Bug 155270 – DFG and FTL should constant-fold RegExpExec, RegExpTest, and StringReplace

• Component: JavaScriptCore
• Status: RESOLVED FIXED

[Filip Pizlo]

Patch forthcoming.

[Filip Pizlo]

Created attachment 273491 [details]
it begins

[Filip Pizlo]

This creates a bunch of problems:

- Yarr needs to be executed on the JIT thread.
- DFG must model what happens in createRegExpMatchesArray, RegExpObject, and RegExpConstructor.
- If the match succeeds, we must still emit code that creates the matches array, and that code should not be any less efficient than what we have in createRegExpMatchesArray.  That code does a one-shot allocation that has enough room for the two names properties and enough room for the indexed properties.  We currently have no DFG node that can do that.

I'm attaching the Yarr concurrency problem just adding locks in various places.

I will make the DFG duplicate the logic of createRegExpMatches array and friends.  This will require creating new nodes, like a node for putting a result summary on the RegExp constructor.

I'm attacking the array allocation problem by extending MaterializeNewObject to be able to do these kinds of one-shot allocations.  Basically it will become the superset of NewArrayWithSize, NewArray, and NewObject.  I'm also extending MaterializeNewObject to be compilable by the DFG tier, since we want to do this folding in DFG too.

[Filip Pizlo]

Created attachment 273590 [details]
OMG this is so insane

[Filip Pizlo]

Created attachment 273602 [details]
I thnk that MaterializeNewObject is done

Now I can write the folding rule.  This will be *SO MUCH FUN*.

[Filip Pizlo]

Created attachment 273621 [details]
added RecordRegExpCachedResult

[Filip Pizlo]

Created attachment 273628 [details]
inching towards victory!

[Filip Pizlo]

Created attachment 273638 [details]
it is written

I will eventually work up the courage to try to compile it.

[Filip Pizlo]

Created attachment 273665 [details]
it compiles!

[Filip Pizlo]

Created attachment 273670 [details]
it does things!

[Filip Pizlo]

Created attachment 273673 [details]
more tests

[Filip Pizlo]

Created attachment 273675 [details]
passes all of the easy tests

[Filip Pizlo]

Created attachment 273678 [details]
maybe the patch

[Filip Pizlo]

Created attachment 273681 [details]
the patch

[WebKit Commit Bot]

Attachment 273681 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1418:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
Total errors found: 9 in 61 files


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

[Filip Pizlo]

Created attachment 273684 [details]
the patch

Contains some more fixes.

[WebKit Commit Bot]

Attachment 273684 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1418:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
Total errors found: 9 in 61 files


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

[Filip Pizlo]

Created attachment 273691 [details]
the patch

[Filip Pizlo]

Still running tests, not yet ready for review.

[WebKit Commit Bot]

Attachment 273691 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1418:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
ERROR: Source/JavaScriptCore/runtime/StringPrototype.h:66:  The parameter name "reg" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 10 in 69 files


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

[Filip Pizlo]

Created attachment 273692 [details]
the patch

[WebKit Commit Bot]

Attachment 273692 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1418:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
ERROR: Source/JavaScriptCore/runtime/StringPrototype.h:66:  The parameter name "reg" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 10 in 69 files


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

[Filip Pizlo]

Created attachment 273695 [details]
the patch

with 32-bit fixes

[WebKit Commit Bot]

Attachment 273695 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:101:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1418:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
ERROR: Source/JavaScriptCore/runtime/StringPrototype.h:66:  The parameter name "reg" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 10 in 69 files


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

[Build Bot]

Comment on attachment 273695 [details]
the patch

Attachment 273695 [details] did not pass mac-debug-ews (mac):
Output: http://webkit-queues.webkit.org/results/959280

Number of test failures exceeded the failure limit.

[Build Bot]

Created attachment 273699 [details]
Archive of layout-test-results from ews114 for mac-yosemite

The attached test failures were seen while running run-webkit-tests on the mac-debug-ews.
Bot: ews114  Port: mac-yosemite  Platform: Mac OS X 10.10.5

[Filip Pizlo]

(In reply to comment #25)
> Created attachment 273699 [details]
> Archive of layout-test-results from ews114 for mac-yosemite
> 
> The attached test failures were seen while running run-webkit-tests on the
> mac-debug-ews.
> Bot: ews114  Port: mac-yosemite  Platform: Mac OS X 10.10.5

LOL, this is because of a stale checkout situation on my end.  I had added an assertion to RegExpMatchesArray.h as part of a previous abortive effort.  That assertion was wrong.  It ended up in my patch.  And it set this bot on fire.

