- 21 Nov, 2017 39 commits
-
-
Bill Budge authored
- Eliminates CommitRegion and UncommitRegion methods, replacing them with calls to SetPermissions. - Makes a similar change to the API of VirtualMemory. - This changes system calls from mmap to mprotect on most POSIX platforms. Bug: chromium:756050 Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng Change-Id: Ib10f8293c9398c6c1e729cd7d686b7c97e6a5d75 Reviewed-on: https://chromium-review.googlesource.com/769679Reviewed-by:
Hannes Payer <hpayer@chromium.org> Reviewed-by:
Michael Lippautz <mlippautz@chromium.org> Commit-Queue: Bill Budge <bbudge@chromium.org> Cr-Commit-Position: refs/heads/master@{#49552}
-
Adam Klein authored
These functions should only be called in case of a parse error, so speed of calling them should not be a concern. In local testing, this saves ~16k of binary size on a release mode build. Bug: v8:7090 Change-Id: I433df81c2a5811ed922885dbab3ce003427f3d1c Reviewed-on: https://chromium-review.googlesource.com/780693Reviewed-by:
Marja Hölttä <marja@chromium.org> Commit-Queue: Adam Klein <adamk@chromium.org> Cr-Commit-Position: refs/heads/master@{#49551}
-
Michael Lippautz authored
Bug: Change-Id: Idc0bbc55713a6e6dcc88db77477bfddf3bd80a02 Reviewed-on: https://chromium-review.googlesource.com/781769Reviewed-by:
Ulan Degenbaev <ulan@chromium.org> Commit-Queue: Michael Lippautz <mlippautz@chromium.org> Cr-Commit-Position: refs/heads/master@{#49550}
-
Martyn Capewell authored
Reduce the deopt table size by reusing the entry address available in a temp register to compute the entry id. Saves ~200kB. Bug: Change-Id: I3a1baf0e4c8cf19a0aa149da2bea623c1349a9ca Reviewed-on: https://chromium-review.googlesource.com/774890Reviewed-by:
Jaroslav Sevcik <jarin@chromium.org> Cr-Commit-Position: refs/heads/master@{#49549}
-
Clemens Hammacher authored
Both can be used to optionally initialize an object, but with base::Optional it will be stack-allocated. R=ahaas@chromium.org Change-Id: I9977e1b2e0532505f8582cc68e27687aaeebd33d Reviewed-on: https://chromium-review.googlesource.com/781920Reviewed-by:
Andreas Haas <ahaas@chromium.org> Commit-Queue: Clemens Hammacher <clemensh@chromium.org> Cr-Commit-Position: refs/heads/master@{#49548}
-
Michal Majewski authored
Filtering by status file split to four parts: 1. Getting outcomes - reads both variant dependent and independent outcomes, no more need to do it twice. 2. Checking unused rules - has a switch to check only variant dependent/independent rules. 3. Reading flags - if outcome starts with '--' it is treated as a flag. 4. Actual filtering. Outcomes removed from the testcase object, can be accessed by call to its testsuite. Bug: v8:6917 Cq-Include-Trybots: master.tryserver.v8:v8_linux_noi18n_rel_ng Change-Id: I35762f891010ddda926250452b88656047433daa Reviewed-on: https://chromium-review.googlesource.com/775160 Commit-Queue: Michał Majewski <majeski@google.com> Reviewed-by:
Sergiy Byelozyorov <sergiyb@chromium.org> Reviewed-by:
Michael Achenbach <machenbach@chromium.org> Cr-Commit-Position: refs/heads/master@{#49547}
-
Mircea Trofin authored
On x64, we optimize out EmbeddedReferences, unless we explicitly indicate serialization is enabled. We serialize js-to-wasm wrappers, which include such references. Bug: v8:7083 Change-Id: I976da4af74bf7ee3245e1465b8e47f2c042ec3b4 Reviewed-on: https://chromium-review.googlesource.com/780207Reviewed-by:
Yang Guo <yangguo@chromium.org> Reviewed-by:
Ben Titzer <titzer@chromium.org> Reviewed-by:
