This CL adds a moving GC-safe, JS-exposed mutex behind the
--harmony-struct flag. It uses a ParkingLot-inspired algorithm and
each mutex manages its own waiter queue.
For more details, please see the design doc: https://docs.google.com/document/d/1QHkmiTF770GKxtoP-VQ1eKF42MpedLUeqiQPfCqus0Y/edit?usp=sharing
Bug: v8:12547
Cq-Include-Trybots: luci.v8.try:v8_linux64_tsan_rel_ng
Change-Id: Ic58f8750d2e14ecd573173d17d5235a136bedef9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3595460
Commit-Queue: Shu-yu Guo <syg@chromium.org>
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Adam Klein <adamk@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80789}
Due to collections with inlined storage, Oilpan still supports on-stack
Members, which are always compressed if pointer compression is enabled.
This CL scans halfwords (together with full words) on stack to find
potential pointers. Since on-heap pointers can only be compressed and
in-construction objects always reside on heap, only halfwords need to be
scanned for them.
The alternative potential followup approaches:
1) Use a separate uncompressed type for pointer in inlined collections;
2) Dynamically register regions of stack containing compressed pointers.
Bug: chromium:1325007
Change-Id: Ia706fd8e7383d30aff11f4014faa9edd3d289a55
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3644959
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Anton Bikineev <bikineev@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80747}
Math between disparate enums is deprecated. Use constexprs instead.
This requires switching some caller code to work with the new non-enum
constants also.
Bug: chromium:1284275
Change-Id: Ifb3c8757ed62e2a0966120f830f0a7e282b53a16
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3661148
Auto-Submit: Peter Kasting <pkasting@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Peter Kasting <pkasting@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80722}
Original CL got reverted, this time the failing test should be fixed.
Bug: v8:12578
Change-Id: Id2d8801f07742e8b00884fefec8200e4270f4250
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3657434
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80668}
A few of LogTests have been crashing intermittently
after they were moved to unittests in this CL:
https://crrev.com/c/3616424
Will re-enable once issue is investigated.
Change-Id: I53435596274c935c028a625b610c54eadda9d1de
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3647092
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Milad Farazmand <mfarazma@redhat.com>
Cr-Commit-Position: refs/heads/main@{#80551}
We currently have a BitVector implementation which is used a lot by the
two (mid-tier and top-tier) register allocators. Their size is the
number of virtual registers or the number of blocks in the function. If
one of those numbers gets huge, the BitVector does not perform well any
more, and it consumes huge amounts of memory (we see up to several GBs
for huge Wasm functions).
This CL introduces a SparseBitVector implementation with a compatible
interface, meant to replace the BitVector implementation. Usages will be
introduced in follow-up CLs, first for the mid-tier allocator, then
top-tier. This will allow us to assess performance changes better, and
revert individual usages.
R=mslekova@chromium.org
Bug: chromium:1313379, v8:12780
Change-Id: I804311e0c188526961f70e88a43dd1ea26497cda
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3634780
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80546}
Bug: v8:12868
This wires up the parser and the decoder interface for stringref. All
of the interfaces throw UNIMPLEMENTED, however.
Change-Id: If8cb131032e425a5672f793c6e4c24ddd188aebc
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3645115
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Commit-Queue: Andy Wingo <wingo@igalia.com>
Cr-Commit-Position: refs/heads/main@{#80545}
This CL adds to the existing experimental implementation of the
object start bitmap, that is evaluated as a mechanism for resolving
inner pointers (behind the flag v8_enable_conservative_stack_scanning).
It fixes method ObjectStartBitmap::FindBasePtr to ensure that the
correct base pointer is returned, even if the bitmap is not fully
populated (e.g., with object evacuation or inline object allocation).
This method now recalculates the part of the bitmap that is
required for returning the correct result, by iterating through
objects of the page. A special constructor has been introduced to the
PagedSpaceObjectIterator for this purpose.
It also moves the existing inline methods of ObjectStartBitmap to a
new -inl.h header file, to avoid circular dependencies.
Bug: v8:12851
Change-Id: Iabd0df020bee3bb63ef9d4888591b25d24d79dd9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3641179
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Nikolaos Papaspyrou <nikolaos@chromium.org>
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80538}
MSVC does not implicitly convert the std::array iterator to a char*,
hence explicitly use the {data()} accessor instead of {begin()}, which
is cleaner anyway.
