73345}
This is the biggest chunk, including
- all of src/wasm,
- torque file for wasm objects,
- torque file for wasm builtins,
- wasm builtins,
- wasm runtime functions,
- int64 lowering,
- simd scala lowering,
- WasmGraphBuilder (TF graph construction for wasm),
- wasm frame types,
- wasm interrupts,
- the JSWasmCall opcode,
- wasm backing store allocation.
Those components are all recursively entangled, so I found no way to
split this change up further.
Some includes that were recursively included by wasm headers needed to
be added explicitly now.
backing-store-unittest.cc is renamed to wasm-backing-store-unittest.cc
because it only tests wasm backing stores. This file is excluded from
no-wasm builds then.
R=jkummerow@chromium.org, jgruber@chromium.org, mlippautz@chromium.org, petermarshall@chromium.org
Bug: v8:11238
Change-Id: I7558f2d12d2dd6c65128c4de7b79173668c80b2b
Cq-Include-Trybots: luci.v8.try:v8_linux64_no_wasm_compile_rel
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2742955
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73344}
Processing StateValues into operands is one of the most costly
parts of instruction selection. As it happens, StateValues are
shared by many nodes, and so we are unecessarily reprocessing
the same StateValues multiple times. This CL introduces caching
for the processed StateValues enabling very fast emitting of
operands for subsiquent instructions with the same StateValue.
The hitrate for the cache is higher than 90% on most optimizations.
BUG=v8:9684
Change-Id: I45db86dcbf22ab972b892f11c608b825aeb3ecf3
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2749634
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73342}
Liftoff defines many signatures of fixed size. This is currently done by
defining a fixed-size array on the stack, and then using this in the
signature definition. This is cumbersome and hard to read, since the
array contains return types and parameter types, and only the signature
definition separates the two. But also the order of those two sizes in
the signature is non-obvious and easy to get wrong.
This CL introduces a helper to define fixed-size signatures in a
"builder style", i.e. parameters and return types can be added
separately. The fixed-size array will be contained in the returned
class, so it will still be stack-allocated like before. The copies to
iteratively build up this array should be completely eliminated by the
compiler, so the binary code should look exactly the same.
R=ahaas@chromium.org
Bug: v8:11384
Change-Id: I167830d6c3429f535b7d1241920730498a9bb4c1
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2747505
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73341}
CsaLoadElimination failed to account for truncation when optimizing
loads. This CL extends the notion of compatible Loads and Stores to
include ({store}, {load}) pairs which both have integral representation
and {store}'s representation is no smaller than {load}'s. In case the
representations are not identical, it truncates and possibly
sign-extends {store} before forwarding it to {load}.
Additional change: Extend ObjectMayAlias with wasm allocating builtin
calls.
Bug: v8:11504
Change-Id: I43f89a13793b54477a33be18aaf346462aefa8e5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2739975
Reviewed-by: Georg Neis <neis@chromium.org>
Commit-Queue: Manos Koukoutos <manoskouk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73340}
Adds GCInfo folding that delegates GCInfo requests to the
parent-most object if finalizer semantics match.
Folding is disabled for builds that want exact object names
as those names are also managed through GCInfo objects.
Bug: chromium:1056170
Change-Id: I783aad930587853741da533d0b9b56ba160d0596
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748588
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73339}
The definition of {wasm::WasmCode} will not be available in no-wasm
builds, hence avoid any accesses to WasmCode for logging.
Drive-by: Inline enumeration of wasm modules for logging of existing
code, to avoid another #if.
R=petermarshall@chromium.org, jgruber@chromium.org
Bug: v8:11238
Change-Id: I3b78cf90f9ad155b5bea64e0941531aed2d4291a
Cq-Include-Trybots: luci.v8.try:v8_linux64_no_wasm_compile_rel
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2739978
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73338}
In https://chromium-review.googlesource.com/c/v8/v8/+/1866771 we added
a static regexp stack area to ensure a stack always exists. We
apparently forgot to update EnsureCapacity s.t. we skip
dynamically-allocating a stack when the static stack suffices.
Found by lizeb@, thanks!
