Otherwise the side effect is already applied before we throw the exception,
already breaking the environment.
Bug: chromium:1043151
Change-Id: I6d6b4706481bd8eaead6de3503aabc4bad9fb6de
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2016597
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Auto-Submit: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65982}
SafeStackFrameIterator::Advance() will stop at wasm_to_js frames, so we
should fully expect that they show up when calling frame() after
Advance().
This fixes some breakages for profiling wasm and asm.js.
Bug: chromium:1010534, v8:10150
Change-Id: Ifc78cd9eba33e777b33fc1cbea19ee376b25299c
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2019485
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65980}
This CL attempts to remove the #undef check
from the Presubmit for .cc files because we
are removing Jumbo support.
Bug: chromium:994387
Change-Id: I1bb6d22393c1b533f23dd586e5028ecc6d254c99
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2013119
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Wez <wez@chromium.org>
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Almothana Athamneh <almuthanna@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65978}
Port b110d4808b
Original Commit Message:
This relands commit 009993adb4.
The fix is in liftoff-assembler-ia32.h, the codegen was incorrect.
Original change's description:
> Implement f32x4.splat and enable handling this in Liftoff.
>
> We add a new macro for defining test cases to run on TurboFan, Liftoff,
> interpreter, and scalar lowering.
>
> Also add an assertion that the execution tier used is what we expected
> it to be. This is useful for Liftoff, because by default it falls back
> to TurboFan when it encounters an unimplemented opcode.
>
> Bug: v8:9909
R=zhin@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=
LOG=N
Change-Id: I0fc54cd04379a6c65aa2bee64bf4cf9dad5b9636
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2019582
Reviewed-by: Joran Siu <joransiu@ca.ibm.com>
Commit-Queue: Milad Farazmand <miladfar@ca.ibm.com>
Cr-Commit-Position: refs/heads/master@{#65977}
As these are simple methods, and for consistency with the GraphAssembler
interface, the getters should not be capitalized.
R=ecmziegler@chromium.org
Bug: v8:10123
Change-Id: I51871115b061bca4483dc23e982303508d9d8f55
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2017471
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65976}
This changes the WasmGraphBuilder to use the GraphAssembler for storing
the current effect and control, instead of having separate fields.
This will allow us to use GraphAssembler methods anywhere in the
WasmGraphBuilder without first transferring the WasmGraphBuilder's
effect and control to the GraphAssembler and storing it back afterwards.
Some methods (StackCheck, BuildCallToRuntimeWithContext) have been
changed to work on the current effect and control instead of receiving
them as parameters. This is necessary to also use the GraphAssembler
there, and it makes the code a bit easier.
R=ecmziegler@chromium.org
Bug: v8:10123
Change-Id: Ia9f6359cfe44a67a875c3eb0fe51f87db8bd17db
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2017470
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65975}
TailCalls have no output effect, similarly to e.g. return nodes. Thus,
don't put them in the effect chain.
The current graph verifier does not complain about this, but once we
switch to the GraphAssembler, we will hit a SLOW_DCHECK there because
we are trying to use a node as effect which has no effect output.
R=jkummerow@chromium.org
Bug: v8:10070
Change-Id: I23d9516eb9bd42a5aca46b09ca74830eafce635c
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2017474
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65973}
We currently print asm.js functions converted to wasm as
"function foo() { [native code] }", even though without asm to wasm
translation we get the proper source code. This is an observable
difference that should not be, and also foozzie finds this frequently in
different variations.
This CL makes us remember the start position (position of the "function"
token) and end position (right behind the closing "}") of each function
we transform to wasm. These offsets, together with the Script that
contained the function, allows us to reconstruct the source code of the
function for the {toString()} method.
R=jkummerow@chromium.org
Bug: chromium:667678
Change-Id: If22471cad4cefdfc67f6d1b8fda85aa0eeb411bd
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2016582
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65972}
The code did originally only handle the int32 case, but was then
extended to all wasm types in https://crrev.com/c/1965577. The comments
and variable names still mention the int32 type only though.
This CL fixes that.
R=ecmziegler@chromium.org
Bug: v8:10070
Change-Id: I3b1fa4047ee7ee9f283dce3758f6b8a39c964b22
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2017723
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65970}
This code is commented out since a long time (at least since 2008), but
I found it to be useful information.
