This adds support for lazy deserialization of JS-linkage (TFJ) builtins,
still gated behind the --lazy-deserialization flag. If enabled, we
proceed as follows:
During isolate initialization, only eager builtins are deserialized. All
references to lazy builtins are replaced by the DeserializeLazy builtin.
In particular, this happens in the builtin table (Builtins::builtins_)
and in SharedFunctionInfo objects.
When calling into a not-yet deserialized function (i.e. the JSFunction's
code object is the DeserializeLazy builtin), the DeserializeLazy builtin
takes over. It checks the builtin table to see if the target builtin
(determined by looking at the builtin id stored on the
SharedFunctionInfo) has already been deserialized. If so, it simply
copies the builtin code object to the JSFunction and SharedFunctionInfo.
Otherwise, we enter Runtime::kDeserializeLazy to deserialize the
builtin.
With --lazy-deserialization, isolate deserialization is 11% faster
(1.5ms vs. 1.7ms), and code_space->Size() is 33% lower (984K vs.
1475K).
Moving relocation infos & handler tables out of the partial snapshot
cache would additionally let us save up to 30K per isolate. Adding code
stubs to that list increases further potential savings to 262K.
Bug: v8:6624
Change-Id: I0ac7d05d165d2466998269bd431ac076a311cbeb
Reviewed-on: https://chromium-review.googlesource.com/649166
Commit-Queue: Jakob Gruber <jgruber@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47818}
This change adapts the Call bytecode handlers such that they don't require
a stack frame. It does this by modifying the call bytecode handler to
tail-call the Call or InterpreterPushArgsAndCall builtins. As a result, the
callee function will return to the InterpreterEntryTrampoline when it returns
(since this is the return address on the interpreter frame), which is
adapted to dispatch to the next bytecode handler. The return bytecode
handler is modified to tail-call a new InterpreterExitTramoline instead
of returning to the InterpreterEntryTrampoline.
Overall this significanlty reduces the amount of stack space required for
interpreter frames, increasing the maximum depth of recursive calls from
around 6000 to around 12,500 on x64.
BUG=chromium:753705
Change-Id: I23328e4cef878df3aca4db763b47d72a2cce664c
Reviewed-on: https://chromium-review.googlesource.com/634364
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Leszek Swirski <leszeks@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47617}
Code size in snapshot can be reduced ~41KB
Contributed by kanghua.yu@intel.com
Bug: None
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng
Change-Id: Ib73af39fe97cd38728affea40c593236f15bf6e5
Reviewed-on: https://chromium-review.googlesource.com/588751
Commit-Queue: Pan Deng <pan.deng@intel.com>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47531}
Code aging is no longer supported by any remaining compilers now
that full codegen has been removed. This CL removes all vestiges of
code aging.
BUG=v8:6409
Change-Id: I945ebcc20c7c55120550c8ee36188bfa042ea65e
Reviewed-on: https://chromium-review.googlesource.com/619153
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Marja Hölttä <marja@chromium.org>
Commit-Queue: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47501}
This reverts commit af37f6b970.
Reason for revert: Reverted dependency fixed.
Original change's description:
> Revert "[wasm] Reference indirect tables as addresses of global handles"
>
> This reverts commit 186099d49f.
>
> Reason for revert: Need to revert:
> https://chromium-review.googlesource.com/c/613880
>
> Original change's description:
> > [wasm] Reference indirect tables as addresses of global handles
> >
> > This sets us up for getting the wasm code generation off the GC heap.
> > We reference tables as global handles, which have a stable address. This
> > requires an extra instruction when attempting to make an indirect call,
> > per table (i.e. one for the signature table and one for the function
> > table).