Bug: v8:11540
Change-Id: Ie63b0b5e5959fbf0768cc3597f63943b1775fbf2
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2749015
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73337}
Resize() is not similar to realloc() in that it allocates a new object
when passed a nullptr object.
Avoid corner cases around Resize(nullptr, size) where size may be
problematic if non-null by just requiring a valid object. The caller
can perform the necesary nullptr check.
Bug: chromium:1056170
Change-Id: Ic05972ae67c2968fc3eb002a6302b44e56b41ab4
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2752147
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73336}
For use at locations where we know, that no GC can happen.
This avoids unnecessary handlifying of objects.
Bug: v8:11420
Change-Id: Ic549c56c4366060a6da3a3772dbd0aae23151eab
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2735394
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Patrick Thier <pthier@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73334}
Splitting a range at an instruction position can lead to incorrect code
generation. See the attached bug for a concrete example of that, in
particular comment 6.
The issue is when we add a gap move to connect the split ranges during
the ConnectLiveRanges phase. If the split position is a gap position,
the move coincides with the start of the range. But if the split
position is an instruction position, the move is inserted in the last
gap position, which is outside of the range. This violates assumptions
made during the main register allocation phase and can invalidate the
use of that register in a different range.
The fix proposed here works by moving the split position backwards to
the previous gap position. This ensures that the connecting gap move is
always at the start of the range that it defines.
R=sigurds@chromium.org
Bug: chromium:1182985
Change-Id: Ic4a9f56d5551f01cc91bece087d5ab3afd9b04fd
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2735396
Reviewed-by: Sigurd Schneider <sigurds@chromium.org>
Commit-Queue: Thibaud Michaud <thibaudm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73333}
Previously `setBreakpointByUrl` and friends would only filter based on
line number to find matching scripts. But that didn't work when there
were multiple scripts in the same line (i.e. minified HTML), and we'd
end up setting multiple breakpoints in different inline scripts, looking
for the next possible break location in each of them individually.
Fixed: chromium:1183664
Also-By: pfaffe@chromium.org, kimanh@chromium.org
Change-Id: I957811d30aa71609a38da75f33a24c0f720116f6
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2749155
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/master@{#73332}
... instead of Code. This is useful because usually the callers are
interested in having just a builtin ID but not the Code object.
This CL also makes Builtins::kNoBuiltinId a part of the Builtins::Name
enum.
Bug: v8:11527
Change-Id: I501e3e52dccc73cc7800f271939e0bf9fd00a975
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2749635
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73331}
OptimizedFrame::Summarize is used by debugger features etc
to inspect the frame of an optimized function (and the virtual frames
of functions that got inlined). It could end up materializing a JSArray
with the same backing store as one that would later get left-trimmed,
resulting in a dangling elements pointer. This CL fixes that by creating
a fresh copy of the elements store instead.
Bug: chromium:1182647
Change-Id: Iaf329464520a927b0ba33166cad2524d3752c450
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748593
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73330}
Robustify %CompileBaseline against fuzzing, and allowlist it in the
fuzzer.
Bug: v8:11420
Change-Id: I44947014c8c9362d80ea98636dbbaa5d07d6a177
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2739643
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73325}
This will make it easier to generate builtin calls that require the
context to be passed in that register, because this can be represented
as a {LiftoffRegister} then.
Drive-by: Fix a typo (x8 -> x28).
R=thibaudm@chromium.org
Bug: v8:11453
Change-Id: I7dcf7a0390dbde9713334d3da28ce6d79413a9f9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748585
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73324}
Building arm64 binaries on arm64 hosts works as long as you set
the correct options in args.gn. This patch teaches gm.py to do
that.
Building 32-bit arm binaries on arm64 hosts requires an extra
definition in snapshot_toolchain.gni (as well as some system
setup to support running 32-bit binaries).
Change-Id: I66c1f8f51932e2f5425033ef09181c31ea5d633e
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743889
Commit-Queue: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73323}
This is the V8 side change for https://crrev.com/c/2744048, which an
explicit "stack" property into the (renamed) "Expression" scope to align
the behavior and appearance of that scope with the other scopes.
