On x86, signalling NaNs get converted to quiet NaNs when they get push
on the stack and popped again. This happens in the code generation for
arm, specifically for the vmov instruction with the immediate parameter.
This CL replaces the vmov function in assembler-arm to take the
immediate as a uint64_t instead of a double, to guarantee that the bit
pattern does not change even if the parameter is a signalling NaN.
BUG=v8:6564
Change-Id: I062559f9a7ba8b0f560628e5c39621ca578c3e7d
Reviewed-on: https://chromium-review.googlesource.com/558964
Commit-Queue: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Reviewed-by: Martyn Capewell <martyn.capewell@arm.com>
Cr-Commit-Position: refs/heads/master@{#46418}
The naming convention in v8 has trivial getters named like the field,
no 'get_' prefix, and dropping the '_' suffix of the field.
BUG=
Review-Url: https://codereview.chromium.org/2958283003
Cr-Commit-Position: refs/heads/master@{#46332}
It appears we actually get a compile time boost, and sometimes a
runtime boost, at the cost of some reloc info growth.
Bug:
Change-Id: I1d1dc48f364e6611f895ebd00f86451199dd8626
Reviewed-on: https://chromium-review.googlesource.com/544713
Commit-Queue: Mircea Trofin <mtrofin@chromium.org>
Reviewed-by: Brad Nelson <bradnelson@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46302}
The fuzzer has already been removed from chromium. In addition I removed
code which was only used by this fuzzer.
BUG=chromium:734550
R=clemensh@chromium.orgCC=mstarzinger@chromium.org
Change-Id: I2ff4614e4d64131412ead759318e5c38e38f5d3d
Reviewed-on: https://chromium-review.googlesource.com/542816
Commit-Queue: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46078}
- Now that there are no boolean vector types, we can directly test the
results of relational ops.
Bug: v8:6020
Change-Id: Id2139133ae3a548a9985a26a3427cbeddc6272a6
Reviewed-on: https://chromium-review.googlesource.com/536176
Reviewed-by: Aseem Garg <aseemgarg@chromium.org>
Commit-Queue: Bill Budge <bbudge@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46075}
- Use correct prefixes for SIMD/Atomics ops
- S128 LoadMem/StoreMem should not use 0xc0/0xc1 opcodes, these are now
being used for sign extension
- S128 LoadMem/StoreMem should use prefixed opcodes
BUG=v8:6020
Review-Url: https://codereview.chromium.org/2943773002
Cr-Commit-Position: refs/heads/master@{#46016}
- Eliminates S32x4Shuffle, S16x8Shuffle opcodes. All shuffles are subsumed
by S8x16Shuffle. This aligns us with the latest WASM SIMD spec.
LOG=N
BUG=v8:6020
Review-Url: https://codereview.chromium.org/2923103003
Cr-Commit-Position: refs/heads/master@{#45929}
This gives the embedder more context for deciding whether code
generation should be allowed or not, or they can chose to include the
code in a report.
BUG=chromium:732736
R=ahaas@chromium.org
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Change-Id: Ibbaa3d0574319d290f15565be3eed2ee4d3dda36
Reviewed-on: https://chromium-review.googlesource.com/532875
Commit-Queue: Jochen Eisinger <jochen@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45905}
Once a buffer has been externalized, V8 is no longer responsible for managing
the memory. The fact that V8 was freeing was leading to double free errors once
Blink's GC got around to freeing the buffer too.
Bug: chromium:730171, chromium:731046
Change-Id: Ib18a7e37cafd51bce0c5a983d5cf8f3e64eb2c13
Reviewed-on: https://chromium-review.googlesource.com/530132
Commit-Queue: Brad Nelson <bradnelson@chromium.org>
Reviewed-by: Brad Nelson <bradnelson@chromium.org>
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45882}
Avoid constructing zones and large zone objects when initializing
WasmCompilationUnit. The main reason we did that is so we can cache
the CEntryStub node, which requires a code object, obtainable only
on the main thread. We need that value, however, on background threads,
which is also where we need the aforementioned large objects. We only
need that for the WasmCompilationUnits being currently compiled, which
is a number proportional to the number of background threads provided
by the embedder. Specifically, one zone is needed only for the duration
of the background compilation, while the second zone needs to survive
past that, so the compilation results may be committed to the GC heap
as Code objects.
