Add a simple, linear-time scheduler to check whether two nodes can be
scheduled to a same basic block without actually building basic blocks.
Bug: v8:12716
Change-Id: I20506f28a9126f881b7e4748f54b12551967ba76
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3388910
Reviewed-by: Almothana Athamneh <almuthanna@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Hao A Xu <hao.a.xu@intel.com>
Cr-Commit-Position: refs/heads/main@{#81015}
Earlier, we had to introduce some temporary workarounds in Turbofan to
enable optimizations for common wasm-gc patterns. These are now not
required, since these optimizations are applied in WasmTyper and WasmGCOperatorReducer.
Bug: v8:7748
Change-Id: I3a7bc4bd2a8023a438ee4620934ff3fcb8bcfc6a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3693999
Commit-Queue: Manos Koukoutos <manoskouk@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Cr-Commit-Position: refs/heads/main@{#81011}
When a 8x16 shuffle matches a packed byte to dword zero extension,
1. input1 is S128Zero after canonicalization,
2. the indices {0,4,8,16} are consecutive value in the range [0-15] and
other indices are in the range [16-31],
the shuffle can be matched to packed byte to dword zero extend. These
shuffles are commonly used in image processing.
Change-Id: I14d1e35401dbc5ecd91f67c46ea9762628835d01
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3547667
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Fanchen Kong <fanchen.kong@intel.com>
Cr-Commit-Position: refs/heads/main@{#80953}
Applying the set of unittest changes from
https://crrev.com/c/3678208 to BE.
Change-Id: I02d0f2f388720e3acc35660042d5c2c76fa589e1
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3687474
Reviewed-by: Manos Koukoutos <manoskouk@chromium.org>
Commit-Queue: Milad Farazmand <mfarazma@redhat.com>
Cr-Commit-Position: refs/heads/main@{#80942}
We introduce a typing phase into the Turbofan compilation pipeline for
wasm-gc. It has two functionalities: (1) to type nodes that were not
typed during code generation (mainly phi nodes) and (2) to narrow types
as much as possible.
The following nodes are handled, which should be enough for our
purposes: TypeGuard, WasmTypeCast, AssertNotNull, Phi, LoadFromObject,
and LoadImmutableFromObject.
Loop phi types are computed by first assigning the type of the
non-recursive input, and updating once we have the type of the recursive
inputs, and repeating this process to a fixed point.
Drive-by: Remove the narrowing of function signatures during wasm
inlining, as it created some issues and should not be needed after this
series of changes.
Bug: v8:7748
Change-Id: I8a72488d5c221c4ae8257fc5abf6f0368cf10e96
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3678208
Commit-Queue: Manos Koukoutos <manoskouk@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80912}
This CL fixes all spots where wasm Turbofan code did not satisfy the
invariant that all nodes with effect outputs are connected to another
node. Also, it enables the related verification for wasm code.
Drive-by:
- Simplify how stack checks are removed during loop unrolling.
- Fix a test declaration in test-gc.cc.
Change-Id: Id32af8584ba0ec281f4bf7757bd2915e6d8bf443
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3676862
Commit-Queue: Manos Koukoutos <manoskouk@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80854}
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}
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}
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}
.. which points back at the corresponding feedback vector slot for each
JumpLoop bytecode.
Bug: v8:12161
Change-Id: I95f4d013544a69e088314655af7eb1dc504a8657
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3596166
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Jakob Linke <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#80048}
Immediate version of the Bitclear instruction can be used for logical
And with some immediates. It can also be used to implement
And(x, Not(imm)) in a single instruction. This patch gives ~0.5% runtime
improvement in one benchmark on Neoverse N1.
Change-Id: Ia926c6746f0c252f81626c6fca21c4dfb41679d9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3160667
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Martyn Capewell <martyn.capewell@arm.com>
Cr-Commit-Position: refs/heads/main@{#80015}
To make the existing mechanism for printing JSON for turbolizer able to
print types other than the ones stored in the nodes (so the verifier can
print its own types here), this CL restructures the printing mechanism
into a single non-private class that can be inherited to override
certain parts of the printing. In this CL only GetType is made virtual
to allow verifier to override it, but additional parts can be made
overridable whenever necessary.
