Bug 183458 – Split DirectArguments into JSValueOOB and JSValueStrict parts

• Status: RESOLVED
• Resolution: FIXED
• Component: JavaScriptCore

[Filip Pizlo]

Patch forthcoming.

[Filip Pizlo]

Created attachment 335312 [details]
work in progress

[Filip Pizlo]

The DFG changes in this patch will break 32-bit DFG because of register starvation.

[Filip Pizlo]

Created attachment 335332 [details]
more

[Filip Pizlo]

Created attachment 335342 [details]
more

[Filip Pizlo]

Created attachment 335345 [details]
maybe it's done

[Filip Pizlo]

Created attachment 335355 [details]
fixing build issues

[Michael Catanzaro]

(In reply to Filip Pizlo from comment #2)
> The DFG changes in this patch will break 32-bit DFG because of register
> starvation.

Guillaume will be the best person to CC. Thanks!

[Filip Pizlo]

Created attachment 335362 [details]
more

[EWS Watchlist]

Attachment 335362 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/DirectArguments.h:121:  Please replace ASSERT_WITH_SECURITY_IMPLICATION() with RELEASE_ASSERT_WITH_SECURITY_IMPLICATION().  [security/assertion] [5]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:43:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:44:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:45:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:46:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:48:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:5186:  Weird number of spaces at line-start.  Are you using a 4-space indent?  [whitespace/indent] [3]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
Total errors found: 9 in 18 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[EWS Watchlist]

Comment on attachment 335362 [details]
more

Attachment 335362 [details] did not pass jsc-ews (mac):
Output: http://webkit-queues.webkit.org/results/6865987