Instead of printing the full script information, just print the {Brief}
version. In a debugger, users can still print the full information based
on the printed pointer value if needed.
R=jkummerow@chromium.org
Change-Id: I79b19f0bbc8797746b916c1f5860f727c0f0ac39
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2015240
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65969}
This cleans up string-util.h a little. Instead of distinguishing
BinaryStringBuffer from StringBufferImpl (which has a legacy
::adopt() interface), provide two functions for making StringBuffer
from either an 8 bit (std::vector<uint8_t>) or a 16 bit (String16)
source, and keep the classes in the .cc file only. Since it's easy
enough, this adds a class also for the empty string buffer as there's
no need to keep an empty vector / String16 this way.
No public API changes here.
Change-Id: Idb25fe24ea94f27f8001d552cede089e04eacd32
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2016015
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Dmitry Gozman <dgozman@chromium.org>
Commit-Queue: Johannes Henkel <johannes@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65964}
This relands commit 009993adb4.
The fix is in liftoff-assembler-ia32.h, the codegen was incorrect.
Original change's description:
> Implement f32x4.splat and enable handling this in Liftoff.
>
> We add a new macro for defining test cases to run on TurboFan, Liftoff,
> interpreter, and scalar lowering.
>
> Also add an assertion that the execution tier used is what we expected
> it to be. This is useful for Liftoff, because by default it falls back
> to TurboFan when it encounters an unimplemented opcode.
>
> Bug: v8:9909
Bug: v8:9909
Change-Id: I7daacbe8b195d9212367190c515b0babbc457a88
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2018043
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65963}
This fixes an issue found by the upcoming change that verifies a match
between 32-bit Torque output from a 32-bit build and 32-bit output from
a 64-bit build: https://crrev.com/c/v8/v8/+/1998078 . 32-bit output
never needs to deal with pointer compression.
Bug: v8:10124
Change-Id: I0e04728dafc4118a8816a3afa98651dddea7a574
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2001460
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#65962}
This change adds support for the postmortem inspection library to show
the content of cached external strings if that content is available. It
also fixes a minor annoyance where strings with unavailable data would
show up as "...". Now, if fetching the very first character fails, we
omit the literal value from the output.
Bug: v8:9376
Change-Id: Id694a774c231ab3467fb59b1c149284729acfb20
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1987922
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Seth Brenith <seth.brenith@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#65961}
This reverts commit 5cfe053e45.
Reason for revert: "liftoff-simd-params" also fails on no-sse :/
Original change's description:
> [wasm-simd][liftoff] Add S128 case for stack movements
>
> The two cases we are fixing here are Construct and
> LoadCallerFrameSlot, which are closely related.
>
> Construct is called during PrepareCall, where we build up
> LiftoffStackSlots when we need to move an arg from caller's stack frame
> into callee's stack frame. LoadCallerFrameSlot is the parallel to
> this, called in ProcessParameter during decoding of the callee's
> function body.
>
> In most cases, Construct needs a new case to handle kWasmS128, and calls
> the relevant assembler to push a s128 onto the stack.
>
> ARM64 requires 16-byte alignment of sp, so we need to Claim the right
> number of kXRegSize slots first, which requires
> us traversing the list of slots to figure out how many s128 values there
> are. This is a straightforward way to fix this, if efficiency is a
> problem, we can change LiftOffStackSlots::Add to sum up the slot sizes.
>
> On IA32, pushing s128 values will require 4 calls to push. Instead, we
> use a sub and two movdqu, which will generate less code in most cases.
>
> On x64, there is no 128-bit push, so we call push twice.
>
> Bug: v8:9909
> Change-Id: I3af35b8462ea9c3b9b2d90800c37d11b5e95be59
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2015945
> Commit-Queue: Zhi An Ng <zhin@chromium.org>
> Reviewed-by: Clemens Backes <clemensb@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#65956}
TBR=clemensb@chromium.org,zhin@chromium.org,joey.gouly@arm.com
Change-Id: Ib3c5a088e2d85baf1d8b143272844fb5ebb33c57
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:9909
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2017724
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65960}
This reverts commit 009993adb4.
