Updates the EffectControlLinerizer to feed all nodes it processes
through the GraphAssembler. This is required to enable the GraphAssembler
to maintain the schedule for TurboProp, but also means we can avoid
keeping track of the current effect and control nodes in the
EffectControlLinearizer and use the GraphAssembler for that instead.
Also modifies EffectControlLinearizer to avoid accessing the basic block
while lowering nodes, since a basic block updating GraphAssembler could
modify the current block. Once lowered, we finalizes GraphAssembler to
provide the updated basic block for which the original control should be
processed.
BUG=v8:9684
Change-Id: Ibe7f396e15f8bebf35b9c50d56c245cbc92547f5
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1842453
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64528}
Specifically string, object, proxy & regexp.
With this CL, the pattern is removed from all torque files.
R=tebbi@chromium.org
Change-Id: Ifcc1efda6053df8f02fc730825055f6cd5644e84
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873691
Commit-Queue: Michael Stanton <mvstanton@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64527}
This is a reland of 5d57f4e143
Breakage addressed by
https://chromium-review.googlesource.com/c/chromium/src/+/1874491
Original change's description:
> [Intl] Ship calendar and numberingSystem options
>
> Ship the "calendar" and "numberingSystem" options for
> Intl.DateTimeFormat (both options) and Intl.NumberFormat (only the later
> one) and support other calendar. Also consider the calendar while
> choosing calendar pattern.
>
> I2L: http://shorturl.at/bgkAH
> I2S: http://shorturl.at/nuKUV
>
> Flags: --harmony-intl-add-calendar-numbering-system
> --harmony-intl-other-calendars
>
> API owner approvals: chrishtr@ yoav@yoav.wsbratell.d@gmail.com
>
> Plan to land into m80 tree and only merge after 10/17 m79 branch off.
>
> Bug: v8:9154, v8:9155, v8:9320
> Change-Id: Ifa209919a40db60465f99405f3620a3b73b10204
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1838436
> Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
> Commit-Queue: Frank Tang <ftang@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#64437}
Bug: v8:9154, v8:9155, v8:9320, chromium:1016909
Change-Id: Ie8eac6283042cb66fc4a98fd2230385c068fa759
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1874089
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Frank Tang <ftang@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64526}
The immediate value was incorrect and-ed with 3.
Also, for palignr, if the immediate is larger that 32
(for 128-bit) or 16 (for 64-bit), it produces a zero result.
In the case of disasm, I don't think we need to do anything.
Change-Id: I258fd16fbe57fa7e00ab306d0fbf1b1b73950566
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1876660
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64524}
Crashkeys are static and non-refcounted, so when one thread clears
a crashkey, it affects all other threads. This means, we cannot set
them in parallel running jobs such as ScavengePage. This change moves
the crashkey about heap collection up the stack into the main thread.
Change-Id: I28f16eaadd9b122c06a68d1d4207f27319994509
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1874384
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Irina Yatsenko <irinayat@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#64523}
Adds the ability for the GraphAssembler to operate on, and maintain, a
scheduled graph. This will be used by TurboProp to maintain the initial
schedule created before effect-control-linearization, by updating this schedule
during effect-control, select and memory lowering stages rather than doing a
later reschedule.
In order to do this, an internal BlockUpdater is added to GraphAssembler,
which is enabled by passing the schedule to the GraphAssembler. The
GraphAssembler is modified to call into the block updater when nodes are added
and updates the schedule with new basic blocks when new control flow is updated.
BUG=v8:9684
Change-Id: I6d428ad21d869c472bb20f43cc8caf44722f090a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1841355
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64519}
This function was only used for the write barrier since the store
buffer only stored slots and needed a way to get to the object's start.
Now that we insert into the remembered set directly from the write
barrier this isn't an issue anymore: the write barrier knows the
object start.
Change-Id: I701465ea40b7c4ee20404ecbcf3750e5fa6fd219
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1876049
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Dominik Inführ <dinfuehr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64518}
The CL fixes the following builtins:
%TypedArray%.prototype.map
Bug: v8:4153
Change-Id: I1db5716d5044788da8a792e4449d501ac7507823
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1876047
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64515}
The size/length limits are still at kSmiMaxValue.
Bug: v8:4153
Change-Id: I6ffda50a3b9f235b97a3718e86df7deadce9f6f8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1874346
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64514}
Added the opcodes kWord32Equal and kChangeTaggedToCompressed to
DecompressionOptimizer.
Currently TaggedEqual node generates ChangeTaggedToCompressed
nodes before the Word32Equal inputs. This means we need to also
propagate kOnly32BitsObserved from kChangeTaggedToCompressed.
Bug: v8:7703
Change-Id: I61fee07269696c60e065e1bbfee3a65c1b1537fc
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1871911
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Commit-Queue: Santiago Aboy Solanes <solanes@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64513}
This changes the way {SaveFPRegsMode} is stored in a {CallDescriptor}.
The value is now encoded in the flags field, similar to all other
boolean properties. It removes the last mutable field from the call
descriptor, making the struct fully immutable.