The problem with these large objects is that the first allocation
in a Zone is at minimum 8KB. We used to allocate 2 zones. For
modules with 200K functions, that means 3.2GB of memory pre-allocated
before any of it is actually needed.
This change attaches a Handle to the CEntryStub on the WasmCompilationUnits,
and delays zone creation to when needed. The change also adds a way to
cache CEntryStubs in a JSGraph from a given Code handle - limited to the
scenario needed by wasm (and removable once we get wasm off the GC heap,
which subsumes removing this dependency on CEntryStubs)
An additional constraint for this change is that we want it to be easily
back-mergeable to address chromium:723899.
For the wasm payload in question, collecting the max memory used by d8
using /usr/bin/time --format='(%Xtext+%Ddata %Mmax)', we get the
following numbers (in KB):
- unchanged: 3307480
- patch 1: 1807140 (45% reduction)
- patch 3: 1230320 (62% reduction from first)
- patch 5/6: 519368 (84% reduction from first)
Bug: chomium:732010, chromium:723899
Change-Id: I45b96792daf8a9c8dc47d45fb52da75945a41401
Reviewed-on: https://chromium-review.googlesource.com/530193
Commit-Queue: Mircea Trofin <mtrofin@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45880}
In many places in WasmModule and contained structs we store references
into the wire bytes as pairs of offset and length.
This CL introduces a WireBytesRef struct which encapsulates these two
connected fields. This makes it easier to pass them and assign them as
one unit.
R=ahaas@chromium.org, mtrofin@chromium.org
BUG=v8:6474
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Change-Id: I4f2a40d848a51dc6f6f599f9253c3c6ed6e51627
Reviewed-on: https://chromium-review.googlesource.com/530687
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45859}
This is a testing-only function, which is semantically equivalent to a
SyncCompile followed by SyncInstantiate.
We add a new SyncCompileAndInstantiate function to do those two steps
in one go, and use this method instead.
For AsmJs modules, a new testing function CompileAndRunAsmWasmModule is
introduced.
This is part of our effort to reduce the number of special paths for
testing. It is connected with
https://chromium-review.googlesource.com/529210, but should not
conflict with it.
After landing both CLs, we can later also get rid of
InstantiateModuleForTesting.
R=ahaas@chromium.org, mtrofin@chromium.org
BUG=v8:6474
Change-Id: I7891e968370d5eb68803076ce2639c65a2799dcc
Reviewed-on: https://chromium-review.googlesource.com/529844
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45852}
This CL removes most occurences of "WASM" from outputs and comments in
the code. They are replaced either by "WebAssembly" or (especially in
comments) "wasm". These are the spellings officially proposed on
http://webassembly.org/.
R=ahaas@chromium.org
BUG=v8:6474
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Change-Id: Id39fa5e25591678263745a4eab266db546e65983
Reviewed-on: https://chromium-review.googlesource.com/529085
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45824}
- Eliminates b1x4, b1x8, and b1x16 as distinct WASM types.
- All vector comparisons return v128 type.
- Eliminates b1xN and, or, xor, not.
- Selects take a v128 mask vector and are now bit-wise.
- Adds a new test for Select, where mask is non-canonical (not 0's and -1's).
LOG=N
BUG=v8:6020
Review-Url: https://codereview.chromium.org/2919203002
Cr-Commit-Position: refs/heads/master@{#45795}
Most prototype implementations are not fully supported in the
interpreter. This is the case at least for exception handling, simd, and
atomics. Any function can be redirected to the interpreter though,
either by passing --wasm-interpret-all, or by dynamically redirecting to
the interpreter for debugging.
Making the flags experimental keeps the fuzzer from playing around with
these flags.
Drive-by: Refactor tests which explicitly set the prototype flag to use
a new scope for that.
R=ahaas@chromium.org
BUG=chromium:727584
Change-Id: I67da79f579f1ac93c67189afef40c6524bdd4430
Reviewed-on: https://chromium-review.googlesource.com/519402
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45639}
Currently SIMD integer comparison ops are implemented using Lt/Le, this is
sub-optimal on Intel, because all compares are done using pcmpgt(d/w/b) that
clobber the destination register, and will need additional instructions to
when using Lt/Le as the base implementation. This CL proposes moving to Gt/Ge
as the underlying implementation as this will only require swapping operands
on MIPS and is consistent with x86/ARM instructions.
