- 07 Jan, 2019 38 commits
-
-
Junliang Yan authored
R=joransiu@ca.ibm.com Drive-by: clean up the macro on s390x since it's not used. Change-Id: I317508c1f8a1520ee8873b4323cacd63b8a7cce2 Reviewed-on: https://chromium-review.googlesource.com/c/1398121Reviewed-by:
Joran Siu <joransiu@ca.ibm.com> Commit-Queue: Junliang Yan <jyan@ca.ibm.com> Cr-Commit-Position: refs/heads/master@{#58599}
-
Junliang Yan authored
PPC/S390 has been droped 32/31-bit support. So adjust the padding size for 64-bit only. Change-Id: I3533ef4a90bee0b1e6f49aeb61498ce3054e85e7 Reviewed-on: https://chromium-review.googlesource.com/c/1397866Reviewed-by:
Jakob Gruber <jgruber@chromium.org> Commit-Queue: Junliang Yan <jyan@ca.ibm.com> Cr-Commit-Position: refs/heads/master@{#58598}
-
Jakob Kummerow authored
When the InstructionSelector doesn't have a valid Isolate, it should avoid using it to look up ExternalReferences. Fortunately, this is easy, because the result is only used for a comparison, which in case of invalid Isolate would always fail anyway. Bug: v8:3770 Change-Id: Ie3d65235a22021b05cf0274bf27d91bb7af21023 Reviewed-on: https://chromium-review.googlesource.com/c/1397702 Commit-Queue: Jakob Kummerow <jkummerow@chromium.org> Reviewed-by:
Michael Starzinger <mstarzinger@chromium.org> Cr-Commit-Position: refs/heads/master@{#58597}
-
Junliang Yan authored
Port ccc068d5 Original Commit Message: This CL does two things: 1. It introduces Call/JumpCodeObject as the bottleneck for all calls to non-heap-constant Code objects; and 2. it dispatches directly to the off-heap entry point for all embedded code. Codegen at runtime remains unchanged to preserve the shorter, branch-less calling sequence. R=jgruber@chromium.org, joransiu@ca.ibm.com, michael_dawson@ca.ibm.com BUG= LOG=N Change-Id: I282a5711fdd481a1fde3569e72f0a6141ebcdf2a Reviewed-on: https://chromium-review.googlesource.com/c/1396501 Commit-Queue: Junliang Yan <jyan@ca.ibm.com> Reviewed-by:
Sigurd Schneider <sigurds@chromium.org> Cr-Commit-Position: refs/heads/master@{#58596}
-
Camillo Bruni authored
- Directly use VisitFunctionLiteral where possible - Take shortcut for StringLiterals in BuildLoadPropertyKey Change-Id: Ib5c3de3d2bdd354acbfeb607415854ba90622e89 Reviewed-on: https://chromium-review.googlesource.com/c/1382750Reviewed-by:
Ross McIlroy <rmcilroy@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#58595}
-
Clemens Hammacher authored
Use the new macro to define lazily initialized leaky singletons. Avoid the clumsy LazyInstance, which we can hopefully remove soon. R=mlippautz@chromium.org Bug: v8:8600 Change-Id: Ib4d23f275c7ff5ca71fa9b47345284935330ead7 Reviewed-on: https://chromium-review.googlesource.com/c/1397711Reviewed-by:
Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Clemens Hammacher <clemensh@chromium.org> Cr-Commit-Position: refs/heads/master@{#58594}
-
Tamer Tas authored
- Removed the old test. - Created a fake test suite and added a test for loading it with a TestConfig R=machenbach@chromium.org CC=yangguo@chromium.org,sergiyb@chromium.org Bug: v8:8174 Change-Id: Ib7587ceec9e31ecd4cb8f45c3158e73c79a9bc5b Reviewed-on: https://chromium-review.googlesource.com/c/1396082Reviewed-by:
Michael Achenbach <machenbach@chromium.org> Commit-Queue: Tamer Tas <tmrts@chromium.org> Cr-Commit-Position: refs/heads/master@{#58593}
-
Sigurd Schneider authored
This CL removes the graph between phase changes. This prevents incorrect path layouting after changing from a phase where a path is displayed that is not a correct path in the phase we change to. Change-Id: Iad80f49efc8d8c71600ad51432981c3a206ef9cb Bug: v8:7327 Reviewed-on: https://chromium-review.googlesource.com/c/1397710Reviewed-by:
Georg Neis <neis@chromium.org> Commit-Queue: Sigurd Schneider <sigurds@chromium.org> Cr-Commit-Position: refs/heads/master@{#58592}
-
Camillo Bruni authored
