DFG should support tuples #11086

kmiller68 · 2023-03-05T18:57:33Z

`f2f3c91`

DFG should support tuples
https://bugs.webkit.org/show_bug.cgi?id=253413

Reviewed by Yusuke Suzuki.

This change adds support for tuples in the DFG. It works similarly to how tuples
work in B3 where there is a side buffer which holds most of the metadata for a
tuple. In the DFG there are three pieces of information held here:
    1) The reference count
    2) The result flags
    3) The virtual register

1) tells us if the ExtractFromTuple Node for a given tuple result still exists
(e.g. it could have been constant folded or dead code eliminated). From that we
will decide to fill (3) the virtual register when allocating registers. If we
didn't have the reference count then we would have no way to know that the
virtual registers isn't going to be consumed thus we will leak it. Lastly we
have (2) the result flags for the node. This tells the ExtractFromTuple and
theoretically (as right now the only ExtractFromTuple users produce Int32s) any
 consumers of the ExtractFromTuple what value format the tuple's result will
 be in. For the DFG, we don't need this because we know there's at most one
 ExtractFromTuple for any given tuple index since we don't duplicate code.

When dumping the DFG graph ExtractFromTuple follows the same pattern as B3
and uses `<<X` to denote the offset we are extracting from. So it will look
something like:

  6  1  0:  D@101:< 1:->    EnumeratorNextUpdateIndexAndMode(Check:Untyped:D@97, Check:Untyped:D@98, Check:Untyped:D@100, Check:Untyped:D@99, VarArgs, SelectUsingPredictions+NonArray+InBounds+AsIs+Read, enumeratorModes = 4, R:World, W:Heap, Exits, ClobbersExit, bc#115, ExitValid)
  7  1  0:  D@102:< 1:->    ExtractFromTuple(Check:Untyped:D@101, Int32|UseAsOther, <<0, bc#115, ExitInvalid)
  8  1  0:  D@103:<!0:->    MovHint(Check:Untyped:D@102, MustGen, loc10, W:SideState, ClobbersExit, bc#115, ExitInvalid)
  9  1  0:  D@104:< 1:->    ExtractFromTuple(Check:Untyped:D@101, Int32|UseAsOther, <<1, bc#115, ExitInvalid)

This patch also adds support for calling operations in both the FTL/B3 via
CCall. CCall can take exactly the tuple of `{ pointerType(), pointerType() }`,
which, for every calling conevention we support, should be returned in both
the return value registers. As the only way to look into a tuple is via the
B3 prodecure, the first Air::Arg of any CCall/ColdCCall Inst is now the
CCallSpecial for the compiling Air::Code. This gives us access to Air::Code
inside CCallCustom::forEachArg and isValidForm.

