- 17 Feb, 2017 19 commits
-
-
jarin authored
Revert of [turbofan] extend escape analysis to reduce CheckMaps (patchset #7 id:140001 of https://codereview.chromium.org/2680973013/ ) Reason for revert: Likely causing http://crbug.com/692254 BUG=chromium:692254 Original issue's description: > [turbofan] extend escape analysis to reduce CheckMaps > > R=bmeurer@chromium.org > > BUG= > > Review-Url: https://codereview.chromium.org/2680973013 > Cr-Commit-Position: refs/heads/master@{#43163} > Committed: https://chromium.googlesource.com/v8/v8/+/f01c8a6e4b24ab38e58bb141fc251bc6b75dbe21 TBR=bmeurer@chromium.org,tebbi@chromium.org # Not skipping CQ checks because original CL landed more than 1 days ago. BUG= Review-Url: https://codereview.chromium.org/2704573003 Cr-Commit-Position: refs/heads/master@{#43282}
-
addaleax authored
Add `ValueSerializer::SetTreatArrayBufferViewsAsHostObjects()` which instructs the `ValueSerializer` to treat ArrayBufferView objects as host objects. BUG=v8:5926 Review-Url: https://codereview.chromium.org/2696133007 Cr-Commit-Position: refs/heads/master@{#43281}
-
Michael Achenbach authored
For results processor to work, we need to maintain the assumption that the cwd is equal to the currently running benchmark directory. NOTRY=true TBR=clemensh@chromium.org Change-Id: I585c301dea846194652645973470786ad43aa280 Reviewed-on: https://chromium-review.googlesource.com/444784Reviewed-by: Michael Achenbach <machenbach@chromium.org> Reviewed-by: Clemens Hammacher <clemensh@chromium.org> Commit-Queue: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#43280}
-
Camillo Bruni authored
BUG=691323 Change-Id: I84f2c90355982567c421639e115745eadd5fcb21 Reviewed-on: https://chromium-review.googlesource.com/441964Reviewed-by: Caitlin Potter <caitp@igalia.com> Reviewed-by: Camillo Bruni <cbruni@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#43279}
-
Marja Hölttä authored
Add some tests which already pass. BUG=v8:5516 R=vogelheim@chromium.org Change-Id: Icf5cf5e40010deb2efb69dea750661854b055566 Reviewed-on: https://chromium-review.googlesource.com/444744 Commit-Queue: Marja Hölttä <marja@chromium.org> Reviewed-by: Daniel Vogelheim <vogelheim@chromium.org> Cr-Commit-Position: refs/heads/master@{#43278}
-
clemensh authored
Instead of placing a runtime call to StackGuard in the compiled wasm code, we just call the builtin, which is cheaper. By passing Smi::kZero as context, we save even more code space and avoid embedding the context in the code. The WasmStackGuard builtin then calls the new WasmStackGuard runtime function, which gets the context from the instance attached to the calling wasm code, and then does the usual StackGuard logic. For the unity benchmark in asm-wasm mode, generated code size reduces from 63.0 to 61.6 MB (-2.1%). R=titzer@chromium.org, ahaas@chromium.org, mstarzinger@chromium.org Review-Url: https://codereview.chromium.org/2691993004 Cr-Commit-Position: refs/heads/master@{#43277}
-
Michael Lippautz authored
A fast promotion mode that works solely on GC heuristics without requiring compiler or other profiler support by implementing zero-copy evacuation for new space. - Once a threshold of survived bytes in the Scavenger is reached the mode is activated. - In fast promotion mode all pages are moved to old space instead of performing a Scavenge. - The inevitable upcoming full MC determines whether the decision whether the mode should stay on or be turned off based on the young generation survival rate. BUG=chromium:693413 Change-Id: Ifdf296092a9bac609f9dcdfb47a24046f3093745 Reviewed-on: https://chromium-review.googlesource.com/442560 Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Reviewed-by: Ulan Degenbaev <ulan@chromium.org> Reviewed-by: Hannes Payer <hpayer@chromium.org> Cr-Commit-Position: refs/heads/master@{#43276}
-
vabr authored
https://codereview.chromium.org/2694003002/ introduced "SyntaxError: Lexical declaration cannot appear in a single-statement context" for the case when let + desctructuring from a list happen. As was pointed out in https://codereview.chromium.org/2694003002/#msg18, the case without destructuring would also benefit from a better message: if a single statement is expected and "let identifier = ..." is seen, the error is indeed again that the lexical declaration is not a statement. However, the current error is "Unexpected identifier", because the parser tries to accept "let" as an identifier in an expression statement, and then gives up seeing the other identifier after "let". This CL ensures that the parser recognises the error properly and reports accordingly. It also renames the existing test, which contains destructuring, and adds the one with a non-destructuring lexical declaration. BUG=v8:5686 Review-Url: https://codereview.chromium.org/2697193007 Cr-Commit-Position: refs/heads/master@{#43275}