[Filip Pizlo]

Created attachment 273746 [details]
the patch

Reverts RegExpMatchesArray.h, in which I had accidentally added a wrong assertion that is unrelated to this patch.  This should fix the debug tests.

[WebKit Commit Bot]

Attachment 273746 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1418:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
ERROR: Source/JavaScriptCore/runtime/StringPrototype.h:66:  The parameter name "reg" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 10 in 68 files


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

[Saam Barati]

Comment on attachment 273746 [details]
the patch

View in context: https://bugs.webkit.org/attachment.cgi?id=273746&action=review

r=me

> Source/JavaScriptCore/ChangeLog:20
> +        properties. Previously it could only handle objects (but not arrays) and naed properties

typo: named

> Source/JavaScriptCore/dfg/DFGLazyJSValue.cpp:120
> +                dataLog("Giving up because of size limit.\n");

remove.

> Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:521
> +                            m_nodeIndex, origin, jsNumber(result.start), UntypedUse));

Why is this untyped use?

> Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:525
> +                    m_graph.m_varArgChildren.append(Edge(stringNode, UntypedUse));

Shouldn't this be StringUse?

> Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:651
> +                int replLen = replace.length();

unsigned.

> Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:681
> +                        m_nodeIndex, origin, jsNumber(0), UntypedUse));

KnownInt32Use?

> Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:6397
> +                HashMap<
> +                    int32_t, LValue, DefaultHash<int32_t>::Hash,
> +                    WTF::UnsignedWithZeroKeyHashTraits<int32_t>> indexMap;

style: I think this can be 1 line.

[Filip Pizlo]

(In reply to comment #29)
> Comment on attachment 273746 [details]
> the patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=273746&action=review
> 
> r=me
> 
> > Source/JavaScriptCore/ChangeLog:20
> > +        properties. Previously it could only handle objects (but not arrays) and naed properties
> 
> typo: named

Fixed.

> 
> > Source/JavaScriptCore/dfg/DFGLazyJSValue.cpp:120
> > +                dataLog("Giving up because of size limit.\n");
> 
> remove.

Fixed.

> 
> > Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:521
> > +                            m_nodeIndex, origin, jsNumber(result.start), UntypedUse));
> 
> Why is this untyped use?

With the exception of the publicLength, vectorLength, and structure arguments to MaterializeNewObject, all of them are UntypedUse because they're just values you're storing into the object.  You can store any JSValue into an object.  Therefore, UntypedUse.

More importantly, its not legal to just use whatever use kind you want anywhere you want.  The signature of a DFG node is its NodeType and its UseKinds.  If you use a UseKind that the node type doesn't recognize, you break the compiler.  MaterializeNewObject does not recognize anything but UntypedUse for anything other than publicLength/vectorLength/structure.

In general, if there is a use edge from A (the user) to B (the thing being used), then you don't pick the use kind based on what B is - you pick it based on what A wants B to be.  If B is an int, you don't tell A to use it as an int unless A knows what to do with that information.

> 
> > Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:525
> > +                    m_graph.m_varArgChildren.append(Edge(stringNode, UntypedUse));
> 
> Shouldn't this be StringUse?

See above.

> 
> > Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:651
> > +                int replLen = replace.length();
> 
> unsigned.

OK.  I just copied what the StringPrototype code was doing.

> 
> > Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:681
> > +                        m_nodeIndex, origin, jsNumber(0), UntypedUse));
> 
> KnownInt32Use?

See above, you can't just pick whatever use kind you like.

> 
> > Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:6397
> > +                HashMap<
> > +                    int32_t, LValue, DefaultHash<int32_t>::Hash,
> > +                    WTF::UnsignedWithZeroKeyHashTraits<int32_t>> indexMap;
> 
> style: I think this can be 1 line.

OK.

