The patch introduces a dedicated dispatching class for JIT code events. It is
set as a helper on the isolate.
This allows classes across v8 to break their dependency on Logger and CpuProfiler.
These two became just regular clients of the dispatcher.
BUG=v8:4789
Review-Url: https://codereview.chromium.org/2061623002
Cr-Commit-Position: refs/heads/master@{#37005}
This CheckBounds simplified operator is similar to the HBoundsCheck in
Crankshaft, and is hooked up to the new type feedback support in the
SimplifiedLowering. We use it to check the index bounds for keyed
property accesses.
Note to perf sheriffs: This will tank quite a few benchmarks, as the
operator makes some redundant branch elimination ineffective for
certain patterns of keyed accesses. This does require more serious
redundancy elimination, which we will do in a separate CL. So ignore
any regressions from this CL, we know there will be a few.
R=jarin@chromium.org
BUG=v8:4470,v8:5100
Committed: https://crrev.com/85e5567dae66a918500ae94c5568221137a0f5d4
Review-Url: https://codereview.chromium.org/2035893004
Cr-Original-Commit-Position: refs/heads/master@{#36947}
Cr-Commit-Position: refs/heads/master@{#37003}
The new approach is that instead of compiling custom handlers for
every global object's PropertyCell it uses single dispatcher that
caches PropertyCells in respective slot of the feedback vector.
Currently the new LoadGlobalIC machinery is disabled.
This CL also removes unused LoadGlobalViaContext* stuff.
BUG=chromium:576312
LOG=Y
Review-Url: https://codereview.chromium.org/2065113002
Cr-Commit-Position: refs/heads/master@{#37002}
These simplified operators are used to perform the hole checks when
loading elements from a holey array. Depending on the CheckHoleMode,
they either return the hole as undefined or some NaN, or deoptimize
if the value is the hole or the hole NaN.
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/2066223002
Cr-Commit-Position: refs/heads/master@{#37001}
Reason for revert:
[Sheriff] Speculative revert for http://crbug.com/620279
Original issue's description:
> Reland: Add a trace-event for each runtime-stats timer (CL 2052523002)
>
> The trace-events will have a high overhead when turned on, but they are in a disabled-by-default category.
>
> As long as the off overhead is negligible, this CL allows us to understand the behavior of V8 rather than its performance at the moment.
>
> The original CL was failing the TSAN builder, the variable in question was intended to be accessed quickly with no guarantee.
> Switched to using an Atomic variable with no barrier read/write.
>
> BUG=v8:5089
>
> patch from issue 2052523002 at patchset 100001 (http://crrev.com/2052523002#ps100001)
>
> Committed: https://crrev.com/fd7080cbefc21f2f890b5db00d4eadf163e2cbbf
> Cr-Commit-Position: refs/heads/master@{#36973}
TBR=cbruni@chromium.org,fmeawad@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=v8:5089
Review-Url: https://codereview.chromium.org/2068143002
Cr-Commit-Position: refs/heads/master@{#36997}
This is to make sure the test in question does not run out of stack
space during bootstrapping on any configuration. Our fuzzers take the
test an run it against a broad spectrum of configuration. The new size
of 100 is used throughout our test suite as "the smallest" stack size.
R=jkummerow@chromium.org
TEST=mjsunit/regress/regress-1132
BUG=chromium:619744
Review-Url: https://codereview.chromium.org/2068993002
Cr-Commit-Position: refs/heads/master@{#36995}
Adds an evacuation mode that allows moving pages within new space without
copying objects.
Basic idea:
a) Move page within new space
b) Sweep page to make iterable and process ArrayBuffers
c) Finish sweep till next scavenge
Threshold is currently 70% live bytes, i.e., the same threshold we use
to determine fragmented pages.
BUG=chromium:581412
LOG=N
CQ_EXTRA_TRYBOTS=tryserver.v8:v8_linux_arm64_gc_stress_dbg,v8_linux_gc_stress_dbg,v8_mac_gc_stress_dbg,v8_linux64_tsan_rel,v8_mac64_asan_rel
Review-Url: https://codereview.chromium.org/1957323003
Cr-Commit-Position: refs/heads/master@{#36990}
Reason for revert:
Blows up on the waterfall.
Original issue's description:
> [turbofan] Stage binop type feedback.
>
> Committed: https://crrev.com/28fbec405548c43088cade1cd17d8407ca948a21
> Cr-Commit-Position: refs/heads/master@{#36988}
TBR=mvstanton@chromium.org,bmeurer@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
Review-Url: https://codereview.chromium.org/2066193002
Cr-Commit-Position: refs/heads/master@{#36989}
If JSAdd or JSSubtract has number feedback, there's no benefit to use
the speculative versions of NumberAdd and/or NumberSubtract. Relying on
the existing operators gives us some nice truncations instead, so let's
stick to that for now.