We plan to store additional information that is not related to scopes. The new name will reflect this fact better. Change-Id: I4ddb1017bc255e6ad271e4448848ed630f367d5b Reviewed-on: https://chromium-review.googlesource.com/c/1388538 Commit-Queue: Camillo Bruni <cbruni@chromium.org> Reviewed-by:
Toon Verwaest <verwaest@chromium.org> Reviewed-by:
Ulan Degenbaev <ulan@chromium.org> Cr-Commit-Position: refs/heads/master@{#58591}
-
peterwmwong authored
- Removes build step and bootstrapping associated with building/loading `EXPERIMENTAL_EXTRAS` JS natives. - Removes `--experimental-extras` flag Bug: v8:7624 Change-Id: I4c45fe70da42847545037d63e9f1da77f5957f8b Reviewed-on: https://chromium-review.googlesource.com/c/1397906Reviewed-by:
Yang Guo <yangguo@chromium.org> Commit-Queue: Peter Wong <peter.wm.wong@gmail.com> Cr-Commit-Position: refs/heads/master@{#58590}
-
Jaroslav Sevcik authored
Bug: chromium:918763 Change-Id: Ic9faaed6b3194269748ba35740cda0dc8dde3241 Reviewed-on: https://chromium-review.googlesource.com/c/1397707Reviewed-by:
-
Clemens Hammacher authored
We currently iterate the list of unexecuted register moves repeatedly, always executing the moves whose destination register is not being used as source register any more. This can lead to quadratic execution times if only a small number of moves is processed in every iteration. This CL refactors this such that we iterate the moves at most three times: Once for executing moves which can be executed right away (fast path) and for computing the source register use counts. A second time to execute all remaining non-cyclic moves, and a third time to execute cyclic moves. During the second and third iteration, whenever we decrement the source register use count, we check whether it drops to zero and execute the respective move right away. R=ahaas@chromium.org Bug: v8:6600, v8:8423 Change-Id: I503328f5ae5f0208e35d53c71b4c289d75799892 Reviewed-on: https://chromium-review.googlesource.com/c/1397703 Commit-Queue: Clemens Hammacher <clemensh@chromium.org> Reviewed-by:
Andreas Haas <ahaas@chromium.org> Cr-Commit-Position: refs/heads/master@{#58588}
-
Clemens Hammacher authored
Stack transfers consist of a number of register moves plus a number of register loads. We currently store both in separate vectors. This CL changes that to be stored in arrays indexed by the destination register (such that it behaves like a map). This avoids any dynamically growing structures. Measured locally, this speeds up stack transfer processing by ~10%, which translates to ~0.5% of overall Liftoff compilation time. R=ahaas@chromium.org Bug: v8:6600, v8:8423 Change-Id: Id532960dcc12f228507ed75e392ad4c57710593f Reviewed-on: https://chromium-review.googlesource.com/c/1396278 Commit-Queue: Clemens Hammacher <clemensh@chromium.org> Reviewed-by:
-
Jakob Kummerow authored
The AsmType class uses a design similar to the old Object* model, where arbitrary values (including 0) are reinterpret_cast to pointers. This yields the following UBSan error, among others: src/asmjs/asm-parser.cc:2000:51: runtime error: member call on null pointer of type 'v8::internal::wasm::AsmType' This patch does the smallest possible fix by turning the affected methods into static functions. Longer-term, we should consider switching the overall class design to a "struct wrapping an Address" model like the new Object definition, which is a bit non-trivial because some AsmType types are ZoneObject subclasses. Bug: v8:3770 Change-Id: Ie2a7cdc9eab32c4c469d699212c84b0419480b4f Reviewed-on: https://chromium-review.googlesource.com/c/1397663Reviewed-by: -
Clemens Hammacher authored
We were missing a few declarations, especially fully-qualified declarations in cc files like here: https://crrev.com/c/1396459 R=mlippautz@chromium.org Bug: v8:8616 No-Try: true Change-Id: Iff93f1cd4fde18ac7fc8391e459e6bdcb4eb8f9b Reviewed-on: https://chromium-review.googlesource.com/c/1397706Reviewed-by:
-