Reason for revert: New test fails, see https://ci.chromium.org/p/v8/builders/ci/V8%20Linux/35534 and https://ci.chromium.org/p/v8/builders/ci/V8%20Win32%20-%20debug/23778
Original change's description:
> [liftoff][wasm-simd] Implement f32x4.splat
>
> Implement f32x4.splat and enable handling this in Liftoff.
>
> We add a new macro for defining test cases to run on TurboFan, Liftoff,
> interpreter, and scalar lowering.
>
> Also add an assertion that the execution tier used is what we expected
> it to be. This is useful for Liftoff, because by default it falls back
> to TurboFan when it encounters an unimplemented opcode.
>
> Bug: v8:9909
> Change-Id: I594955fce778173191fc44c38c4f956a05e77839
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2014753
> Commit-Queue: Zhi An Ng <zhin@chromium.org>
> Reviewed-by: Clemens Backes <clemensb@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#65954}
TBR=clemensb@chromium.org,zhin@chromium.org
Change-Id: Ie6970a8c29baab149150dd734a95f89be5fd89ff
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:9909
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2017722
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65958}
The following instructions are affected: memory.init, data.drop,
table.init, table.drop. A segment index should be decoded as an unsigned
number, but these instructions were decoding as signed. This works
properly up to 63, but fails at 64 (which is decoded as -64 =
4294967232).
Bug: v8:10151
Change-Id: I742b74cf0bcadf2ff2f606beb65b7bae3e816530
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2015960
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Ben Smith <binji@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65957}
The two cases we are fixing here are Construct and
LoadCallerFrameSlot, which are closely related.
Construct is called during PrepareCall, where we build up
LiftoffStackSlots when we need to move an arg from caller's stack frame
into callee's stack frame. LoadCallerFrameSlot is the parallel to
this, called in ProcessParameter during decoding of the callee's
function body.
In most cases, Construct needs a new case to handle kWasmS128, and calls
the relevant assembler to push a s128 onto the stack.
ARM64 requires 16-byte alignment of sp, so we need to Claim the right
number of kXRegSize slots first, which requires
us traversing the list of slots to figure out how many s128 values there
are. This is a straightforward way to fix this, if efficiency is a
problem, we can change LiftOffStackSlots::Add to sum up the slot sizes.
On IA32, pushing s128 values will require 4 calls to push. Instead, we
use a sub and two movdqu, which will generate less code in most cases.
On x64, there is no 128-bit push, so we call push twice.
Bug: v8:9909
Change-Id: I3af35b8462ea9c3b9b2d90800c37d11b5e95be59
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2015945
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65956}
Implement f32x4.splat and enable handling this in Liftoff.
We add a new macro for defining test cases to run on TurboFan, Liftoff,
interpreter, and scalar lowering.
Also add an assertion that the execution tier used is what we expected
it to be. This is useful for Liftoff, because by default it falls back
to TurboFan when it encounters an unimplemented opcode.
Bug: v8:9909
Change-Id: I594955fce778173191fc44c38c4f956a05e77839
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2014753
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65954}
Port 7b79a02d2b
Original Commit Message:
This CL adds a --debug-in-liftoff flag, which takes another path in
{WasmScript::SetBreakPointForFunction}, and sets the breakpoint via
{wasm::DebugInfo} (Liftoff-related) instead of {WasmDebugInfo} (C++
interpreter related).
Actual breakpoint support is not there yet, so the new test which sets
this flag does not currently break anywhere. This will change with a
future CL.
R=clemensb@chromium.org, joransiu@ca.ibm.com, jyan@ca.ibm.com, michael_dawson@ca.ibm.com
BUG=
LOG=N
Change-Id: I632e4c945c88f58d8caba8688356f66406bc04aa
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2016915
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Milad Farazmand <miladfar@ca.ibm.com>
Cr-Commit-Position: refs/heads/master@{#65953}
Load and store external references using the root register rather than
generating a constant and dereferencing it. This typically uses 1
instruction rather than up to 4.
Also adds external reference store optimisation for arm64.
Bug: v8:7844
Change-Id: I5f73728e7a72e366a31bfb694581e2e7d8250947
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2007270
Commit-Queue: Dan Elphick <delphick@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65950}
The TPH needs to support the idea of 'code' objects, and the ability to query whether an object is of this type.