Bug: v8:12619
Change-Id: Idf31f8cdb49eb6c3204c6abfbb74fc981330d6d2
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3571818
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Nico Hartmann <nicohartmann@chromium.org>
Cr-Commit-Position: refs/heads/main@{#79909}
Change the unittest runner to no longer uncondtionally set up a default
platform in the "environment", but to instead make platform set-up part
of the "mixin" framework for test fixtures.
Requires modifying some tests that expect the platform to be available,
and all flag implications resolved, before the mixin constructors run.
We still keep the environment for setting up the process for cppgc. This
process setup can only be done once per process, so it can no longer use
the platform -- that's ok though, the page allocator used by cppgc's
process initialisation doesn't have to be the same as the platform's so
we can just pass in a separate new one.
Change-Id: Ic8ccf39722e8212962c5bba87350c4b304388a7c
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3571886
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/main@{#79820}
- For y = x & 0xFF, we could use movzxbq y, x.
- For y = x & 0xFFFF, we could use movzxwq y, x.
- For y = x & 0xFFFFFFFF, we could use movl y, x.
- For y = x & immediate and immediate fits into uint32,
we could use andl x, immediate.
Bug: v8:12337
Change-Id: I31f04fa9058c6acabb210f0fce61ac713ed1a382
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3518913
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/main@{#79668}
This is a reland of 6b690a6b48.
The previous version of this CL was a bit too aggressive in the
duplication of branch conditions. This caused an increase in
register pressure in some cases, thus reducing performance.
In fact, duplicating branch conditions that require an "== 0" to be
added provides no benefits. We are thus now a bit less aggressive, and
only duplicate comparisons.
Original change's description:
> Reland [compiler] Simplify "==0" branches in MachineOperatorReducer
>
> This is a reland of 48b443f692.
>
> While fixing the initial CL, we stumbled upon a few bugs that
> we had to fix:
>
> - CommonOperatorReducer and SimplifiedOperatorReducer were applied
> before and after SimplifiedLowering, but always assumed that it
> was before SimplifiedLowering, and thus had the wrong semantics
> for branches in some cases. They now have an added parameter to
> know which semantics of branch they should use.
>
> - The lowering of StaticAssert was wrong and could leave kHeapConstant
> in the assert (instead of machine Booleans).
>
> Original change's description:
> > [compiler] Simplify "==0" branches in MachineOperatorReducer
> >
> > Bug: v8:12484
> > Change-Id: I0667c7464c0dd71338bc199a24a69248a7a0a525
> > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3497303
> > Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> > Owners-Override: Tobias Tebbi <tebbi@chromium.org>
> > Commit-Queue: Darius Mercadier <dmercadier@chromium.org>
> > Cr-Commit-Position: refs/heads/main@{#79379}
>
> Bug: v8:12484
> Change-Id: Ibbf5df96fce5ccb04868dc517539479bf69f5703
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3516869
> Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> Commit-Queue: Darius Mercadier <dmercadier@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#79528}
Bug: v8:12484
Change-Id: I31f575a59811a83c7c1acb4c14bf5ded63a8f536
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3540102
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Darius Mercadier <dmercadier@chromium.org>
Cr-Commit-Position: refs/heads/main@{#79560}
This reverts commit 6b690a6b48.
Reason for revert: causes a few regressions here https://chromeperf.appspot.com/group_report?rev=79528
Original change's description:
> Reland [compiler] Simplify "==0" branches in MachineOperatorReducer
>
> This is a reland of 48b443f692.
>
> While fixing the initial CL, we stumbled upon a few bugs that
> we had to fix:
>
> - CommonOperatorReducer and SimplifiedOperatorReducer were applied
> before and after SimplifiedLowering, but always assumed that it
> was before SimplifiedLowering, and thus had the wrong semantics
> for branches in some cases. They now have an added parameter to
> know which semantics of branch they should use.