Fixed: chromium:1159310
Bug: chromium:1071432, chromium:1159307
Change-Id: Ic070c50b674d8c1cff4a93538f708cc431c3f2cb
Screenshot: https://imgur.com/a/ryFiOGA.png
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748591
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Philip Pfaffe <pfaffe@chromium.org>
Auto-Submit: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Philip Pfaffe <pfaffe@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73321}
If the try block cannot throw, we don't expect an exception on the stack
when we exit the block.
R=clemensb@chromium.org
Bug: v8:11537
Change-Id: I5f2ea6df2d14205b3ad2cd61a5132bc0b3db6cc8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748583
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Thibaud Michaud <thibaudm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73320}
This fixes a handle leaking out of
V8HeapExplorer::ExtractPropertyReferences when enabling
v8_dict_mode_prototypes, which breaks the
test-heap-profiler/NoHandleLeaks cctest.
Bug: v8:11388
Change-Id: I911f60e3701277540ae5a63fa59f3ac2f52879ef
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748580
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Marja Hölttä <marja@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Frank Emrich <emrich@google.com>
Cr-Commit-Position: refs/heads/master@{#73319}
Reporting marked bytes after atomic sweeping means we might be missing
allocations in case finalizers are allocating during sweeping.
Instead report marked bytes and marking time directly to
LocalEmbedderHeapTracer as soon as marking is done.
Bug: chromium:1056170
Change-Id: Ie770f077d2eec10dea182a503a7cd514d3b66baf
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748579
Commit-Queue: Omer Katz <omerkatz@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73318}
Loops will always execute a stack check in the header, so having the
instance cached in a register is handy. Instead of clearing it before
entering a loop, ensure that backward jumps to the loop header move the
instance into the right register.
R=thibaudm@chromium.org
Bug: v8:11336
Change-Id: I16cb13457438b7a1603182d56a3d2ea99d670911
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743892
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73317}
Fix to make the macro actually use the relaxed accessors instead of the
atomic ones.
Bug: v8:7790, v8:11539
Change-Id: Ic74ee4084a76176feac138716e10b88a3a5e1a1d
Fixed: v8:11539
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748088
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Auto-Submit: Santiago Aboy Solanes <solanes@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73316}
Some ValueKinds were defined based on the size of a system pointer or
the size of a tagged value. In order to be able to reuse those
definitions in both LiftoffAssembler and LiftoffCompiler, define them as
public constants on LiftoffAssembler.
Also, avoid the "ValueType" suffix, since they are not ValueTypes, but
ValueKinds.
R=jkummerow@chromium.org
Change-Id: I38f9c9b6c4e6592d31ee58466b786bf24a55f19c
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743890
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73315}
Bug: v8:7790
Change-Id: I1e7448c6583a36b2311474e14f4611935e2aa79f
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2748076
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Santiago Aboy Solanes <solanes@chromium.org>
Commit-Queue: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73314}
In contrast to wasm modules, asm.js modules have an empty source URL.
Thus loosen a DCHECK and handle the nullptr source_url correctly.
Also add regression tests that check that we don't crash. Those can
later be extended to check that the profile looks as expected; for now
they only check that we terminate.
R=bmeurer@chromium.org
Bug: chromium:1185919
Change-Id: I6b879f540a2c3647920ad2970efcf7c94712d8c7
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2745895
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73313}
Rolling v8/build: 077f859..389d3f3
Rolling v8/buildtools: a5cc843..368c7dd
Rolling v8/third_party/aemu-linux-x64: dBlHF6-1NU-vr6DU068Y8_WQHCWdk_yovRmg225wIr0C..oXIWsntGxFugYHN4Qb0Ve-IRPq4Rq1FxFYNiuEKnP0kC
Rolling v8/third_party/depot_tools: 75c9832..57af8c3
Rolling v8/third_party/icu: e6379c8..d879aac
Rolling v8/tools/clang: 630ab8a..fc54e6bTBR=v8-waterfall-sheriff@grotations.appspotmail.com
Change-Id: I23f92a7e958b90228f898df85e3efa87c9429a73
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2746751
Reviewed-by: v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com>
Commit-Queue: v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com>
Cr-Commit-Position: refs/heads/master@{#73311}
This is done with a script that does something like:
files=$(ag 'v\d\d?x\d\d?[._]?all_?true' -l)
sed -i 's/V\(8x16\|16x8\|32x4\|64x2\)\([._]\?\)\([aA]ll_\?[tT]rue\)/I\1\2\3/g' $(files)
sed -i 's/v\(8x16\|16x8\|32x4\|64x2\)\([._]\?\)\([aA]ll_\?[tT]rue\)/i\1\2\3/g' $(files)
And manual fixups in test-run-wasm-simd.cc and wasm-opcodes-inl.h.
