This is the next in the series of simplifying the logic to collect feedback
in compare bytecode handlers. An earlier cl (
https://chromium-review.googlesource.com/c/483399/) modified StrictEquals
bytecode handler. This cl inlines the type feedback collection for the
Equalbytecode handler.
Bug: v8:4280
Change-Id: I36210a2412bb84a3fdb982aabccf8cdefe87e30e
Reviewed-on: https://chromium-review.googlesource.com/498447
Commit-Queue: Mythri Alle <mythria@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45183}
This reverts commit 4fa473cb75.
Reason for revert: Problems when v8 isn't located in a folder called "v8".
Original change's description:
> [test] Don't flatten testcfg globals
>
> This loads each test's testcfg.py as a unique module rather than flattening all into testcfg. Other than accessing LoadTestSuite there should be no references into testcfg files.
>
> Bug: v8:6375
> Change-Id: If863c1b35096b2589111e8091bb7d68f135da674
> Reviewed-on: https://chromium-review.googlesource.com/498807
> Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
> Commit-Queue: Michael Achenbach <machenbach@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#45178}
TBR=jkummerow@chromium.org,machenbach@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Bug: v8:6375
Change-Id: I3600b54279c0d98a39475432c5b2163f510153f0
Reviewed-on: https://chromium-review.googlesource.com/500130
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45180}
This patch also makes concurrent marking visitor loads atomic.
BUG=chromium:694255
Review-Url: https://codereview.chromium.org/2872443003
Cr-Commit-Position: refs/heads/master@{#45179}
This loads each test's testcfg.py as a unique module rather than flattening all into testcfg. Other than accessing LoadTestSuite there should be no references into testcfg files.
Bug: v8:6375
Change-Id: If863c1b35096b2589111e8091bb7d68f135da674
Reviewed-on: https://chromium-review.googlesource.com/498807
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45178}
This CL prevents problems with library libicui18n.so during execution
inspector tests when component is defined as shared library.
TEST=inspector/*
BUG=
Review-Url: https://codereview.chromium.org/2863383003
Cr-Commit-Position: refs/heads/master@{#45176}
This reverts commit 662aa425ba.
Reason for revert: Crashing on Canary
BUG=chromium:718891
Original change's description:
> Reland: [TypeFeedbackVector] Store optimized code in the vector
>
> Since the feedback vector is itself a native context structure, why
> not store optimized code for a function in there rather than in
> a map from native context to code? This allows us to get rid of
> the optimized code map in the SharedFunctionInfo, saving a pointer,
> and making lookup of any optimized code quicker.
>
> Original patch by Michael Stanton <mvstanton@chromium.org>
>
> BUG=v8:6246
> TBR=yangguo@chromium.org,ulan@chromium.org
>
> Change-Id: Ic83e4011148164ef080c63215a0c77f1dfb7f327
> Reviewed-on: https://chromium-review.googlesource.com/494487
> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
> Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#45084}
TBR=ulan@chromium.org,rmcilroy@chromium.org,yangguo@chromium.org,jarin@chromium.org
# Not skipping CQ checks because original CL landed > 1 day ago.
BUG=v8:6246
Change-Id: Idab648d6fe260862c2a0e35366df19dcecf13a82
Reviewed-on: https://chromium-review.googlesource.com/498633
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45174}
This reverts commit f7c25da680.
Reason for revert: Fixed
Original change's description:
> Revert "Introducing an event loop mechanism for d8."
>
> This reverts commit de964dbe57.
>
> Reason for revert:
> https://build.chromium.org/p/client.v8/builders/V8%20Linux/builds/17958
>
> Original change's description:
> > Introducing an event loop mechanism for d8.
> >
> > This mechanism ensures APIs like wasm async complete their work,
> > without requiring use of natives (%APIs).
> >
> > The mechanism is similar to the one used in content_shell,
> > which should allow us to easily port tests in that environment.
> >
> > Review-Url: https://codereview.chromium.org/2842843005
> > Cr-Original-Commit-Position: refs/heads/master@{#44908}
> > Bug:
> > Change-Id: I9deee0d256a600c60b42902fc8ef8478e5546344
> > Reviewed-on: https://chromium-review.googlesource.com/494968
> > Commit-Queue: Mircea Trofin <mtrofin@google.com>
> > Reviewed-by: Jochen Eisinger <jochen@chromium.org>
> > Cr-Commit-Position: refs/heads/master@{#45165}
>
> TBR=bradnelson@chromium.org,mtrofin@chromium.org,mtrofin@google.com,jochen@chromium.org
> NOPRESUBMIT=true
> NOTREECHECKS=true
> NOTRY=true
>
> Change-Id: Iafec2615d705d1990c57229cab3a988c00b5e12f
> Reviewed-on: https://chromium-review.googlesource.com/498630
> Reviewed-by: Michael Achenbach <machenbach@chromium.org>
> Commit-Queue: Michael Achenbach <machenbach@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#45166}
TBR=bradnelson@chromium.org,machenbach@chromium.org,mtrofin@chromium.org,mtrofin@google.com,jochen@chromium.org,v8-reviews@googlegroups.com
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: Ic3c782e918326e291a6cb9bb349c609e9a340b09
Reviewed-on: https://chromium-review.googlesource.com/498430
Reviewed-by: Mircea Trofin <mtrofin@chromium.org>
Commit-Queue: Mircea Trofin <mtrofin@google.com>
Cr-Commit-Position: refs/heads/master@{#45172}
Intrinsic and generic lowering for generator object creation. In a follow-on, create lowering will be addressed.