> >
> > Bug:
> > Change-Id: I83743ba0f1dfdeba9aee5d27232f8823981288f8
> > Reviewed-on: https://chromium-review.googlesource.com/612322
> > Commit-Queue: Mircea Trofin <mtrofin@chromium.org>
> > Reviewed-by: Brad Nelson <bradnelson@chromium.org>
> > Cr-Commit-Position: refs/heads/master@{#47444}
>
> TBR=bradnelson@chromium.org,titzer@chromium.org,mtrofin@chromium.org
>
> Change-Id: Ic3dff87410a51a2072ddc16cfc83a230526d4c56
> No-Presubmit: true
> No-Tree-Checks: true
> No-Try: true
> Reviewed-on: https://chromium-review.googlesource.com/622568
> Reviewed-by: Michael Achenbach <machenbach@chromium.org>
> Commit-Queue: Michael Achenbach <machenbach@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#47450}
TBR=bradnelson@chromium.org,machenbach@chromium.org,titzer@chromium.org,mtrofin@chromium.org
Change-Id: I3dc5dc8be26b5462703edac954cbedbb8f504c1e
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/622035
Reviewed-by: Mircea Trofin <mtrofin@chromium.org>
Commit-Queue: Mircea Trofin <mtrofin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47455}
This reverts commit 186099d49f.
Reason for revert: Need to revert:
https://chromium-review.googlesource.com/c/613880
Original change's description:
> [wasm] Reference indirect tables as addresses of global handles
>
> This sets us up for getting the wasm code generation off the GC heap.
> We reference tables as global handles, which have a stable address. This
> requires an extra instruction when attempting to make an indirect call,
> per table (i.e. one for the signature table and one for the function
> table).
>
> Bug:
> Change-Id: I83743ba0f1dfdeba9aee5d27232f8823981288f8
> Reviewed-on: https://chromium-review.googlesource.com/612322
> Commit-Queue: Mircea Trofin <mtrofin@chromium.org>
> Reviewed-by: Brad Nelson <bradnelson@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#47444}
TBR=bradnelson@chromium.org,titzer@chromium.org,mtrofin@chromium.org
Change-Id: Ic3dff87410a51a2072ddc16cfc83a230526d4c56
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Reviewed-on: https://chromium-review.googlesource.com/622568
Reviewed-by: Michael Achenbach <machenbach@chromium.org>
Commit-Queue: Michael Achenbach <machenbach@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47450}
This sets us up for getting the wasm code generation off the GC heap.
We reference tables as global handles, which have a stable address. This
requires an extra instruction when attempting to make an indirect call,
per table (i.e. one for the signature table and one for the function
table).
Bug:
Change-Id: I83743ba0f1dfdeba9aee5d27232f8823981288f8
Reviewed-on: https://chromium-review.googlesource.com/612322
Commit-Queue: Mircea Trofin <mtrofin@chromium.org>
Reviewed-by: Brad Nelson <bradnelson@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47444}
This removes:
- CodeBreakIterator for FCG code.
- RelocModes for debug breaks.
- Code generator for debug break slots.
- GC support for debug break slots.
- Code flag to indicate code with debug break slots.
- Builtin type DBG.
- Mechanisms to replace FCG code in the debugger and LiveEdit.
- Runtime entry to the debugger from debug break slots.
R=bmeurer@chromium.org, rmcilroy@chromium.org, ulan@chromium.org
Bug: v8:6409
Change-Id: I5662c8800e3ef1b1584ad107bfe0aae26c9d8abb
Reviewed-on: https://chromium-review.googlesource.com/613263
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Commit-Queue: Yang Guo <yangguo@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47364}
The way we access wasm addresses or sizes is the same, on
a platform. We have 2 size parameters - memory and table - and
2 addresses - globals and memory.
The CL also renames for generality the address setting API.