R=mlippautz@chromium.org
Bug: chromium:1323177
Change-Id: I65c6836889eb57a18b655cb9e6c98008a8fbcaac
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3644794
Auto-Submit: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Deepti Gandluri <gdeepti@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80533}
This is more consistent with similar features, for example
V8_ENABLE_WEBASSEMBLY or V8_ENABLE_MAGLEV.
Drive-by: remove V8_SANDBOX_IS_AVAILABLE as it's no longer needed.
Bug: v8:10391
Change-Id: I8658c5b0c331a4c73892737083b2c2f9b8f84056
Cq-Include-Trybots: luci.v8.try:v8_linux64_heap_sandbox_dbg_ng,v8_linux_arm64_sim_heap_sandbox_dbg_ng
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3647355
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Auto-Submit: Samuel Groß <saelo@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80530}
This introduces a class which can be used for formatting dynamic values
into a constant-size, stack-allocated array. You get ostream-style code
but printf-style performance, and in particular no dynamic allocation.
This makes this class also suitable to be used in OOM or other fatal
situations where we cannot rely on dynamic memory allocation to still
work.
Using FormattedString will automatically compute the format string
depending on the types. It also computes the maximum size of the output.
Last but not least, it makes the code a lot more readable than
traditional printf style printing.
R=mlippautz@chromium.org
Bug: chromium:1323177
Change-Id: I47228b3603c694c1fa23516dd3f1c57e39c0ca35
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3644622
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80529}
Now that we require C++17 support, we can just use the standard
static_assert without message, instead of our STATIC_ASSERT macro.
R=leszeks@chromium.org
Bug: v8:12425
Change-Id: I1d4e39c310b533bcd3a4af33d027827e6c083afe
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3647353
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Hannes Payer <hpayer@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80524}
With this CL, the decompression simply becomes:
movsxd rax, edi
add rax, rax
and rax, qword ptr fs:[base@TPOFF]
Bug: chromium:1325007
Change-Id: I931e4e667a9b9697671bccf14575420f8cb705e8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3629871
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Anton Bikineev <bikineev@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80521}
With caged heap enabled, we can halve Member<> by storing only the least
significant half. The base of the heap is stored in a thread local
variable. The feature has therefore an implication that only single heap
is allowed per thread.
The feature is gated by the new GN arg:
cppgc_enable_pointer_compression.
Bug: chromium:1325007
Change-Id: Ic7f1ecb7b9ded57caad63d95bbc8e8ad6ad65031
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2739979
Reviewed-by: Almothana Athamneh <almuthanna@chromium.org>
Commit-Queue: Almothana Athamneh <almuthanna@chromium.org>
Commit-Queue: Tamer Tas <tmrts@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Tamer Tas <tmrts@chromium.org>
Auto-Submit: Anton Bikineev <bikineev@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80518}
Add new opcode numbers for relaxed-simd opcodes as in:
https://github.com/WebAssembly/relaxed-simd/pull/61, and updates
for decoded opcodes larger than two bytes.
Bug: v8:12284
Change-Id: I0bff22812243a39599bfeee3c0bc69171dd3c7d2
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3625835
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Deepti Gandluri <gdeepti@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80432}
Many uses of "volatile" are deprecated in C++20 because they don't
actually do anything. Remove "volatile" in these cases.
Bug: chromium:1284275
Change-Id: I64a3989d73f25e0cd933375dd6fa0b3f2b3acb54
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3630343
Auto-Submit: Peter Kasting <pkasting@chromium.org>
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Commit-Queue: Nico Hartmann <nicohartmann@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80428}
Allocation granularity (8 bytes on 64bit platforms) is already subsumed
by whatever range size we choose.
Bug: chromium:1029379
Change-Id: Iab95e6f36955c9ffbbe9ea0c98cb9d1f8dd0337b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3629869
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Auto-Submit: Anton Bikineev <bikineev@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80415}
- Repurpose flag `treat_global_objects_as_roots` when taking a heap
snapshot for toggling whether internals should be exposed (to
`hide_internals`).
- Use the toggle in creating heap snapshots for exposing class names
as object names for C++ objects that have not explicitly been given a
name.