New failing tests:
stress/arguments-custom-properties-gc.js.dfg-eager
stress/typedarray-sort.js.dfg-eager
stress/typedarray-some.js.dfg-eager-no-cjit-validate
stress/typedarray-map.js.ftl-eager
stress/get-stack-mapping.js.dfg-eager
stress/arguments-custom-properties-gc.js.no-cjit-collect-continuously
stress/typedarray-findIndex.js.ftl-eager-no-cjit
stress/typedarray-map.js.dfg-eager
stress/async-arrow-functions-lexical-arguments-binding.js.dfg-eager-no-cjit-validate
stress/arguments-bizarre-behavior.js.ftl-eager-no-cjit
stress/eval-func-decl-block-with-var-and-remove.js.dfg-eager-no-cjit-validate
stress/typedarray-findIndex.js.no-cjit-collect-continuously
microbenchmarks/call-spread-apply.js.dfg-eager-no-cjit-validate
stress/eval-func-decl-block-with-var-sinthesize.js.ftl-eager
microbenchmarks/call-spread-apply.js.dfg-eager
stress/generator-arguments.js.no-cjit-collect-continuously
stress/regress-170661.js.ftl-eager-no-cjit
microbenchmarks/v8-raytrace-with-empty-try-catch.js.no-cjit-collect-continuously
stress/typedarray-forEach.js.ftl-eager-no-cjit
stress/arrowfunction-lexical-bind-this-8.js.ftl-eager-no-cjit
stress/arguments-interference-cfg.js.ftl-eager-no-cjit
stress/rest-parameter-inlined.js.dfg-eager-no-cjit-validate
v8-v6/v8-raytrace.js.ftl-eager
stress/typedarray-fill.js.dfg-eager
stress/arguments-length-always-dont-enum.js.ftl-eager-no-cjit
stress/sorting-boolean-result-comparator.js.dfg-eager
stress/typedarray-reduce.js.dfg-eager-no-cjit-validate
stress/typedarray-subarray.js.ftl-eager
stress/typedarray-findIndex.js.dfg-eager
stress/reflect-set.js.ftl-eager-no-cjit
stress/tagged-template-tdz.js.ftl-eager
stress/typedarray-fill.js.ftl-eager
stress/async-iteration-yield-star.js.ftl-eager-no-cjit
stress/typedarray-copyWithin.js.dfg-eager-no-cjit-validate
stress/string-raw.js.dfg-eager-no-cjit-validate
stress/get-stack-mapping.js.no-cjit-collect-continuously
stress/typedarray-find.js.dfg-eager-no-cjit-validate
stress/async-iteration-yield-star.js.no-cjit-collect-continuously
stress/typedarray-copyWithin.js.ftl-eager
stress/typedarray-slice.js.ftl-eager
stress/string-raw.js.ftl-eager-no-cjit
stress/string-raw.js.no-cjit-collect-continuously
stress/typedarray-map.js.ftl-eager-no-cjit
stress/async-iteration-yield-promise.js.no-cjit-collect-continuously
stress/typedarray-sort.js.dfg-eager-no-cjit-validate
stress/arguments-bizarre-behaviour-disable-enumerability.js.dfg-eager-no-cjit-validate
stress/regress-170661.js.dfg-eager-no-cjit-validate
stress/string-raw.js.dfg-eager
stress/dfg-create-arguments-inline-alloc.js.dfg-eager-no-cjit-validate
stress/sorting-boolean-result-comparator.js.no-cjit-collect-continuously
stress/arguments-length-always-dont-enum.js.ftl-eager
stress/arrowfunction-lexical-bind-arguments-non-strict-1.js.dfg-eager-no-cjit-validate
stress/typedarray-set.js.ftl-eager-no-cjit
stress/generator-arguments.js.ftl-eager-no-cjit
stress/async-arrow-functions-lexical-arguments-binding.js.ftl-eager-no-cjit
stress/typedarray-filter.js.no-cjit-collect-continuously
microbenchmarks/call-spread-apply.js.ftl-eager
stress/typedarray-indexOf.js.ftl-eager-no-cjit
stress/arguments-interference-cfg.js.no-cjit-collect-continuously
stress/promise-finally.js.ftl-eager-no-cjit
stress/spread-calling.js.no-cjit-collect-continuously
stress/typedarray-subarray.js.ftl-eager-no-cjit
stress/typedarray-filter.js.ftl-eager
stress/arrowfunction-lexical-bind-this-2.js.dfg-eager
stress/typedarray-lastIndexOf.js.ftl-eager-no-cjit
stress/typedarray-forEach.js.dfg-eager-no-cjit-validate
stress/typedarray-sort.js.ftl-eager-no-cjit
microbenchmarks/call-spread-call.js.dfg-eager-no-cjit-validate
stress/arguments-copy-register-array-backing-store.js.ftl-eager-no-cjit
stress/direct-arguments-osr-entry.js.dfg-eager-no-cjit-validate
stress/typedarray-fill.js.ftl-eager-no-cjit
stress/typedarray-slice.js.dfg-eager-no-cjit-validate
stress/spread-calling.js.ftl-eager-no-cjit
stress/eval-func-decl-block-with-var-and-remove.js.no-cjit-collect-continuously
v8-v6/v8-earley-boyer.js.dfg-eager-no-cjit-validate
stress/typedarray-reduce.js.ftl-eager-no-cjit
stress/typedarray-reduce.js.ftl-eager
microbenchmarks/call-spread-call.js.ftl-eager-no-cjit
stress/load-varargs-then-inlined-call-and-exit.js.dfg-eager-no-cjit-validate
stress/typedarray-lastIndexOf.js.dfg-eager
stress/direct-arguments-osr-entry.js.ftl-eager
stress/typedarray-forEach.js.ftl-eager
microbenchmarks/deltablue-varargs.js.dfg-eager
stress/async-iteration-yield-star.js.dfg-eager
v8-v6/v8-earley-boyer.js.ftl-eager-no-cjit
v8-v6/v8-earley-boyer.js.ftl-eager
stress/load-varargs-then-inlined-call-inlined.js.dfg-eager
stress/typedarray-find.js.ftl-eager
stress/async-iteration-async-from-sync.js.dfg-eager-no-cjit-validate
microbenchmarks/v8-raytrace-with-empty-try-catch.js.ftl-eager-no-cjit
stress/typedarray-every.js.dfg-eager
stress/async-iteration-async-from-sync.js.ftl-eager
stress/async-iteration-async-from-sync.js.ftl-eager-no-cjit
stress/typedarray-indexOf.js.dfg-eager
stress/arguments-length-always-dont-enum.js.dfg-eager
stress/tagged-template-tdz.js.no-cjit-collect-continuously
stress/arguments-custom-properties-gc.js.ftl-eager-no-cjit
stress/typedarray-set.js.dfg-eager
stress/typedarray-subarray.js.no-cjit-collect-continuously
stress/typedarray-every.js.ftl-eager-no-cjit
stress/dfg-create-arguments-inline-alloc.js.ftl-eager-no-cjit
stress/const-not-strict-mode.js.ftl-eager-no-cjit
microbenchmarks/v8-raytrace-with-empty-try-catch.js.ftl-eager
stress/arguments-copy-register-array-backing-store.js.dfg-eager-no-cjit-validate
stress/async-iteration-yield-star.js.dfg-eager-no-cjit-validate
stress/tagged-template-tdz.js.dfg-eager-no-cjit-validate
stress/arguments-interference-cfg.js.dfg-eager-no-cjit-validate
stress/async-iteration-yield-promise.js.dfg-eager
stress/typedarray-filter.js.ftl-eager-no-cjit
stress/typedarray-map.js.dfg-eager-no-cjit-validate
stress/typedarray-copyWithin.js.no-cjit-collect-continuously
stress/typedarray-slice.js.no-cjit-collect-continuously
stress/const-not-strict-mode.js.dfg-eager
stress/async-iteration-yield-promise.js.ftl-eager-no-cjit
stress/typedarray-some.js.ftl-eager-no-cjit