* JSTests/stress/for-in-redefine-enumerable.js:
(shouldBe):
* Source/JavaScriptCore/b3/B3LowerToAir.cpp:
* Source/JavaScriptCore/b3/B3Type.h:
(JSC::B3::pointerType):
(JSC::B3::registerType):
* Source/JavaScriptCore/b3/B3Validate.cpp:
* Source/JavaScriptCore/b3/air/AirCCallingConvention.cpp:
(JSC::B3::Air::cCallResultCount):
(JSC::B3::Air::cCallArgumentRegisterWidth):
(JSC::B3::Air::cCallResult):
* Source/JavaScriptCore/b3/air/AirCCallingConvention.h:
* Source/JavaScriptCore/b3/air/AirCustom.cpp:
(JSC::B3::Air::CCallCustom::isValidForm):
* Source/JavaScriptCore/b3/air/AirCustom.h:
(JSC::B3::Air::CCallCustom::forEachArg):
* Source/JavaScriptCore/b3/air/AirLowerAfterRegAlloc.cpp:
(JSC::B3::Air::lowerAfterRegAlloc):
* Source/JavaScriptCore/b3/air/AirLowerMacros.cpp:
(JSC::B3::Air::lowerMacros):
* Source/JavaScriptCore/b3/air/AirOpcode.opcodes:
* Source/JavaScriptCore/b3/testb3.h:
* Source/JavaScriptCore/b3/testb3_3.cpp:
(addCallTests):
* Source/JavaScriptCore/b3/testb3_5.cpp:
(JSC_DEFINE_JIT_OPERATION):
(testCallPairResult):
(testCallPairResultRare):
* Source/JavaScriptCore/dfg/DFGAbstractInterpreter.h:
(JSC::DFG::AbstractInterpreter::setTupleConstant):
* Source/JavaScriptCore/dfg/DFGAbstractInterpreterInlines.h:
(JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects):
* Source/JavaScriptCore/dfg/DFGAtTailAbstractState.cpp:
(JSC::DFG::AtTailAbstractState::AtTailAbstractState):
* Source/JavaScriptCore/dfg/DFGAtTailAbstractState.h:
(JSC::DFG::AtTailAbstractState::forTupleNode):
(JSC::DFG::AtTailAbstractState::clearForTupleNode):
(JSC::DFG::AtTailAbstractState::setForTupleNode):
(JSC::DFG::AtTailAbstractState::setTypeForTupleNode):
(JSC::DFG::AtTailAbstractState::setNonCellTypeForTupleNode):
(JSC::DFG::AtTailAbstractState::makeBytecodeTopForTupleNode):
(JSC::DFG::AtTailAbstractState::makeHeapTopForTupleNode):
* Source/JavaScriptCore/dfg/DFGByteCodeParser.cpp:
(JSC::DFG::ByteCodeParser::addToGraph):
(JSC::DFG::ByteCodeParser::parseBlock):
* Source/JavaScriptCore/dfg/DFGClobberize.h:
(JSC::DFG::clobberize):
* Source/JavaScriptCore/dfg/DFGDoesGC.cpp:
(JSC::DFG::doesGC):
* Source/JavaScriptCore/dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupNode):
* Source/JavaScriptCore/dfg/DFGGenerationInfo.h:
(JSC::DFG::GenerationInfo::initFromTupleResult):
* Source/JavaScriptCore/dfg/DFGGraph.cpp:
(JSC::DFG::Graph::dump):
* Source/JavaScriptCore/dfg/DFGGraph.h:
* Source/JavaScriptCore/dfg/DFGInPlaceAbstractState.cpp:
(JSC::DFG::InPlaceAbstractState::InPlaceAbstractState):
* Source/JavaScriptCore/dfg/DFGInPlaceAbstractState.h:
(JSC::DFG::InPlaceAbstractState::forTupleNode):
(JSC::DFG::InPlaceAbstractState::clearForTupleNode):
(JSC::DFG::InPlaceAbstractState::setForTupleNode):
(JSC::DFG::InPlaceAbstractState::setTypeForTupleNode):
(JSC::DFG::InPlaceAbstractState::setNonCellTypeForTupleNode):
(JSC::DFG::InPlaceAbstractState::makeBytecodeTopForTupleNode):
(JSC::DFG::InPlaceAbstractState::makeHeapTopForTupleNode):
* Source/JavaScriptCore/dfg/DFGMayExit.cpp:
* Source/JavaScriptCore/dfg/DFGNode.h:
(JSC::DFG::Node::isTuple const):
(JSC::DFG::Node::setTupleOffset):
(JSC::DFG::Node::tupleOffset const):
(JSC::DFG::Node::hasExtractOffset const):
(JSC::DFG::Node::extractOffset const):
(JSC::DFG::Node::tupleIndex const):
(JSC::DFG::Node::tupleSize const):
(JSC::DFG::Node::hasVirtualRegister):
(JSC::DFG::Node::virtualRegister):
(JSC::DFG::Node::setVirtualRegister):
* Source/JavaScriptCore/dfg/DFGNodeType.h:
* Source/JavaScriptCore/dfg/DFGOperations.cpp:
(JSC::DFG::JSC_DEFINE_JIT_OPERATION):
* Source/JavaScriptCore/dfg/DFGOperations.h:
(JSC::DFG::makeUGPRPair):
* Source/JavaScriptCore/dfg/DFGPredictionPropagationPhase.cpp:
* Source/JavaScriptCore/dfg/DFGSafeToExecute.h:
(JSC::DFG::safeToExecute):
* Source/JavaScriptCore/dfg/DFGScoreBoard.h:
(JSC::DFG::ScoreBoard::use):
* Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp:
* Source/JavaScriptCore/dfg/DFGSpeculativeJIT.h:
(JSC::DFG::SpeculativeJIT::strictInt32TupleResultWithoutUsingChildren):
* Source/JavaScriptCore/dfg/DFGSpeculativeJIT32_64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* Source/JavaScriptCore/dfg/DFGSpeculativeJIT64.cpp:
(JSC::DFG::SpeculativeJIT::compile):
* Source/JavaScriptCore/dfg/DFGValidate.cpp:
* Source/JavaScriptCore/dfg/DFGVirtualRegisterAllocationPhase.cpp:
(JSC::DFG::VirtualRegisterAllocationPhase::run):
* Source/JavaScriptCore/ftl/FTLCapabilities.cpp:
(JSC::FTL::canCompile):
* Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:
(JSC::FTL::DFG::LowerDFGToB3::LowerDFGToB3):
(JSC::FTL::DFG::LowerDFGToB3::compileNode):
(JSC::FTL::DFG::LowerDFGToB3::compileCompareStrictEq):
* Source/JavaScriptCore/ftl/FTLOutput.cpp:
(JSC::FTL::Output::extract):
* Source/JavaScriptCore/ftl/FTLOutput.h:
* Source/JavaScriptCore/wasm/WasmAirIRGeneratorBase.h:
(JSC::Wasm::AirIRGeneratorBase::emitCCall):

Canonical link: https://commits.webkit.org/262068@main

ab3884c

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🛠 gtk
	~~🧪 api-ios~~	✅ 🧪 mac-wk1	✅ 🧪 gtk-wk2
✅ 🛠 🧪 jsc	✅ 🛠 tv	✅ 🧪 mac-wk2	⏳ 🧪 api-gtk
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	~~🧪 mac-AS-debug-wk2~~	✅ 🛠 jsc-armv7
	✅ 🛠 watch	✅ 🧪 mac-wk2-stress	✅ 🧪 jsc-armv7-tests
✅ 🛠 🧪 merge	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			✅ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-05T18:57:38Z

EWS run on previous version of this PR (hash 340db27)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~
✅ 🧪 webkitperl	⏳ 🧪 ios-wk2	~~🧪 api-mac~~	✅ 🛠 gtk
	~~🧪 api-ios~~	~~🧪 mac-wk1~~	~~🧪 gtk-wk2~~
⏳ 🛠 🧪 jsc	✅ 🛠 tv	~~🧪 mac-wk2~~	~~🧪 api-gtk~~
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	✅ 🧪 mac-AS-debug-wk2	❌ 🛠 jsc-armv7
	✅ 🛠 watch	~~🧪 mac-wk2-stress~~	❌ 🧪 jsc-armv7-tests
	✅ 🛠 watch-sim		❌ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-05T19:04:19Z