BUG=v8:6020
R=bbudge@chromium.org, bmeurer@chromium.org, bradnelson@chromium.org
Review-Url: https://codereview.chromium.org/2874403002
Cr-Commit-Position: refs/heads/master@{#45440}
The WebAssembly code now uses these new APIs to allocate memory with guard
regions. Guarded array buffers are no longer always external, which eliminates
a lot of special cases around WebAssembly memory.
Bug: chromium:720302
Change-Id: I355b74ac30a05a18c8b363bd256d57458742849f
Reviewed-on: https://chromium-review.googlesource.com/505715
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Reviewed-by: Brad Nelson <bradnelson@chromium.org>
Commit-Queue: Eric Holk <eholk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45436}
This makes sure that the order of exports as they appear in asm.js
modules is maintained globally (not just per function) while being
translated to a WASM module.
R=clemensh@chromium.org
TEST=mjsunit/asm/asm-validation
BUG=chromium:720586
Change-Id: I8b26d717ae2f88467d41670bced901f196c7b3fc
Reviewed-on: https://chromium-review.googlesource.com/503708
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45277}
- Currently if GrowMemory is called with pages = 0, an attempt is made to
unregister the ArrayBuffer even if it is external. Cleanup so all Detaching
of ArrayBuffer is centralized to one method, and can only be called fromJS.
- Gate creating WeakHandles to the memory on the buffer having guard pages
enabled. Currently creating a WeakHandle is gated only on if the buffer
is_external true. If a buffer is marked is_external = true to begin with,
the WeakHandle is created and the Finalizer is run causing the program to
crash.
BUG=chromium:717647
Review-Url: https://codereview.chromium.org/2867233002
Cr-Commit-Position: refs/heads/master@{#45238}
- S32x4Shuffle by decomposing into s-register moves if no patterns match.
- S16x8Shuffle, S8x16Shuffle implemented with vtbl if no patterns match.
LOG=N
BUG=v8:6020
Review-Url: https://codereview.chromium.org/2856363003
Cr-Commit-Position: refs/heads/master@{#45210}
- Removes primitive shuffle opcodes.
- Adds Shuffle opcode for S32x4, S16x8, S8x16.
- Adds code to ARM instruction selector to pick best opcodes for some
common shuffle patterns.
LOG=N
BUG=v8:6020
Review-Url: https://codereview.chromium.org/2847663005
Cr-Commit-Position: refs/heads/master@{#45104}
During computation of the side table, ignore stack effects of
instructions following any unconditional jump in the same block
(|unreachable|, |br|, |br_table| or |return| jump out of the block).
Without this fix, the current stack height might underflow, or we compute an
unnecessarily large max_stack_height_. Note that those instruction will
never get executed anyway.
Hence, we don't need to store any side table information for such
unreachable code.
R=rossberg@chromium.org
BUG=chromium:716936, chromium:715990
Change-Id: I282f7f18ba1b972a112210e692f6cd05cf32308c
Reviewed-on: https://chromium-review.googlesource.com/493266
Reviewed-by: Andreas Rossberg <rossberg@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45059}
Executing the |end| opcode of a loop assumed that the stack height was
being reset to the height at start of the loop. Hence we were ignoring
the arity of the loop.
During computation of the side table, the arity of the label associated
with the loop was explicitly set to 0, such that a |br| instruction to
that label would not transfer any values.
It turns out though that we need to remember the arity in order to
precompute the correct stack height when executing the |end| opcode of
a loop.
Also, add a regression test.
R=rossberg@chromium.org
BUG=chromium:716936
Change-Id: Ib3a559998f1ce5f8fcd7b94af1426637b3e48f86
Reviewed-on: https://chromium-review.googlesource.com/493286
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Rossberg <rossberg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45041}
This avoids generating redundant code for different template
instantiations.
I also introduce getters instead of accessing the fields directly.
R=ahaas@chromium.org
BUG=v8:6325
Change-Id: I3e0eca9ef6a01e0a3ebb73f4f357bcb59e120f43
Reviewed-on: https://chromium-review.googlesource.com/490166
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44976}
This reduces the amount of special paths for testing.
Setup the memory used for testing exactly the same way as in real world.
Also, always connect the interpreter to the instance being executed,
and to the existing WasmInstance struct. This keeps information
synchronized between interpreter and test runner.