stress/regress-170661.js.ftl-eager
stress/eval-func-decl-block-with-var-and-remove.js.ftl-eager
stress/typedarray-forEach.js.dfg-eager
stress/typedarray-some.js.no-cjit-collect-continuously
stress/arguments-length-always-dont-enum.js.dfg-eager-no-cjit-validate
stress/typedarray-forEach.js.no-cjit-collect-continuously
stress/arguments-copy-register-array-backing-store.js.dfg-eager
stress/regress-170661.js.dfg-eager
stress/typedarray-findIndex.js.ftl-eager
stress/eval-func-decl-block-with-var-and-remove.js.dfg-eager
stress/eval-func-decl-block-with-var-sinthesize.js.no-cjit-collect-continuously
stress/typedarray-find.js.dfg-eager
stress/async-arrow-functions-lexical-arguments-binding.js.ftl-eager
stress/const-not-strict-mode.js.no-cjit-collect-continuously
stress/async-iteration-yield-star-interface.js.no-cjit-collect-continuously
stress/tailCallForwardArguments.js.no-cjit-collect-continuously
stress/const-not-strict-mode.js.dfg-eager-no-cjit-validate
stress/async-await-syntax.js.ftl-eager
stress/arguments-custom-properties-gc.js.ftl-eager
stress/typedarray-lastIndexOf.js.dfg-eager-no-cjit-validate
stress/promise-finally.js.no-cjit-collect-continuously
stress/typedarray-find.js.ftl-eager-no-cjit
stress/arguments-bizarre-behaviour-disable-enumerability.js.dfg-eager
stress/typedarray-set.js.dfg-eager-no-cjit-validate
stress/async-await-syntax.js.ftl-eager-no-cjit
stress/typedarray-sort.js.no-cjit-collect-continuously
stress/tagged-template-tdz.js.dfg-eager
ChakraCore.yaml/ChakraCore/test/Miscellaneous/HasOnlyWritableDataPropertiesCache.js.default
stress/get-stack-mapping-with-dead-get-stack.js.no-cjit-collect-continuously
stress/typedarray-fill.js.no-cjit-collect-continuously
stress/tagged-template-tdz.js.ftl-eager-no-cjit
stress/typedarray-reduceRight.js.ftl-eager
stress/typedarray-reduceRight.js.dfg-eager
stress/typedarray-reduceRight.js.no-cjit-collect-continuously
stress/typedarray-slice.js.ftl-eager-no-cjit
stress/spread-calling.js.dfg-eager-no-cjit-validate
stress/eval-func-decl-block-with-var-and-remove.js.ftl-eager-no-cjit
stress/arguments-interference.js.no-cjit-collect-continuously
stress/typedarray-copyWithin.js.dfg-eager
stress/eval-func-decl-block-with-var-sinthesize.js.ftl-eager-no-cjit
microbenchmarks/deltablue-varargs.js.dfg-eager-no-cjit-validate
stress/sorting-boolean-result-comparator.js.ftl-eager-no-cjit
stress/get-stack-mapping.js.dfg-eager-no-cjit-validate
stress/arguments-bizarre-behavior.js.ftl-eager
microbenchmarks/v8-raytrace-with-empty-try-catch.js.dfg-eager-no-cjit-validate
stress/arguments-copy-register-array-backing-store.js.no-cjit-collect-continuously
stress/eval-func-decl-in-eval-within-block-with-let.js.ftl-eager
stress/typedarray-every.js.no-cjit-collect-continuously
stress/arguments-bizarre-behaviour-disable-enumerability.js.ftl-eager
stress/async-await-syntax.js.dfg-eager-no-cjit-validate
microbenchmarks/deltablue-varargs.js.ftl-eager-no-cjit
stress/arrowfunction-lexical-bind-this-2.js.dfg-eager-no-cjit-validate
microbenchmarks/call-spread-apply.js.no-cjit-collect-continuously
stress/typedarray-find.js.no-cjit-collect-continuously
stress/typedarray-indexOf.js.no-cjit-collect-continuously
microbenchmarks/call-spread-call.js.no-cjit-collect-continuously
stress/rest-parameter-inlined.js.ftl-eager-no-cjit
stress/typedarray-set.js.no-cjit-collect-continuously
stress/promise-finally.js.dfg-eager-no-cjit-validate
stress/typedarray-indexOf.js.dfg-eager-no-cjit-validate
stress/typedarray-sort.js.ftl-eager
v8-v6/v8-earley-boyer.js.no-cjit-collect-continuously
stress/typedarray-lastIndexOf.js.no-cjit-collect-continuously
stress/get-stack-mapping-with-dead-get-stack.js.dfg-eager
microbenchmarks/call-spread-call.js.dfg-eager
stress/typedarray-lastIndexOf.js.ftl-eager
microbenchmarks/call-spread-call.js.ftl-eager
stress/typedarray-reduceRight.js.ftl-eager-no-cjit
stress/typedarray-map.js.no-cjit-collect-continuously
stress/async-iteration-yield-star-interface.js.ftl-eager
stress/async-iteration-yield-promise.js.ftl-eager
stress/async-iteration-yield-star-interface.js.dfg-eager-no-cjit-validate
stress/typedarray-every.js.dfg-eager-no-cjit-validate
stress/typedarray-findIndex.js.dfg-eager-no-cjit-validate
stress/typedarray-subarray.js.dfg-eager
stress/sorting-boolean-result-comparator.js.ftl-eager
stress/regress-170661.js.no-cjit-collect-continuously
microbenchmarks/deltablue-varargs.js.no-cjit-collect-continuously
stress/get-my-argument-by-val-creates-arguments.js.dfg-eager-no-cjit-validate
stress/typedarray-reduce.js.no-cjit-collect-continuously
stress/typedarray-subarray.js.dfg-eager-no-cjit-validate
stress/typedarray-filter.js.dfg-eager
stress/typedarray-filter.js.dfg-eager-no-cjit-validate
stress/spread-calling.js.ftl-eager
stress/typedarray-slice.js.dfg-eager
stress/spread-calling.js.dfg-eager
stress/arguments-interference.js.dfg-eager-no-cjit-validate
microbenchmarks/v8-raytrace-with-empty-try-catch.js.dfg-eager
microbenchmarks/v8-raytrace-with-try-catch.js.dfg-eager
stress/get-stack-mapping-with-dead-get-stack.js.dfg-eager-no-cjit-validate
stress/typedarray-indexOf.js.ftl-eager
v8-v6/v8-earley-boyer.js.dfg-eager
stress/tailCallForwardArguments.js.dfg-eager
stress/async-iteration-async-from-sync.js.dfg-eager
stress/typedarray-every.js.ftl-eager
stress/typedarray-reduceRight.js.dfg-eager-no-cjit-validate
stress/typedarray-copyWithin.js.ftl-eager-no-cjit
stress/arrowfunction-lexical-bind-this-8.js.ftl-eager
stress/typedarray-reduce.js.dfg-eager
stress/direct-arguments-osr-entry.js.no-cjit-collect-continuously
stress/typedarray-set.js.ftl-eager
stress/ftl-get-my-argument-by-val-inlined-and-not-inlined.js.no-cjit-collect-continuously
stress/typedarray-some.js.ftl-eager
stress/sorting-boolean-result-comparator.js.dfg-eager-no-cjit-validate
stress/arguments-custom-properties-gc.js.dfg-eager-no-cjit-validate
stress/arrowfunction-lexical-bind-this-2.js.no-cjit-collect-continuously
stress/typedarray-some.js.dfg-eager
stress/generator-arguments.js.ftl-eager
stress/load-varargs-then-inlined-call.js.no-cjit-collect-continuously
stress/tailCallForwardArguments.js.dfg-eager-no-cjit-validate
stress/typedarray-fill.js.dfg-eager-no-cjit-validate

