This proceeds the experiment using the legacy bots, since luci bots
are blocked on https://crbug.com/776734.
NOTRY=true
Bug: chromium:772816, chromium:776734
Change-Id: I7532e04e98a96d69dbeefcd6edf66fb0cddbe111
Reviewed-on: https://chromium-review.googlesource.com/743365
Reviewed-by: Sergiy Byelozyorov <sergiyb@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49034}
Following up on adding n-ary nodes, this extends the parser and
interpreter to support n-ary logical operations.
Bug: v8:6964
Bug: chromium:731861
Change-Id: Ife2141c389b9abccd917ab2aaddf399c436ef777
Reviewed-on: https://chromium-review.googlesource.com/735497
Reviewed-by: Adam Klein <adamk@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49029}
The bytecode interpreter nor TF rely on the explicit return value of store ICs anymore, so we can just return whatever is in the result slot. It won't be visible to JS anyway.
Bug:
Change-Id: I389615d1d77c5b050832f23a08e3d3bc07d9cbc6
Reviewed-on: https://chromium-review.googlesource.com/743366
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49027}
This is preparation for supporting OOB loads from strings, which
requires the KeyedLoadIC to track this information.
Bug: v8:6936, v8:7014
Change-Id: Ide132244ee523397dd418d21fe3377976f6633fd
Reviewed-on: https://chromium-review.googlesource.com/743481
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49026}
This reverts commit 32141e93ff.
Reason for revert: Got merged to stable end of last week.
Original change's description:
> Disable --turbo-inline-array-builtins by default.
>
> Current chrome stable has a high number of crashes due to bugs in
> this feature. These bugs are already fixed but the fixes are hard
> to merge back. Therefore we decided to disable the feature in stable.
> This CL is intended to be merged to stable and then reverted in tot.
>
> Bug: chromium:762020
> Change-Id: Ibd5a08e3b303a204fb84a408271a1c0f97cc5b7b
> Reviewed-on: https://chromium-review.googlesource.com/738176
> Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
> Commit-Queue: Georg Neis <neis@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#48931}
TBR=jarin@chromium.org,neis@chromium.org
# Not skipping CQ checks because original CL landed > 1 day ago.
Bug: chromium:762020
Change-Id: Id1fe2f417e09a7b451484fc16710b09469cdb74d
Reviewed-on: https://chromium-review.googlesource.com/743321
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49025}
The CheckSeqString operator unnecessarily materialized two bits for the
separate conditions, but we can actually check for SeqString with just a
single mask instead, which saves the intermediate computations and two
registers.
Bug: v8:5267
Change-Id: Ib937a5a4eebdc271a2bbbc6f074f5992e06b2fa9
Reviewed-on: https://chromium-review.googlesource.com/741702
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49023}
This was discovered on the babel test of the web-tooling-benchmark,
which suffers from a high %KeyedGetProperty overhead, and most of
these calls come from the fact that the KeyedLoadIC_Megamorphic bails
out to the runtime call for all String instance types. Just handling
in-bound accesses to characters reduces the overhead incurred by
%KeyedGetProperty from roughly 9% to roughly 2% only.
This reduces the number of runs per second on the babel test by around
7-8% on average.
Bug: v8:6936, v8:7014
Change-Id: I0dc247d7d6457c7032636d2852cb54cef1b24979
Reviewed-on: https://chromium-review.googlesource.com/743012
Reviewed-by: Yang Guo <yangguo@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49021}
Interpreter entries may be called from wasm functions, when debugging.
That means that, when moving on the native heap, interpreter entries
need to be native, too. That means they cannot reference movable GC
objects.
The only such reference was to the instance object, which is needed
in the WasmRunInterpreter runtime function. We can fetch the instance
using GetWasmInstanceOnTop() instead.