Bug:
Change-Id: Ib66c3aff6a0ab4313391528cd2692749bb389559
Reviewed-on: https://chromium-review.googlesource.com/612597
Commit-Queue: Brad Nelson <bradnelson@chromium.org>
Reviewed-by: Brad Nelson <bradnelson@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47350}
Bug: 749486
The feature is off by default, and could be turned on via
`v8_enable_csa_write_barrier = true`. With this CL, only x64 uses this
feature
Change-Id: Ie024f08b7d796a4cc4d55285dc9fe796780f0e53
Reviewed-on: https://chromium-review.googlesource.com/588891
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Tobias Tebbi <tebbi@chromium.org>
Reviewed-by: Ulan Degenbaev <ulan@chromium.org>
Commit-Queue: Albert Mingkun Yang <albertnetymk@google.com>
Cr-Commit-Position: refs/heads/master@{#47122}
Until now, when generating a builtin, it can only embed builtins
(as call targets) that have already been generated. This is either
achieved by reordering the builtins list, or by loading the call
target at runtime from the builtins list (see
MacroAssembler::TailCallBuiltin).
This patch works around this issue by filling the builtins list
with dummy code objects, which are later replaced with the completed
actual builtins. In release mode, this adds around 3ms to 140ms we
previously needed to populate the builtins list.
Change-Id: I7d451b3c09a1db4b9e755548102a80c7f0dfada2
Reviewed-on: https://chromium-review.googlesource.com/586531
Commit-Queue: Yang Guo <yangguo@chromium.org>
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47062}
Register configuration data is not the same as frame configuration data.
This CL moves the last remnants of register configuration into
the assembler files, to be with the other register configuration
macros.
Next step: extract this register configuration data into
platform-specific files that can be included independent of the
assembler.
R=mstarzinger@chromium.org
Bug:
Change-Id: I10933b5090be94e90e2a1442197528dfe30bb566
Reviewed-on: https://chromium-review.googlesource.com/595590
Commit-Queue: Ben Titzer <titzer@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#47044}
This switches the "code entry" field on JSFunction to no longer be an
inner pointer into a Code object (i.e. to the start of the instruction
stream), but a properly tagged pointer instead.
Motivation behind this is the ability to treat this field regularly as
part of escape analysis in the optimizing compiler. Also simplifies the
object visitation for JSFunction objects.
R=bmeurer@chromium.org
Change-Id: Ib53a3fc5f3d783a6fed06dbcab319f5568632acc
Reviewed-on: https://chromium-review.googlesource.com/577890
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Yang Guo <yangguo@chromium.org>
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46844}
This makes {NeedsDebugHookCheck} the default for all invocations, as
there is no call-site left that doesn't perform said check. All other
pieces of the {CallWrapper} are dead since Crankshafts removal.
R=jgruber@chromium.org
Change-Id: I158b816c089ede42972e8a7bdfc6ef0c02053a6b
Reviewed-on: https://chromium-review.googlesource.com/577531
Reviewed-by: Jakob Gruber <jgruber@chromium.org>
Commit-Queue: Michael Starzinger <mstarzinger@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46758}
The tail call implementation is hidden behind the --harmony-tailcalls
flag, which is off-by-default (and has been unstaged since February).
It is known to be broken in a variety of cases, including clusterfuzz
security issues (see sample Chromium issues below). To avoid letting
the implementation bitrot further on trunk, this patch removes it.
Bug: v8:4698, chromium:636914, chromium:724746
Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng;master.tryserver.v8:v8_linux_noi18n_rel_ng
Change-Id: I9cb547101456a582374fdf7b1a3f044a9ef33e5c
Reviewed-on: https://chromium-review.googlesource.com/569069
Commit-Queue: Adam Klein <adamk@chromium.org>
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Reviewed-by: Igor Sheludko <ishell@chromium.org>
Reviewed-by: Ross McIlroy <rmcilroy@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46651}
TF will instantiate the Assembler when we're already on a background
thread, so it's not safe to read out the heap's max_old_generation_size
(it can change). This CL simply removes the use of that value from the
assembler. If the buffer gets too large we will fail when creating the
actual code object.
Bug: v8:6048
Change-Id: Ifb8a64c90222e4516117d237b001779fae060d28
Reviewed-on: https://chromium-review.googlesource.com/567921
Reviewed-by: Michael Lippautz <mlippautz@chromium.org>
Reviewed-by: Michael Starzinger <mstarzinger@chromium.org>
Commit-Queue: Georg Neis <neis@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46581}
The use of double variables to store bit patterns may lead to bit flips
when the stored bit pattern is a signaling NaN (sNaN). Operations on a
sNaN variable (even just returning the variable from a function) may
turn it into a quiet NaN (qNaN), flipping the signaling bit and
affecting the information stored in the variable.