[EWS Watchlist]

Comment on attachment 335362 [details]
more

Attachment 335362 [details] did not pass ios-sim-ews (ios-simulator-wk2):
Output: http://webkit-queues.webkit.org/results/6866080

New failing tests:
perf/typing-at-end-of-line.html
perf/htmlcollection-last-item.html
imported/w3c/web-platform-tests/WebCryptoAPI/sign_verify/ecdsa.worker.html
perf/svg-path-appenditem.html
perf/array-reverse.html
imported/w3c/web-platform-tests/streams/readable-streams/default-reader.serviceworker.https.html
perf/set-attribute.html
perf/htmlcollection-backwards-iteration.html
perf/table-rows-length-caching.html

[EWS Watchlist]

Created attachment 335375 [details]
Archive of layout-test-results from ews121 for ios-simulator-wk2

The attached test failures were seen while running run-webkit-tests on the ios-sim-ews.
Bot: ews121  Port: ios-simulator-wk2  Platform: Mac OS X 10.12.6

[Filip Pizlo]

(In reply to Michael Catanzaro from comment #7)
> (In reply to Filip Pizlo from comment #2)
> > The DFG changes in this patch will break 32-bit DFG because of register
> > starvation.
> 
> Guillaume will be the best person to CC. Thanks!

I ended up doing the 32-bit support.  We're not going to kill 32-bit DFG yet, as much as I wish we could.

[Filip Pizlo]

Created attachment 335445 [details]
more

I think it passes 64-bit tests.  It's perf neutral there.  Still working on 32-bit.

[EWS Watchlist]

Attachment 335445 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/DirectArguments.h:121:  Please replace ASSERT_WITH_SECURITY_IMPLICATION() with RELEASE_ASSERT_WITH_SECURITY_IMPLICATION().  [security/assertion] [5]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:43:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:44:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:45:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:46:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:48:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:5189:  Weird number of spaces at line-start.  Are you using a 4-space indent?  [whitespace/indent] [3]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
Total errors found: 9 in 19 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[EWS Watchlist]

Comment on attachment 335445 [details]
more

Attachment 335445 [details] did not pass win-ews (win):
Output: http://webkit-queues.webkit.org/results/6881506

Number of test failures exceeded the failure limit.

[EWS Watchlist]

Created attachment 335464 [details]
Archive of layout-test-results from ews206 for win-future

The attached test failures were seen while running run-webkit-tests on the win-ews.
Bot: ews206  Port: win-future  Platform: CYGWIN_NT-6.1-2.9.0-0.318-5-3-x86_64-64bit

[Filip Pizlo]

Created attachment 335481 [details]
more

getting 32-bit to work

[EWS Watchlist]

Attachment 335481 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/DirectArguments.h:121:  Please replace ASSERT_WITH_SECURITY_IMPLICATION() with RELEASE_ASSERT_WITH_SECURITY_IMPLICATION().  [security/assertion] [5]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:43:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:44:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:45:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:46:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:48:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:5189:  Weird number of spaces at line-start.  Are you using a 4-space indent?  [whitespace/indent] [3]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
Total errors found: 9 in 19 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Filip Pizlo]

Created attachment 335494 [details]
the patch

[EWS Watchlist]

Attachment 335494 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/DirectArguments.h:121:  Please replace ASSERT_WITH_SECURITY_IMPLICATION() with RELEASE_ASSERT_WITH_SECURITY_IMPLICATION().  [security/assertion] [5]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:43:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:44:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:45:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:46:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:48:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:5189:  Weird number of spaces at line-start.  Are you using a 4-space indent?  [whitespace/indent] [3]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
Total errors found: 9 in 19 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Filip Pizlo]

Created attachment 335496 [details]
the patch

[EWS Watchlist]

Attachment 335496 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/DirectArguments.h:121:  Please replace ASSERT_WITH_SECURITY_IMPLICATION() with RELEASE_ASSERT_WITH_SECURITY_IMPLICATION().  [security/assertion] [5]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:43:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:44:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:45:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:46:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:48:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:5189:  Weird number of spaces at line-start.  Are you using a 4-space indent?  [whitespace/indent] [3]
ERROR: Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsSlowPathGenerator.h:47:  Wrong number of spaces before statement. (expected: 12)  [whitespace/indent] [4]
Total errors found: 9 in 19 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Yusuke Suzuki]

Comment on attachment 335496 [details]
the patch

View in context: https://bugs.webkit.org/attachment.cgi?id=335496&action=review

Is my understanding about the threat model cared in this patch correct? If so, r=me.

> Source/JavaScriptCore/ChangeLog:12
> +        window, a write could appear to be made speculatively and rolled out later. This means that:

My understanding is that, if you have length in the object (like DirectArguments), speculative execution of `m_length = xxx` (maybe in an indirected way) in the untaken path can be done and this updated `m_length` can be used for speculative loading. Is it correct?

if (unmitigated structure check) {
    // any read and write to this JSObject inline fields should not allow attackers to speculatively read secret memory region into the cache.
    // But DirectArguments is breaking this rule right now.
}

> Source/JavaScriptCore/ChangeLog:14
> +        - JSValue objects cannot have lengths, masks, or anything else inline.

So, if this is used for spectre mitigation, attacker can perform spectre attack with speculative execution of writes to these fields.

> Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsSlowPathGenerator.h:39
>  public:

It restores the value of lengthGPR and storageGPR.

> Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:37
> +// This calls operationCreateDirectArguments but then restores the value of lengthGPR.

It restores storageGPR.

> Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:5184
> +        fastStorage = m_out.add(fastStorage, m_out.constIntPtr(DirectArguments::storageHeaderSize()));
> +        m_out.store32(length.value, fastStorage, m_heaps.DirectArguments_Storage_length);
> +        m_out.store32(m_out.constInt32(minCapacity), fastStorage, m_heaps.DirectArguments_Storage_minCapacity);
> +        

OK, we may store the pointer to the middle of auxiliary buffer in the stack/register and perform GC. But it's OK since it is not JSCell.

> Source/JavaScriptCore/runtime/DirectArguments.h:127
> +        return *preciseIndexMaskPtr(
> +            offset.offset(),
> +            std::max(storageHeader(storage).length, storageHeader(storage).minCapacity),
> +            ptr);

It would be nice if we have a comment what preciseIndexMaskPtr does in wtf/MathExtras.h.
It masks the given ptr with 0xffffffffffffffff (ptrwidth) if `index < length`. Otherwise, it masks ptr with 0. Similar to Linux kernel's array_ptr.

[Radar WebKit Bug Importer]

<rdar://problem/38357169>

[Filip Pizlo]

(In reply to Yusuke Suzuki from comment #24)
> Comment on attachment 335496 [details]
> the patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=335496&action=review
> 
> Is my understanding about the threat model cared in this patch correct? 

Yes! :-)

> If
> so, r=me.
> 
> > Source/JavaScriptCore/ChangeLog:12
> > +        window, a write could appear to be made speculatively and rolled out later. This means that:
> 
> My understanding is that, if you have length in the object (like
> DirectArguments), speculative execution of `m_length = xxx` (maybe in an
> indirected way) in the untaken path can be done and this updated `m_length`
> can be used for speculative loading. Is it correct?
> 
> if (unmitigated structure check) {
>     // any read and write to this JSObject inline fields should not allow
> attackers to speculatively read secret memory region into the cache.
>     // But DirectArguments is breaking this rule right now.
> }
> 
> > Source/JavaScriptCore/ChangeLog:14
> > +        - JSValue objects cannot have lengths, masks, or anything else inline.
> 
> So, if this is used for spectre mitigation, attacker can perform spectre
> attack with speculative execution of writes to these fields.
> 
> > Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsSlowPathGenerator.h:39
> >  public:
> 
> It restores the value of lengthGPR and storageGPR.
> 
> > Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:37
> > +// This calls operationCreateDirectArguments but then restores the value of lengthGPR.
> 
> It restores storageGPR.
> 
> > Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:5184
> > +        fastStorage = m_out.add(fastStorage, m_out.constIntPtr(DirectArguments::storageHeaderSize()));
> > +        m_out.store32(length.value, fastStorage, m_heaps.DirectArguments_Storage_length);
> > +        m_out.store32(m_out.constInt32(minCapacity), fastStorage, m_heaps.DirectArguments_Storage_minCapacity);
> > +        
> 
> OK, we may store the pointer to the middle of auxiliary buffer in the
> stack/register and perform GC. But it's OK since it is not JSCell.
> 
> > Source/JavaScriptCore/runtime/DirectArguments.h:127
> > +        return *preciseIndexMaskPtr(
> > +            offset.offset(),
> > +            std::max(storageHeader(storage).length, storageHeader(storage).minCapacity),
> > +            ptr);
> 
> It would be nice if we have a comment what preciseIndexMaskPtr does in
> wtf/MathExtras.h.
> It masks the given ptr with 0xffffffffffffffff (ptrwidth) if `index <
> length`. Otherwise, it masks ptr with 0. Similar to Linux kernel's array_ptr.

[Filip Pizlo]

(In reply to Yusuke Suzuki from comment #24)
> Comment on attachment 335496 [details]
> the patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=335496&action=review
> 
> Is my understanding about the threat model cared in this patch correct? If
> so, r=me.
> 
> > Source/JavaScriptCore/ChangeLog:12
> > +        window, a write could appear to be made speculatively and rolled out later. This means that:
> 
> My understanding is that, if you have length in the object (like
> DirectArguments), speculative execution of `m_length = xxx` (maybe in an
> indirected way) in the untaken path can be done and this updated `m_length`
> can be used for speculative loading. Is it correct?
> 
> if (unmitigated structure check) {
>     // any read and write to this JSObject inline fields should not allow
> attackers to speculatively read secret memory region into the cache.
>     // But DirectArguments is breaking this rule right now.
> }
> 
> > Source/JavaScriptCore/ChangeLog:14
> > +        - JSValue objects cannot have lengths, masks, or anything else inline.
> 
> So, if this is used for spectre mitigation, attacker can perform spectre
> attack with speculative execution of writes to these fields.
> 
> > Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsSlowPathGenerator.h:39
> >  public:
> 
> It restores the value of lengthGPR and storageGPR.

Fixed.

> 
> > Source/JavaScriptCore/dfg/DFGCallCreateDirectArgumentsWithKnownLengthSlowPathGenerator.h:37
> > +// This calls operationCreateDirectArguments but then restores the value of lengthGPR.
> 
> It restores storageGPR.

Fixed.

> 
> > Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:5184
> > +        fastStorage = m_out.add(fastStorage, m_out.constIntPtr(DirectArguments::storageHeaderSize()));
> > +        m_out.store32(length.value, fastStorage, m_heaps.DirectArguments_Storage_length);
> > +        m_out.store32(m_out.constInt32(minCapacity), fastStorage, m_heaps.DirectArguments_Storage_minCapacity);
> > +        
> 
> OK, we may store the pointer to the middle of auxiliary buffer in the
> stack/register and perform GC. But it's OK since it is not JSCell.