Bug: v8:6876
Change-Id: I7198a849cc0e441b057537a570a3dfa6f3197149
Reviewed-on: https://chromium-review.googlesource.com/742391
Commit-Queue: Mircea Trofin <mtrofin@chromium.org>
Reviewed-by: Clemens Hammacher <clemensh@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49019}
This eliminates the AstValue class, effectively moving its
implementation into the Literal AstNode. This should cause
no difference in behavior, but it does signal some shifts
in the underlying system. Biggest changes include:
- Reduction in AST memory usage
- No duplicate HeapNumbers in Ignition constant pools
- Non-String values are allocated either at constant pool
creation time (or at boilerplate creation time for literals),
rather than at AstValueFactory::Internalize() time.
There are a variety of test-only/debug-only changes due to these
switches as well.
Bug: v8:6984
Change-Id: I5f178040ce2796d4e7370c24d1063419e1c843a1
Reviewed-on: https://chromium-review.googlesource.com/731111
Commit-Queue: Adam Klein <adamk@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Marja Hölttä <marja@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49013}
This happens when RCS are enabled dynamically and the callsite is inside
the background parser.
BUG=chromium:760649
Change-Id: I216b955ed91d9c663ce3027aaa8ffb515bfe13ab
Reviewed-on: https://chromium-review.googlesource.com/740911
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Commit-Queue: Alexei Filippov <alph@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49010}
We have a check on maximum number of levels that can be inlined. This
in some cases causes performance cliffs, when we cannot inline a small
function because it has exceeded the number of levels. This cl removes
that check. The intuition is that, having gone down several levels in
a particular line stopping inlining that chain and exploring a new
call site may not be beneficial.
Bug: v8:6871
Change-Id: I120056db38e78ce48dff010b6cf994259238582a
Reviewed-on: https://chromium-review.googlesource.com/741705
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49009}
Instead of creating a runtime call for the static class field
initializer in the AST, we do it in the bytecode generator.
This adds the initializer function to the ClassLiteral AST node.
Bug: v8:5367
Change-Id: Iffaa6531511023812011ee19fc96cea9e5c9d3f3
Reviewed-on: https://chromium-review.googlesource.com/736315
Reviewed-by: Georg Neis <neis@chromium.org>
Reviewed-by: Mythri Alle <mythria@chromium.org>
Reviewed-by: Adam Klein <adamk@chromium.org>
Commit-Queue: Sathya Gunasekaran <gsathya@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49008}
To avoid accessing the heap during asm.js compilation, use the pending
error handler to store the pending warnings such that they can be reported
later during finalization.
As part of this change, refactor PendingCompilationErrorHandler to have a
MessageDetails class holding details of either error or warning messages.
BUG=v8:5203
Change-Id: I5b09254f8899b8dc57d94f1986c7183da847eae3
Reviewed-on: https://chromium-review.googlesource.com/735607
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49007}
This moves the handling of the receiver and the final spread argument
into Generate_InterpreterPushArgs and merges the stack allocation to a
single claim operation for the whole argument setup. When we start
padding arguments in TF, we will simply need to claim one extra slot
and store padreg to it in Generate_InterpreterPushArgs (hence the
remaining TODO).
This also replaces the single use of Generate_CheckStackOverflow with
Generate_StackOverflowCheck and removes the former. This change is also
done for arm, ia32 and x64.
Bug: v8:6644
Change-Id: I8d06dda96dbc4a6f219b73c711f894320c2f6cdf
Reviewed-on: https://chromium-review.googlesource.com/738031
Commit-Queue: Georgia Kouveli <georgia.kouveli@arm.com>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49005}
The status-file flags and the flags from the test case's source code
must always overwrite extra flags set by bots.
Bug: v8:6924
Cq-Include-Trybots: master.tryserver.v8:v8_linux_noi18n_rel_ng
Change-Id: I0e2aabb69da7cfb8ba6c1c79bd3851462071a6ac
Reviewed-on: https://chromium-review.googlesource.com/732656
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Reviewed-by: Sergiy Byelozyorov <sergiyb@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#49001}
(This is a reland of cb84b6f624)
This works in the ICs since compiled handlers are not shared anymore.