-
bmeurer authored
When trying to infer receiver maps for property accesses, go hunting in the effect chain to find an earlier CheckMaps or StoreField for the receiver. BUG=v8:5267 R=jarin@chromium.org Review-Url: https://codereview.chromium.org/2704563005 Cr-Commit-Position: refs/heads/master@{#43274}
-
bmeurer authored
In the JSInliner we shouldn't introduce JSConvertReceiver needlessly if the input is already known to produce a receiver, i.e. if it's a JSCreate or something like that. BUG=v8:5267 R=jarin@chromium.org Review-Url: https://codereview.chromium.org/2697513010 Cr-Commit-Position: refs/heads/master@{#43273}
-
mvstanton authored
Set flag --mark-shared-functions-for-tier-up to true. R=leszeks@chromium.org Review-Url: https://codereview.chromium.org/2694363006 Cr-Commit-Position: refs/heads/master@{#43272}
-
Michael Achenbach authored
This reverts commit e1f7ae33. Reason for revert: Breaks on the bot: https://build.chromium.org/p/client.v8/builders/V8%20Linux64%20-%20cfi/builds/8773/steps/compile/logs/stdio Original change's description: > [swarming] Dogfood exparchive feature > > The new swarming archiving method requires isolate files to be copied to the build dir. > > BUG=chromium:598990 > NOTRY=true > TBR=tandrii@chromium.org,tansell@chromium.org,vogelheim@chromium.org > > Change-Id: I87633725d9ce715684f761ce0c8d2691642c5389 > Reviewed-on: https://chromium-review.googlesource.com/444644 > Reviewed-by: Michael Achenbach <machenbach@chromium.org> > Commit-Queue: Michael Achenbach <machenbach@chromium.org> > Cr-Commit-Position: refs/heads/master@{#43266} TBR=machenbach@chromium.org,vogelheim@chromium.org,tandrii@chromium.org,tansell@chromium.org,v8-reviews@googlegroups.com NOPRESUBMIT=true NOTREECHECKS=true NOTRY=true BUG=chromium:598990 Change-Id: Ie25205f8df69981fde223a232348534487460808 Reviewed-on: https://chromium-review.googlesource.com/444225 Commit-Queue: Michael Achenbach <machenbach@chromium.org> Reviewed-by: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#43271}
-
Michael Achenbach authored
Before this change, presubmit on upload/commit would miss checking status files when e.g. test files were deleted. But the status file check in CI will enforce that all referenced test files exist. NOTRY=true Change-Id: I6069563a0a4e98406977dbce2ae44b299f7cd4b0 Reviewed-on: https://chromium-review.googlesource.com/443467Reviewed-by: Jakob Kummerow <jkummerow@chromium.org> Commit-Queue: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#43270}
-
Toon Verwaest authored
BUG= Change-Id: Ie7a8327fac62a0608b60cf9f0f072f979ac4fd76 Reviewed-on: https://chromium-review.googlesource.com/443528Reviewed-by: Benedikt Meurer <bmeurer@chromium.org> Commit-Queue: Toon Verwaest <verwaest@google.com> Cr-Commit-Position: refs/heads/master@{#43269}
-
dcheng authored
TestWithIsolate and TestWithContext already instantiate the scopers. BUG=none Review-Url: https://codereview.chromium.org/2690303009 Cr-Commit-Position: refs/heads/master@{#43268}
-
bmeurer authored
The CodeStubAssembler::SubString method should not try to be defensive about non-string inputs, but really enforce String receiver. Passing a non-String input will result in crash in the %SubString runtime function anyway. R=jgruber@chromium.org Review-Url: https://codereview.chromium.org/2700103002 Cr-Commit-Position: refs/heads/master@{#43267}
-
Michael Achenbach authored
The new swarming archiving method requires isolate files to be copied to the build dir. BUG=chromium:598990 NOTRY=true TBR=tandrii@chromium.org,tansell@chromium.org,vogelheim@chromium.org Change-Id: I87633725d9ce715684f761ce0c8d2691642c5389 Reviewed-on: https://chromium-review.googlesource.com/444644Reviewed-by: Michael Achenbach <machenbach@chromium.org> Commit-Queue: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#43266}
-
jyan authored
1. use ltr/ltgr when possible 2. combine compares with possible load R=joransiu@ca.ibm.com, bjaideep@ca.ibm.com Review-Url: https://codereview.chromium.org/2696343002 Cr-Commit-Position: refs/heads/master@{#43265}