EWS run on previous version of this PR (hash 7e7c552)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🛠 gtk
	✅ 🧪 api-ios	✅ 🧪 mac-wk1	✅ 🧪 gtk-wk2
✅ 🛠 🧪 jsc	✅ 🛠 tv	❌ 🧪 mac-wk2	✅ 🧪 api-gtk
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	✅ 🧪 mac-AS-debug-wk2	❌ 🛠 jsc-armv7
	✅ 🛠 watch	⏳ 🧪 mac-wk2-stress	❌ 🧪 jsc-armv7-tests
	✅ 🛠 watch-sim		❌ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-06T15:50:27Z

EWS run on previous version of this PR (hash bfc0ba5)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~
✅ 🧪 webkitperl	⏳ 🧪 ios-wk2	~~🧪 api-mac~~	✅ 🛠 gtk
	~~🧪 api-ios~~	~~🧪 mac-wk1~~	~~🧪 gtk-wk2~~
⏳ 🛠 🧪 jsc	✅ 🛠 tv	~~🧪 mac-wk2~~	~~🧪 api-gtk~~
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	~~🧪 mac-AS-debug-wk2~~	✅ 🛠 jsc-armv7
	✅ 🛠 watch	⏳ 🧪 mac-wk2-stress	~~🧪 jsc-armv7-tests~~
	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			~~🧪 jsc-mips-tests~~

webkit-early-warning-system · 2023-03-06T15:54:57Z

EWS run on previous version of this PR (hash 5edde02)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🛠 gtk
	✅ 🧪 api-ios	✅ 🧪 mac-wk1	✅ 🧪 gtk-wk2
✅ 🛠 🧪 jsc	✅ 🛠 tv	✅ 🧪 mac-wk2	✅ 🧪 api-gtk
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	✅ 🧪 mac-AS-debug-wk2	✅ 🛠 jsc-armv7
	✅ 🛠 watch	✅ 🧪 mac-wk2-stress	✅ 🧪 jsc-armv7-tests
	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			✅ 🧪 jsc-mips-tests

Source/JavaScriptCore/dfg/DFGAbstractInterpreterInlines.h

Constellation · 2023-03-08T23:15:30Z

Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp

    flushRegisters();
    JSValueRegsFlushedCallResult result(this);
    JSValueRegs resultRegs = result.regs();
+    // FIXME: We should teach callOperation how to handle more than one result.


I think we support two result GPRs. See extractResult etc.

extractResult doesn't quite work because it does a single register on 64-bit. I guess I could do:

setupArguments<decltype(operationEnumeratorNextUpdateIndexAndMode)>(LinkableConstant::globalObject(*this, node), baseRegs, indexGPR, modeGPR, enumeratorGPR); appendCallSetResult(operationEnumeratorNextUpdateIndexAndMode, resultGPR1, resultGPR2);

Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp

Constellation · 2023-03-09T01:35:57Z

Source/JavaScriptCore/dfg/DFGPredictionPropagationPhase.cpp

-            setPrediction(SpecInt32Only);
+        case ExtractFromTuple: {
+            // Use mergePrediction because ExtractFromTuple doesn't know if the prediction could change.
+            mergePrediction(m_tupleSpeculations[m_currentNode->tupleIndex()]);


How about having this prediction in TupleData? This can be useful in FixupPhase too.

Why would you do that here? This is the prediction on the extract node.

Constellation · 2023-03-09T01:39:09Z

Source/JavaScriptCore/dfg/DFGGraph.h

+    struct TupleData {
+        uint16_t refCount { 0 };
+        uint16_t resultFlags { 0 };
+        VirtualRegister virtualRegister;
+    };


I think we can extend this further, like having m_prediction etc. Basically, this can be a data structure containing most of DFG::Node.

You don't really need the prediction here because consumers of the node will see the prediction via the ExtractFromTuple. I could see adding OpInfos here for extra metadata as needed but it seems like we should do that when we need it.

Constellation · 2023-03-09T01:43:31Z

Source/JavaScriptCore/dfg/DFGNode.h

+    unsigned tupleOffset() const
+    {
+        ASSERT(isTuple());
+        return m_virtualRegister.toLocal();


This is a bit unfortunate, but yeah, it works since we never allocate VirtualRegister for tuple.

Yeah, I figured saving the OpInfos for other data seemed more useful since we know that we're not going to use the virtualRegister for anything.

Source/JavaScriptCore/dfg/DFGGraph.cpp

Constellation · 2023-03-09T02:03:57Z

Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp

+    ASSERT(m_graph.m_tupleData.at(node->tupleIndex()).virtualRegister == node->virtualRegister());
+    VirtualRegister virtualRegister = node->virtualRegister();
+    GenerationInfo& info = generationInfoFromVirtualRegister(virtualRegister);
+    info.stealForNode(node);


Maybe, we should have a bit better name for that, like, initFromTupleElement?
And don't we need to setup this GenerationInfo based on resultFlags of tuple element? (As the same way to FTL)
Maybe, the producer side already initialize this GenerationInfo with appropriate types, and since we are using the same VirtualRegister here, we do not need to initialize it much, but maybe we should assert here to ensure that the resulted GenerationInfo matches against the expected Tuple element's type.

