Registers cannot be used as a merge destination if they have more than
one use, otherwise the merge will unexpectedly affect other uses of that
register.
R=ahaas@chromium.org,clemensb@chromium.org
Bug: chromium:1084151
Change-Id: I0d6ad97c585920357a37d95361e0320d32c71f4b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2208851
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Thibaud Michaud <thibaudm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67904}
Lookups and updates to the executable_memory_ unordered_map need to be
protected with mutex.
Bug: v8:10315, v8:10546
Change-Id: Ic17e19d1e4fda18b99103a96052940e68e970586
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2208867
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67902}
The typed function references proposal allows an optional second
parameter to Table.grow containing the initialization value for the
newly added entries for tables that do not support null defaults.
This CL adds this functionality but hides it behind a newly added
experimental flag --experimental-wasm-typed-funcref.
R=ahaas@chromium.orgCC=jkummerow@chromium.orgCC=manoskouk@chromium.org
Bug: v8:9495
Change-Id: Ia156aeacf95bc36a9fc182990f315c42075cbb7b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207184
Commit-Queue: Emanuel Ziegler <ecmziegler@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67900}
Do not lock mutex for the full duration of the slow path of allocation
on background threads. For example sweeping and allocation of a new page
is now performed outside of the lock.
Bug: v8:10315
Change-Id: Ifee9bc1569d1ec53fdced965bd222dd1eab11b24
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207131
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67898}
This is a reland of 3cc981cb7a with a
fix for data race detected by TSan.
Original change's description:
> [wasm][debug] Fix tier down during streaming compilation
>
> If the debugger is enabled while streaming compilation is happening, we
> won't correctly tier down to Liftoff. This is because during streaming
> compilation, we always compile for no debugging. Fixing that is a bit
> tricky, since when the debugger is enabled, functions can either already
> have finished compiling, or they are currently being compiled, or their
> wire bytes are not received yet.
> Instead of handling this correctly while streaming compilation is
> running, we just recompile the whole module with Liftoff after streaming
> compilation finished.
>
> For testing this, we use the existing tests for async compilation, and
> enable --wasm-test-streaming, which compiles via the streaming decoder
> even in the async compilation case.
>
> R=thibaudm@chromium.org
>
> Bug: v8:10531
> Change-Id: I0177248a9ad2e90f83faee965d6746de05423f1f
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207133
> Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
> Commit-Queue: Clemens Backes <clemensb@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#67882}
Bug: v8:10531, v8:10544
Change-Id: I884922b6ac55543e6ff9b1046438f6b3abab6f64
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207187
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67896}
Fix underflow when the break type list is empty, and do not try to
reuse the first value to generate the wanted type.
Bug: chromium:1084452
Change-Id: Ia9855a267730bb9f427518c27157f449475fb6ba
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2208858
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Thibaud Michaud <thibaudm@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67895}
For single-threaded, we can not tier up in the background. Hence we can
either tier up in foreground (which is pointless), or not tier up.
This CL disables Liftoff, so will compile TurboFan code right away.
R=ahaas@chromium.org
Bug: v8:10530
Change-Id: Icfdc26643ab219d617f844c2ea8e149870168dbc
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2208853
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67894}
TSan handles SIGPROF incorrectly, hence we don't get the samples we
expect on that test, and it quite frequently times out on TSan bots.
Skip it for now, until the TSan issue is fixed.
R=petermarshall@chromium.org
Bug: v8:9869
Change-Id: I27c65cdd10c53b441faad216e00d6c60b42229e6
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2208857
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67892}
Allocation observers are not thread-safe, so the allocation steps should
be invoked only on the main thread.
Bug: v8:10536
Change-Id: I90e809fa230d18cce135f4bea67b62ef616ca593
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207189
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Commit-Queue: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67891}
Call_WithFeedback
CallWithArrayLike_WithFeedback
CallWithSpread_WithFeedback
ConstructWithArrayLike_WithFeedback
ConstructWithSpread_WithFeedback
These are used in generic lowering if --turbo-nci is passed.
Bug: v8:8888
Change-Id: I78b56a1f358fa7c213e375eeb2feaa65432adfdb
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2199352
Reviewed-by: Georg Neis <neis@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67888}
Return type still needs TNodification. It can be Object or Float64T.
Bug: v8:10506, v8:6949
Change-Id: I71e7ae28f3039b2e1c5c2a5c2e383c335cdec38a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2204281
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67887}
Used to be named FastFixedArrayForEachBody. Rename to remove the 'fixed'
part since it can also be a PropertyArray (i.e not a FixedArray).
Rename BuildFastFixedArrayForEach to also remove the 'fixed'.
Bug: v8:10506, v8:6949
Change-Id: I840edf802d334b9ca930f3022553c476e2dca34a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2202910
Reviewed-by: Dan Elphick <delphick@chromium.org>
Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67886}
This reverts commit 3cc981cb7a.