-
v8-autoroll authored
Rolling v8/build: https://chromium.googlesource.com/chromium/src/build/+log/c8fd116..1ecc6e7 Rolling v8/third_party/catapult: https://chromium.googlesource.com/external/github.com/catapult-project/catapult/+log/574285d..36a5082 Rolling v8/tools/clang: https://chromium.googlesource.com/chromium/src/tools/clang/+log/7764539..799a6a3 TBR=machenbach@chromium.org,vogelheim@chromium.org,hablich@chromium.org Change-Id: I1652a02293550cdface44d4dc5d0fd930b480ebe Reviewed-on: https://chromium-review.googlesource.com/444285Reviewed-by: v8 autoroll <v8-autoroll@chromium.org> Commit-Queue: v8 autoroll <v8-autoroll@chromium.org> Cr-Commit-Position: refs/heads/master@{#43264}
-
- 16 Feb, 2017 21 commits
-
-
dcheng authored
https://crrev.com/2500363002 updated FunctionTemplate::HasInstance to follow the hidden prototype chain of a global proxy to the global object. However, remote contexts don't have a global object to check; instead, teach the instance check knows about the conventions of global proxy setup and have it also check the constructor's prototype. Similarly, also teach Object::FindInstanceInPrototypeChain about the unusual conventions for remote contexts. BUG=527190 Review-Url: https://codereview.chromium.org/2698683003 Cr-Commit-Position: refs/heads/master@{#43263}
-
jwolfe authored
For functions declared in source code, the .toString() representation will be an excerpt of the source code. * For functions declared with the "function" keyword, the excerpt starts at the "function" or "async" keyword and ends at the final "}". The previous behavior would start the excerpt at the "(" of the parameter list, and prepend a canonical `"function " + name` or similar, which would discard comments and formatting surrounding the function's name. Anonymous functions declared as function expressions no longer get the name "anonymous" in their toString representation. * For methods, the excerpt starts at the "get", "set", "*" (for generator methods), or property name, whichever comes first. Previously, the toString representation for methods would use a canonical prefix before the "(" of the parameter list. Note that any "static" keyword is omitted. * For arrow functions and class declarations, the excerpt is unchanged. For functions created with the Function, GeneratorFunction, or AsyncFunction constructors: * The string separating the parameter text and body text is now "\n) {\n", where previously it was "\n/*``*/) {\n" or ") {\n". * At one point, newline normalization was required by the spec here, but that was removed from the spec, and so this CL does not do it. Included in this CL is a fix for CreateDynamicFunction parsing. ')' and '`' characters in the parameter string are no longer disallowed, and Function("a=function(", "}){") is no longer allowed. BUG=v8:4958, v8:4230 Review-Url: https://codereview.chromium.org/2156303002 Cr-Commit-Position: refs/heads/master@{#43262}
-
jkummerow authored
Without relying on the stub cache. Review-Url: https://codereview.chromium.org/2696993002 Cr-Commit-Position: refs/heads/master@{#43261}
-
Daniel Clifford authored
Before this patch, the registers needed for bytecode dispatch in interpreter handlers were inconsistently stored in the interpreter frame and/or kept in values that remained live across calls. After this patch, these registers are explicitly reloaded after calls, making it possible to elide the spills of those registers before the call in many cases. Some highlights from the CL: * Added methods to the CSA and InterpreterAssembler to efficiently store and load Smis values and Smi interpreter registers on x64 without explicit tagging/untagging. * Created Variables for all of the interpreter-internal values that need to be reloaded before bytecode dispatch at the end of an interpreter handler. * The bytecode offset can be written out early in a handler by marking it has having a call along it's critical path. By moving this early in a handler, it becomes possible to use memory operands for pushes used to marshall parameters when making calls. Change-Id: Icf8d7798789f88a4489e06a7092616bbbb881577 Reviewed-on: https://chromium-review.googlesource.com/442566 Commit-Queue: Daniel Clifford <danno@chromium.org> Reviewed-by: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by: Michael Starzinger <mstarzinger@chromium.org> Cr-Commit-Position: refs/heads/master@{#43260}