Yeah, that's right I was relying on the fact that we are just repurposing the layout of the tuple's result. Sure, I can add those ASSERTs.

webkit-early-warning-system · 2023-03-13T21:55:36Z

EWS run on previous version of this PR (hash e00e56f)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	⏳ 🛠 wincairo
	✅ 🛠 ios-sim	⏳ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~
✅ 🧪 webkitperl	~~🧪 ios-wk2~~	~~🧪 api-mac~~	✅ 🛠 gtk
	~~🧪 api-ios~~	~~🧪 mac-wk1~~	~~🧪 gtk-wk2~~
⏳ 🛠 🧪 jsc	⏳ 🛠 tv	✅ 🧪 mac-wk2	~~🧪 api-gtk~~
❌ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	⏳ 🧪 mac-AS-debug-wk2	✅ 🛠 jsc-armv7
	⏳ 🛠 watch	✅ 🧪 mac-wk2-stress	❌ 🧪 jsc-armv7-tests
	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-14T00:00:27Z

EWS run on previous version of this PR (hash 96e7b43)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🛠 gtk
	✅ 🧪 api-ios	✅ 🧪 mac-wk1	❌ 🧪 gtk-wk2
✅ 🛠 🧪 jsc	✅ 🛠 tv	✅ 🧪 mac-wk2	~~🧪 api-gtk~~
⏳ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	~~🧪 mac-AS-debug-wk2~~	✅ 🛠 jsc-armv7
	✅ 🛠 watch	✅ 🧪 mac-wk2-stress	✅ 🧪 jsc-armv7-tests
	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-14T13:20:17Z

EWS run on previous version of this PR (hash 3aadb9a)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🛠 gtk
	✅ 🧪 api-ios	✅ 🧪 mac-wk1	✅ 🧪 gtk-wk2
✅ 🛠 🧪 jsc	✅ 🛠 tv	✅ 🧪 mac-wk2	~~🧪 api-gtk~~
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	~~🧪 mac-AS-debug-wk2~~	✅ 🛠 jsc-armv7
	✅ 🛠 watch	✅ 🧪 mac-wk2-stress	✅ 🧪 jsc-armv7-tests
	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-14T23:40:38Z

EWS run on previous version of this PR (hash 224d40e)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🛠 gtk
	✅ 🧪 api-ios	✅ 🧪 mac-wk1	✅ 🧪 gtk-wk2
✅ 🛠 🧪 jsc	✅ 🛠 tv	✅ 🧪 mac-wk2	❌ 🧪 api-gtk
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	✅ 🧪 mac-AS-debug-wk2	✅ 🛠 jsc-armv7
	✅ 🛠 watch	✅ 🧪 mac-wk2-stress	❌ 🧪 jsc-armv7-tests
	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

Constellation

r=me

Constellation · 2023-03-15T00:59:46Z

Source/JavaScriptCore/b3/B3Validate.cpp

@@ -662,6 +662,12 @@ class Validater {
                VALIDATE(!value->kind().hasExtraBits(), ("At ", *value));
                VALIDATE(value->numChildren() >= 1, ("At ", *value));
                VALIDATE(value->child(0)->type() == pointerType(), ("At ", *value));
+                if (value->type().isTuple()) {


Let's add comment that currently we are only supporting two-element case.

Constellation · 2023-03-15T21:04:51Z

Source/JavaScriptCore/b3/air/AirCustom.cpp

+    if (!inst.args[0].isSpecial())
+        return false;
+


Why do we need this check twice?

I forgot to delete it here. Will delete.

Constellation · 2023-03-15T21:06:53Z

Source/JavaScriptCore/b3/air/AirCustom.cpp

+    for (unsigned i = inst.args.size() - 1; i > 1; --i) {
+        Arg arg = inst.args[i];


This does not visit args[1]. i starts with size - 1, and we run this loop if i > 1, this means i needs to be 2 or more. So, it misses the visit to callee, is there any reason?

Ah yeah, it should include callee. I think I confused myself with the args layout for CCallSpecial when we lower to that.

Constellation · 2023-03-15T21:19:41Z

Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp

-
-        jsValueResult(resultRegs, node);
+        useChildren(node);
+        strictInt32TupleResultWithoutUsingChildren(newIndex.gpr(), node, 0);


Don't we need to check refCount for this first tuple element?

We don't bother constant folding it because it's the index. If we did you'd need to check yeah but you'd trigger an assert to notice you did it wrong.

Constellation · 2023-03-15T21:20:52Z

Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp

        return;
    }

    if (node->enumeratorMetadata() == JSPropertyNameEnumerator::OwnStructureMode && baseEdge.useKind() == CellUse) {
        SpeculateCellOperand base(this, baseEdge);
-        JSValueRegsTemporary result(this);
+        GPRTemporary newIndex(this);
+        GPRTemporary newMode(this, Reuse, mode);
        GPRReg baseGPR = base.gpr();


Let's first declare all GPRs here like,

GPRReg newIndexGPR = newIndex.gpr(); GPRReg newModeGPR = newMode.gpr();

since gpr() is side-effectful operation. If we jump over that, it causes the bug. So it is always nice if we get all GPRs first.

I think it used to be side-effectful but .*Temporary.gpr() is not anymore, which IMO causes a bit of an anti pattern where the operands you declared can get spilled before they are filled. .*Operands still get filled on first use so those have to be initialized early.