Georg Neis authored
Bug: v8:7790 Change-Id: Ie8825227048a00892117e98cd4e591b3e5e06930 Reviewed-on: https://chromium-review.googlesource.com/c/1396090Reviewed-by:
Maya Lekova <mslekova@chromium.org> Reviewed-by:
Jaroslav Sevcik <jarin@chromium.org> Commit-Queue: Georg Neis <neis@chromium.org> Cr-Commit-Position: refs/heads/master@{#58584}
-
Georg Neis authored
Bug: v8:7790 Change-Id: I8cc88aadaaacca4cc6b87a6f5bead9129b8dfa14 Reviewed-on: https://chromium-review.googlesource.com/c/1394550Reviewed-by:
-
Maya Lekova authored
Design doc: https://docs.google.com/document/d/1vCQYhtFPqXafSMweSnGD8l0TKEIB6cPV5UGMHJtpy8k/edit?ts=5bf7d341 This CL only introduces a skeleton of the new phase that implements a bytecode walker. The SUPPORTED_BYTECODE_LIST is supposed to be filled in gradually. Bug:v8:7790 R=jarin@chromium.org, neis@chromium.org Change-Id: I57fea91c55dca888581f2490bdf7b831fc61eda4 Reviewed-on: https://chromium-review.googlesource.com/c/1386872Reviewed-by:
-
Georg Neis authored
R=jarin@chromium.org Change-Id: I08ecc5c86aa098a566c57e8ac8575504c9c36361 Reviewed-on: https://chromium-review.googlesource.com/c/1397667Reviewed-by:
-
Jakob Kummerow authored
The C++ spec does not guarantee IEEE-754 behavior for doubles, in particular it says that dividing by zero is undefined behavior, and UBSan complains about it when it happens. Bug: v8:3770 Change-Id: I79e52c0e11ebfb581191f6f1c3ff95eb747dd97f Reviewed-on: https://chromium-review.googlesource.com/c/1391751 Commit-Queue: Jakob Kummerow <jkummerow@chromium.org> Reviewed-by:
-
Sigurd Schneider authored
This CL simplifies the keydown handling code and fixes several issues: - Input to the search box was not reliably working, because the SVG keydown handler was attached to the window and its repeat-key detection was supressing key events. - Selecting the input of a node via keys 1-9 did not select the input, but always enabled the corresponding input node. 1-9 now select the input node, and CTRL+1 through CTRL+9 can be used to toggle the input edge. Bug: v8:7327 Notry: true Change-Id: Ifedc8b703f6552e101ad00fee2f3c50f29b325b5 Reviewed-on: https://chromium-review.googlesource.com/c/1397666 Commit-Queue: Sigurd Schneider <sigurds@chromium.org> Reviewed-by:
-
Sigurd Schneider authored
This improves readability and encapsulation of the code. Change-Id: Ifbca8441941a1776797937c973a064153818c859 Notry: true Bug: v8:7327 Reviewed-on: https://chromium-review.googlesource.com/c/1396423 Commit-Queue: Sigurd Schneider <sigurds@chromium.org> Reviewed-by:
-
peterwmwong authored
- Removes the last `CORE` JS native script: `prologue.js`. - Removes build step and bootstrapping associated with building/loading `CORE` JS natives. - Removes `natives_utils_object` from context. - Deprecates `--expose-natives-as` flag. - Ports extra utils functions to C++ (`uncurryThis`) or Torque (`createPrivateSymbol`, `markPromiseAsHandled`, and `promiseState`). - Move extra utils constants initialization into bootstrapper (`kPROMISE_PENDING`, `kPROMISE_FULFILLED`, `kPROMISE_REJECTED`). - Removes unused extra utils functions `log` and `logStackTrace`. Drive-by: Added test coverage for Array#includes being an unscopeable. Bug: v8:7624 Change-Id: I5d983f8d11b76cb4dd3c2c67592ce1dc88364cd9 Reviewed-on: https://chromium-review.googlesource.com/c/1381672Reviewed-by:
-
Sigurd Schneider authored
This CL enables noImplicitReturns and noImplicitThis warnings in TypeScript, another step on the road to stricter types. Drive-by: Fix bug in search function. Change-Id: Iafb528b5f0e7ccc8774bc218fd0dcdb206a0de31 Notry: true Bug: v8:7327 Reviewed-on: https://chromium-review.googlesource.com/c/1396422 Commit-Queue: Sigurd Schneider <sigurds@chromium.org> Reviewed-by:
-
Sigurd Schneider authored