Change-Id: I77d71babfdfe53269964fe81ed985037a431c28b
Bug: chromium:1321620
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3623740
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80391}
This prevents ambiguity errors in C++20 due to ADL when casting types in
std::, which gains std::bit_cast<>().
Bug: chromium:1284275
Change-Id: I25046d1952a9304852e481ad8b84049c6769c289
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3625838
Auto-Submit: Peter Kasting <pkasting@chromium.org>
Reviewed-by: Adam Klein <adamk@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Adam Klein <adamk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80378}
Before this CL, the heap object name of unnamed objects(those not
inheriting from NameProvider) would be solely determined by whether the
build-time configuration cppgc_enable_object_names is enabled.
This patch adds a way to override that value at runtime. This is
useful for preserving default behavior with custom builds but at the
same time allow them to still enable the feature.
Bug: chromium:1321620
Change-Id: I3aa06db15e58d9ba9773be6797572f17f007e9ee
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3620279
Reviewed-by: Anton Bikineev <bikineev@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Omer Katz <omerkatz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80338}
To be consistent with the all the other tiers and avoid confusion, we
rename --opt to ---turbofan, and --always-opt to --always-turbofan.
Change-Id: Ie23dc8282b3fb4cf2fbf73b6c3d5264de5d09718
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3610431
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Camillo Bruni <cbruni@chromium.org>
Reviewed-by: Jakob Linke <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80336}
This fixes:
runtime error: upcast of misaligned address 0x000000000001
for type 'cppgc::internal::(anonymous namespace)::SimpleGCed<64>', which
requires 8 byte alignment
Bug: chromium:1029379
Change-Id: Id03ce022e72fc07a18c171d4cf5e42f50cb684f1
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3622918
Auto-Submit: Anton Bikineev <bikineev@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80332}
Due to the consistent overhead of snapshot checksum verification
we ideally want to avoid it all-together. However there are still enough
devices out there that suffer from corrupted snapshots that might
cause hard to debug heap corruptions.
This CL exposes the calculated (dummy value for now) and the expected
snapshot checksum as a crash key, so it can be easily consulted during
investigation.
Note: The calculated crash key contains 0x0 for now as a dummy value. We
will come up with a strategy later-on to limit the overhead of
calculating the checksum.
Bug: v8:12195
Change-Id: I6da6d74c035cb6f9b0edae212a36e6c41c048a5b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3605813
Reviewed-by: Jakob Linke <jgruber@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Camillo Bruni <cbruni@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80325}
This is a reland of commit 9145388055
Fixed: properly reference the ClearedValue in CSA (i.e. without
the cage_base upper 32 bits).
Original change's description:
> Reland "[osr] Use the new OSR cache"
>
> This is a reland of commit 91da38831d
>
> Fixed: Use an X register for JumpIfCodeTIsMarkedForDeoptimization
> on arm64.
>
> Original change's description:
> > [osr] Use the new OSR cache
> >
> > This CL switches over our OSR system to be based on the feedback
> > vector osr caches.
> >
> > - OSRing to Sparkplug is fully separated from OSR urgency. If
> > SP code exists, we simply jump to it, no need to maintain an
> > installation request.
> > - Each JumpLoop checks its dedicated FeedbackVector cache slot.
> > If a valid target code object exists, we enter it *without*
> > calling into runtime to fetch the code object.
> > - Finally, OSR urgency still remains as the heuristic for
> > requesting Turbofan OSR compile jobs. Note it no longer has a
> > double purpose of being a generic untargeted installation
> > request.
> >
> > With the new system in place, we can remove now-unnecessary
> > hacks:
> >
> > - Early OSR tierup is replaced by the standard OSR system. Any
> > present OSR code is automatically entered.
> > - The synchronous OSR compilation fallback is removed. With
> > precise installation (= per-JumpLoop-bytecode) we no longer
> > have the problem of 'getting unlucky' with JumpLoop/cache entry
> > mismatches. Execution has moved on while compiling? Simply spawn
> > a new concurrent compile job.
> > - Remove the synchronous (non-OSR) Turbofan compile request now
> > that we always enter available OSR code as early as possible.
> > - Tiering into Sparkplug no longer messes with OSR state.