[Saam Barati]

(In reply to comment #30)
> (In reply to comment #29)
> > Comment on attachment 273746 [details]
> > the patch
> > 
> > View in context:
> > https://bugs.webkit.org/attachment.cgi?id=273746&action=review
> > 
> > r=me
> > 
> > > Source/JavaScriptCore/ChangeLog:20
> > > +        properties. Previously it could only handle objects (but not arrays) and naed properties
> > 
> > typo: named
> 
> Fixed.
> 
> > 
> > > Source/JavaScriptCore/dfg/DFGLazyJSValue.cpp:120
> > > +                dataLog("Giving up because of size limit.\n");
> > 
> > remove.
> 
> Fixed.
> 
> > 
> > > Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:521
> > > +                            m_nodeIndex, origin, jsNumber(result.start), UntypedUse));
> > 
> > Why is this untyped use?
> 
> With the exception of the publicLength, vectorLength, and structure
> arguments to MaterializeNewObject, all of them are UntypedUse because
> they're just values you're storing into the object.  You can store any
> JSValue into an object.  Therefore, UntypedUse.
> 
> More importantly, its not legal to just use whatever use kind you want
> anywhere you want.  The signature of a DFG node is its NodeType and its
> UseKinds.  If you use a UseKind that the node type doesn't recognize, you
> break the compiler.  MaterializeNewObject does not recognize anything but
> UntypedUse for anything other than publicLength/vectorLength/structure.
> 
> In general, if there is a use edge from A (the user) to B (the thing being
> used), then you don't pick the use kind based on what B is - you pick it
> based on what A wants B to be.  If B is an int, you don't tell A to use it
> as an int unless A knows what to do with that information.
> 
Makes sense. Thanks for clarifying.

> > 
> > > Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:525
> > > +                    m_graph.m_varArgChildren.append(Edge(stringNode, UntypedUse));
> > 
> > Shouldn't this be StringUse?
> 
> See above.
> 
> > 
> > > Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:651
> > > +                int replLen = replace.length();
> > 
> > unsigned.
> 
> OK.  I just copied what the StringPrototype code was doing.
> 
> > 
> > > Source/JavaScriptCore/dfg/DFGStrengthReductionPhase.cpp:681
> > > +                        m_nodeIndex, origin, jsNumber(0), UntypedUse));
> > 
> > KnownInt32Use?
> 
> See above, you can't just pick whatever use kind you like.
> 
> > 
> > > Source/JavaScriptCore/ftl/FTLLowerDFGToB3.cpp:6397
> > > +                HashMap<
> > > +                    int32_t, LValue, DefaultHash<int32_t>::Hash,
> > > +                    WTF::UnsignedWithZeroKeyHashTraits<int32_t>> indexMap;
> > 
> > style: I think this can be 1 line.
> 
> OK.

[Filip Pizlo]

Created attachment 274169 [details]
rebased patch

[WebKit Commit Bot]

Attachment 274169 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1401:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
ERROR: Source/JavaScriptCore/runtime/StringPrototype.h:66:  The parameter name "reg" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 10 in 67 files


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

[Filip Pizlo]

Created attachment 274530 [details]
rebased patch

[WebKit Commit Bot]

Attachment 274530 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1414:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
ERROR: Source/JavaScriptCore/runtime/StringPrototype.h:66:  The parameter name "reg" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 10 in 67 files


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

[Filip Pizlo]

Created attachment 274929 [details]
rebased patch

[WebKit Commit Bot]

Attachment 274929 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1414:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
ERROR: Source/JavaScriptCore/runtime/StringPrototype.h:66:  The parameter name "reg" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 10 in 67 files


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

[Filip Pizlo]

Created attachment 275633 [details]
rebased patch

[WebKit Commit Bot]

Attachment 275633 [details] did not pass style-queue:


ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "exec" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/runtime/JSObject.h:102:  The parameter name "structure" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGInsertionSet.h:132:  The parameter name "block" adds no information, so it should be removed.  [readability/parameter_name] [5]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:39:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:40:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:41:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:43:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOpInfo.h:45:  Should be indented on a separate line, with the colon or comma first on that line.  [whitespace/indent] [4]
ERROR: Source/JavaScriptCore/dfg/DFGOperations.cpp:1414:  Declaration has space between type name and * in length * sizeof  [whitespace/declaration] [3]
ERROR: Source/JavaScriptCore/runtime/StringPrototype.h:66:  The parameter name "reg" adds no information, so it should be removed.  [readability/parameter_name] [5]
Total errors found: 10 in 67 files


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

[Filip Pizlo]

Landed in http://trac.webkit.org/changeset/199075