We observed this behaviour on ia32 architectures and therefore in the
simulator builds for other platforms. The use of the wrapper class
Double should prevent this behaviour.
R=ahaas@chromium.org
Change-Id: Ibd1119924a59db771fd4c250689ad9c2a35fff75
Reviewed-on: https://chromium-review.googlesource.com/562771
Reviewed-by: Jaideep Bajwa <bjaideep@ca.ibm.com>
Reviewed-by: Andreas Haas <ahaas@chromium.org>
Commit-Queue: Enrico Bacis <enricobacis@google.com>
Cr-Commit-Position: refs/heads/master@{#46533}
This introduces a new builtin (MapLookupHashIndex) and uses it
in Turbofan to compute Map.p.get and Map.p.has.
I have also refactored the existing CSA builtins for Map.p.get and
Map.p.has to use the new builtin under the hood.
The code for the lookup has been also improved.
- Specialized lookups for smis, strings, heap numbers and everything else.
- the advantage is that we can use fast equalities for the lookup.
- strings can likely be optimized further if we care about the
internalized string fast case.
- Instead of a call to runtime to get the hash code, we now call C directly.
In the Turbofan implementation itself, there are no special optimizations yet.
The next step is to teach load elimination to reuse the indexes from
previous calls of MapLookupHashIndex.
BUG=v8:6410
Change-Id: I0b1a70493eb031d444e51002f6b2cc1f30ea2b68
Reviewed-on: https://chromium-review.googlesource.com/560169
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Jaroslav Sevcik <jarin@chromium.org>
Cr-Commit-Position: refs/heads/master@{#46510}
On x86, signalling NaNs get converted to quiet NaNs when they get push
on the stack and popped again. This happens in the code generation for
arm, specifically for the vmov instruction with the immediate parameter.
This CL replaces the vmov function in assembler-arm to take the
immediate as a uint64_t instead of a double, to guarantee that the bit
pattern does not change even if the parameter is a signalling NaN.
BUG=v8:6564
Change-Id: I062559f9a7ba8b0f560628e5c39621ca578c3e7d
Reviewed-on: https://chromium-review.googlesource.com/558964
Commit-Queue: Andreas Haas <ahaas@chromium.org>
Reviewed-by: Ben Titzer <titzer@chromium.org>
Reviewed-by: Martyn Capewell <martyn.capewell@arm.com>
Cr-Commit-Position: refs/heads/master@{#46418}
OrderedHashSet doesn't need a ::Get, so we can move it to
OrderedHashMap.
Bug: v8:5717
Change-Id: I9606d8c4608473f9daecf8a87b4dd2e3b9570246
Reviewed-on: https://chromium-review.googlesource.com/522348
Reviewed-by: Camillo Bruni <cbruni@chromium.org>
Commit-Queue: Sathya Gunasekaran <gsathya@chromium.org>
Cr-Commit-Position: refs/heads/master@{#45740}
Instead of allocating and embedding certain heap numbers into the code
during code assembly, emit dummies but record the allocation requests.
Later then, in Assembler::GetCode, allocate the heap numbers and patch
the code by replacing the dummies with the actual objects. The
RelocInfos for the embedded objects are already recorded correctly when
emitting the dummies.
R=jarin@chromium.org
BUG=v8:6048
Review-Url: https://codereview.chromium.org/2900683002
Cr-Commit-Position: refs/heads/master@{#45635}
This patch also adds sharing of code target entries, which requires
sharing the RelocInfo for those entries as well. The disassembler
is also modified in order to print comments for the RelocInfo that
is now shared.
This improves the snapshot size for arm by about 4%.