Bug: v8:10946
Change-Id: Ib5dad388dd6dd9cd0fb575ad961dffc189a2e6ef
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2740488
Reviewed-by: Bill Budge <bbudge@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73309}
They were using a fixed offset of pc+2, but since the instructions can
be multiple bytes long (leb128 encoded), it should be using *len.
Drive-by fix to add missing instructions to wasm-module-builder.js.
Bug: chromium:1185323,chromium:1185492
Change-Id: I12f396cc2969ecc284aba35b94b1bc5640f12277
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2745977
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73308}
The test ensures that in-construction objects that have been found
through a write barrier are properly processed (marked + trace) when
finalizing the collection conservatively with a different stack.
This is a test for https://crrev.com/c/2744074
Bug: chromium:1056170
Change-Id: I8099bca1fb9025a315a8f0a3530aac822d1c45d2
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2745334
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73306}
It could happen that a background thread expands the heap by one page,
but by the time the thread tries to allocate on it the space is already
used by other background threads. If this happens three times in a row,
V8 would crash with an OOM error. This CL prevents such situations by
always allocating the object immediately at area_start().
Bug: v8:10315
Change-Id: I6390c84e742bf4105e70e930c21557ff1f4d952d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743881
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73305}
The wasm instance will initially be in kWasmInstanceRegister, and for
each call we also need to put it in that register. Hence, when getting a
new register to cache the instance, prefer that register, if it is
available.
R=thibaudm@chromium.org
Bug: v8:11336
Change-Id: Ie7026c4c7c5e4b825b9ab310839f0273bd3ce7f1
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743885
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73303}
... and after creating the debug side table entry. The safepoint and the
debug side table entry should be generated right after the call, so that
they are associated with the return address of the previous call.
R=clemensb@chromium.org
Bug: v8:11453
Change-Id: I71395851c5a7f4e2c873907454245c9d04f972f1
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2739629
Commit-Queue: Thibaud Michaud <thibaudm@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73302}
The generated Torque files class-verifiers.cc and class-debug-readers.cc
currently include files which are part of v8_initializers, despite being
used in unrelated build targets. This change removes the unnecessary
inclusions. There is still a lot of code included via all-objects-inl.h,
but that's because these files require full class definitions for every
object type.
Bug: v8:11528
Change-Id: Ib26496f2a30ef576f1101636e0aca2cafbfd1f37
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743087
Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73301}
This CL removes the caching of the stack slot used for the fallback mechanism
in V8, as the current implementation is incorrect and needs to be reworked.
Bug: chromium:1185753
Change-Id: I9f77bc42bfd649e0dbcd294b000b48c928cf99d0
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743886
Commit-Queue: Maya Lekova <mslekova@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73300}
Lazy native accessors require special handling to rewrite
the accessor into a data property, so transition to a
slow handler for this case.
Bug: v8:11485
Change-Id: I01636c6e624562619a216fea5e836ae85c7da93f
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743882
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Sathya Gunasekaran <gsathya@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73299}
This ensures that after the function-entry stack check, the instance
will still be available in a register. The cost is having to reload it
from the stack in the OOL code for the stack check, even though it is
not clear if that register will still be used.
This does not affect code size significantly (~0.25% reduction), but can
improve performance a little bit if there are memory accesses or other
instructions that require the instance right at the beginning of the
function.
R=thibaudm@chromium.org
Bug: v8:11336
Change-Id: Ib72db172813d55120f527b31014b69a734934ff3
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2743878
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#73298}