Yeah, the GC will mark the auxiliary buffer. But it will not scan it for outgoing pointers. So, it's safe to store things to it, provided you don't try to use it as the only root for other pointers.

> 
> > Source/JavaScriptCore/runtime/DirectArguments.h:127
> > +        return *preciseIndexMaskPtr(
> > +            offset.offset(),
> > +            std::max(storageHeader(storage).length, storageHeader(storage).minCapacity),
> > +            ptr);
> 
> It would be nice if we have a comment what preciseIndexMaskPtr does in
> wtf/MathExtras.h.
> It masks the given ptr with 0xffffffffffffffff (ptrwidth) if `index <
> length`. Otherwise, it masks ptr with 0. Similar to Linux kernel's array_ptr.

Yeah, I added such a comment.

[Filip Pizlo]

Landed in https://trac.webkit.org/changeset/229518/webkit

[Ryan Haddad]

This change caused 800 test failures on the 32-bit JSC bot:

https://build.webkit.org/builders/Apple%20High%20Sierra%2032-bit%20JSC%20%28BuildAndTest%29/builds/1392

stress/varargs-exit.js.default: ASSERTION FAILED: m_data[index].lockCount
stress/varargs-exit.js.default: ./dfg/DFGRegisterBank.h(175) : void JSC::DFG::RegisterBank<JSC::GPRInfo>::retain(RegID, JSC::VirtualRegister, SpillHint) [BankInfo = JSC::GPRInfo]
stress/varargs-exit.js.default: 1   0x14eb9da WTFCrash
stress/varargs-exit.js.default: 2   0xa02996 JSC::DFG::RegisterBank<JSC::GPRInfo>::retain(JSC::X86Registers::RegisterID, JSC::VirtualRegister, unsigned int)
stress/varargs-exit.js.default: 3   0xa05b46 JSC::DFG::SpeculativeJIT::cellResult(JSC::X86Registers::RegisterID, JSC::DFG::Node*, JSC::DFG::SpeculativeJIT::UseChildrenMode)
stress/varargs-exit.js.default: 4   0xa5abaf JSC::DFG::SpeculativeJIT::compileCreateDirectArguments(JSC::DFG::Node*)
stress/varargs-exit.js.default: 5   0xa21510 JSC::DFG::SpeculativeJIT::compile(JSC::DFG::Node*)
stress/varargs-exit.js.default: 6   0xa0f1c6 JSC::DFG::SpeculativeJIT::compileCurrentBlock()
stress/varargs-exit.js.default: 7   0xa24fbf JSC::DFG::SpeculativeJIT::compile()
stress/varargs-exit.js.default: 8   0x84b27c JSC::DFG::JITCompiler::compileBody()
stress/varargs-exit.js.default: 9   0x850893 JSC::DFG::JITCompiler::compileFunction()
stress/varargs-exit.js.default: 10  0x903318 JSC::DFG::Plan::compileInThreadImpl()
stress/varargs-exit.js.default: 11  0x900485 JSC::DFG::Plan::compileInThread(JSC::DFG::ThreadData*)
stress/varargs-exit.js.default: 12  0x7c60e8 JSC::DFG::compileImpl(JSC::VM&, JSC::CodeBlock*, JSC::CodeBlock*, JSC::DFG::CompilationMode, unsigned int, JSC::Operands<JSC::JSValue> const&, WTF::Ref<JSC::DeferredCompilationCallback, WTF::DumbPtrTraits<JSC::DeferredCompilationCallback> >&&)
stress/varargs-exit.js.default: 13  0x7c5b09 JSC::DFG::compile(JSC::VM&, JSC::CodeBlock*, JSC::CodeBlock*, JSC::DFG::CompilationMode, unsigned int, JSC::Operands<JSC::JSValue> const&, WTF::Ref<JSC::DeferredCompilationCallback, WTF::DumbPtrTraits<JSC::DeferredCompilationCallback> >&&)
stress/varargs-exit.js.default: 14  0xdcb055 operationOptimize
stress/varargs-exit.js.default: 15  0x4565883e
stress/varargs-exit.js.default: 16  0x2fdf2b llint_entry
stress/varargs-exit.js.default: 17  0x2f8230 vmEntryToJavaScript
stress/varargs-exit.js.default: 18  0xdc1818 JSC::JITCode::execute(JSC::VM*, JSC::ProtoCallFrame*)
stress/varargs-exit.js.default: 19  0xd57f93 JSC::Interpreter::executeProgram(JSC::SourceCode const&, JSC::ExecState*, JSC::JSObject*)
stress/varargs-exit.js.default: 20  0x107eb72 JSC::evaluate(JSC::ExecState*, JSC::SourceCode const&, JSC::JSValue, WTF::NakedPtr<JSC::Exception>&)
stress/varargs-exit.js.default: 21  0xf7731 runWithOptions(GlobalObject*, CommandLine&)
stress/varargs-exit.js.default: 22  0xca48c jscmain(int, char**)::$_3::operator()(JSC::VM&, GlobalObject*) const
stress/varargs-exit.js.default: 23  0xafbff int runJSC<jscmain(int, char**)::$_3>(CommandLine, bool, jscmain(int, char**)::$_3 const&)
stress/varargs-exit.js.default: 24  0xae3bf jscmain(int, char**)
stress/varargs-exit.js.default: 25  0xae2d7 main
stress/varargs-exit.js.default: 26  0xa759f6e1 start
stress/varargs-exit.js.default: 27  0x8
stress/varargs-exit.js.default: test_script_11047: line 2: 43850 Segmentation fault: 11  ( "$@" ../../.vm/JavaScriptCore.framework/Resources/jsc --useFTLJIT\=false --useFunctionDotArguments\=true --validateExceptionChecks\=true --useDollarVM\=true --maxPerThreadStackUsage\=1572864 --useFTLJIT\=true varargs-exit.js )
stress/varargs-exit.js.default: ERROR: Unexpected exit code: 139