> >
> > Bug: v8:12161
> > Change-Id: I0a85e53d363504b7dac174dbaf69c03c35e66700
> > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3596167
> > Commit-Queue: Jakob Linke <jgruber@chromium.org>
> > Auto-Submit: Jakob Linke <jgruber@chromium.org>
> > Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> > Commit-Queue: Leszek Swirski <leszeks@chromium.org>
> > Cr-Commit-Position: refs/heads/main@{#80147}
>
> Bug: v8:12161
> Change-Id: Ib3597cf1d99cdb5d0f2c5ac18e311914f376231d
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3606232
> Auto-Submit: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> Commit-Queue: Leszek Swirski <leszeks@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80167}
Bug: v8:12161,chromium:1320189
Change-Id: Ibd9a2ab61f51ebb32a3f5a66f7c602faead71c3e
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3620273
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Jakob Linke <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80306}
The CL prepares the sources and the tests for enabling
cppgc_enable_young_generation by default. The static initializer
in YoungGenerationEnabler (due to v8::base::Mutex) changed to be lazy.
The tests are now checking the runtime flag.
Bug: chromium:1029379
Change-Id: I1497a3dd2b8d62c1acd48496821f07324b7944d5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3616726
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Auto-Submit: Anton Bikineev <bikineev@chromium.org>
Commit-Queue: Anton Bikineev <bikineev@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80304}
This reverts commit 9145388055.
Reason for revert: Breaking the Fuchsia Deterministic Builder
Original change's description:
> Reland "[osr] Use the new OSR cache"
>
> This is a reland of commit 91da38831d
>
> Fixed: Use an X register for JumpIfCodeTIsMarkedForDeoptimization
> on arm64.
>
> Original change's description:
> > [osr] Use the new OSR cache
> >
> > This CL switches over our OSR system to be based on the feedback
> > vector osr caches.
> >
> > - OSRing to Sparkplug is fully separated from OSR urgency. If
> > SP code exists, we simply jump to it, no need to maintain an
> > installation request.
> > - Each JumpLoop checks its dedicated FeedbackVector cache slot.
> > If a valid target code object exists, we enter it *without*
> > calling into runtime to fetch the code object.
> > - Finally, OSR urgency still remains as the heuristic for
> > requesting Turbofan OSR compile jobs. Note it no longer has a
> > double purpose of being a generic untargeted installation
> > request.
> >
> > With the new system in place, we can remove now-unnecessary
> > hacks:
> >
> > - Early OSR tierup is replaced by the standard OSR system. Any
> > present OSR code is automatically entered.
> > - The synchronous OSR compilation fallback is removed. With
> > precise installation (= per-JumpLoop-bytecode) we no longer
> > have the problem of 'getting unlucky' with JumpLoop/cache entry
> > mismatches. Execution has moved on while compiling? Simply spawn
> > a new concurrent compile job.
> > - Remove the synchronous (non-OSR) Turbofan compile request now
> > that we always enter available OSR code as early as possible.
> > - Tiering into Sparkplug no longer messes with OSR state.
> >
> > Bug: v8:12161
> > Change-Id: I0a85e53d363504b7dac174dbaf69c03c35e66700
> > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3596167
> > Commit-Queue: Jakob Linke <jgruber@chromium.org>
> > Auto-Submit: Jakob Linke <jgruber@chromium.org>
> > Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> > Commit-Queue: Leszek Swirski <leszeks@chromium.org>
> > Cr-Commit-Position: refs/heads/main@{#80147}
>
> Bug: v8:12161
> Change-Id: Ib3597cf1d99cdb5d0f2c5ac18e311914f376231d
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3606232
> Auto-Submit: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> Commit-Queue: Leszek Swirski <leszeks@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80167}
Bug: v8:12161
Change-Id: I73e2d98660e9edfbe07a152a14402380ea9227de
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3615219
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Deepti Gandluri <gdeepti@chromium.org>
Owners-Override: Deepti Gandluri <gdeepti@chromium.org>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Cr-Commit-Position: refs/heads/main@{#80287}
The CL uses the different scheme to enable the generational barrier. The
separate global counter (is_enabled_) keeps track of the number of heaps
that enable generational GC. If at least one of the heaps enables the
generational GC, the counter will enable the write barrier. Technically,
the counter could be merged with WriteBarrier::is_enabled_, but having a
separate variable allows us to keep DCHECKs if generational barrier is
enabled.