Constellation · 2023-03-15T21:21:56Z

Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp

        return;
    }

    JSValueOperand base(this, baseEdge);
+#if USE(JSVALUE64)
+    GPRTemporary newMode(this, Reuse, mode);
+#endif
    JSValueRegs baseRegs = base.regs();


Ditto. Let's get all GPRs here first.

Constellation · 2023-03-15T21:23:23Z

Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp

+    strictInt32TupleResultWithoutUsingChildren(indexResult.gpr(), node, 0);
+    strictInt32TupleResultWithoutUsingChildren(modeResult.gpr(), node, 1);


Don't we need to check refCount?

I think I'll just change it so we check in strictInt32TupleResultWithoutUsingChildren.

Constellation · 2023-03-15T21:24:27Z

Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp

+    ASSERT(m_graph.m_tupleData.at(node->tupleIndex()).virtualRegister == node->virtualRegister());
+    VirtualRegister virtualRegister = node->virtualRegister();
+    GenerationInfo& info = generationInfoFromVirtualRegister(virtualRegister);
+    if (ASSERT_ENABLED) {


IIRC, this causes some build failures. Let's just use #if ASSERT_ENABLED

Constellation · 2023-03-15T21:35:17Z

Source/JavaScriptCore/b3/air/AirCustom.cpp

+    Code& code = special->code();
+
+    size_t resultCount = cCallResultCount(code, value);
+    size_t expectedArgCount = resultCount + 1; // first Arg is always CCallSpecial.
    for (Value* child : value->children()) {


OK, it is including callee itself.

Constellation · 2023-03-15T21:38:41Z

Source/JavaScriptCore/b3/air/AirCustom.h

+        for (size_t n = cCallResultCount(code, value); n; --n) {
+            Type type = value->type().isTuple() ? code.proc().typeAtOffset(value->type(), n - 1) : value->type();


n is decremented. But next's index is incremented. Is this correct?

I guess it doesn't matter right now because if there's more than one result you get the same registers anyway. In the future it could matter though so I'll fix.

webkit-early-warning-system · 2023-03-16T20:26:40Z

EWS run on previous version of this PR (hash 1ae27f9)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🛠 gtk
	✅ 🧪 api-ios	✅ 🧪 mac-wk1	✅ 🧪 gtk-wk2
✅ 🛠 🧪 jsc	✅ 🛠 tv	✅ 🧪 mac-wk2	~~🧪 api-gtk~~
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	✅ 🧪 mac-AS-debug-wk2	✅ 🛠 jsc-armv7
	✅ 🛠 watch	✅ 🧪 mac-wk2-stress	❌ 🧪 jsc-armv7-tests
	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-23T18:29:49Z

EWS run on previous version of this PR (hash 500b65f)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	❌ 🛠 ios	❌ 🛠 mac	~~🛠 wpe~~	~~🛠 wincairo~~
	❌ 🛠 ios-sim	❌ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	❌ 🧪 api-mac	❌ 🛠 gtk
	❌ 🧪 api-ios	❌ 🧪 mac-wk1	❌ 🧪 gtk-wk2
⏳ 🛠 🧪 jsc	❌ 🛠 tv	❌ 🧪 mac-wk2	❌ 🧪 api-gtk
❌ 🛠 🧪 jsc-arm64	❌ 🛠 tv-sim	❌ 🧪 mac-AS-debug-wk2	❌ 🛠 jsc-armv7
	⏳ 🛠 watch	❌ 🧪 mac-wk2-stress	❌ 🧪 jsc-armv7-tests
	❌ 🛠 watch-sim		❌ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-23T18:39:30Z

EWS run on previous version of this PR (hash 4e4244e)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	❌ 🛠 ios	❌ 🛠 mac	❌ 🛠 wpe	🛠 wincairo
	❌ 🛠 ios-sim	❌ 🛠 mac-AS-debug	❌ 🧪 wpe-wk2
✅ 🧪 webkitperl	❌ 🧪 ios-wk2	❌ 🧪 api-mac	❌ 🛠 gtk
	❌ 🧪 api-ios	❌ 🧪 mac-wk1	❌ 🧪 gtk-wk2
⏳ 🛠 🧪 jsc	❌ 🛠 tv	❌ 🧪 mac-wk2	❌ 🧪 api-gtk
❌ 🛠 🧪 jsc-arm64	❌ 🛠 tv-sim	❌ 🧪 mac-AS-debug-wk2	❌ 🛠 jsc-armv7
	❌ 🛠 watch	❌ 🧪 mac-wk2-stress	❌ 🧪 jsc-armv7-tests
	❌ 🛠 watch-sim		❌ 🛠 jsc-mips
			❌ 🧪 jsc-mips-tests

webkit-early-warning-system · 2023-03-23T18:59:44Z

EWS run on previous version of this PR (hash 0303753)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	~~🧪 wpe-wk2~~
✅ 🧪 webkitperl	~~🧪 ios-wk2~~	✅ 🧪 api-mac	✅ 🛠 gtk
	~~🧪 api-ios~~	✅ 🧪 mac-wk1	~~🧪 gtk-wk2~~
⏳ 🛠 🧪 jsc	✅ 🛠 tv	✅ 🧪 mac-wk2	⏳ 🧪 api-gtk
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	~~🧪 mac-AS-debug-wk2~~	✅ 🛠 jsc-armv7
	✅ 🛠 watch	✅ 🧪 mac-wk2-stress	~~🧪 jsc-armv7-tests~~
	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			~~🧪 jsc-mips-tests~~

webkit-early-warning-system · 2023-03-23T19:47:14Z

EWS run on current version of this PR (hash ab3884c)