Bug: v8:9533
Change-Id: Ic36b235bd9640aa675a3ef0d8c3f6c98dd8cd862
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2013116
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Steve Blackburn <steveblackburn@google.com>
Cr-Commit-Position: refs/heads/master@{#65948}
Remove trace id mangling since it's been deprecated (and is also being
removed) in Chromium [1] and the actual id mangling operation was never
implemented in V8's version.
This patch doesn't introduce the upstream replacement
(TRACE_ID_{LOCAL,GLOBAL}), since nothing in V8 needs it and it'll be
shortly brought in with the transition to the Perfetto client library[2].
Bug: chromium:639003
[1] crbug.com/639003
[2] https://docs.google.com/document/d/1f7tt4cb-JcA5bQFR1oXk60ncJPpkL02_Hi_Bc6MfTQk/
Change-Id: Ifabda63b9c56918fafcc24dfc589b8e513a3f29b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2016592
Commit-Queue: Sami Kyöstilä <skyostil@chromium.org>
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Auto-Submit: Sami Kyöstilä <skyostil@chromium.org>
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65946}
For CompareOperations, we only care about Null or Undefined feedback
for oddballs. Other oddballs like Booleans should be treated as "Any"
feedback. There was a bug in the way we handled feedback when lhs is
undefined or null and rhs is anything other than JSReceiver. This cl
fixes it by removing an unnecessary check. This fixes a deopt loop in
TurboProp on maps benchmark.
Bug: v8:9684
Change-Id: I6a09c4b5637bacdcd7f7f96e1afe603b6653c7c6
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2016591
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Commit-Queue: Mythri Alle <mythria@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65945}
Did some drive-by cleanups, like removing the Sloppy-ness
Change-Id: Ia11ddc6752168ad10cad578bb8d839ab9b7473c8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2015021
Reviewed-by: Mythri Alle <mythria@chromium.org>
Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65944}
I hit a case where the 'V8_Dcheck' was only the sixth stack frame. Thus
increase the limit from 5 to 7.
R=ahaas@chromium.org
No-Try: true
Change-Id: I1ea37f07ff08ab5acffdfcc89d01ff102750a1e0
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2016589
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65943}
The NativeModule that was serialized in another isolate might be
temporarily kept alive by background threads. By keeping a weak pointer
to the module, we can wait for it to really die.
This happens very rarely, and the module will die pretty quickly, so
busy-waiting is fine in this case.
R=thibaudm@chromium.org
Bug: v8:10148
Change-Id: I8c4645acfccd04a820ef3f694cad3eb15e75acb6
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2016585
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65942}
The instruction selector assumed for Word32AtomicPairBinop nodes that if
there exists a Projection(1) user, then there also exists a
Projection(0) user. This, however, is not the case, because TurboFan
eliminates unreachable nodes. The missing projection node lead to a
failed DCHECK in the register allocator.
With this CL we allocate the right registers for the existing
projections, and allocate the other needed registers as temp registers.
R=gdeepti@chromium.org
Bug: v8:10140
Change-Id: I22331cae58f933e89dac6993fe3b21ff6502838a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2011829
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65941}
This CL implements load_extend with 2 lanes and all load_splat
operations on IA32. The necessary assemblers together with their
corresponding disassemblers and tests are also added in this CL.
The newly added opcodes include: S8x16LoadSplat, S16x8LoadSplat,
S32x4LoadSplat, S64x2LoadSplat, I64x2Load32x2S, I64x2Load32x2U.
Bug: v8:9886
Change-Id: I0a5dae0a683985c14c433ba9d85acbd1cee6705f
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1982989
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Reviewed-by: Zhi An Ng <zhin@chromium.org>
Commit-Queue: Zhiguo Zhou <zhiguo.zhou@intel.com>
Cr-Commit-Position: refs/heads/master@{#65937}
The feature has been on-by-default in Chrome for nearly a year now,
and is an established part of the ECMAScript standard.
Change-Id: Icf9d424e5fe9139c12fc26b41603b4e39f79ea54
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2015942
Reviewed-by: Ben Smith <binji@chromium.org>
Commit-Queue: Adam Klein <adamk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#65935}