Bug: chromium:1029379
Change-Id: Iafaa76f96acb18a73f8bde7231434e68c04cb683
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3616518
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Anton Bikineev <bikineev@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80281}
The CL introduces a new option --cppgc-young-generation. This option
can't be enabled statically, because V8 options are parsed after heap
initialization. The CL changes minor GC so that it can be enabled
dynamically. The way it works is as follows:
- the user calls YoungGenerationEnabler::Enable();
- a heap checks in the next atomic pause whether the flag was enabled;
- if so, the heap enables young generation for itself.
To avoid barrier regressions without young-generation enabled, the CL changes the meaning of the global flag is-any-incremental-or-concurrent-marking to is-barrier-enabled.
The runtime option would enable us to test young generation on try-
and performance-bots.
Bug: chromium:1029379
Change-Id: I664cccdcd208225ffcbf9901f1284b56d088c5c3
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3607993
Commit-Queue: Anton Bikineev <bikineev@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80274}
This is a reland of commit 9d31f8663a
There were issues with --future flag implications on M1.
Original change's description:
> [rwx][mac] Support fast W^X permission switching on Apple Silicon (M1)
>
> ... for V8 code space. The feature is currently disabled.
>
> In order to use fast W^X permission switching we must allocate
> executable pages with readable writable executable permissions (RWX).
> However, MacOS on ARM64 ("Apple M1"/Apple Silicon) prohibits further
> permission changing of RWX memory pages. This means that the code page
> headers must be allocated with RWX permissions too because otherwise
> it wouldn't be possible to allocate a large code page over the freed
> regular code page and vice versa.
>
> When enabled, the new machinery works as follows:
>
> 1) when memory region is reserved for allocating executable pages, the
> whole region is committed with RWX permissions and then decommitted,
> 2) since reconfiguration of RWX page permissions is not allowed on
> MacOS on ARM64 ("Apple M1"/Apple Silicon), there must be no attempts
> to change them,
> 3) the request to set RWX permissions in the executable page region
> just recommits the pages without changing permissions (see (1), they
> were already allocated as RWX and then discarded),
> 4) in order to make executable pages inaccessible one must use
> OS::DiscardSystemPages() instead of OS::DecommitPages() or
> setting permissions to kNoAccess because the latter two are not
> allowed by the MacOS (see (2)).
> 5) since code space page headers are allocated as RWX pages it's also
> necessary to switch between W^X modes when updating the data in the
> page headers (i.e. when marking, updating stats, wiring pages in
> lists, etc.). The new CodePageHeaderModificationScope class is used
> in the respective places. On unrelated configurations it's a no-op.
>
> The fast permission switching can't be used for V8 configuration with
> enabled pointer compression and disabled external code space because
> a) the pointer compression cage has to be reserved with MAP_JIT flag
> which is too expensive,
> b) in case of shared pointer compression cage if the code range will
> be deleted while the cage is still alive then attempt to configure
> permissions of pages that were previously set to RWX will fail.
>
> This also CL extends the unmapper unit tests with permissions tracking
> for discarded pages.
>
> Bug: v8:12797
> Change-Id: Idb28cbc481306477589eee9962d2e75167d87c61
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3579303
> Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
> Reviewed-by: Clemens Backes <clemensb@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Commit-Queue: Igor Sheludko <ishell@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80238}
Bug: v8:12797
Change-Id: I0fe86666f31bad37d7074e217555c95900d2afba
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3610433
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80259}
There are three ways to parse /proc/self/maps in platform-linux.cc,
remove one to use common code. In the process, add a unit test, and fix
some issues in the latest iteration of /proc/self/maps parsing.
Change-Id: I4701ea49fe8cce53aea0179e194dc48fbebb2ff5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3605226
Commit-Queue: Benoit Lize <lizeb@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80258}
For some reason the compiler was optimizing away the reference to the
object in WeakContainerTest.ConservativeGCTracesWeakContainer and thus
not finding it conservatively.
This CL revises the tests such that the compiler is no longer able to
optimize references away.
Bug: v8:12824
Change-Id: Ie598a1cf1124c2983a6c61fd4e990734d36f5832
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3610627
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Omer Katz <omerkatz@chromium.org>
Reviewed-by: Anton Bikineev <bikineev@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80257}
This reverts commit 9d31f8663a.