Misc	iOS, tvOS & watchOS	macOS	Linux	Windows
✅ 🧪 style	✅ 🛠 ios	✅ 🛠 mac	✅ 🛠 wpe	✅ 🛠 wincairo
	✅ 🛠 ios-sim	✅ 🛠 mac-AS-debug	✅ 🧪 wpe-wk2
✅ 🧪 webkitperl	✅ 🧪 ios-wk2	✅ 🧪 api-mac	✅ 🛠 gtk
	~~🧪 api-ios~~	✅ 🧪 mac-wk1	✅ 🧪 gtk-wk2
✅ 🛠 🧪 jsc	✅ 🛠 tv	✅ 🧪 mac-wk2	⏳ 🧪 api-gtk
✅ 🛠 🧪 jsc-arm64	✅ 🛠 tv-sim	~~🧪 mac-AS-debug-wk2~~	✅ 🛠 jsc-armv7
	✅ 🛠 watch	✅ 🧪 mac-wk2-stress	✅ 🧪 jsc-armv7-tests
✅ 🛠 🧪 merge	✅ 🛠 watch-sim		✅ 🛠 jsc-mips
			✅ 🧪 jsc-mips-tests

https://bugs.webkit.org/show_bug.cgi?id=253413 Reviewed by Yusuke Suzuki. This change adds support for tuples in the DFG. It works similarly to how tuples work in B3 where there is a side buffer which holds most of the metadata for a tuple. In the DFG there are three pieces of information held here: 1) The reference count 2) The result flags 3) The virtual register 1) tells us if the ExtractFromTuple Node for a given tuple result still exists (e.g. it could have been constant folded or dead code eliminated). From that we will decide to fill (3) the virtual register when allocating registers. If we didn't have the reference count then we would have no way to know that the virtual registers isn't going to be consumed thus we will leak it. Lastly we have (2) the result flags for the node. This tells the ExtractFromTuple and theoretically (as right now the only ExtractFromTuple users produce Int32s) any consumers of the ExtractFromTuple what value format the tuple's result will be in. For the DFG, we don't need this because we know there's at most one ExtractFromTuple for any given tuple index since we don't duplicate code. When dumping the DFG graph ExtractFromTuple follows the same pattern as B3 and uses `<<X` to denote the offset we are extracting from. So it will look something like: 6 1 0: D@101:< 1:-> EnumeratorNextUpdateIndexAndMode(Check:Untyped:D@97, Check:Untyped:D@98, Check:Untyped:D@100, Check:Untyped:D@99, VarArgs, SelectUsingPredictions+NonArray+InBounds+AsIs+Read, enumeratorModes = 4, R:World, W:Heap, Exits, ClobbersExit, bc#115, ExitValid) 7 1 0: D@102:< 1:-> ExtractFromTuple(Check:Untyped:D@101, Int32|UseAsOther, <<0, bc#115, ExitInvalid) 8 1 0: D@103:<!0:-> MovHint(Check:Untyped:D@102, MustGen, loc10, W:SideState, ClobbersExit, bc#115, ExitInvalid) 9 1 0: D@104:< 1:-> ExtractFromTuple(Check:Untyped:D@101, Int32|UseAsOther, <<1, bc#115, ExitInvalid) This patch also adds support for calling operations in both the FTL/B3 via CCall. CCall can take exactly the tuple of `{ pointerType(), pointerType() }`, which, for every calling conevention we support, should be returned in both the return value registers. As the only way to look into a tuple is via the B3 prodecure, the first Air::Arg of any CCall/ColdCCall Inst is now the CCallSpecial for the compiling Air::Code. This gives us access to Air::Code inside CCallCustom::forEachArg and isValidForm. * JSTests/stress/for-in-redefine-enumerable.js: (shouldBe): * Source/JavaScriptCore/b3/B3LowerToAir.cpp: * Source/JavaScriptCore/b3/B3Type.h: (JSC::B3::pointerType): (JSC::B3::registerType): * Source/JavaScriptCore/b3/B3Validate.cpp: * Source/JavaScriptCore/b3/air/AirCCallingConvention.cpp: (JSC::B3::Air::cCallResultCount): (JSC::B3::Air::cCallArgumentRegisterWidth): (JSC::B3::Air::cCallResult): * Source/JavaScriptCore/b3/air/AirCCallingConvention.h: * Source/JavaScriptCore/b3/air/AirCustom.cpp: (JSC::B3::Air::CCallCustom::isValidForm): * Source/JavaScriptCore/b3/air/AirCustom.h: (JSC::B3::Air::CCallCustom::forEachArg): * Source/JavaScriptCore/b3/air/AirLowerAfterRegAlloc.cpp: (JSC::B3::Air::lowerAfterRegAlloc): * Source/JavaScriptCore/b3/air/AirLowerMacros.cpp: (JSC::B3::Air::lowerMacros): * Source/JavaScriptCore/b3/air/AirOpcode.opcodes: * Source/JavaScriptCore/b3/testb3.h: * Source/JavaScriptCore/b3/testb3_3.cpp: (addCallTests): * Source/JavaScriptCore/b3/testb3_5.cpp: (JSC_DEFINE_JIT_OPERATION): (testCallPairResult): (testCallPairResultRare): * Source/JavaScriptCore/dfg/DFGAbstractInterpreter.h: (JSC::DFG::AbstractInterpreter::setTupleConstant): * Source/JavaScriptCore/dfg/DFGAbstractInterpreterInlines.h: (JSC::DFG::AbstractInterpreter<AbstractStateType>::executeEffects): * Source/JavaScriptCore/dfg/DFGAtTailAbstractState.cpp: (JSC::DFG::AtTailAbstractState::AtTailAbstractState): * Source/JavaScriptCore/dfg/DFGAtTailAbstractState.h: (JSC::DFG::AtTailAbstractState::forTupleNode): (JSC::DFG::AtTailAbstractState::clearForTupleNode): (JSC::DFG::AtTailAbstractState::setForTupleNode): (JSC::DFG::AtTailAbstractState::setTypeForTupleNode): (JSC::DFG::AtTailAbstractState::setNonCellTypeForTupleNode): (JSC::DFG::AtTailAbstractState::makeBytecodeTopForTupleNode): (JSC::DFG::AtTailAbstractState::makeHeapTopForTupleNode): * Source/JavaScriptCore/dfg/DFGByteCodeParser.cpp: (JSC::DFG::ByteCodeParser::addToGraph): (JSC::DFG::ByteCodeParser::parseBlock): * Source/JavaScriptCore/dfg/DFGClobberize.h: (JSC::DFG::clobberize): * Source/JavaScriptCore/dfg/DFGDoesGC.cpp: (JSC::DFG::doesGC): * Source/JavaScriptCore/dfg/DFGFixupPhase.cpp: (JSC::DFG::FixupPhase::fixupNode): * Source/JavaScriptCore/dfg/DFGGenerationInfo.h: (JSC::DFG::GenerationInfo::initFromTupleResult): * Source/JavaScriptCore/dfg/DFGGraph.cpp: (JSC::DFG::Graph::dump): * Source/JavaScriptCore/dfg/DFGGraph.h: * Source/JavaScriptCore/dfg/DFGInPlaceAbstractState.cpp: (JSC::DFG::InPlaceAbstractState::InPlaceAbstractState): * Source/JavaScriptCore/dfg/DFGInPlaceAbstractState.h: (JSC::DFG::InPlaceAbstractState::forTupleNode): (JSC::DFG::InPlaceAbstractState::clearForTupleNode): (JSC::DFG::InPlaceAbstractState::setForTupleNode): (JSC::DFG::InPlaceAbstractState::setTypeForTupleNode): (JSC::DFG::InPlaceAbstractState::setNonCellTypeForTupleNode): (JSC::DFG::InPlaceAbstractState::makeBytecodeTopForTupleNode): (JSC::DFG::InPlaceAbstractState::makeHeapTopForTupleNode): * Source/JavaScriptCore/dfg/DFGMayExit.cpp: * Source/JavaScriptCore/dfg/DFGNode.h: (JSC::DFG::Node::isTuple const): (JSC::DFG::Node::setTupleOffset): (JSC::DFG::Node::tupleOffset const): (JSC::DFG::Node::hasExtractOffset const): (JSC::DFG::Node::extractOffset const): (JSC::DFG::Node::tupleIndex const): (JSC::DFG::Node::tupleSize const): (JSC::DFG::Node::hasVirtualRegister): (JSC::DFG::Node::virtualRegister): (JSC::DFG::Node::setVirtualRegister): * Source/JavaScriptCore/dfg/DFGNodeType.h: * Source/JavaScriptCore/dfg/DFGOperations.cpp: (JSC::DFG::JSC_DEFINE_JIT_OPERATION): * Source/JavaScriptCore/dfg/DFGOperations.h: (JSC::DFG::makeUGPRPair): * Source/JavaScriptCore/dfg/DFGPredictionPropagationPhase.cpp: * Source/JavaScriptCore/dfg/DFGSafeToExecute.h: (JSC::DFG::safeToExecute): * Source/JavaScriptCore/dfg/DFGScoreBoard.h: (JSC::DFG::ScoreBoard::use): * Source/JavaScriptCore/dfg/DFGSpeculativeJIT.cpp: * Source/JavaScriptCore/dfg/DFGSpeculativeJIT.h: (JSC::DFG::SpeculativeJIT::strictInt32TupleResultWithoutUsingChildren): * Source/JavaScriptCore/dfg/DFGSpeculativeJIT32_64.cpp: (JSC::DFG::SpeculativeJIT::compile): * Source/JavaScriptCore/dfg/DFGSpeculativeJIT64.cpp: (JSC::DFG::SpeculativeJIT::compile): * Source/JavaScriptCore/dfg/DFGValidate.cpp: * Source/JavaScriptCore/dfg/DFGVirtualRegisterAllocationPhase.cpp: (JSC::DFG::VirtualRegisterAllocationPhase::run): * Source/JavaScriptCore/ftl/FTLCapabilities.cpp: (JSC::FTL::canCompile): * Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp: (JSC::FTL::DFG::LowerDFGToB3::LowerDFGToB3): (JSC::FTL::DFG::LowerDFGToB3::compileNode): (JSC::FTL::DFG::LowerDFGToB3::compileCompareStrictEq): * Source/JavaScriptCore/ftl/FTLOutput.cpp: (JSC::FTL::Output::extract): * Source/JavaScriptCore/ftl/FTLOutput.h: * Source/JavaScriptCore/wasm/WasmAirIRGeneratorBase.h: (JSC::Wasm::AirIRGeneratorBase::emitCCall): Canonical link: https://commits.webkit.org/262068@main

webkit-commit-queue · 2023-03-24T12:46:35Z

Committed 262068@main (f2f3c91): https://commits.webkit.org/262068@main

Reviewed commits have been landed. Closing PR #11086 and removing active labels.