[Filip Pizlo]

(In reply to Ryan Haddad from comment #29)
> This change caused 800 test failures on the 32-bit JSC bot:
> 
> https://build.webkit.org/builders/Apple%20High%20Sierra%2032-
> bit%20JSC%20%28BuildAndTest%29/builds/1392
> 
> stress/varargs-exit.js.default: ASSERTION FAILED: m_data[index].lockCount
> stress/varargs-exit.js.default: ./dfg/DFGRegisterBank.h(175) : void
> JSC::DFG::RegisterBank<JSC::GPRInfo>::retain(RegID, JSC::VirtualRegister,
> SpillHint) [BankInfo = JSC::GPRInfo]
> stress/varargs-exit.js.default: 1   0x14eb9da WTFCrash
> stress/varargs-exit.js.default: 2   0xa02996
> JSC::DFG::RegisterBank<JSC::GPRInfo>::retain(JSC::X86Registers::RegisterID,
> JSC::VirtualRegister, unsigned int)
> stress/varargs-exit.js.default: 3   0xa05b46
> JSC::DFG::SpeculativeJIT::cellResult(JSC::X86Registers::RegisterID,
> JSC::DFG::Node*, JSC::DFG::SpeculativeJIT::UseChildrenMode)
> stress/varargs-exit.js.default: 4   0xa5abaf
> JSC::DFG::SpeculativeJIT::compileCreateDirectArguments(JSC::DFG::Node*)
> stress/varargs-exit.js.default: 5   0xa21510
> JSC::DFG::SpeculativeJIT::compile(JSC::DFG::Node*)
> stress/varargs-exit.js.default: 6   0xa0f1c6
> JSC::DFG::SpeculativeJIT::compileCurrentBlock()
> stress/varargs-exit.js.default: 7   0xa24fbf
> JSC::DFG::SpeculativeJIT::compile()
> stress/varargs-exit.js.default: 8   0x84b27c
> JSC::DFG::JITCompiler::compileBody()
> stress/varargs-exit.js.default: 9   0x850893
> JSC::DFG::JITCompiler::compileFunction()
> stress/varargs-exit.js.default: 10  0x903318
> JSC::DFG::Plan::compileInThreadImpl()
> stress/varargs-exit.js.default: 11  0x900485
> JSC::DFG::Plan::compileInThread(JSC::DFG::ThreadData*)
> stress/varargs-exit.js.default: 12  0x7c60e8 JSC::DFG::compileImpl(JSC::VM&,
> JSC::CodeBlock*, JSC::CodeBlock*, JSC::DFG::CompilationMode, unsigned int,
> JSC::Operands<JSC::JSValue> const&,
> WTF::Ref<JSC::DeferredCompilationCallback,
> WTF::DumbPtrTraits<JSC::DeferredCompilationCallback> >&&)
> stress/varargs-exit.js.default: 13  0x7c5b09 JSC::DFG::compile(JSC::VM&,
> JSC::CodeBlock*, JSC::CodeBlock*, JSC::DFG::CompilationMode, unsigned int,
> JSC::Operands<JSC::JSValue> const&,
> WTF::Ref<JSC::DeferredCompilationCallback,
> WTF::DumbPtrTraits<JSC::DeferredCompilationCallback> >&&)
> stress/varargs-exit.js.default: 14  0xdcb055 operationOptimize
> stress/varargs-exit.js.default: 15  0x4565883e
> stress/varargs-exit.js.default: 16  0x2fdf2b llint_entry
> stress/varargs-exit.js.default: 17  0x2f8230 vmEntryToJavaScript
> stress/varargs-exit.js.default: 18  0xdc1818 JSC::JITCode::execute(JSC::VM*,
> JSC::ProtoCallFrame*)
> stress/varargs-exit.js.default: 19  0xd57f93
> JSC::Interpreter::executeProgram(JSC::SourceCode const&, JSC::ExecState*,
> JSC::JSObject*)
> stress/varargs-exit.js.default: 20  0x107eb72 JSC::evaluate(JSC::ExecState*,
> JSC::SourceCode const&, JSC::JSValue, WTF::NakedPtr<JSC::Exception>&)
> stress/varargs-exit.js.default: 21  0xf7731 runWithOptions(GlobalObject*,
> CommandLine&)
> stress/varargs-exit.js.default: 22  0xca48c jscmain(int,
> char**)::$_3::operator()(JSC::VM&, GlobalObject*) const
> stress/varargs-exit.js.default: 23  0xafbff int runJSC<jscmain(int,
> char**)::$_3>(CommandLine, bool, jscmain(int, char**)::$_3 const&)
> stress/varargs-exit.js.default: 24  0xae3bf jscmain(int, char**)
> stress/varargs-exit.js.default: 25  0xae2d7 main
> stress/varargs-exit.js.default: 26  0xa759f6e1 start
> stress/varargs-exit.js.default: 27  0x8
> stress/varargs-exit.js.default: test_script_11047: line 2: 43850
> Segmentation fault: 11  ( "$@"
> ../../.vm/JavaScriptCore.framework/Resources/jsc --useFTLJIT\=false
> --useFunctionDotArguments\=true --validateExceptionChecks\=true
> --useDollarVM\=true --maxPerThreadStackUsage\=1572864 --useFTLJIT\=true
> varargs-exit.js )
> stress/varargs-exit.js.default: ERROR: Unexpected exit code: 139

Looks easy to fix.  Looking at it now.