-
adamk authored
This changes most callers of ParseScopedStatement to call a new, simpler form of ParseStatement, which takes only |labels| and |ok|. This allows us to remove the |legacy| attribute from ParseScopedStatement. The only remaining caller of ParseScopedStatement is ParseIfStatement. This patch is a strict refactoring, and should change no behavior. R=littledan@chromium.org Review-Url: https://codereview.chromium.org/2699793002 Cr-Commit-Position: refs/heads/master@{#43259}
-
vabr authored
ES2017 forbids the sequence of tokens "let [" in in expression statements [1]. This CL makes ParserBase report those instances as SyntaxError. It also adds a customised error message for that, because the standard "Unexpected token" is not applicable: "let" itself is not forbidden in those context, only the sequence of "let [". [1] https://tc39.github.io/ecma262/#sec-expression-statement BUG=v8:5686 Review-Url: https://codereview.chromium.org/2694003002 Cr-Commit-Position: refs/heads/master@{#43258}
-
Michael Achenbach authored
This adds the possibility to specify a python script for post-processing stdout. This also adds some system tests for testing the new feature. NOTRY=true Change-Id: I0383afb3e23513629508feeb639ed2dfce56b54a Reviewed-on: https://chromium-review.googlesource.com/443449Reviewed-by: Clemens Hammacher <clemensh@chromium.org> Commit-Queue: Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#43257}
-
Michael Starzinger authored
This improves the filter deciding whether a checkpoint needs to be created. We now keep track of whether a node having an observable side-effect has been created, allowing to elide checkpoint that are provably effect-dominated by another checkpoint already. By now the initial graphs contain an increasing amount of nodes marked with {Operator::kNoWrite}, making this optimization worthwhile. R=jarin@chromium.org Change-Id: Ie7ffb67e1ab081ef7aa3017675afbe5f9e7601ab Reviewed-on: https://chromium-review.googlesource.com/443466Reviewed-by: Jaroslav Sevcik <jarin@chromium.org> Commit-Queue: Michael Starzinger <mstarzinger@chromium.org> Cr-Commit-Position: refs/heads/master@{#43256}
-
Michael Achenbach authored
This upgrades to a precompiled plugin version including: https://chromium.googlesource.com/v8/v8/+/4b0edcf7 BUG=v8:5970 TBR=clemensh@chromium.org,mstarzinger@chromium.org Change-Id: I28ecdd568e4bc075533b3d14b7946a4a7ce5f9e0 Reviewed-on: https://chromium-review.googlesource.com/443648 Commit-Queue: Michael Achenbach <machenbach@chromium.org> Reviewed-by: Michael Achenbach <machenbach@chromium.org> Reviewed-by: Clemens Hammacher <clemensh@chromium.org> Cr-Commit-Position: refs/heads/master@{#43255}
-
gsathya authored
Errors are swallowed by promises, so just exit with stack trace. Review-Url: https://codereview.chromium.org/2693383004 Cr-Commit-Position: refs/heads/master@{#43254}
-
littledan authored
The default locale can be changed in some environments with environment variables. These environment variables used to allow the system to get into an invalid state, where the default locale was unsupported. This patch detects that case and falls back to "und" as the default locale if there is an Intl service which does not support the locale that ICU reports as the default. It also has a slight cleanup of surrounding code. I haven't gone through the work to set up an automated test, as triggering the case requires setting environment variables, which our tests don't tend to do, but I tested interactively as follows: dehrenberg@dehrenberg:~/v8/v8$ LC_ALL="tlh-FR" rlwrap out/Release/d8 V8 version 5.7.0 (candidate) d8> new Intl.NumberFormat("foo").resolvedOptions().locale "und" d8> new Intl.NumberFormat().resolvedOptions().locale "und" d8> dehrenberg@dehrenberg:~/v8/v8$ LC_ALL="de" rlwrap out/Release/d8 V8 version 5.7.0 (candidate) d8> new Intl.NumberFormat().resolvedOptions().locale "de" d8> new Intl.NumberFormat("foo").resolvedOptions().locale "de" d8> BUG=v8:4216 Review-Url: https://codereview.chromium.org/2646593002 Cr-Commit-Position: refs/heads/master@{#43253}
-
mvstanton authored