>
> - The lowering of StaticAssert was wrong and could leave kHeapConstant
> in the assert (instead of machine Booleans).
>
> Original change's description:
> > [compiler] Simplify "==0" branches in MachineOperatorReducer
> >
> > Bug: v8:12484
> > Change-Id: I0667c7464c0dd71338bc199a24a69248a7a0a525
> > Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3497303
> > Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> > Owners-Override: Tobias Tebbi <tebbi@chromium.org>
> > Commit-Queue: Darius Mercadier <dmercadier@chromium.org>
> > Cr-Commit-Position: refs/heads/main@{#79379}
>
> Bug: v8:12484
> Change-Id: Ibbf5df96fce5ccb04868dc517539479bf69f5703
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3516869
> Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> Commit-Queue: Darius Mercadier <dmercadier@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#79528}
Bug: v8:12484
Change-Id: I457464d793e9c5af8448564aa3b46be863b96fbb
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3540148
Bot-Commit: Rubber Stamper <rubber-stamper@appspot.gserviceaccount.com>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Darius Mercadier <dmercadier@chromium.org>
Cr-Commit-Position: refs/heads/main@{#79552}
This is a reland of 48b443f692.
While fixing the initial CL, we stumbled upon a few bugs that
we had to fix:
- CommonOperatorReducer and SimplifiedOperatorReducer were applied
before and after SimplifiedLowering, but always assumed that it
was before SimplifiedLowering, and thus had the wrong semantics
for branches in some cases. They now have an added parameter to
know which semantics of branch they should use.
- The lowering of StaticAssert was wrong and could leave kHeapConstant
in the assert (instead of machine Booleans).
Original change's description:
> [compiler] Simplify "==0" branches in MachineOperatorReducer
>
> Bug: v8:12484
> Change-Id: I0667c7464c0dd71338bc199a24a69248a7a0a525
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3497303
> Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> Owners-Override: Tobias Tebbi <tebbi@chromium.org>
> Commit-Queue: Darius Mercadier <dmercadier@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#79379}
Bug: v8:12484
Change-Id: Ibbf5df96fce5ccb04868dc517539479bf69f5703
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3516869
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Darius Mercadier <dmercadier@chromium.org>
Cr-Commit-Position: refs/heads/main@{#79528}
Modernise the RegList interface to be a proper class, rather than a
typedef to an integer, and add proper methods onto it rather than ad-hoc
bit manipulation.
In particular, this makes RegList typesafe, adding a DoubleRegList for
DoubleRegisters.
The Arm64 CPURegList isn't updated to use (or extend) the new RegList
interface, because of its weird type-erasing semantics (it can store
Registers and VRegisters). Maybe in the future we'll want to get rid of
CPURegList entirely and use RegList/DoubleRegList directly.
Change-Id: I3cb2a4d386cb92a4dcd2edbdd3fba9ef71f354d6
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3516747
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/main@{#79460}
vector register has different register file from float register in Risc64 rvv extension.
So this cl add third FPalising kind INDEPENDENT to allocate independently simd register.
Bug: v8:11976
doc: https://docs.google.com/document/d/1UwmUwOI3eeIMYzZFRmeXmfyNXRFHNZAQ4BcN0ODdMmo/edit?usp=sharing
Change-Id: I0fb8901294b4bc44b0bee55e630b60460e42bef2
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3383513
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Auto-Submit: Yahan Lu <yahan@iscas.ac.cn>
Commit-Queue: Yahan Lu <yahan@iscas.ac.cn>
Cr-Commit-Position: refs/heads/main@{#79449}
Previously, the accumulator was at the end of liveness bitvectors, which
meant that checking for accumulator liveness required a length lookup.
This CL moves it to the start of the bitvector, with registers starting
at index 1 -- the assumption is that the addition of 1 to the index on
register liveness access can be constant folded away.