Turbolizer only remembered the expansion state of the panes, but not their widths. This CL remembers the relative widths, and restores them upon reload. This is also useful when the size of the Turbolizer window changes. Change-Id: I0fd81c1266bfbddded86da16e2241420cdf73f4e Notry: true Bug: v8:7327 Reviewed-on: https://chromium-review.googlesource.com/c/1396421 Commit-Queue: Sigurd Schneider <sigurds@chromium.org> Reviewed-by:
-
Hajime Hoshi authored
This CL fixes the problem that the platform was set at MockPlatform's constructor but was not reset. Bug: v8:8527 Change-Id: I21c3b19320885b1b38999161db7cc1b8f15d798e Reviewed-on: https://chromium-review.googlesource.com/c/1397821Reviewed-by:
-
Sigurd Schneider authored
Drive-by: Fix z-index (stacking) issue with hovers and pane resize bars. Change-Id: Ibd028a666681face1ccd6ec36b388f1ef25b1393 Notry: true Bug: v8:7327 Reviewed-on: https://chromium-review.googlesource.com/c/1396420 Commit-Queue: Sigurd Schneider <sigurds@chromium.org> Reviewed-by:
-
Sigurd Schneider authored
This is a step towards removing all instances of implicit any types from turbolizer. This CL also replaces var with const/let. This improves readability and warnings. Change-Id: I67c2974df209f857e67dfdbb743ce695ce861982 Notry: true Bug: v8:7327 Reviewed-on: https://chromium-review.googlesource.com/c/1396419 Commit-Queue: Sigurd Schneider <sigurds@chromium.org> Reviewed-by:
-
Sigurd Schneider authored
Refactor NodeLabel from GNode, which saves memory and is a step towards decoupling the node layout from the graph structure. Change-Id: I095a2f7a7ab28067161deffbc37952ae15410e0a Notry: true Bug: v8:7327 Reviewed-on: https://chromium-review.googlesource.com/c/1396418 Commit-Queue: Sigurd Schneider <sigurds@chromium.org> Reviewed-by:
-
Toon Verwaest authored
This simplifies NextArrowFunctionInfo, allows us to Scope::Snapshot::Reparent directly rather than moving it, and allows us to skip reparenting in the simple parameter arrow function cases. This is a reland of https://chromium-review.googlesource.com/c/v8/v8/+/1397664, simply splitting out the arrow-function-name-inferring part. Change-Id: I640d911a9607edc3bbb0e5ff3bf992094e4159e4 Reviewed-on: https://chromium-review.googlesource.com/c/1397701Reviewed-by:
-
Clemens Hammacher authored
This is a fully semantically preserving CL, it just moves code around a bit to make the follow-up CL easier to review. R=ahaas@chromium.org Bug: v8:6600, v8:8423 Change-Id: I0de80b18faa3ae570894cbd8073c495a5731d255 Reviewed-on: https://chromium-review.googlesource.com/c/1396096Reviewed-by:
-
Ross McIlroy authored
Real world websites don't benifit from aborting preparsing to eagerly compile long trivial functions, and it adds unecessary complexity to the parser and doesn't work well with bytecode flushing, so we remove it. Perf Sheriffs: this is expected to regress the MandreelLatency benchmark on Octane. BUG=v8:8395 Change-Id: Ia60cd67d4dd100376d2a366939a1d2a97cbc2b0d Reviewed-on: https://chromium-review.googlesource.com/c/1394297 Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> Reviewed-by:
Marja Hölttä <marja@chromium.org> Reviewed-by:
-
Toon Verwaest authored
This is a reland of 2963f1b2 Original change's description: > [parser] Simplify ParseFunctionBody > > - Merge is_simple branches at the top > - Remove block around inner_body parsing. Always merge fully at the end. > - Remove conditional inner block adding to outer body. Simply add it to the > inner body making merge push it to the parent. > > Change-Id: I1f062918a7abac354b949136463517bd0440984f > Reviewed-on: https://chromium-review.googlesource.com/c/1386111 > Commit-Queue: Toon Verwaest <verwaest@chromium.org> > Reviewed-by: Igor Sheludko <ishell@chromium.org> > Cr-Commit-Position: refs/heads/master@{#58390} Change-Id: I145f0cb1eda1dca4dd047b55e54b2b1bb704ecf8 Reviewed-on: https://chromium-review.googlesource.com/c/1397662Reviewed-by:
-
cjihrig authored
