We need the shared function info of inlined functions to prevent code
flushing for their unoptimized code, and also to make sure that liveedit
can find the proper functions to deoptimize.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1156403002
Cr-Commit-Position: refs/heads/master@{#28677}
The control flow optimization should work independent of the JSGraph. We
used the JSGraph there because it was convinient, not because it was
necessary.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1160863003
Cr-Commit-Position: refs/heads/master@{#28674}
The SimplifiedOperatorReducer is (mostly) unused, except for the very
rough store elimination, and just eats compilation time.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1162563002
Cr-Commit-Position: refs/heads/master@{#28673}
The list of inlined functions is used in exactly two places - for live
edit and to prevent code flushing for inlined functions - and those are
fine with SharedFunctionInfo and don't require a closure.
This is one additional step towards inlining based on SharedFunctionInfo
instead of JSFunction.
R=jarin@chromium.org
Review URL: https://codereview.chromium.org/1145893003
Cr-Commit-Position: refs/heads/master@{#28672}
This test is little-endian specific and therefore it can't run
correctly on big-endian platforms.
BUG=
Review URL: https://codereview.chromium.org/1155113008
Cr-Commit-Position: refs/heads/master@{#28667}
Port 44e9810345
Original commit message:
When we enter a method that needs access to the [[HomeObject]]
we allocate a local variable `.home_object` and assign it the
value from the [[HomeObject]] private symbol. Something along
the lines of:
method() {
var .home_object = %ThisFunction()[home_object_symbol];
...
}
R=arv@chromium.org, dstence@us.ibm.com, michael_dawson@ca.ibm.com
BUG=
Review URL: https://codereview.chromium.org/1161603004
Cr-Commit-Position: refs/heads/master@{#28665}
Implements the strong mode proposal's restrictions on the ability of user code
to modify the prototype of strong objects.
Setting the strong bit is still wip, so this change will only affect those
objects that have the bit correctly set. The tests reflect this, and will be
expanded as more objects can be marked as strong.
BUG=v8:3956
LOG=N
Review URL: https://codereview.chromium.org/1143623002
Cr-Commit-Position: refs/heads/master@{#28664}
The DCHECK was a lie. The idea was that - when a bookmark is set -
the scanner must clearly be at a character boundary and hence the
bookmark does not need to save a 'partial' UTF-8 code point. The
first part is true - the Scanner is always at a character boundary -
but the 'partial' UTF-8 code point is at the end of a block, not at
the current character position of the Scanner.
Hence, the 'partial' character needs to be saved as well.
jkummerow: Thanks for noticing.
BUG=chromium:470930
R=jochen@chromium.org, jkummerow@chromium.org
LOG=N
Review URL: https://codereview.chromium.org/1154773004
Cr-Commit-Position: refs/heads/master@{#28661}
Also adapt code generation to pass the slot to the
store/keyed-store ic. AST nodes ObjectLiteral, Assignment,
ForEach, Call and CountOperation now include one or more
feedback vector ic slot ids.
BUG=
Review URL: https://codereview.chromium.org/1161623002
Cr-Commit-Position: refs/heads/master@{#28659}
Strong Object/Array literals are currently being created with incorrect
internal prototypes. This CL fixes this and extends the test suite to check.
BUG=
LOG=N
Review URL: https://codereview.chromium.org/1158933002
Cr-Commit-Position: refs/heads/master@{#28655}
Before we used to scale the growing factor based on freed global handles (which may have caused jank when many global handles got freed on site navigation).
BUG=
Review URL: https://codereview.chromium.org/1158433003
Cr-Commit-Position: refs/heads/master@{#28652}
This lifts the sqrt(n) limit on number of evacuation candidates,
replaces O(n * sqrt(n)) algorithm with O(n*log(n)) algorithm, and
removes hard-coded constants.
Evacuation candidates are selected as follows:
1) Sort pages from the most free to the least free.
2) Select the first m pages as evacuation candidates such that m is as
large as possible under the two conditions:
- The total size of live objects in the first m pages does not exceed
the given limit. This is based on the assumption that the evacuation cost is
proportional to the total size of moved objects.
- The fragmentation of the (m+1)-th page does not exceed the given
limit.
Review URL: https://codereview.chromium.org/1038313003
Cr-Commit-Position: refs/heads/master@{#28651}
And delete remnants of non-vectorized LoadICs from the type feedback oracle
Review URL: https://codereview.chromium.org/1147253004
Cr-Commit-Position: refs/heads/master@{#28650}
This simplifies inlining, in that we only need to update uses of Start
and inputs of End instead of walking the whole inlinee to update all
outer frame states.