Reason for revert: TSan failures: https://ci.chromium.org/p/v8/builders/ci/V8%20Linux64%20TSAN/31572
Original change's description:
> [wasm][debug] Fix tier down during streaming compilation
>
> If the debugger is enabled while streaming compilation is happening, we
> won't correctly tier down to Liftoff. This is because during streaming
> compilation, we always compile for no debugging. Fixing that is a bit
> tricky, since when the debugger is enabled, functions can either already
> have finished compiling, or they are currently being compiled, or their
> wire bytes are not received yet.
> Instead of handling this correctly while streaming compilation is
> running, we just recompile the whole module with Liftoff after streaming
> compilation finished.
>
> For testing this, we use the existing tests for async compilation, and
> enable --wasm-test-streaming, which compiles via the streaming decoder
> even in the async compilation case.
>
> R=thibaudm@chromium.org
>
> Bug: v8:10531
> Change-Id: I0177248a9ad2e90f83faee965d6746de05423f1f
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207133
> Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
> Commit-Queue: Clemens Backes <clemensb@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#67882}
TBR=clemensb@chromium.org,thibaudm@chromium.org
Change-Id: I26e750c6c6d0783b5e4a0f19a5462a5fbe99a742
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:10531
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207186
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67885}
We implement the conversions via C functions (external references)
because 64-bit conversions are difficult to implement on ia32 and
would not be significantly faster.
Bug: v8:10520
Change-Id: I3573d30bf6b3d558f708c7e0273f027c30a1a771
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2202988
Commit-Queue: Richard Stotz <rstz@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67883}
If the debugger is enabled while streaming compilation is happening, we
won't correctly tier down to Liftoff. This is because during streaming
compilation, we always compile for no debugging. Fixing that is a bit
tricky, since when the debugger is enabled, functions can either already
have finished compiling, or they are currently being compiled, or their
wire bytes are not received yet.
Instead of handling this correctly while streaming compilation is
running, we just recompile the whole module with Liftoff after streaming
compilation finished.
For testing this, we use the existing tests for async compilation, and
enable --wasm-test-streaming, which compiles via the streaming decoder
even in the async compilation case.
R=thibaudm@chromium.org
Bug: v8:10531
Change-Id: I0177248a9ad2e90f83faee965d6746de05423f1f
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207133
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67882}
So far this is mainly a readability improvement to specify
expectations on the packed argument. In the future we should also
check signedness during bytecode generation.
Drive-by: Update DCHECK to allow signed args to
CHECK_CURRENT_POSITION.
Bug: chromium:1083450
Change-Id: I9376ec691b51eb251c972309ad65dd6c04eec3ae
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207137
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67880}
Verify that `current` always points inside, or just past the end of the
subject string.
Bug: chromium:1083450
Change-Id: I27ba49cbfd0aa93cd2e305efafc23b155c98a49b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2207136
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67879}
The lowering for anytrue was assuming that the input nodes are all
integers. The regression test added in https://crrev.com/c/2194471 calls
anytrue with float operands, this was causing the lowering to generate
cmpl instructions with a float register and an immediate, which is
wrong.
The fix is to use GetReplacementsWithType on the input nodes, but
only if the input were floats, since we use Word32Equal.
Drive-by clean up of comments in the aforementioned regression test.
Bug: v8:10535
Change-Id: I4de89516c178e9003a4c745808d831be87918381
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2203400
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Bill Budge <bbudge@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67878}
The codegen uses a bunch of vpmax to try and keep set bits around. The
datatype for vpmax does not need to change for each instruction, since
vpmax U32 will persist set bits just as well. This simplifies the
instruction sequences for S1x8 and S1x16 anytrue.
I added a test to check a special case when a f64x2 contains -0.0 (top
bit set). A previous attempt to optimize codegen used floating point
compare, which does not distinguish between 0.0 and -0.0. So -0.0 will
compare equals to 0.0, and incorrect return 0 for anytrue.
Change-Id: I66013796af08a666009e6b2d774ea7ee7bdfe1ad
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2203113
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Bill Budge <bbudge@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67875}
This is a reland of edf90ee828
The fix here is to call GetUnusedRegister with {}, in
liftoff-assembler-ia32.h emit_i8x16_shri_u.
Original change's description:
> [wasm-simd][liftoff][ia32][x64] Implement i8x16 shr
>
> The code sequence is the same as TurboFan, only wrapped in a template to
> share the implementation.
>
> Bug: v8:9909
> Change-Id: I9c1b37bbfafe91d1bd8edd7f9dafd86ff1c07623
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2202723
> Commit-Queue: Zhi An Ng <zhin@chromium.org>
> Reviewed-by: Clemens Backes <clemensb@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#67842}
Bug: v8:9909
Change-Id: Id56a612cd6580c68a5129e71d7a0e7b29d64b368
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2204080
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67873}
Making them private was a way to hide the constructor, we can
explicitly delete them, which give a better compilation error message as
well.