Bug: chromium:759734
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Change-Id: I3b7d6ae34bb4cfa213f679354fac56fb51bfd5ed
Reviewed-on: https://chromium-review.googlesource.com/739822
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48999}
Because this test uses heap verification, it is quite slow. Split it
into 4 smaller tests to avoid test timeout and allow them to be run
in parallel.
R=ahaas@chromium.org
Bug:
Change-Id: Ie4ac841d1d8215019bb5cfcc335daea6b10ab789
Reviewed-on: https://chromium-review.googlesource.com/738146
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Ben Titzer <titzer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48998}
This should avoid changes in code serializer in a follow-up class literals
improvement CL.
Bug: v8:5799
Change-Id: I4e4386333bcc7abb30399d4e9abba4eee80ca932
Reviewed-on: https://chromium-review.googlesource.com/733037
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48996}
This adds two tests to verify that the --liftoff flag has the indented
effect, and that Liftoff compilation is off by default.
R=titzer@chromium.org
Bug: v8:6600
Change-Id: Ie7e13184b5068f572b78dbdf7abbcded6d859fc5
Reviewed-on: https://chromium-review.googlesource.com/733561
Commit-Queue: Clemens Hammacher <clemensh@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48995}
We now represent the SameValue operation explicitly in TurboFan and the
operation can thus participate in all kinds of optimizations. Especially
we get rid of the JSCall node in the general case, which blocks several
optimizations across the call. The general, baseline performance is now
always on par with StrictEqual.
Once the StrictEqual operator is also a simplified operator, we should
start unifying the type based optimizations in SimplifiedLowering.
In the micro-benchmark we go from
testStrictEqual: 1422 ms.
testObjectIs: 1520 ms.
testManualSameValue: 1759 ms.
to
testStrictEqual: 1426 ms.
testObjectIs: 1357 ms.
testManualSameValue: 1766 ms.
which gives the expected result.
Bug: v8:7007
Change-Id: I0de3ff6ff6209ab4c3edb69de6a16e387295a9c8
Reviewed-on: https://chromium-review.googlesource.com/741228
Reviewed-by: Jaroslav Sevcik <jarin@chromium.org>
Commit-Queue: Benedikt Meurer <bmeurer@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48994}
... and use them to complete the BigInt function.
Bug: v8:6791
Change-Id: Ia36db86b92d1a0cfcb783516e04d6c0e3750f194
Reviewed-on: https://chromium-review.googlesource.com/737643
Commit-Queue: Georg Neis <neis@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48993}
This is an intermediate CL to move the complete --trace-map infrastructure to
a log-based version.
Change-Id: I0673052b1b87fe338e38dc609434a52af6a0652d
Reviewed-on: https://chromium-review.googlesource.com/738835
Commit-Queue: Camillo Bruni <cbruni@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48988}
This is a reland of c34042cc64
Original change's description:
> [heap-profiler] Do not treat WeakMap values as weak.
>
> For the WeakHashTable objects only mark keys as weak while leaving values as strong references.
>
> BUG=chomium:773722
>
> Change-Id: Iabd5ba293d05fe68a2af6503fcdd711ecc182482
> Reviewed-on: https://chromium-review.googlesource.com/730771
> Commit-Queue: Alexei Filippov <alph@chromium.org>
> Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#48985}
TBR=ulan@chromium.org
Bug: chomium:773722
Change-Id: Icede16fde528d147cde5c3f6c72f2029876b099f
Reviewed-on: https://chromium-review.googlesource.com/740722
Reviewed-by: Alexei Filippov <alph@chromium.org>
Commit-Queue: Alexei Filippov <alph@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48987}
For the WeakHashTable objects only mark keys as weak while leaving values as strong references.
BUG=chomium:773722
Change-Id: Iabd5ba293d05fe68a2af6503fcdd711ecc182482
Reviewed-on: https://chromium-review.googlesource.com/730771
Commit-Queue: Alexei Filippov <alph@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#48985}