As a cleanup, replace all the custom liveness printing code with a
single unified ToString. This places the accumulator at the end of the
printed liveness, to avoid having to change test expectations (also, the
position of the accumulator is now an implementation detail). As a
similar cleanup, change StateValue node building to use the
BytecodeLivenessState interface rather than the underlying bitvector.
These two cleanups allow us to remove the raw bitvector accessor from
liveness entirely.
Change-Id: Ic2744b5e8e16b8527e6a4e8d3b4ddad7096289d9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3455144
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Auto-Submit: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Cr-Commit-Position: refs/heads/main@{#79066}
Now that concurrent inlining is shipping on stable, remove support
--no-concurrent-inlining.
Note that it's still possible to run Turbofan exclusively on the
main thread by passing --no-concurrent-recompilation.
Bug: v8:7790, v8:12142, chromium:1240585
Change-Id: I1943bbbcad7dea7e3a3c337c239f14f7d96c23cd
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3308798
Reviewed-by: Liviu Rau <liviurau@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/main@{#78644}
This commit allows using unaligned load/store, which is more efficient
for 2 bytes,4 bytes and 8 bytes memory access.
Use RISCV_HAS_NO_UNALIGNED to control whether enable the fast path or not.
Change-Id: I1d321e6e5fa5bc31541c8dbfe582881d80743483
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3329803
Reviewed-by: ji qiu <qiuji@iscas.ac.cn>
Commit-Queue: Yahan Lu <yahan@iscas.ac.cn>
Cr-Commit-Position: refs/heads/main@{#78427}
In each wasm CallDescriptor, we store the signature of the call based on
the real parameters passed to the call. This signature is more precise
than the formal function signature. We use this signature in inlining
to enable more optimizations.
Changes:
- Add wasm_sig_ field to CallDescriptor.
- Construct the real signature in {DoCall} and {DoReturnCall} in
graph-builder-interface, and pass it to all call-related functions in
WasmGraphBuilder.
- Update {ReplaceTypeInCallDescriptorWith} to use ValueType over
MachineType. Construct the updated function signature.
- In wasm-inlining, kill the Call node after inlining.
- Add two tests.
Bug: v8:11510
Change-Id: Ica711b6b4d83945ecb7201be26577eab7db3c060
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3270539
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Commit-Queue: Manos Koukoutos <manoskouk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77889}
This is a reland of 16df1dfa13
No changes have been made to this reland as previous commit was reverted
due to a new test revealing an existing bug. This bug has now been fixed.
Original change's description:
> [arm64][wasm-simd] Use Cm(0) for integer comparison with 0
>
> Use an immediate zero operand for integer comparison when possible. This
> gives ~1% runtime performance improvement in some benchmarks on Neoverse
> N1.
>
> Change-Id: I727a8104f8e6ca3d122d6b5b8b3d38d7bdd76c47
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3158327
> Reviewed-by: Zhi An Ng <zhin@chromium.org>
> Commit-Queue: Martyn Capewell <martyn.capewell@arm.com>
> Cr-Commit-Position: refs/heads/main@{#76847}
Change-Id: I77d6923d79407a83becbd39970c6a3f62d3a304d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3178482
Reviewed-by: Zhi An Ng <zhin@chromium.org>
Commit-Queue: Rodolph Perfetta <rodolph.perfetta@arm.com>
Cr-Commit-Position: refs/heads/main@{#77260}
We implement two optimizations for trap conditionals for patterns that
come up in wasm-gc.
In case of a Merge followed by a trap, where the path conditions of all
branches of the Merge contain the trap condition, we lift the trap into
the branches of the Merge.
In case of a Branch whose IfTrue branch is followed by a TrapIf with the
same condition, we replace it with the trap followed by the IfFalse
branch. Symmetrically for IfFalse and TrapUnless.
Bug: v8:7748
Change-Id: I43040aebe60eab7b2230fc3130e3b8250e8b2f45
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3190109
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Commit-Queue: Manos Koukoutos <manoskouk@chromium.org>
Cr-Commit-Position: refs/heads/main@{#77181}
Due to MIPS64 ISA feature, 32-bit values should be sign-extended
in 64-bit registers, no matter it's signed or unsigned.