Reason for revert: crashes on Mac/arm64 bots:
https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Mac%20-%20arm64%20-%20debug/5923/overview
Original change's description:
> [rwx][mac] Support fast W^X permission switching on Apple Silicon (M1)
>
> ... for V8 code space. The feature is currently disabled.
>
> In order to use fast W^X permission switching we must allocate
> executable pages with readable writable executable permissions (RWX).
> However, MacOS on ARM64 ("Apple M1"/Apple Silicon) prohibits further
> permission changing of RWX memory pages. This means that the code page
> headers must be allocated with RWX permissions too because otherwise
> it wouldn't be possible to allocate a large code page over the freed
> regular code page and vice versa.
>
> When enabled, the new machinery works as follows:
>
> 1) when memory region is reserved for allocating executable pages, the
> whole region is committed with RWX permissions and then decommitted,
> 2) since reconfiguration of RWX page permissions is not allowed on
> MacOS on ARM64 ("Apple M1"/Apple Silicon), there must be no attempts
> to change them,
> 3) the request to set RWX permissions in the executable page region
> just recommits the pages without changing permissions (see (1), they
> were already allocated as RWX and then discarded),
> 4) in order to make executable pages inaccessible one must use
> OS::DiscardSystemPages() instead of OS::DecommitPages() or
> setting permissions to kNoAccess because the latter two are not
> allowed by the MacOS (see (2)).
> 5) since code space page headers are allocated as RWX pages it's also
> necessary to switch between W^X modes when updating the data in the
> page headers (i.e. when marking, updating stats, wiring pages in
> lists, etc.). The new CodePageHeaderModificationScope class is used
> in the respective places. On unrelated configurations it's a no-op.
>
> The fast permission switching can't be used for V8 configuration with
> enabled pointer compression and disabled external code space because
> a) the pointer compression cage has to be reserved with MAP_JIT flag
> which is too expensive,
> b) in case of shared pointer compression cage if the code range will
> be deleted while the cage is still alive then attempt to configure
> permissions of pages that were previously set to RWX will fail.
>
> This also CL extends the unmapper unit tests with permissions tracking
> for discarded pages.
>
> Bug: v8:12797
> Change-Id: Idb28cbc481306477589eee9962d2e75167d87c61
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3579303
> Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
> Reviewed-by: Clemens Backes <clemensb@chromium.org>
> Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
> Commit-Queue: Igor Sheludko <ishell@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80238}
Bug: v8:12797
Change-Id: Ic07948e036db36326d464a2a901d052aa060a406
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3611665
Commit-Queue: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Auto-Submit: Adam Klein <adamk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80239}
... for V8 code space. The feature is currently disabled.
In order to use fast W^X permission switching we must allocate
executable pages with readable writable executable permissions (RWX).
However, MacOS on ARM64 ("Apple M1"/Apple Silicon) prohibits further
permission changing of RWX memory pages. This means that the code page
headers must be allocated with RWX permissions too because otherwise
it wouldn't be possible to allocate a large code page over the freed
regular code page and vice versa.
When enabled, the new machinery works as follows:
1) when memory region is reserved for allocating executable pages, the
whole region is committed with RWX permissions and then decommitted,
2) since reconfiguration of RWX page permissions is not allowed on
MacOS on ARM64 ("Apple M1"/Apple Silicon), there must be no attempts
to change them,
3) the request to set RWX permissions in the executable page region
just recommits the pages without changing permissions (see (1), they
were already allocated as RWX and then discarded),
4) in order to make executable pages inaccessible one must use
OS::DiscardSystemPages() instead of OS::DecommitPages() or
setting permissions to kNoAccess because the latter two are not
allowed by the MacOS (see (2)).
5) since code space page headers are allocated as RWX pages it's also
necessary to switch between W^X modes when updating the data in the
page headers (i.e. when marking, updating stats, wiring pages in
lists, etc.). The new CodePageHeaderModificationScope class is used
in the respective places. On unrelated configurations it's a no-op.
The fast permission switching can't be used for V8 configuration with
enabled pointer compression and disabled external code space because
a) the pointer compression cage has to be reserved with MAP_JIT flag
which is too expensive,
b) in case of shared pointer compression cage if the code range will
be deleted while the cage is still alive then attempt to configure
permissions of pages that were previously set to RWX will fail.