R=mstarzinger@chromium.org
Review URL: https://codereview.chromium.org/1146403008
Cr-Commit-Position: refs/heads/master@{#28649}
If both inputs to JSStrictEqual/JSStrictNotEqual are unique values (i.e.
values with a canonical representation), we can lower the comparison to
ReferenceEqual instead of StringEqual or CompareIC.
Review URL: https://codereview.chromium.org/1154303002
Cr-Commit-Position: refs/heads/master@{#28646}
When we enter a method that needs access to the [[HomeObject]]
we allocate a local variable `.home_object` and assign it the
value from the [[HomeObject]] private symbol. Something along
the lines of:
method() {
var .home_object = %ThisFunction()[home_object_symbol];
...
}
BUG=v8:3867, v8:4031
LOG=N
Review URL: https://codereview.chromium.org/1135243004
Cr-Commit-Position: refs/heads/master@{#28644}
These are similar to the Map/Set constructors when called with an array,
except that they are guaranteed to be side-effect free if called with
a packed array.
This will be useful in implementing structured clone which, as
specified in HTML, speaks in terms of the internal [[MapData]]
and [[SetData]] slots without going through the exposed iteration
ES semantics.
BUG=v8:3340
LOG=y
Review URL: https://codereview.chromium.org/1155893003
Cr-Commit-Position: refs/heads/master@{#28642}
The April 14 2015 final draft of the ES6 specification states that the
`prototype` property of generator function instances should be writable.
BUG=v8:4140, v8:4140
LOG=N
R=arv@chromium.org
Review URL: https://codereview.chromium.org/1153633003
Cr-Commit-Position: refs/heads/master@{#28641}
These return arrays representing the current contents of the given
Map/Set. They are similar to what would be returned by the JS code:
Array.from(collection)
except that they are guaranteed side-effect free.
This will be useful in implementing structured clone which, as
specified in HTML, speaks in terms of the internal [[MapData]]
and [[SetData]] slots without going through the exposed iteration
ES semantics.
BUG=v8:3340
LOG=y
Review URL: https://codereview.chromium.org/1148383007
Cr-Commit-Position: refs/heads/master@{#28640}
Only supports constructing new objects and returning size.
Followup patch will need to add ability to retrieve and
set contents in order to support structured clone.
Also removes a bunch of outdated "experimental" markers from v8.h.
BUG=v8:3340
LOG=y
Review URL: https://codereview.chromium.org/1157453002
Cr-Commit-Position: refs/heads/master@{#28637}
Port eca5b5d7ab
Original commit message:
* Hash code is now just done with a private own symbol instead of the hidden string, which predates symbols.
* In the long run we should do all hidden properties this way and get rid of the
hidden magic 0-length string with the zero hash code. The advantages include
less complexity and being able to do things from JS in a natural way.
* Initially, the performance of weak set regressed, because it's a little harder
to do the lookup in C++. Instead of heroics in C++ to make things faster I
moved some functionality into JS and got the performance back. JS is supposed to be good at looking up named properties on objects.
* This also changes hash codes of Smis so that they are always Smis.
Performance figures are in the comments to the code review. Summary: Most of js-perf-test/Collections is neutral. Set and Map with object keys are 40-50% better. WeakMap is -5% and WeakSet is +9%. After the measurements, I fixed global proxies, which cost 1% on most tests and 5% on the weak ones :-(.
In the code review comments is a patch with an example of the heroics we could do in C++ to make lookup faster (I hope we don't have to do this. Instead of checking for the property, then doing a new lookup to insert it, we could do one lookup and handle the addition immediately). With the current benchmarks above this buys us nothing, but if we go back to doing more lookups in C++ instead of in stubs and JS then it's a win.
In a similar vein we could give the magic zero hash code to the hash code
symbol. Then when we look up the hash code we would sometimes see the table
with all the hidden properties. This dual use of the field for either the hash
code or the table with all hidden properties and the hash code is rather ugly,
and this CL gets rid of it. I'd be loath to bring it back. On the benchmarks quoted above it's slightly slower than moving the hash code lookup to JS like in this CL.
One worry is that the benchmark results above are more monomorphic than real
world code, so may be overstating the performance benefits of moving to JS. I
think this is part of a general issue we have with handling polymorphic code in
JS and any solutions there will benefit this solution, which boils down to
regular property access. Any improvement there will lift all boats.
R=erikcorry@chromium.org, dstence@us.ibm.com, michael_dawson@ca.ibm.com
BUG=
Review URL: https://codereview.chromium.org/1157123002
Cr-Commit-Position: refs/heads/master@{#28634}
This configures *san in v8 just like in chromium's
common.gypi. I also addresses compilation problems with ICU
and usage of instrumented libc++.
TBR=svenpanne@chromium.org
Review URL: https://codereview.chromium.org/1146863006
Cr-Commit-Position: refs/heads/master@{#28631}