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/2063073004
Cr-Commit-Position: refs/heads/master@{#36987}
Named capture groups may be specified using the /(?<name>pattern)/u
syntax, with named backreferences specified as /\k<name>/u. They're
hidden behind the --harmony-regexp-named-captures flag, and are only
enabled for unicode regexps.
R=yangguo@chromium.org
BUG=
Review-Url: https://codereview.chromium.org/2050343002
Cr-Commit-Position: refs/heads/master@{#36986}
The Vector type is deprecated, and new code should use ZoneVector
instead. This new overload of NewStringFromTwoByte will be used in an
upcoming regexp CL.
R=bmeurer@chromium.org
BUG=
Review-Url: https://codereview.chromium.org/2065053002
Cr-Commit-Position: refs/heads/master@{#36985}
Now that we have the PlainPrimitiveToNumber operator(s), we can unify
all the places where we expect a number, but can also safely handle any
plain-primitive (via ToNumber truncation).
Drive-by-fix: Also handle Math.min consistently with Math.max.
R=jarin@chromium.org
Review-Url: https://codereview.chromium.org/2064953004
Cr-Commit-Position: refs/heads/master@{#36984}
Rolling v8/build to 696616b539a8ab8cf400a4cfaead772154c6645c
Rolling v8/tools/clang to 792bfbf9f3126cb51eddd5c6f2f9e9a67393687b
Rolling v8/tools/mb to 5c0ee173872849b18b0837590c094a8579bc0013
TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org
Review-Url: https://codereview.chromium.org/2064113003
Cr-Commit-Position: refs/heads/master@{#36983}
The last 4 test cases in test/cctest/wasm/test-run-wasm-asmjs.cc added by the CL 36911 (https://codereview.chromium.org/2061583002) use float_t and double_t type for WasmRunner.
For examples: At line 249: WasmRunner<float_t> r(&module, MachineType::Uint32());
But float_t and double_t depends on FLT_EVAL_METHOD macro of compiler. FLT_EVAL_METHOD is variant on different platform, if the FLT_EVAL_METHOD is 2, both float_t and double_t will be long
double and gcc or clang will met error when compiling WasmRunner<long double> r(&module,MachineType::Uint32());
For more details, please refer:
float_t: http://www.cplusplus.com/reference/cmath/float_t/
FLT_EVAL_METHOD: https://en.wikipedia.org/wiki/C99 check the IEEE 754 floating point support section directly.
This CL used float and double to replace float_t and double_t to avoid this issue.
BUG=
Review-Url: https://codereview.chromium.org/2066703003
Cr-Commit-Position: refs/heads/master@{#36982}
The modifications were necessary to properly represent asm types:
1) fround is no longer an overloaded function.
2) the constructor for MinMaxTypes now takes a return type.
3) Adds pseudo-types for representing the Load/Store types for fp heap views.
BUG=
Review-Url: https://codereview.chromium.org/2069443002
Cr-Commit-Position: refs/heads/master@{#36980}
Reporting use counts by invoking a callback once per occurrence has
a large overhead cost in certain situations, for example when it needs
to be dispatched to a different thread (which is the case for Web Workers).
Parsing large scripts can produce a lot of occurrences (strict/sloppy mode
once per function).
Chromium (the only known user of UseCounters so far) does not actually care
about number of occurrences, but simply whether they happened at least once.
This commit changes behavior to report features at most once, which dramatically
improves performance for impacted use cases, and should not affect the only
known real world usage.
R=littledan@chromium.org
BUG=chromium:614775
Review-Url: https://codereview.chromium.org/2062203002
Cr-Commit-Position: refs/heads/master@{#36979}
Support for relocatable globals, to facilitate compilation before
instantiation.
BUG=v8:5072
Review-Url: https://codereview.chromium.org/2062003002
Cr-Commit-Position: refs/heads/master@{#36978}
Reason for revert:
Does not fix the bug it intended to fix.
Original issue's description:
> Fix scope flags for default parameters
>
> R=rossberg,adamk
> BUG=chromium:616386
>
> Committed: https://crrev.com/4cc1331c341f5bfbeee54fec521f682a8a406af4
> Cr-Commit-Position: refs/heads/master@{#36755}
TBR=adamk@chromium.org,rossberg@chromium.org,jochen@chromium.org
# Not skipping CQ checks because original CL landed more than 1 days ago.
BUG=chromium:616386
Review-Url: https://codereview.chromium.org/2062593002
Cr-Commit-Position: refs/heads/master@{#36976}
The trace-events will have a high overhead when turned on, but they are in a disabled-by-default category.
As long as the off overhead is negligible, this CL allows us to understand the behavior of V8 rather than its performance at the moment.
The original CL was failing the TSAN builder, the variable in question was intended to be accessed quickly with no guarantee.
Switched to using an Atomic variable with no barrier read/write.
BUG=v8:5089
patch from issue 2052523002 at patchset 100001 (http://crrev.com/2052523002#ps100001)
Review-Url: https://codereview.chromium.org/2063853002
Cr-Commit-Position: refs/heads/master@{#36973}