This also CL extends the unmapper unit tests with permissions tracking
for discarded pages.
Bug: v8:12797
Change-Id: Idb28cbc481306477589eee9962d2e75167d87c61
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3579303
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80238}
tests where moved from cctest to unittests under this cl:
https://crrev.com/c/3607370
Bug: v8:12781
Change-Id: If625e0dda51034e731c5e7fe87d591dce9804888
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3611182
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Milad Farazmand <mfarazma@redhat.com>
Cr-Commit-Position: refs/heads/main@{#80227}
Add flag --always-use-string-forwarding-table to always use the
forwarding table (usually only used for shared strings) instead of
ThinString migrations initially (during GC strings will be migrated
to normal ThinStrings). The goal is to get more coverage of this code
that is designed for shared strings.
Bug: v8:12007
Change-Id: I7eb2e5ccf0018c4ac349611aebe337d8288de5c8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3536650
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Shu-yu Guo <syg@chromium.org>
Commit-Queue: Patrick Thier <pthier@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80206}
Introduce a bottleneck for right-trimming an object. In a subsequent CL
we will use this method to update the cached size of invalidated
objects.
This CL also tries to clean-up the various CreateFillerObjectAt
methods. CreateFillerObjectAtRaw is now the internal method for all
these methods. After moving right-trimming to NotifyObjectSizeChange,
both CreateFillerObjectAt and CreateFillerObjectAtBackground don't need
those arguments for clearing slots or memory anymore.
Bug: v8:12578, chromium:1316289
Change-Id: I6ff0bfaced3e0a1765152700e68a4ad33a155723
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3607992
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80200}
It's necessary to support fast W^X permission switching on MacOS on
ARM64 ("Apple M1"/Apple Silicon) where permission modification of RWX
pages to anything else is prohibited.
On all the other architectures/platforms RecommitPages() is equivalent
to SetPermissions().
The new API will be used in a follow-up CLs.
Bug: v8:12797
Change-Id: Id0d8b8c42c81b80cd8fa6b47c227680d7d1f9b10
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3606231
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Samuel Groß <saelo@chromium.org>
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80190}
When setting a range for a newly allocated lab, consider adjacent cards.
If either is young, don't mark it as kMixed.
Bug: chromium:1029379
Change-Id: If7d1d920dd5769679de68800eae61f3a8dc1eb17
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3584116
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Commit-Queue: Anton Bikineev <bikineev@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80175}
This is a reland of commit 91da38831d
Fixed: Use an X register for JumpIfCodeTIsMarkedForDeoptimization
on arm64.
Original change's description:
> [osr] Use the new OSR cache
>
> This CL switches over our OSR system to be based on the feedback
> vector osr caches.
>
> - OSRing to Sparkplug is fully separated from OSR urgency. If
> SP code exists, we simply jump to it, no need to maintain an
> installation request.
> - Each JumpLoop checks its dedicated FeedbackVector cache slot.
> If a valid target code object exists, we enter it *without*
> calling into runtime to fetch the code object.
> - Finally, OSR urgency still remains as the heuristic for
> requesting Turbofan OSR compile jobs. Note it no longer has a
> double purpose of being a generic untargeted installation
> request.
>
> With the new system in place, we can remove now-unnecessary
> hacks:
>
> - Early OSR tierup is replaced by the standard OSR system. Any
> present OSR code is automatically entered.
> - The synchronous OSR compilation fallback is removed. With
> precise installation (= per-JumpLoop-bytecode) we no longer
> have the problem of 'getting unlucky' with JumpLoop/cache entry
> mismatches. Execution has moved on while compiling? Simply spawn
> a new concurrent compile job.
> - Remove the synchronous (non-OSR) Turbofan compile request now
> that we always enter available OSR code as early as possible.
> - Tiering into Sparkplug no longer messes with OSR state.
>
> Bug: v8:12161
> Change-Id: I0a85e53d363504b7dac174dbaf69c03c35e66700
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3596167
> Commit-Queue: Jakob Linke <jgruber@chromium.org>
> Auto-Submit: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> Commit-Queue: Leszek Swirski <leszeks@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80147}
Bug: v8:12161
Change-Id: Ib3597cf1d99cdb5d0f2c5ac18e311914f376231d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3606232
Auto-Submit: Jakob Linke <jgruber@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80167}
This reverts commit 91da38831d.