Also see: https://stackoverflow.com/q/55205874
Bug: v8:10488
Change-Id: I8a116637608fcc6a93d6fc4f5ee014d2db863669
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2204156
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67872}
- Add a default, move, and copy constructor for SsaEnv and use them
when possible.
- Bring the Split method signature in line with Steal and
Unimplemented.
Changes:
Change-Id: I626142fbd1c1ed15c1852b9bfc8b39a0e936b839
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2204278
Commit-Queue: Manos Koukoutos <manoskouk@chromium.org>
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67871}
Asynchronicity can be tricky, in particular if the debugger is enabled
while wasm compilation is happening.
We seem to have open issues in streaming compilation there. As a first
step, which CL adds more tests for async compilation (non-streaming).
R=thibaudm@chromium.org
Bug: v8:10531
Change-Id: Idf16790a91aad437ceb981485512a2f52b791bac
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2206736
Reviewed-by: Thibaud Michaud <thibaudm@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67865}
This is a reland of 69f42d408a
Original change's description:
> [nci] Add Construct_WithFeedback builtin
>
> This builtin essentially acts like InterpreterAssembler::Construct. It
> collects full feedback, then tail-calls either the array constructor
> or the construct builtin.
>
> For now, it is only used in generic lowering, if the --turbo-nci flag
> is passed. One of the next steps will be to measure performance impact
> of feedback collection. If minimal, we may want to enable it
> unconditionally in generic lowering.
>
> Bug: v8:8888
> Change-Id: I8a460a2b5954c26fa72658045a8423c5eee6b611
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2198775
> Commit-Queue: Jakob Gruber <jgruber@chromium.org>
> Reviewed-by: Georg Neis <neis@chromium.org>
> Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> Auto-Submit: Jakob Gruber <jgruber@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#67850}
Tbr: neis@chromium.org
Bug: v8:8888
Change-Id: Ib1a81da998c848d63c0119b3a4e90fc917b15e94
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2206738
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67864}
There are some 'mode' still pending removal.
Bug: v8:9708, v8:10506, v8:6949
Change-Id: Ie1dc0397241bf046ac9737666e700b98018b2ff5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2202906
Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org>
Reviewed-by: Mythri Alle <mythria@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67861}
This reverts commit 69f42d408a.
Reason for revert: The last PS introduced a bug https://ci.chromium.org/p/v8/builders/ci/V8%20Linux64%20-%20fyi/15897.
Original change's description:
> [nci] Add Construct_WithFeedback builtin
>
> This builtin essentially acts like InterpreterAssembler::Construct. It
> collects full feedback, then tail-calls either the array constructor
> or the construct builtin.
>
> For now, it is only used in generic lowering, if the --turbo-nci flag
> is passed. One of the next steps will be to measure performance impact
> of feedback collection. If minimal, we may want to enable it
> unconditionally in generic lowering.
>
> Bug: v8:8888
> Change-Id: I8a460a2b5954c26fa72658045a8423c5eee6b611
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2198775
> Commit-Queue: Jakob Gruber <jgruber@chromium.org>
> Reviewed-by: Georg Neis <neis@chromium.org>
> Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
> Auto-Submit: Jakob Gruber <jgruber@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#67850}
TBR=neis@chromium.org,jgruber@chromium.org,tebbi@chromium.org
Change-Id: I3af168373978d773385b9eda9bc1e243e3cbea09
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: v8:8888
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2206737
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67860}
... and other unused methods.
Bug: v8:9708
Change-Id: Ie3658878024fcddcd5503c6462e5ad873eba19b0
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2206733
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Auto-Submit: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67859}
We constantly fight against scrambled output with --print-wasm-code and
other flags. Passing --single-threaded only partially mitigates this,
because there could still be multiple isolates (e.g. Workers), and we
sometimes failed to really execute in a single thread if that flag was
set.
Hence this CL solves the problem in a more fundamental way: Whenever a
{StdoutStream} is constructed, it implicitly takes a global recursive
mutex. The recursive mutex is needed because we still have some printing
methods that don't take a stream as parameter, and instead create their
own instance of {StdoutStream}, which should not crash of course.
The overhead of taking a mutex should be acceptable, since output to
stdout mostly happens if special tracing flags have been passed, and is
slow anyway.
This CL ensures that the {StdoutStream} is used at least for
--print-code, --print-wasm-code, and --trace-turbo-graph.
More flags can later be ported on demand.
The {JSHeapBroker} class was modified to not contain a {StdoutStream},
but instead create one on demand.
R=mlippautz@chromium.org, tebbi@chromium.orgCC=ahaas@chromium.org
Bug: v8:10506
Change-Id: Ib9cf8d76aa79553b4215bb7775e6d47a8179aafa
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/2201767
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#67855}