Besides, LoongArch64 also has this feature, and a similar change
has been made before loong64 port's land in V8. This CL also make
a small fix for loong64.
Change-Id: Ib284662931082365f727925af61781e3653debc8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3193595
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Liu yu <liuyu@loongson.cn>
Cr-Commit-Position: refs/heads/main@{#77154}
The intent of the RISC-V ISA is that 32-bit C values are stored sign extended in registers, even for unsigned types.
So we skip cctest case RunLoadStoreZeroExtend64/RunUnalignedLoadStoreZeroExtend64 due to sign extend uint32
Change-Id: Icfe727916b1c04aad5681902ec4782cc98906964
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3184560
Reviewed-by: Ji Qiu <qiuji@iscas.ac.cn>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Ji Qiu <qiuji@iscas.ac.cn>
Cr-Commit-Position: refs/heads/main@{#77112}
When the input to F64x2PromoteLowF32x4 is a S128Load64Zero, we can skip
the load + promote, and promote directly with a memory operand. The
tricky bit here is that on systems that rely on OOB trap handling, the
load is not eliminatable, so we always visit the S128Load64Zero, even
though after instruction-selector pattern-matching, it is unused. We
mark it as defined to skip visiting it, only if we matched it.
Bug: v8:12189
Change-Id: I0a805a3fce65c56ec52082b3625e1712ea1ee7cf
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3154347
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76917}
When checking for operand interference, if both operands are slots and
one of them is 128 bit wide, check that the slot ranges don't intersect.
R=nicohartmann@chromium.org
Bug: chromium:1248817
Change-Id: Ib18b6e596dbb23427508b7cc07947a0ab4665e85
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3162141
Reviewed-by: Nico Hartmann <nicohartmann@chromium.org>
Commit-Queue: Thibaud Michaud <thibaudm@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76904}
This reverts commit 16df1dfa13.
Reason for revert: Multiple failures, e.g. https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux/43844/overview
Original change's description:
> [arm64][wasm-simd] Use Cm(0) for integer comparison with 0
>
> Use an immediate zero operand for integer comparison when possible. This
> gives ~1% runtime performance improvement in some benchmarks on Neoverse
> N1.
>
> Change-Id: I727a8104f8e6ca3d122d6b5b8b3d38d7bdd76c47
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3158327
> Reviewed-by: Zhi An Ng <zhin@chromium.org>
> Commit-Queue: Martyn Capewell <martyn.capewell@arm.com>
> Cr-Commit-Position: refs/heads/main@{#76847}
Tbr: zhin@chromium.org
Change-Id: I7039106d885c59aecad24dd8dda4d151b8e1f022
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3162053
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Auto-Submit: Clemens Backes <clemensb@chromium.org>
Owners-Override: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76851}
Use an immediate zero operand for integer comparison when possible. This
gives ~1% runtime performance improvement in some benchmarks on Neoverse
N1.
Change-Id: I727a8104f8e6ca3d122d6b5b8b3d38d7bdd76c47
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3158327
Reviewed-by: Zhi An Ng <zhin@chromium.org>
Commit-Queue: Martyn Capewell <martyn.capewell@arm.com>
Cr-Commit-Position: refs/heads/main@{#76847}
By delegating to the macro-assembler, emit AVX instructions for some
float opcodes (float sqrt, round, conversions to and from int,
extract/insert/load word).
Since they now support AVX, we rename the instruction ops to remove the
SSE prefix, changing it to be IA32.
Bug: v8:12148
Change-Id: Ib488f03928756e7d85ab78e6cb28eb869e0641f9
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3131374
Reviewed-by: Adam Klein <adamk@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76755}
Use an immediate zero operand for floating point comparison nodes when
possible. This results in up to 20-25% runtime improvement in some
microbenchmarks, as well as 1-1.5% runtime improvement in some
real-use benchmarks on Cortex-A55 and Neoverse N1.