Reason for revert: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux64%20-%20arm64%20-%20sim%20-%20pointer%20compression%20-%20builder/21150/overview
Original change's description:
> [osr] Use the new OSR cache
>
> This CL switches over our OSR system to be based on the feedback
> vector osr caches.
>
> - OSRing to Sparkplug is fully separated from OSR urgency. If
> SP code exists, we simply jump to it, no need to maintain an
> installation request.
> - Each JumpLoop checks its dedicated FeedbackVector cache slot.
> If a valid target code object exists, we enter it *without*
> calling into runtime to fetch the code object.
> - Finally, OSR urgency still remains as the heuristic for
> requesting Turbofan OSR compile jobs. Note it no longer has a
> double purpose of being a generic untargeted installation
> request.
>
> With the new system in place, we can remove now-unnecessary
> hacks:
>
> - Early OSR tierup is replaced by the standard OSR system. Any
> present OSR code is automatically entered.
> - The synchronous OSR compilation fallback is removed. With
> precise installation (= per-JumpLoop-bytecode) we no longer
> have the problem of 'getting unlucky' with JumpLoop/cache entry
> mismatches. Execution has moved on while compiling? Simply spawn
> a new concurrent compile job.
> - Remove the synchronous (non-OSR) Turbofan compile request now
> that we always enter available OSR code as early as possible.
> - Tiering into Sparkplug no longer messes with OSR state.
>
> Bug: v8:12161
> Change-Id: I0a85e53d363504b7dac174dbaf69c03c35e66700
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3596167
> Commit-Queue: Jakob Linke <jgruber@chromium.org>
> Auto-Submit: Jakob Linke <jgruber@chromium.org>
> Reviewed-by: Leszek Swirski <leszeks@chromium.org>
> Commit-Queue: Leszek Swirski <leszeks@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#80147}
Bug: v8:12161
Change-Id: I4a6955f4f20b6f3b13e98d5600c7c6a5205915bc
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3605608
Auto-Submit: Nico Hartmann <nicohartmann@chromium.org>
Owners-Override: Nico Hartmann <nicohartmann@chromium.org>
Reviewed-by: Nico Hartmann <nicohartmann@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/main@{#80148}
This CL switches over our OSR system to be based on the feedback
vector osr caches.
- OSRing to Sparkplug is fully separated from OSR urgency. If
SP code exists, we simply jump to it, no need to maintain an
installation request.
- Each JumpLoop checks its dedicated FeedbackVector cache slot.
If a valid target code object exists, we enter it *without*
calling into runtime to fetch the code object.
- Finally, OSR urgency still remains as the heuristic for
requesting Turbofan OSR compile jobs. Note it no longer has a
double purpose of being a generic untargeted installation
request.
With the new system in place, we can remove now-unnecessary
hacks:
- Early OSR tierup is replaced by the standard OSR system. Any
present OSR code is automatically entered.
- The synchronous OSR compilation fallback is removed. With
precise installation (= per-JumpLoop-bytecode) we no longer
have the problem of 'getting unlucky' with JumpLoop/cache entry
mismatches. Execution has moved on while compiling? Simply spawn
a new concurrent compile job.
- Remove the synchronous (non-OSR) Turbofan compile request now
that we always enter available OSR code as early as possible.
- Tiering into Sparkplug no longer messes with OSR state.
Bug: v8:12161
Change-Id: I0a85e53d363504b7dac174dbaf69c03c35e66700
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3596167
Commit-Queue: Jakob Linke <jgruber@chromium.org>
Auto-Submit: Jakob Linke <jgruber@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80147}
There is currently a bug in docker where fstat may not
return the correct device id and as a result a check under
`OS::RemapPages, stat_buf.st_dev != enclosing_region.dev`
fails, details on the bug:
https://github.com/moby/moby/issues/43512
Platform specific page sizes are also defined for kMaxPageSize
to fix compilation errors.
Change-Id: I026609329aa6432eda4f1880a0f586c0c2162461
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3601211
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Junliang Yan <junyan@redhat.com>
Commit-Queue: Milad Farazmand <mfarazma@redhat.com>
Cr-Commit-Position: refs/heads/main@{#80111}