Eric Holk <eholk@chromium.org> Commit-Queue: Mircea Trofin <mtrofin@chromium.org> Cr-Commit-Position: refs/heads/master@{#49546}
-
Andreas Haas authored
Typically the interpreter returns 0xdeadbeef to indicate an exception. However, for stack overflows a normal exception is used. The interpreter requires an activation, however, to deal with normal exceptions. With this CL we start an activation before we execute the fuzzer input in the interpreter. R=clemensh@chromium.org Bug: chromium:781103 Change-Id: I4fc3a18bfc2076aab9ff7d2324a3311fe222954a Reviewed-on: https://chromium-review.googlesource.com/776835 Commit-Queue: Andreas Haas <ahaas@chromium.org> Reviewed-by:
Clemens Hammacher <clemensh@chromium.org> Cr-Commit-Position: refs/heads/master@{#49545}
-
jgruber authored
A few coverage tests depend on a GC run that deterministically frees all dead objects, but Runtime::kCollectGarbage did not explicitly disable incremental marking. Incremental marking makes liveness detection timing-dependent and thus should be disabled here. NOTRY=true NOPRESUBMIT=true TBR=mlippautz@chromium.org Bug: v8:7108 Change-Id: I8bebe612bbc2126b8ad778af15f08442ccc91a35 Reviewed-on: https://chromium-review.googlesource.com/781865 Commit-Queue: Jakob Gruber <jgruber@chromium.org> Reviewed-by:
Jakob Gruber <jgruber@chromium.org> Cr-Commit-Position: refs/heads/master@{#49544}
-
Michael Starzinger authored
R=machenbach@chromium.org BUG=v8:7106 Change-Id: I89ff17393848c88ffd885af3fec8b6703c75d073 Reviewed-on: https://chromium-review.googlesource.com/781942Reviewed-by:
-
Igor Sheludko authored
The unused properties fields number is calculatable via used in-object properties count and we can drop it now. Bug: chromium:774644 Change-Id: I7388af7772a8e793593fabc46527886cf2e36095 Reviewed-on: https://chromium-review.googlesource.com/781465Reviewed-by:
-
Toon Verwaest authored
Bug: Change-Id: Iccc0241859970ad4e04b9a6e5f206491aa7c442e Reviewed-on: https://chromium-review.googlesource.com/781939Reviewed-by:
Camillo Bruni <cbruni@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#49541}
-
Clemens Hammacher authored
The wasm context will always be stored on the stack, and will be loaded from there whenever needed (for accessing globals or the memory). We can still improve this later by caching the context address or specific information loaded from it. R=titzer@chromium.org Bug: v8:6600 Change-Id: Idd7fb1ccff28a73beaf545997e3dfdb74757b686 Reviewed-on: https://chromium-review.googlesource.com/779145 Commit-Queue: Clemens Hammacher <clemensh@chromium.org> Reviewed-by:
-
Georg Neis authored
This introduces a JSBitwiseNot operator and lowers it either to a speculative xor with -1 (when we have Number feedback) or to a stub call. The stub is also new. Bug: v8:6791 Change-Id: I362e52de8a741dc5db044c406543878e407eb2ed Reviewed-on: https://chromium-review.googlesource.com/778839 Commit-Queue: Georg Neis <neis@chromium.org> Reviewed-by:
-
Georg Neis authored
This introduces a JSNegate operator and lowers it either to a speculative multiplication with -1 (when we have Number feedback) or to a stub call. The stub is also new. R=jarin@chromium.org Bug: v8:6791 Change-Id: I8e20333fe49cc6088d2d10777be982e42eed2412 Reviewed-on: https://chromium-review.googlesource.com/774718 Commit-Queue: Georg Neis <neis@chromium.org> Reviewed-by:
-
Toon Verwaest authored
Bug: Change-Id: Ic72ebff8c106ef200875de352972c5b324dcb5a0 Reviewed-on: https://chromium-review.googlesource.com/781466 Commit-Queue: Toon Verwaest <verwaest@chromium.org> Reviewed-by:
Ross McIlroy <rmcilroy@chromium.org> Cr-Commit-Position: refs/heads/master@{#49537}
-
Hannes Payer authored
This works because we pool regular non-executable pages on a lower level. Executable pages are currently not supported by the pooling mechanism. If this regresses we should fix it. Change-Id: Ief3484d59f1f1f4bc63f8e718482e4174bedc012 Reviewed-on: https://chromium-review.googlesource.com/778939Reviewed-by:
-
Michael Starzinger authored
This introduces {Heap::write_protect_code_memory} as a race-free copy of the underlying {FLAG_write_protect_code_memory} flag. Since this flag is checked from the parallel sweeper, subsequent flag implications might be racing against the read. This ensures race-free reads. R=hpayer@chromium.org BUG=v8:6792,chromium:774108,v8:7106 Change-Id: I1a1073f11e91bebd60f8d5da440845452ec67c50 Reviewed-on: https://chromium-review.googlesource.com/781662Reviewed-by: -
Clemens Hammacher authored
Currently the SourcePositionTableBuilder requires a Zone because it holds a ZoneVector<byte> of the encoded entries. Since ZoneVector is a suboptimal data structure anyway, and for Liftoff we don't even have a Zone allocated currently, this CL replaces the ZoneVector by std::vector. R=mstarzinger@chromium.org Bug: v8:6600 Change-Id: I8010143e917e2351664e2b53746753b597f4407a Reviewed-on: https://chromium-review.googlesource.com/779181Reviewed-by:
Michael Starzinger <mstarzinger@chromium.org> Reviewed-by:
-
Camillo Bruni authored
In the case of a corrupted snapshot we fall back to initializing the isolate from scratch. Howver, we don't ship the full SetupIsolateDelegate. This causes spurious failures during later initialization. This CL mostly turns the DCHECKs in SetupIsolateDelegate into hard CHECKs making it easier to spot these kind of failures. Bug: chromium:767846 Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng Change-Id: Ibe8a5beece27433439b1b09412f6110be703ff86 Reviewed-on: https://chromium-review.googlesource.com/779189Reviewed-by:
-
Michael Starzinger authored
This makes sure the {CodeSpaceMemoryModificationScope} for the common allocation path is inside the {Heap} component. This will in turn enable finer-grained control of the modification scope in the future. R=hpayer@chromium.org BUG=v8:6792 Change-Id: I6c3bc457bac641e79b2786cf78557b26aa2027e7 Reviewed-on: https://chromium-review.googlesource.com/779399 Commit-Queue: Michael Starzinger <mstarzinger@chromium.org> Reviewed-by: -
Clemens Hammacher authored
This is a reland of 712fa675. Original change's description: > [test] Add Liftoff variant > > Add a variant for testing the current state of the Liftoff > implementation. > This variant will only run on a subset of the bots, just like the > --future variant. > > R=machenbach@chromium.org, hablich@chromium.org > > Bug: v8:7088, v8:6600 > Change-Id: If49fad3a8ed579356504b821a787326754f24e78 > Reviewed-on: https://chromium-review.googlesource.com/779420 > Commit-Queue: Clemens Hammacher <clemensh@chromium.org> > Reviewed-by: Michael Achenbach <machenbach@chromium.org> > Cr-Commit-Position: refs/heads/master@{#49504} TBR=machenbach@chromium.org CC=hablich@chromium.org Bug: v8:7088, v8:6600 Change-Id: Ieb20020f07c70acaa64bb421763a41aa163a261b Reviewed-on: https://chromium-review.googlesource.com/781499 Commit-Queue: Clemens Hammacher <clemensh@chromium.org> Reviewed-by:
-
jgruber authored