BUG=
Review-Url: https://codereview.chromium.org/2869683004
Cr-Commit-Position: refs/heads/master@{#45497}
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}
In the simulators, the ExternalReference constructor rewrites external
addresses, which involves mutating a linked list rooted in the isolate.
We already construct external references concurrently (at least in Turbofan),
but the list mutation was not thread-safe (though no crashes are known). This
CL adds the necessary locking.
BUG=v8:6048
Review-Url: https://codereview.chromium.org/2852983002
Cr-Commit-Position: refs/heads/master@{#45014}
The StringIndexOf fast path used to be very narrow, only allowing
one-byte single-char search strings (and a one-byte subject string).
This changes the CSA fast path to call into our internal SearchString C++
function instead (after attempting to unpack both Strings), and can handle
strings of arbitrary length and encoding. The only remaining runtime call is
when either string needs to be flattened.
BUG=
Review-Url: https://codereview.chromium.org/2814373002
Cr-Commit-Position: refs/heads/master@{#44718}
Taking the slow runtime path for every non-internalized string key
can be avoided by doing optimistic string table lookups: if there
is a matching entry, use that; if there isn't, then no existing
object has a property with that name.
The hashing/internalizing logic is in C++ and called directly.
Review-Url: https://codereview.chromium.org/2811333002
Cr-Commit-Position: refs/heads/master@{#44650}
The last CL https://chromium-review.googlesource.com/c/456707/ caused
some pretty heavy performance regressions. After experimenting, it
seems the easiest and most straight-forward way to copy the elements
into the new typed array is to do it in JS.
Adds a fast path for typed arrays, where the source typed array has
the same elements kind, in which case we can just copy the backing
store using memcpy.
This CL also removes regression test 319120 which is from a pwn2own
vulnerability. The old code path enforced a maximum byte_length
that was too low, which this change removes. The length property of
the typed array must be a Smi, but the byte_length, which can be up
to 8x larger than length for a Float64Array, can be a heap number.
We can also re-use some of the logic from ConstructByLength when
deciding whether to allocate the buffer on- or off-heap, so that
is factored out into InitializeBasedOnLength. We can also re-use
the DoInitialize helper instead of calling into the runtime,
meaning we can remove InitializeFromArrayLike.
BUG=v8:5977,chromium:705503,chromium:705394
Change-Id: I63372652091d4bdf3a9491acef9b4e3ac793a755
Reviewed-on: https://chromium-review.googlesource.com/459621
Reviewed-by: Benedikt Meurer <bmeurer@chromium.org>
Commit-Queue: Peter Marshall <petermarshall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#44301}
The dotAll flag changes behavior of the dot '.' character to match every
possible single character instead of excluding certain line terminators.
The implementation is staged behind --harmony-regexp-dotall.
Spec proposal: https://github.com/mathiasbynens/es-regexp-dotall-flag
BUG=v8:6172
Review-Url: https://codereview.chromium.org/2780173002
Cr-Commit-Position: refs/heads/master@{#44295}
This is a first step towards moving Turbofan code generation off the main thread.
Summary of the changes:
- AssemblerBase no longer has a pointer to the isolate. Instead, its
constructor receives the few things that it needs from the isolate (on most
architectures this is just the serializer_enabled flag).
- RelocInfo no longer has a pointer to the isolate. Instead, the functions
that need it take it as an argument. (There are currently still a few that
implicitly access the isolate through a HeapObject.)
- The MacroAssembler now explicitly holds a pointer to the isolate (before, it
used to get it from the Assembler).
- The jit_cookie also moved from AssemblerBase to the MacroAssemblers, since
it's not used at all in the Assemblers.
- A few architectures implemented parts of the Assembler with the help
of a Codepatcher that is based on MacroAssembler. Since the Assembler no
longer has the isolate, but the MacroAssembler still needs it, this doesn't
work anymore. Instead, these Assemblers now use a new PatchingAssembler.
BUG=v8:6048
Review-Url: https://codereview.chromium.org/2732273003
Cr-Commit-Position: refs/heads/master@{#43890}