BUG=v8:6352
Review-Url: https://codereview.chromium.org/2862213002
Cr-Commit-Position: refs/heads/master@{#45171}
This patch expands scope analysis to skip hole initialization
when it can be determined statically that no hole checks will
be generated at runtime.
Two conditions must be met to safely eliminate hole initialization:
- There must not exist a VariableProxy referencing this Variable
whose HoleCheckMode is kRequired
- The Variable must be stack allocated; any other allocation implies
that it may be accessed from not-yet-analyzed scopes (other modules,
inner functions, or eval code) and that code may require
hole checks.
The new logic required removing debug code in full-codegen which is
now incorrect in some cases.
Also fixed Variable's bitfield helpers to take no more space than needed.
Bug: chromium:651637
Change-Id: Ie5ac326af4e05b7a5c3c37cd4d0afba6a51a504d
Reviewed-on: https://chromium-review.googlesource.com/494006
Commit-Queue: Adam Klein <adamk@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45170}
This reverts commit ec619cbd89.
Reason for revert: Crashing on Canary
BUG=chromium:718891
Original change's description:
> [Interpreter] Transition JSFunctions to call optimized code when possible.
>
> Now that the optimized code hangs off the feedback vector, it is possible
> to check whether a function has optimized code available every time it's
> called in the interpreter entry trampoline. If optimized code exists, the
> interpreter entry trampoline 'self-heals' the closure to point to the
> optimized code and links the closure into the optimized code list.
>
> BUG=v8:6246
>
> Change-Id: If1bd7c555bb0551bfe04b36baa6bcf949604717e
> Reviewed-on: https://chromium-review.googlesource.com/488026
> Reviewed-by: Michael Stanton <mvstanton@chromium.org>
> Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#45103}
TBR=rmcilroy@chromium.org,mvstanton@chromium.org
# Not skipping CQ checks because original CL landed > 1 day ago.
BUG=v8:6246
Change-Id: Ibda719be90fddf1d116c03a2a0c3018bcbe76018
Reviewed-on: https://chromium-review.googlesource.com/498632
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45169}
The code for UMA stats (in counters.h) is not thread safe, and can
lead to using pointers with uninitialized values.
Therefore, this CL turns them off when compiling asynchronously.
It also turns back on several UMA stats that were previously turned
off, but no longer need to because the code now knows if it is
running synchronously.
BUG=v8:6361
Review-Url: https://codereview.chromium.org/2864583004
Cr-Commit-Position: refs/heads/master@{#45168}
This reverts commit de964dbe57.
Reason for revert:
https://build.chromium.org/p/client.v8/builders/V8%20Linux/builds/17958
Original change's description:
> Introducing an event loop mechanism for d8.
>
> This mechanism ensures APIs like wasm async complete their work,
> without requiring use of natives (%APIs).
>
> The mechanism is similar to the one used in content_shell,
> which should allow us to easily port tests in that environment.
>
> Review-Url: https://codereview.chromium.org/2842843005
> Cr-Original-Commit-Position: refs/heads/master@{#44908}
> Bug:
> Change-Id: I9deee0d256a600c60b42902fc8ef8478e5546344
> Reviewed-on: https://chromium-review.googlesource.com/494968
> Commit-Queue: Mircea Trofin <mtrofin@google.com>
> Reviewed-by: Jochen Eisinger <jochen@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#45165}
TBR=bradnelson@chromium.org,mtrofin@chromium.org,mtrofin@google.com,jochen@chromium.org
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Change-Id: Iafec2615d705d1990c57229cab3a988c00b5e12f
Reviewed-on: https://chromium-review.googlesource.com/498630
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45166}
This mechanism ensures APIs like wasm async complete their work,
without requiring use of natives (%APIs).
The mechanism is similar to the one used in content_shell,
which should allow us to easily port tests in that environment.