R=neis@chromium.org
Change-Id: I6bca118c7ce0c5bc77e7aa2fcc6541439bd27b3b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1868616
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64512}
These accessors do not make any use of the interpreter, hence we can
define them on the WasmInstanceObject alone. This will allow to reuse
them for other (non-interpreted) frames.
R=mstarzinger@chromium.org
Bug: v8:9676
Change-Id: Iff8b665a4c25581b934c25b66a13cebe044cb02c
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1875097
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64511}
Add a new PagedSpace called OffThreadSpace. This space will be used for
off-thread allocation -- it never marks or sweeps, and always expands
into fresh pages. Once allocation completes, this space's pages can be
merged into the old space.
The space is similar to the CompactionSpace, and merging for both is
identical, so we intrduce a new LocalSpace base class that both extend.
They differ in interaction with the sweeper and in how new pages are
acquired.
This patch adds the new space and uses it in a few unittests. Future
work will use it in the main source code.
Bug: chromium:1011762
Change-Id: Ia008cc95c6e1ef1d1e7ae305fa80fbfc1ff4be2d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873690
Commit-Queue: Leszek Swirski <leszeks@chromium.org>
Reviewed-by: Dominik Inführ <dinfuehr@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64509}
Previously, embedders had to fetch the 'exec' property off the RegExp
prototype in order to call exec (and such calls involve two
transitions between C++ and JS).
This CL exposes a convenient RegExp::Exec method through the API.
Bug: v8:9695
Change-Id: I57a9174626143d26f2ea34676b8e55fef64932cf
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1864940
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64508}
The new ToIndex() must eventually replace ToSmiIndex().
The CL fixes the following abstract operations:
GetViewValue(view, requestIndex, isLittleEndian, type)
SetViewValue(view, requestIndex, isLittleEndian, type, value)
and the following builtins:
DataView.prototype.getXXX
DataView.prototype.setXXX
where XXX are all typed elements.
Bug: v8:4153
Change-Id: Ic2f33e91b59426deb0efa28bb4c15253e80a299c
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1874345
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Jakob Kummerow <jkummerow@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64506}
Port f22837dbf1R=xwafish@gmail.com
Change-Id: Iece021bb832618cac0d111639b388d94f7da7028
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1868773
Commit-Queue: Mu Tao <pamilty@gmail.com>
Auto-Submit: Mu Tao <pamilty@gmail.com>
Reviewed-by: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64505}
The {InstructionCode} is only used to store plain (non-negative) values
of the {ArchOpcode} enum, or additionally encodes {BitField} values. The
underlying base type 'U' of a {BitField} is uint32_t. To avoid all the
numerous implicit conversions between int32_t and uint32_t, this is
changing the {InstructionCode} so that uint32_t is used exclusively.
R=neis@chromium.org
BUG=v8:9872
Change-Id: If64107ad9298011e219b4827735eafb51465beb0
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1869193
Reviewed-by: Georg Neis <neis@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64503}
This operator implicitly reads the stack pointer register and is thus
not pure. This CL removes it from the list of pure operators and
inserts it into the effect chain at its use sites.
Drive-by: Alpha-sort the list in CanAllocate.
Bug: v8:9534
Change-Id: Icf96fb3e308600dbacec3dbfb7386a4f0d0bdc39
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1875098
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Georg Neis <neis@chromium.org>
Auto-Submit: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64502}
The CL refactors the following builtins:
String.prototype.startsWith
String.prototype.endsWith
to use ClampToIndexRange(x, len) instead of NumberMin(NumberMax(x, 0), len).
Bug: v8:8996
Change-Id: I20ab42088168e517840385cc2db435361004d9c0
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873702
Commit-Queue: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64500}
Port 26afd8f314
Change-Id: I4033e6bb50166d59679dddd8ba0480b4bc4b821d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1874969
Commit-Queue: Mu Tao <pamilty@gmail.com>
Auto-Submit: Mu Tao <pamilty@gmail.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64499}
The same change on the {CopyChars} function gave an unexpected 10-20%
speedup on microbenchmarks across platforms (ia32, x64, Atom_x64).
This CL explores whether a similar change generally speeds up {MemCopy}
and {MemMove} (only on x64 for now). If this is the case, we might be
able to carefully extend the same pattern to other platforms and remove
custom assembly implementations without too much regression.
R=leszeks@chromium.org
Bug: v8:9810
Change-Id: Ib9674807b67cd2a463680b97a91ae1b41c3be65b
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1871607
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Commit-Queue: Clemens Backes <clemensb@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64498}
Port 434f96812f
Change-Id: I9e9cb8b9ca56d5af3b5ffffa5908501c9214752d
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1874968
Commit-Queue: Mu Tao <pamilty@gmail.com>
Auto-Submit: Mu Tao <pamilty@gmail.com>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64497}
These counters track how often the slow path of these two builtins is
hit. Exec is very permissive, its fast-path check doesn't look at the
regexp prototype at all. Replace is strict; any change on the
prototype will trigger the slow path.