kmiller68 requested a review from a team as a code owner March 5, 2023 18:57

kmiller68 self-assigned this Mar 5, 2023

kmiller68 changed the title ~~Need a short description (OOPS!).~~ DFG should support tuples Mar 5, 2023

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 340db27 to 7e7c552 Compare March 5, 2023 19:04

kmiller68 added the JavaScriptCore For bugs in JavaScriptCore, the JS engine used by WebKit, other than kxmlcore issues. label Mar 5, 2023

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 7e7c552 to bfc0ba5 Compare March 6, 2023 15:50

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from bfc0ba5 to 5edde02 Compare March 6, 2023 15:54

Constellation reviewed Mar 9, 2023

View reviewed changes

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 5edde02 to e00e56f Compare March 13, 2023 21:51

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from e00e56f to 96e7b43 Compare March 13, 2023 23:59

webkit-ews-buildbot added the merging-blocked Applied to prevent a change from being merged label Mar 14, 2023

kmiller68 removed the merging-blocked Applied to prevent a change from being merged label Mar 14, 2023

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 96e7b43 to 3aadb9a Compare March 14, 2023 13:19

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 3aadb9a to 224d40e Compare March 14, 2023 23:39

Constellation approved these changes Mar 15, 2023

View reviewed changes

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 224d40e to 1ae27f9 Compare March 16, 2023 14:38

webkit-ews-buildbot added the merging-blocked Applied to prevent a change from being merged label Mar 17, 2023

kmiller68 removed the merging-blocked Applied to prevent a change from being merged label Mar 23, 2023

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 1ae27f9 to 500b65f Compare March 23, 2023 18:28

webkit-ews-buildbot added the merging-blocked Applied to prevent a change from being merged label Mar 23, 2023

kmiller68 removed the merging-blocked Applied to prevent a change from being merged label Mar 23, 2023

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 500b65f to 4e4244e Compare March 23, 2023 18:37

webkit-ews-buildbot added the merging-blocked Applied to prevent a change from being merged label Mar 23, 2023

kmiller68 removed the merging-blocked Applied to prevent a change from being merged label Mar 23, 2023

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 4e4244e to 0303753 Compare March 23, 2023 18:57

kmiller68 force-pushed the eng/DFG-should-support-tuples branch from 0303753 to ab3884c Compare March 23, 2023 19:46

kmiller68 added the merge-queue Applied to send a pull request to merge-queue label Mar 24, 2023

webkit-commit-queue force-pushed the eng/DFG-should-support-tuples branch from ab3884c to f2f3c91 Compare March 24, 2023 12:46

webkit-commit-queue removed the merge-queue Applied to send a pull request to merge-queue label Mar 24, 2023

webkit-commit-queue merged commit f2f3c91 into WebKit:main Mar 24, 2023

		for (unsigned i = inst.args.size() - 1; i > 1; --i) {
		Arg arg = inst.args[i];

		strictInt32TupleResultWithoutUsingChildren(indexResult.gpr(), node, 0);
		strictInt32TupleResultWithoutUsingChildren(modeResult.gpr(), node, 1);

		for (size_t n = cCallResultCount(code, value); n; --n) {
		Type type = value->type().isTuple() ? code.proc().typeAtOffset(value->type(), n - 1) : value->type();

DFG should support tuples #11086

DFG should support tuples #11086

Conversation

kmiller68 commented Mar 5, 2023 • edited by webkit-early-warning-system

webkit-early-warning-system commented Mar 5, 2023 • edited

webkit-early-warning-system commented Mar 5, 2023 • edited

webkit-early-warning-system commented Mar 6, 2023 • edited

webkit-early-warning-system commented Mar 6, 2023 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

webkit-early-warning-system commented Mar 13, 2023 • edited

webkit-early-warning-system commented Mar 14, 2023 • edited

webkit-early-warning-system commented Mar 14, 2023 • edited

webkit-early-warning-system commented Mar 14, 2023 • edited

Constellation left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

webkit-early-warning-system commented Mar 16, 2023 • edited

webkit-early-warning-system commented Mar 23, 2023 • edited

webkit-early-warning-system commented Mar 23, 2023 • edited

webkit-early-warning-system commented Mar 23, 2023 • edited

webkit-early-warning-system commented Mar 23, 2023 • edited

webkit-commit-queue commented Mar 24, 2023

kmiller68 commented Mar 5, 2023 •

edited by webkit-early-warning-system

webkit-early-warning-system commented Mar 5, 2023 •

edited

webkit-early-warning-system commented Mar 5, 2023 •

edited

webkit-early-warning-system commented Mar 6, 2023 •

edited

webkit-early-warning-system commented Mar 6, 2023 •

edited

webkit-early-warning-system commented Mar 13, 2023 •

edited

webkit-early-warning-system commented Mar 14, 2023 •

edited

webkit-early-warning-system commented Mar 14, 2023 •

edited

webkit-early-warning-system commented Mar 14, 2023 •

edited

webkit-early-warning-system commented Mar 16, 2023 •

edited

webkit-early-warning-system commented Mar 23, 2023 •

edited

webkit-early-warning-system commented Mar 23, 2023 •

edited

webkit-early-warning-system commented Mar 23, 2023 •

edited

webkit-early-warning-system commented Mar 23, 2023 •

edited