[Filip Pizlo]

(In reply to Filip Pizlo from comment #30)
> (In reply to Ryan Haddad from comment #29)
> > This change caused 800 test failures on the 32-bit JSC bot:
> > 
> > https://build.webkit.org/builders/Apple%20High%20Sierra%2032-
> > bit%20JSC%20%28BuildAndTest%29/builds/1392
> > 
> > stress/varargs-exit.js.default: ASSERTION FAILED: m_data[index].lockCount
> > stress/varargs-exit.js.default: ./dfg/DFGRegisterBank.h(175) : void
> > JSC::DFG::RegisterBank<JSC::GPRInfo>::retain(RegID, JSC::VirtualRegister,
> > SpillHint) [BankInfo = JSC::GPRInfo]
> > stress/varargs-exit.js.default: 1   0x14eb9da WTFCrash
> > stress/varargs-exit.js.default: 2   0xa02996
> > JSC::DFG::RegisterBank<JSC::GPRInfo>::retain(JSC::X86Registers::RegisterID,
> > JSC::VirtualRegister, unsigned int)
> > stress/varargs-exit.js.default: 3   0xa05b46
> > JSC::DFG::SpeculativeJIT::cellResult(JSC::X86Registers::RegisterID,
> > JSC::DFG::Node*, JSC::DFG::SpeculativeJIT::UseChildrenMode)
> > stress/varargs-exit.js.default: 4   0xa5abaf
> > JSC::DFG::SpeculativeJIT::compileCreateDirectArguments(JSC::DFG::Node*)
> > stress/varargs-exit.js.default: 5   0xa21510
> > JSC::DFG::SpeculativeJIT::compile(JSC::DFG::Node*)
> > stress/varargs-exit.js.default: 6   0xa0f1c6
> > JSC::DFG::SpeculativeJIT::compileCurrentBlock()
> > stress/varargs-exit.js.default: 7   0xa24fbf
> > JSC::DFG::SpeculativeJIT::compile()
> > stress/varargs-exit.js.default: 8   0x84b27c
> > JSC::DFG::JITCompiler::compileBody()
> > stress/varargs-exit.js.default: 9   0x850893
> > JSC::DFG::JITCompiler::compileFunction()
> > stress/varargs-exit.js.default: 10  0x903318
> > JSC::DFG::Plan::compileInThreadImpl()
> > stress/varargs-exit.js.default: 11  0x900485
> > JSC::DFG::Plan::compileInThread(JSC::DFG::ThreadData*)
> > stress/varargs-exit.js.default: 12  0x7c60e8 JSC::DFG::compileImpl(JSC::VM&,
> > JSC::CodeBlock*, JSC::CodeBlock*, JSC::DFG::CompilationMode, unsigned int,
> > JSC::Operands<JSC::JSValue> const&,
> > WTF::Ref<JSC::DeferredCompilationCallback,
> > WTF::DumbPtrTraits<JSC::DeferredCompilationCallback> >&&)
> > stress/varargs-exit.js.default: 13  0x7c5b09 JSC::DFG::compile(JSC::VM&,
> > JSC::CodeBlock*, JSC::CodeBlock*, JSC::DFG::CompilationMode, unsigned int,
> > JSC::Operands<JSC::JSValue> const&,
> > WTF::Ref<JSC::DeferredCompilationCallback,
> > WTF::DumbPtrTraits<JSC::DeferredCompilationCallback> >&&)
> > stress/varargs-exit.js.default: 14  0xdcb055 operationOptimize
> > stress/varargs-exit.js.default: 15  0x4565883e
> > stress/varargs-exit.js.default: 16  0x2fdf2b llint_entry
> > stress/varargs-exit.js.default: 17  0x2f8230 vmEntryToJavaScript
> > stress/varargs-exit.js.default: 18  0xdc1818 JSC::JITCode::execute(JSC::VM*,
> > JSC::ProtoCallFrame*)
> > stress/varargs-exit.js.default: 19  0xd57f93
> > JSC::Interpreter::executeProgram(JSC::SourceCode const&, JSC::ExecState*,
> > JSC::JSObject*)
> > stress/varargs-exit.js.default: 20  0x107eb72 JSC::evaluate(JSC::ExecState*,
> > JSC::SourceCode const&, JSC::JSValue, WTF::NakedPtr<JSC::Exception>&)
> > stress/varargs-exit.js.default: 21  0xf7731 runWithOptions(GlobalObject*,
> > CommandLine&)
> > stress/varargs-exit.js.default: 22  0xca48c jscmain(int,
> > char**)::$_3::operator()(JSC::VM&, GlobalObject*) const
> > stress/varargs-exit.js.default: 23  0xafbff int runJSC<jscmain(int,
> > char**)::$_3>(CommandLine, bool, jscmain(int, char**)::$_3 const&)
> > stress/varargs-exit.js.default: 24  0xae3bf jscmain(int, char**)
> > stress/varargs-exit.js.default: 25  0xae2d7 main
> > stress/varargs-exit.js.default: 26  0xa759f6e1 start
> > stress/varargs-exit.js.default: 27  0x8
> > stress/varargs-exit.js.default: test_script_11047: line 2: 43850
> > Segmentation fault: 11  ( "$@"
> > ../../.vm/JavaScriptCore.framework/Resources/jsc --useFTLJIT\=false
> > --useFunctionDotArguments\=true --validateExceptionChecks\=true
> > --useDollarVM\=true --maxPerThreadStackUsage\=1572864 --useFTLJIT\=true
> > varargs-exit.js )
> > stress/varargs-exit.js.default: ERROR: Unexpected exit code: 139
> 
> Looks easy to fix.  Looking at it now.

Should be fixed in https://trac.webkit.org/changeset/229545/webkit