These changes allow us to execute e.g. GrowMemory from cctests either
in the interpreter or in compiled code.
R=ahaas@chromium.org
Change-Id: Id4726d061f3cdba789275350f500d769d27d2d63
Reviewed-on: https://chromium-review.googlesource.com/488561
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44966}
When branching to a loop header, we were trying to copy over {arity}
values from the value stack. This is correct for block labels, but not
for loops. When branching back to a loop header, no values need to be
transferred.
R=ahaas@chromium.org
BUG=chromium:715454
Change-Id: I90d806de63d039abf8dcac1abec057860c8f69ca
Reviewed-on: https://chromium-review.googlesource.com/488146
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44949}
This reverts commit d7cdea6fa2.
Reason for revert: Flakiness on bots
Original change's description:
> [wasm] Add guard pages before Wasm Memory
>
> Although Wasm memory indices are all unsigned, they sometimes get assembled
> as 32-bit signed immediates. Values in the top half of the Wasm memory space
> will then get sign extended, causing Wasm to access in front of its memory
> buffer.
>
> Usually this region is not mapped anyway, so faults still happen as they are
> supposed to. This change protects this region with guard pages so we are
> guaranteed to always fault when this happens.
>
> Bug: v8:5277
> Change-Id: Id791fbe2a5ac1b1d75460e65c72b5b9db2a47ee7
> Reviewed-on: https://chromium-review.googlesource.com/484747
> Commit-Queue: Eric Holk <eholk@chromium.org>
> Reviewed-by: Mircea Trofin <mtrofin@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#44905}
TBR=bradnelson@chromium.org,gdeepti@chromium.org,mtrofin@chromium.org,eholk@chromium.org,mseaborn@chromium.org,adamk@chromium.org,v8-reviews@googlegroups.com,wasm-v8@google.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: Ia1d3e5dbf4f518815a9fd4197047077bc8e42816
Reviewed-on: https://chromium-review.googlesource.com/487828
Reviewed-by: Adam Klein <adamk@chromium.org>
Commit-Queue: Adam Klein <adamk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44907}
Although Wasm memory indices are all unsigned, they sometimes get assembled
as 32-bit signed immediates. Values in the top half of the Wasm memory space
will then get sign extended, causing Wasm to access in front of its memory
buffer.
Usually this region is not mapped anyway, so faults still happen as they are
supposed to. This change protects this region with guard pages so we are
guaranteed to always fault when this happens.
Bug: v8:5277
Change-Id: Id791fbe2a5ac1b1d75460e65c72b5b9db2a47ee7
Reviewed-on: https://chromium-review.googlesource.com/484747
Commit-Queue: Eric Holk <eholk@chromium.org>
Reviewed-by: Mircea Trofin <mtrofin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44905}
The only users of the LoadStoreOpcodeOf function were a number of
macros in wasm-macro-gen.h, and three test functions using it directly.
This CL refactors those functions to also use the macros.
In one case, this requires storing the value in a local variable first.
R=ahaas@chromium.org
Change-Id: Ia2fbf67a3831fafc9345e155eb240cf1bf6feb5d
Reviewed-on: https://chromium-review.googlesource.com/486842
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44885}
This header file is only used from tests.
Also, move the LoadStoreOpcodeOf method (only used in tests) from
wasm-opcodes.h to wasm-macro-gen.h.
R=ahaas@chromium.org
Change-Id: I8d4691be494b5c1fbe3084441329850930bad647
Reviewed-on: https://chromium-review.googlesource.com/486861
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44845}
wasm-macro-gen.h is mainly used from tests, but LocalDeclEncoder is
also used from various other places.
This CL moves the LocalDeclEncoder to an own compilation unit. We want
to later move wasm-macro-gen.h to the tests folder.
It also refactors the LocalDeclEncoder to reuse the
LEBHelper::write_u32v and LEBHelper::sizeof_u32v methods instead of
reimplementing it.
R=ahaas@chromium.org
Change-Id: Ia4651436f0544578da7c1c43596d343571942e97
Reviewed-on: https://chromium-review.googlesource.com/486724
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44838}
- Adds new F32x4AddHoriz, I32x4AddHoriz, etc. to WASM opcodes.
- Implements them for ARM.
LOG=N
BUG=v8:6020
Review-Url: https://codereview.chromium.org/2804883008
Cr-Commit-Position: refs/heads/master@{#44812}