Chromium CL: https://crrev.com/c/1875250
Bug: v8:5577
Change-Id: I9807f43829981445b75b5c5d29800cbdac9bc26a
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873698
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Peter Marshall <petermarshall@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64496}
This CL integrates the "v8_nosnapshot" source set into the "mksnapshot"
binary. The "mksnapshot" binary is the last use-site after unconditionally
enabling snapshot and embedded builtins.
Bug: v8:8519
Change-Id: I8e737e10a3f7729e18ac71733b811bf399ebf330
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873685
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Simon Zünd <szuend@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64494}
Port a0b95232a9R=xwafish@gmail.com
Change-Id: I15c385a676e574f36d8c25c6f6294ac344b0af43
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1868557
Reviewed-by: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Auto-Submit: Mu Tao <pamilty@gmail.com>
Cr-Commit-Position: refs/heads/master@{#64492}
Port afbbfcbe1cR=xwafish@gmail.com
Change-Id: Iab3a9f32d8bccddcdca8d9a874869e62ae961948
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1868558
Auto-Submit: Mu Tao <pamilty@gmail.com>
Reviewed-by: Zhi An Ng <zhin@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64491}
Port 0089006fc5
Original Commit Message:
.. similar to how it is applied in the interpreter. We reserve a stack
slot for the backtrack count, increment it on each backtrack, and fail
if the limit is hit.
Change-Id: I51879e6cafb3c77d635fc6e84e7e2c5ce3a088e6
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873445
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Auto-Submit: Mu Tao <pamilty@gmail.com>
Cr-Commit-Position: refs/heads/master@{#64490}
This is a reland of 08b26f53c6
Fixed the original crash, by removing a disasm for psllq and psrlq
that is now handled by the macro list.
Original change's description:
> Clean up macros
>
> Move some instruction definitions into sse-instr, which is used to
> generate some disasm tests, so we can remove some cases there.
>
> Bug: v8:9810
> Change-Id: I0615ec823396da08bc5d234cf1dabca6afd3f052
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1866965
> Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
> Commit-Queue: Zhi An Ng <zhin@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#64441}
Bug: v8:9810
Change-Id: I69335a889f5f72b76a79e4e9860835232e6e38a8
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1872298
Reviewed-by: Deepti Gandluri <gdeepti@chromium.org>
Commit-Queue: Zhi An Ng <zhin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64487}
Port 3cad6bf5d7
Original Commit Message:
This is a reland of c7c47c68f2.
This makes TSAN happy in addition to:
Previously I presumed that the context read from a frame in the profiler was
a valid context. Turns out that on non-intel we're not guaranteed that the
frame is properly set up. In the case we looked at, the profiler took a
sample right before writing the frame marker indicating a builtin frame,
causing the "context" pointer from that frame to be a bytecode array. Since
we'll read random garbage on the stack as a possible context pointer, I made
the code reading the native context from it a little more defensive.
Bug: v8:9860
Tbr: ulan@chromium.org, neis@chromium.org, ishell@chromium.org
Original change's description:
> [runtime] Move Context::native_context to the map
>
> Remove the native context slot from contexts by making context maps
> native-context-specific. Now we require 2 loads to go from a context to the
> native context, but we have 1 field fewer to store when creating contexts.
>
> Change-Id: I3c0d7c50c94060c4129db684f46a567de6f30e8d
> Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1859629
> Commit-Queue: Toon Verwaest <verwaest@chromium.org>
> Reviewed-by: Igor Sheludko <ishell@chromium.org>
> Reviewed-by: Peter Marshall <petermarshall@chromium.org>
> Reviewed-by: Maya Lekova <mslekova@chromium.org>
> Reviewed-by: Georg Neis <neis@chromium.org>
> Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
> Reviewed-by: Toon Verwaest <verwaest@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#64296}
R=xwafish@gmail.com
Change-Id: I6496a8c5be8cbabf48cddc2d59111410f31eb75f
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1868774
Auto-Submit: Mu Tao <pamilty@gmail.com>
Commit-Queue: Bill Budge <bbudge@chromium.org>
Reviewed-by: Toon Verwaest <verwaest@chromium.org>
Reviewed-by: Bill Budge <bbudge@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64486}
Remove inactive contributors (cira and mnita) and add ftang and jshin.
Change-Id: Ibd1e4f11553b78e1cd938a7ce8958fe5341829d2
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1872540
Reviewed-by: Frank Tang <ftang@chromium.org>
Commit-Queue: Adam Klein <adamk@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64484}
Otherwise the value cannot be used in gtests in EXPECT_GT expressions.
The reason is that EXPECT_GT takes values by reference and not by
value, so the declaration in v8.h is not enough.
R=clemensb@chromium.org
Bug: v8:9810
Change-Id: I6f85024d483701fcc65d1ccfc5254dbdbcce16c1
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1871606
Reviewed-by: Clemens Backes <clemensb@chromium.org>
Commit-Queue: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64482}
The new predicate indicates whether the backing store was created for
an ArrayBuffer or a SharedArrayBuffer. It is useful for some embedders.
Bug: v8:9380
Change-Id: I804063bb8c4c17815defd6538ce6a1b32f6a4531
Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/1873689
Commit-Queue: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Cr-Commit-Position: refs/heads/master@{#64479}