The integer value denoting the number of captures (and thus the size of the list of captures created in @@replace [0]) can be controlled by the user. This CL ensures we don't overflow and respect Code::kMaxArguments, but note that it is still possible to trigger OOMs through large lists. Bug: chromium:786573 Change-Id: I19c88908c594487818d083b2ba423764ef91eae0 Reviewed-on: https://chromium-review.googlesource.com/779001Reviewed-by:
-
Sergiy Byelozyorov authored
TBR=machenbach@chromium.org Bug: chromium:748002, chromium:773606, chromium:748003, chromium:748035 Change-Id: Ibe24a76598daa8af1fe5530859a8a657c54c3b9b Reviewed-on: https://chromium-review.googlesource.com/779434 Commit-Queue: Sergiy Byelozyorov <sergiyb@chromium.org> Reviewed-by:
-
Georg Neis authored
TBR: rmcilroy@chromium.org Bug: v8:6791 Change-Id: I4ac2bdce353d987a2fe45149d8556b6591569a01 Reviewed-on: https://chromium-review.googlesource.com/771191 Commit-Queue: Georg Neis <neis@chromium.org> Reviewed-by:
-
Mike Stanton authored
Bug: Change-Id: Ib9ca5458554ed23e94e85e2a9cb42d816649c881 Reviewed-on: https://chromium-review.googlesource.com/781661Reviewed-by:
-
Jakob Gruber authored
This enables lazy bytecode handler deserialization by default. Design doc: http://goo.gl/QxZBL2 TBR=hablich@chromium.org Bug: v8:6624 Change-Id: I89a17e16a4ee18ee1bfca75a1d61f7a11cdc211f Reviewed-on: https://chromium-review.googlesource.com/776836Reviewed-by:
Michael Hablich <hablich@chromium.org> Commit-Queue: Jakob Gruber <jgruber@chromium.org> Cr-Commit-Position: refs/heads/master@{#49526}
-
Andreas Haas authored
R=clemensh@chromium.org Change-Id: Ifb5c02698b5ad9189283e227e89fa020f92186a8 Reviewed-on: https://chromium-review.googlesource.com/781720Reviewed-by:
-
Michael Achenbach authored
Change-Id: I3f54db032526e4253adacbff8c50393392aad852 Reviewed-on: https://chromium-review.googlesource.com/781629 Commit-Queue: Michael Achenbach <machenbach@chromium.org> Reviewed-by:
-
Andreas Haas authored
The index of a function in the WasmModule data structure is offset by the number of imported functions in the module. The {DecodeFunctionBody} function of the module decoder, however, requires the function index without this offset. The streaming processor mixed up these two ranges of function indices. This is fixed in this CL. R=clemensh@chromium.org Bug: chromium:781507 Change-Id: Ie3e0c4703b06ecb923c98ffb961844915323197c Reviewed-on: https://chromium-review.googlesource.com/776680 Commit-Queue: Andreas Haas <ahaas@chromium.org> Reviewed-by: -
Sergiy Byelozyorov authored
R=machenbach@chromium.org Bug: chromium:748002, chromium:773606, chromium:748003, chromium:748035 No-Try: true Change-Id: I80ec2a4bf8cb1193c4f31ae1dc84e7f88a2011b4 Reviewed-on: https://chromium-review.googlesource.com/779431 Commit-Queue: Sergiy Byelozyorov <sergiyb@chromium.org> Reviewed-by:
-
Hannes Payer authored
[heap] Initialize the write_unprotect_counter_ of new code pages with the number of open CodeSpaceMemoryModificationScopes. Bug: chromium:774108,v8:6792 Change-Id: Ib5306075aeff61160762c685c343970ceb32b66a Reviewed-on: https://chromium-review.googlesource.com/779201Reviewed-by:
-
Michael Achenbach authored
V8 passes the command explicitly to each swarming task, hence it's not necessary to store the command in the isolate. This drops the Chromium-specific code in MB that creates the swarming command based on Chromium test features. This also makes the swarming targets option a no-op to allow activating it on the infra side without disruption. Bug: chromium:669910 Change-Id: I6cb03f05d034092a25d879d52b4d64952493f55b Reviewed-on: https://chromium-review.googlesource.com/779148Reviewed-by:
-
Ulan Degenbaev authored
Currently the size of compaction spaces is not taken into account in the Heap::CanExpandOldGeneration predicate. This can push the heap size over the hard limit in some cases. This patch makes Heap::CanExpandOldGeneration stricter and also fixes the SelectGarbageCollector to prefer Mark-Compact near the hard limit. Bug: chromium:784077 Change-Id: I00c7295eba8794a342dd6277a45f995529054b64 Reviewed-on: https://chromium-review.googlesource.com/779265Reviewed-by:
-
Marja Hölttä authored
This code was confusing, since "target" declared in one of the subscopes shadowed a parameter with the same name. Change-Id: Ibf694c94f0a26ca65609cb80d22c40a8fa98f4f3 Reviewed-on: https://chromium-review.googlesource.com/779261Reviewed-by:
Benedikt Meurer <bmeurer@chromium.org> Reviewed-by:
-
Michael Starzinger authored
R=jarin@chromium.org BUG=v8:6792 Change-Id: Id97c7a9911eb2c0606f8ea25d0a2c8ebcb4c8ccd Reviewed-on: https://chromium-review.googlesource.com/753729Reviewed-by:
-
Camillo Bruni authored
Not resetting the ticks regresses optimization time without substantial performance benenfits on twitter, facebook, youtube, linkedin and wikipedia. There was no net positive effect visible otherwise. Bug: chromium:786908 Change-Id: I98237dee170e7a387f09ccfbad178793361d4a67 Reviewed-on: https://chromium-review.googlesource.com/779435Reviewed-by:
Jakob Kummerow <jkummerow@chromium.org> Commit-Queue: Camillo Bruni <cbruni@chromium.org> Cr-Commit-Position: refs/heads/master@{#49516}
-
v8-autoroll authored
Rolling v8/build: https://chromium.googlesource.com/chromium/src/build/+log/5718716..6167734 Rolling v8/tools/clang: https://chromium.googlesource.com/chromium/src/tools/clang/+log/509676b..bd3b137 TBR=machenbach@chromium.org,hablich@chromium.org,sergiyb@chromium.org Change-Id: I70c12249d7946044d7049f5cf03357c5a2d1fbaf Reviewed-on: https://chromium-review.googlesource.com/780865Reviewed-by:
v8 autoroll <v8-autoroll@chromium.org> Commit-Queue: v8 autoroll <v8-autoroll@chromium.org> Cr-Commit-Position: refs/heads/master@{#49515}
-
Alexei Filippov authored
BUG=v8:7070 Cq-Include-Trybots: master.tryserver.chromium.linux:linux_chromium_rel_ng Change-Id: I92d7c61afa88e0a52c8eed0c2a44a4e49847e83b Reviewed-on: https://chromium-review.googlesource.com/767618Reviewed-by:
Adam Klein <adamk@chromium.org> Commit-Queue: Alexei Filippov <alph@chromium.org> Cr-Commit-Position: refs/heads/master@{#49514}
-
- 20 Nov, 2017 1 commit
-
-
Ben Smith authored
The strex (Store Exclusive) instruction has the form: strex rd, rt, [rn] It stores the value in register rt at the address in register rn. If the store succeeds, then 0 is stored in rd, otherwise 1 is stored. The ARM manual says that behavior is "unpredictable" if d == n || d == t (i.e. those registers are aliased). We were not checking for this behavior in the assembler or simulator, and as a result were generating output where it occurred. This didn't always break; the tests we run on ARM hardware run this instruction and pass. BUG: chromium:786168 Change-Id: I57fe3a1db406eac96eb04ef2246f6970548d3cf9 Reviewed-on: https://chromium-review.googlesource.com/777777Reviewed-by:Mircea Trofin <mtrofin@chromium.org> Commit-Queue: Ben Smith <binji@chromium.org> Cr-Commit-Position: refs/heads/master@{#49513}
-