Change-Id: I39d10871a08a037dbe8c0877d789d110476e1a58
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3133143
Reviewed-by: Zhi An Ng <zhin@chromium.org>
Commit-Queue: Martyn Capewell <martyn.capewell@arm.com>
Cr-Commit-Position: refs/heads/main@{#76749}
This is a reland of 65515ddd3e
Fix is to use AddWithWraparound for signed additions to avoid UB.
Original change's description:
> [wasm-simd][arm64] Fuse add and extmul
>
> We can select a better instruction for add+extmul, using one of the
> multiply-long-accumulate instruction.
>
> Define a helper struct to pattern match Add(x, OP(y, z)) and
> Add(OP(x, y) z), and ensure that the matched OP is always on the
> LHS, to simplify checking for matches.
>
> Bug: v8:11548
> Change-Id: I7ab488b262aa9f749785f973549ccd9fad72f4c8
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2826725
> Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
> Commit-Queue: Zhi An Ng <zhin@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#76708}
Bug: v8:11548
Change-Id: I675ab8b78d9c6c30b82a8c96c8e7098a548c6a60
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3144379
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76712}
This reverts commit 65515ddd3e.
Reason for revert: https://ci.chromium.org/ui/p/v8/builders/ci/V8%20Linux64%20UBSan/18117/overview
Original change's description:
> [wasm-simd][arm64] Fuse add and extmul
>
> We can select a better instruction for add+extmul, using one of the
> multiply-long-accumulate instruction.
>
> Define a helper struct to pattern match Add(x, OP(y, z)) and
> Add(OP(x, y) z), and ensure that the matched OP is always on the
> LHS, to simplify checking for matches.
>
> Bug: v8:11548
> Change-Id: I7ab488b262aa9f749785f973549ccd9fad72f4c8
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2826725
> Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
> Commit-Queue: Zhi An Ng <zhin@chromium.org>
> Cr-Commit-Position: refs/heads/main@{#76708}
Bug: v8:11548
Change-Id: Ic1560616e7ee6df917fcedbb6ad139a1a9773d68
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3144377
Auto-Submit: Zhi An Ng <zhin@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@{#76709}
We can select a better instruction for add+extmul, using one of the
multiply-long-accumulate instruction.
Define a helper struct to pattern match Add(x, OP(y, z)) and
Add(OP(x, y) z), and ensure that the matched OP is always on the
LHS, to simplify checking for matches.
Bug: v8:11548
Change-Id: I7ab488b262aa9f749785f973549ccd9fad72f4c8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2826725
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76708}
Handle all 4 selects that wasm-compiler generates.
Also modify unittest to allow optional operations (select
operations are not supported on all archs).
Bug: v8:12136
Change-Id: Ia54d7a71cffaa1c5cc8203520a1f3d812997bbb1
Fixed: v8:12136
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3119991
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Maya Lekova <mslekova@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76539}
1. Move Abspd, Negpd from MacroAssembler into TurboAssembler so that we
can use it in code-generator
2. Add Absps and Negps (float32 versions of the instructions in 1)
3. Refactor SSE/AVX float32/float64 abs/neg to use these macro-assembler
helpers.
4. Use these helpers in Liftoff too
This has the benefit of not requiring to set up the masks in a temporary
register, and loading the constants via an ExternalReference instead.
It does require (in ins-sel) to have the input be in a Register, since
the ExternalReference is an operand (and the instruction can only have 1
operand input).
Bug: v8:11589
Change-Id: I68fafaf31b19ab05ee391aa3d54c45d547a85b34
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3123635
Reviewed-by: Adam Klein <adamk@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76520}
This removes 4 arch opcodes.
Bug: v8:11217
Change-Id: Idff04fb205c7d7d1577ce123cc2160d678dfe39a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3114599
Reviewed-by: Adam Klein <adamk@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76473}
This removes 8 arch opcodes.
Bug: v8:11217
Change-Id: I2c7a73b032ba5fa21f9843ebb4325e226a22550a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/3114590
Reviewed-by: Adam Klein <adamk@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/main@{#76442}