This is a workaround for the fact that %SetCode can "lose" the script for a js native. If the js native is re-initialized (for a Realm or something), then the source SharedFunctionInfo won't have a script anymore. Nonetheless, we may want to optimize the function. If we've compiled bytecode, then we can compile optimized code without a script. Here, we carve out a special exception for this case, so that we can turn on the --mark-shared-functions-for-tier-up. BUG=v8:5946 R=leszeks@chromium.org Review-Url: https://codereview.chromium.org/2684033007 Cr-Original-Commit-Position: refs/heads/master@{#43240} Committed: https://chromium.googlesource.com/v8/v8/+/4123a3dd790495c40cf839990318a85c146e057d Review-Url: https://codereview.chromium.org/2684033007 Cr-Commit-Position: refs/heads/master@{#43252}
-
clemensh authored
Enforce the invariant that unreachable implicates an empty live set and fix the implementation of |= and &= operator. This is a fix-up for http://crrev.com/2694103005. R=vegorov@chromium.org CC=mstarzinger@chromium.org, machenbach@chromium.org BUG=v8:5970 Review-Url: https://codereview.chromium.org/2691103008 Cr-Commit-Position: refs/heads/master@{#43251}
-
rossberg authored
We were looking at the unreachable flag or stack_depth of the target frame instead of the current one in a couple of places (most notably BreakTo). This change fixes these bugs and makes us pass the latest spec tests for br_table validation. Also need to ensure that br_table targets have consistent types, which is not implied if the stack is polymorphic. R=titzer@chromium.org BUG= Review-Url: https://codereview.chromium.org/2696813002 Cr-Commit-Position: refs/heads/master@{#43250}
-
jbroman authored
The serializer won't ever write a more complex object. Not validating this allows other things to be used as keys, and converted to string when the property set actually occurs. It turns out this gives an opportunity to trigger OOM by giving an object a key which is a very large sparse array (whose string representation is very large). This case is now rejected by the deserializer. BUG=chromium:686511 Review-Url: https://codereview.chromium.org/2697023002 Cr-Commit-Position: refs/heads/master@{#43249}
-
neis authored
The SpeculativeNumberOp helper lives now in js-type-hint-lowering.cc and is no longer needed in js-typed-lowering.cc. R=mstarzinger@chromium.org BUG= Review-Url: https://codereview.chromium.org/2701643002 Cr-Commit-Position: refs/heads/master@{#43248}
-
Peter Marshall authored
BUG=v8:5974 Change-Id: If79ff5c29bea79ebf8019c4a8e72d2bd7c6b9029 Reviewed-on: https://chromium-review.googlesource.com/443448 Commit-Queue: Benedikt Meurer <bmeurer@chromium.org> Reviewed-by: Benedikt Meurer <bmeurer@chromium.org> Cr-Commit-Position: refs/heads/master@{#43247}
-
littledan authored
BUG=v8:5880 CQ_INCLUDE_TRYBOTS=master.tryserver.v8:v8_linux_noi18n_rel_ng Review-Url: https://codereview.chromium.org/2689283008 Cr-Commit-Position: refs/heads/master@{#43246}
-
Andreas Haas authored
One optimization in the machine-operator-reducer did not consider that that word32 shift left instructions only consider the last 5 bits of the shift input. The issue only occurs for WebAssembly because in JavaScript we always add a "& 0xf" on the shift value to the TurboFan graph. For additional background: The JavaScript and WebAssembly spec both say that only the last 5 bits of the shift value are used in the word32-shift-left operation. This means that an "x << 0x29", in the code is actually executed as "x << 0x09". Therefore the changes in this CL are okay because they mask the last 5 bit of the shift value. BUG=chromium:689450 Change-Id: Id92f298ed6d7f1714b109b3f4fbcecd5ac6d30f7 Reviewed-on: https://chromium-review.googlesource.com/439312Reviewed-by: Ben Titzer <titzer@chromium.org> Commit-Queue: Andreas Haas <ahaas@chromium.org> Cr-Commit-Position: refs/heads/master@{#43245}
-
jgruber authored
Map OOM breaks generated by inspector to DebugEvent.OOM. This avoids generating unintentional DebugEvent.Break events. Also be more future-proof in event categorization. On a related note, this CL also fixes a DCHECK in Runtime::GetFrameDetails. The receiver needs to be grabbed from the inlined frame, not the outer optimized frame. Optimized frames only provide the receiver on a best-effort basis. BUG=v8:5950 Review-Url: https://codereview.chromium.org/2696173002 Cr-Commit-Position: refs/heads/master@{#43244}
-
Andreas Haas authored
R=eholk@chromium.org Change-Id: Ieb88f807275e1cc31cc7715270e316c427b212d4 Reviewed-on: https://chromium-review.googlesource.com/442425 Commit-Queue: Andreas Haas <ahaas@chromium.org> Reviewed-by: Clemens Hammacher <clemensh@chromium.org> Cr-Commit-Position: refs/heads/master@{#43243}
-