Review-Url: https://codereview.chromium.org/2842843005
Cr-Original-Commit-Position: refs/heads/master@{#44908}
Bug:
Change-Id: I9deee0d256a600c60b42902fc8ef8478e5546344
Reviewed-on: https://chromium-review.googlesource.com/494968
Commit-Queue: Mircea Trofin <mtrofin@google.com>
Reviewed-by: Jochen Eisinger <jochen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45165}
Another fix for more explicit management of ownership. The
CompilationHelper now always owns the WasmModule, and transfers
ownership to the generated WasmModuleWrapper (a Managed<WasmModule>)
once that object is created. Since the stored uniqe_ptr cannot be
accessed any more after this transfer, the creation of the
WasmModuleWrapper is delayed until it is really needed (step 5 in async
compilation).
R=ahaas@chromium.org
Change-Id: I22dea2e14a364ddf76751d97bd0d736a4c0ceff4
Reviewed-on: https://chromium-review.googlesource.com/498507
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45164}
Reason for revert:
Should define its own NO_HARNESS_PATTERN. See comments.
Original issue's description:
> [test] add --no-harness option to debugger tests.
>
> Review-Url: https://codereview.chromium.org/2831083003
> Cr-Commit-Position: refs/heads/master@{#44774}
> Committed: 43c20d4cc5TBR=caitp@igalia.com,yangguo@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
Review-Url: https://codereview.chromium.org/2871593002
Cr-Commit-Position: refs/heads/master@{#45163}
Since the wasm module is verified before starting execution with lazy
compilation, the compilation of individual functions should not fail
later.
This CL changes the implementation to check this condition earlier
and removes unused error paths.
R=ahaas@chromium.org, mstarzinger@chromium.org
BUG=chromium:719286
Change-Id: If4bab457a47f214b457b2e2bc8570cba8c8bbcfd
Reviewed-on: https://chromium-review.googlesource.com/497755
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45161}
Make ModuleResult and FunctionResult return Result<std::unique_ptr<X>>.
This makes memory ownership and transfer of ownership more clear and
avoids a lot of manual releases of the referenced native heap object.
R=ahaas@chromium.org
Change-Id: I7a3f5bd7761b6ae1ebdc7d17ff1b96a8df599871
Reviewed-on: https://chromium-review.googlesource.com/498352
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45160}
We don't currently depend on this, but it might improve performance by
avoiding intermediate copies. The functions are already set up for
perfect forwarding, but without declaring the parameters as forwarding
references, this does not work as expected.
R=ahaas@chromium.org
Change-Id: I2c4d96ea1108b3f884d3e581e74c20aafd232934
Reviewed-on: https://chromium-review.googlesource.com/497409
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45158}
Inside the CompilationHelper, we were creating another compilation
helper to execute sequential or parallel compilation.
I don't see the reason to do so.
R=ahaas@chromium.org
Change-Id: Ib2c4486296a8f923e7e38620879c02963fff7d60
Reviewed-on: https://chromium-review.googlesource.com/497754
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45157}
With this CL we share code among the wasm fuzzers which construct a
module and run it in the interpreter and as compiled code.The fuzzers
themselves only contain the code now which creates the module and the
parameters.
BUG=v8:6325
R=eholk@chromium.org
Change-Id: I1c2d8b013531c86cb27837f1b8ec89d2688c536b
Reviewed-on: https://chromium-review.googlesource.com/490048
Commit-Queue: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Brad Nelson <bradnelson@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45156}
It was replaced by more generic handling in 1320666798, which
is functionally fine, but for performance it makes sense to keep
the fast path.
Review-Url: https://codereview.chromium.org/2864463004
Cr-Commit-Position: refs/heads/master@{#45155}
Also make the macro name more scary, so people don't add new calls
BUG=v8:5830
R=jgruber@chromium.org
Change-Id: I06760110b7f0429d7775345b414c75c8df5e503a
Reviewed-on: https://chromium-review.googlesource.com/497451
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jochen Eisinger <jochen@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45154}
In its destructor, the ErrorThrower already reifies exceptions and
throws them if an error has been set.
R=mtrofin@chromium.org
Change-Id: I17d7a6d300fe4a5860431f214746d053eaf9f104
Reviewed-on: https://chromium-review.googlesource.com/497467
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45153}
Bug:v8:5510
R=yangguo@chromium.org,jgruber@chromium.org
Change-Id: Ieb355110bd858efe2495a6271ffeda67d41af129
Reviewed-on: https://chromium-review.googlesource.com/497153
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Loo Rong Jie <loorongjie@gmail.com>
Cr-Commit-Position: refs/heads/master@{#45151}
History has shown that 99.93% (or more) of all memory allocations are less
than 1 megabyte, and they all appear in the same UMA stat entry.
To give perspective, the entry for <= 1Mb is about 20,000 times larger
than any other entry in the table. This makes the distribution in the
table hard to see.