This commit updates gen-postmortem-metadata.py to handle the new ACCESSORS2 macro. Once that migration is complete, ACCESSORS2 can be dropped from this script. A constant is also added for SharedFunctionInfo's kFunctionDataOffset, which was broken in a55803a1. See: https://github.com/nodejs/node-v8/issues/95 Change-Id: I5c3f960b4fd739a76f96d0ece9543574ff96be0f Reviewed-on: https://chromium-review.googlesource.com/c/1392449Reviewed-by:
-
Georg Neis authored
Change-Id: If8763d2a46a13e75722f2f5314ba50dc9df281e0 Reviewed-on: https://chromium-review.googlesource.com/c/1396089Reviewed-by:
Benedikt Meurer <bmeurer@chromium.org> Commit-Queue: Georg Neis <neis@chromium.org> Cr-Commit-Position: refs/heads/master@{#58565}
-
Clemens Hammacher authored
This reverts commit dcd75706. Reason for revert: Breaks layout tests, blocks roll, see https://crrev.com/c/1396602; there are wasm CLs in this range too, but this CL looks like the most likely culprit. Original change's description: > Reland: [Compiler] Ensure unoptimized code generation is context independent. > > Now that Asm.js code is also context independent, move code to ensure context independence > from BytecodeGenerator to FinalizeUnoptimizedCode. > > Reland of CL: https://chromium-review.googlesource.com/c/v8/v8/+/1349236 > > Change-Id: I718090850870c61733e0719d4091ec60bc080ebb > Reviewed-on: https://chromium-review.googlesource.com/c/1396201 > Commit-Queue: Ross McIlroy <rmcilroy@chromium.org> > Reviewed-by: Dan Elphick <delphick@chromium.org> > Cr-Commit-Position: refs/heads/master@{#58558} TBR=rmcilroy@chromium.org,delphick@chromium.org # Not skipping CQ checks because original CL landed > 1 day ago. Change-Id: I5f547319f31f87777165361747dd42d223fc0b0e Reviewed-on: https://chromium-review.googlesource.com/c/1396427Reviewed-by:
Clemens Hammacher <clemensh@chromium.org> Commit-Queue: Clemens Hammacher <clemensh@chromium.org> Cr-Commit-Position: refs/heads/master@{#58564}
-
Benedikt Meurer authored
We cannot take the fast-path if the user messed with the Symbol.species property on the Promise.prototype, as that makes the internal promises observable. Bug: chromium:917076 Change-Id: I928e0bd17836ca78cf88591610526aa7bc1d293c Reviewed-on: https://chromium-review.googlesource.com/c/1396426 Commit-Queue: Benedikt Meurer <bmeurer@chromium.org> Reviewed-by:
-
Yang Guo authored
R=jkummerow@chromium.org Change-Id: I54c6137a3c6e14d4102188f154aa7216e7414dbc Reviewed-on: https://chromium-review.googlesource.com/c/1388533Reviewed-by:
Jakob Kummerow <jkummerow@chromium.org> Commit-Queue: Yang Guo <yangguo@chromium.org> Cr-Commit-Position: refs/heads/master@{#58562}
-
- 06 Jan, 2019 2 commits
-
-
Mathias Bynens authored
This patch ensures each HTML page has a DOCTYPE (to trigger standards mode as opposed to quirks mode), a <meta charset="utf-8">, and a <title>. Additionally, it removes redundant attribute/value pairs such as `type="text/javascript"` on <script> elements or `type="text/css"` on <style> or <link rel="stylesheet"> elements. [1] Finally, it removes the optional solidus for self-closing HTML elements. [2] [1] https://mathiasbynens.be/notes/html5-levels#type-attributes [2] https://mathiasbynens.be/notes/html5-levels#solidus Change-Id: I66d2700be120dc8fd52bdf38f9d34749f55e1e7f Reviewed-on: https://chromium-review.googlesource.com/c/1396084Reviewed-by:
-
v8-ci-autoroll-builder authored
Rolling v8/build: https://chromium.googlesource.com/chromium/src/build/+log/de2f586..fd2243f Rolling v8/buildtools: https://chromium.googlesource.com/chromium/buildtools/+log/0e1cbc4..7d88270 TBR=machenbach@chromium.org,hablich@chromium.org,sergiyb@chromium.org Change-Id: Iad30ca6331cf69d1fbc33399f4c3ebc929fd8970 Reviewed-on: https://chromium-review.googlesource.com/c/1396742Reviewed-by:
v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Commit-Queue: v8-ci-autoroll-builder <v8-ci-autoroll-builder@chops-service-accounts.iam.gserviceaccount.com> Cr-Commit-Position: refs/heads/master@{#58560}
-