And, for allocation failures at this size, the percentage of failures
(when compared to number of requests) is soo small (millions to one)
that little data can be gleamed from the <= 1Mb entry.
Note: requires CL https://codereview.chromium.org/2867483002
BUG=chrome:704922
R=bradnelson@chromium.org, bbudge@chromium.org, isherman@chromium.org
Review-Url: https://codereview.chromium.org/2856663002
Cr-Commit-Position: refs/heads/master@{#45148}
After the recent fast-property deletion changes, there can be a
non-empty out-of-object backing store (that previously held properties)
even though the next double property will be stored in-object.
BUG=chromium:718779
Review-Url: https://codereview.chromium.org/2861093004
Cr-Commit-Position: refs/heads/master@{#45146}
As a first step towards improving generator creation, create a builtin
that can do it without a call to the runtime. Thread that builtin
into the interpreter via an intrinsic.
BUG=v8:6352
R=bmeurer@chromium.org
Review-Url: https://codereview.chromium.org/2861983002
Cr-Commit-Position: refs/heads/master@{#45145}
The concurrent marker now colors objects grey and black using
the atomic markbit operations.
The heap visitor is changed in two ways:
1) It iterates the map pointer of each object.
2) It guards object visitation with a predicate, which is overridden
in the concurrent marker with the result of GreyToBlack transition.
BUG=chromium:694255
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_chromium_rel_ng
Review-Url: https://codereview.chromium.org/2866473003
Cr-Commit-Position: refs/heads/master@{#45144}
At this point, last_index is definitely a canonicalized non-negative
number, which implies that any non-Smi last_index is greater than
the maximal string length. That in turn means that the matcher will
definitely fail, and we can avoid the expensive runtime call.
BUG=v8:6365,v8:6344
Review-Url: https://codereview.chromium.org/2863643004
Cr-Commit-Position: refs/heads/master@{#45143}
Adds a flag (--disable-in-process-stack-traces) to not install
signal handlers so that e.g. ASan signal handlers will work.
This flag mirrors chromium's one.
R=jochen@chromium.org
BUG=chromium:716235
Review-Url: https://codereview.chromium.org/2854173002
Cr-Commit-Position: refs/heads/master@{#45142}
This CL migrates the CPP builtin to CSA with fast paths for strings
that can be unpacked to direct one-byte strings. Short strings are
handled directly in CSA, others need to call into C for conversion.
Microbenchmarks for "abcd".toLowerCase() show speedups of 2.5x.
BUG=v8:6353,v8:6344
Review-Url: https://codereview.chromium.org/2859203002
Cr-Commit-Position: refs/heads/master@{#45141}
Forcing an object into dictionary mode now requires deleting
a property that's not the last property.
BUG=chromium:718816
NOTRY=true
Review-Url: https://codereview.chromium.org/2860403002
Cr-Commit-Position: refs/heads/master@{#45140}
Currently the VisitObject function iterates the object and then colors
it black. This does not work well with concurrent marking. The function
should instead first try to mark the object black and iterate its body
only if the color transition succeeds.
BUG=chromium:694255
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_chromium_rel_ng
Review-Url: https://codereview.chromium.org/2863933002
Cr-Commit-Position: refs/heads/master@{#45139}
concurrent marking is enabled.
This patch adds kAtomicity flag to IncrementalMarking that is set
depending on the concurrent marking compile time flag.
BUG=chromium:694255
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_chromium_rel_ng
Review-Url: https://codereview.chromium.org/2858343003
Cr-Commit-Position: refs/heads/master@{#45138}
With this adaptation build V8 in Debug mode will be finished without
error messages.
TEST=
BUG=
Review-Url: https://codereview.chromium.org/2861143002
Cr-Commit-Position: refs/heads/master@{#45137}
Now non-atomic color transition operations return a boolean indicating
whether the transition succeeded or not.
This allows to replace color check and transition operations with a
single transition operation. For example:
if (IsWhite(object)) {
WhiteToBlack(object);
Foo();
}
becomes
if (WhiteToBlack(object)) {
Foo();
}
BUG=chromium:694255
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_chromium_rel_ng
Review-Url: https://codereview.chromium.org/2863703004
Cr-Commit-Position: refs/heads/master@{#45136}
CodeStubAssembler::StrictEqual is modified to collect type feedback when
requested (https://chromium-review.googlesource.com/c/483399/). There
are a couple of Gotos introduced in that cl, which are only necessary
when collecting type feedback. Changed it to generate these Gotos only
whencollecting type feedback.
Bug: v8:4280
Change-Id: I322a6065efcfb601bd9f5dc25be6e6bb00b2d1c0
Reviewed-on: https://chromium-review.googlesource.com/496268
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Commit-Queue: Mythri Alle <mythria@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45135}
