Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.
Bug 324384 - CDT editor hangs up when you execute auto completion
Summary: CDT editor hangs up when you execute auto completion
Status: RESOLVED FIXED
Alias: None
Product: CDT
Classification: Tools
Component: cdt-parser (show other bugs)
Version: 7.0   Edit
Hardware: PC Windows XP
: P3 critical (vote)
Target Milestone: 7.0.1   Edit
Assignee: Markus Schorn CLA
QA Contact: Mike Kucera CLA
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-09-03 05:07 EDT by Victor Yagusevich CLA
Modified: 2010-09-08 10:56 EDT (History)
1 user (show)

See Also:

Attachments
testcase + fix (2.31 KB, patch)
2010-09-06 04:15 EDT, Markus Schorn CLA 		mschorn.eclipse: iplog-
Details | Diff
fix + testcase (1.94 KB, patch)
2010-09-06 10:14 EDT, Markus Schorn CLA 		mschorn.eclipse: iplog-
Details | Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Victor Yagusevich CLA 2010-09-03 05:07:32 EDT 
Build Identifier: 20100617-1415

Auto completion hangs up for macros and functions but it works good for local variables. 
It is easy to reproduce for the following snippet. If you try to execute auto completion in the place marked by !{cursor}! Eclipse will hang up.
I tried it out with the latest release of CDT and even with CDT compiled from sources. This bug is reproduced on several machine and on different platforms (Windows XP, Windows 7).

#include <stdio.h>
#include "macros.h"

DBG_TAG_MODULE(MAIN)
DBG_TAG_!{cursor}!SUBMODULE(main)

int main() {
  return 0;
}

Content of my eclipse.ini
-startup
plugins/org.eclipse.equinox.launcher_1.1.0.v20100507.jar
--launcher.library
plugins/org.eclipse.equinox.launcher.win32.win32.x86_1.1.0.v20100503
-product
org.eclipse.epp.package.rcp.product
--launcher.defaultAction
openFile
--launcher.XXMaxPermSize
256M
-showsplash
org.eclipse.platform
--launcher.XXMaxPermSize
256m
--launcher.defaultAction
openFile
-vmargs
-Dosgi.requiredJavaVersion=1.5
-Xms40m
-Xmx256m

Reproducible: Always
Comment 1 Anton Leherbauer CLA 2010-09-03 08:54:59 EDT 
The parser seems to get into an endless loop in countKnRCParms:

Thread [main] (Suspended)	
	GNUCSourceParser.countKnRCParms() line: 1894	
	GNUCSourceParser.functionDeclarator(DeclarationOptions) line: 1490	
	GNUCSourceParser.declarator(List<IASTPointerOperator>, IASTName, IASTDeclarator, int, int, DeclarationOptions) line: 1408	
	GNUCSourceParser.declarator(IASTDeclSpecifier, DeclarationOptions) line: 1343	
	GNUCSourceParser.initDeclarator(IASTDeclSpecifier, DeclarationOptions) line: 1302	
	GNUCSourceParser(AbstractGNUSourceCodeParser).declSpecifierSequence_initDeclarator(DeclarationOptions, boolean) line: 1608	
	GNUCSourceParser.simpleDeclaration(DeclarationOptions) line: 309	
	GNUCSourceParser.declaration(DeclarationOptions) line: 292	
	GNUCSourceParser(AbstractGNUSourceCodeParser).declarationList(IASTDeclarationListOwner, DeclarationOptions, boolean, int) line: 1315	
	GNUCSourceParser(AbstractGNUSourceCodeParser).parseTranslationUnit() line: 1253	
	GNUCSourceParser(AbstractGNUSourceCodeParser).translationUnit() line: 1248	
	GNUCSourceParser(AbstractGNUSourceCodeParser).parse() line: 645	
	GCCLanguage(AbstractCLikeLanguage).getCompletionNode(FileContent, IScannerInfo, IncludeFileContentProvider, IIndex, IParserLogService, int) line: 171	
	WorkingCopy(TranslationUnit).getCompletionNode(IIndex, int, int) line: 897	
	CContentAssistInvocationContext.getCompletionNode() line: 159	
	DOMCompletionProposalComputer(ParsingBasedProposalComputer).computeCompletionProposals(ContentAssistInvocationContext, IProgressMonitor) line: 47	
	CompletionProposalComputerDescriptor.computeCompletionProposals(ContentAssistInvocationContext, IProgressMonitor) line: 300	
	CompletionProposalCategory.computeCompletionProposals(ContentAssistInvocationContext, String, SubProgressMonitor) line: 261	
	CContentAssistProcessor(ContentAssistProcessor).collectProposals(ITextViewer, int, IProgressMonitor, ContentAssistInvocationContext) line: 267	
	CContentAssistProcessor(ContentAssistProcessor).computeCompletionProposals(ITextViewer, int) line: 220	
	ContentAssistant.computeCompletionProposals(ITextViewer, int) line: 1834	
	CompletionProposalPopup.computeProposals(int) line: 556	
	CompletionProposalPopup.access$16(CompletionProposalPopup, int) line: 553	
	CompletionProposalPopup$14.run() line: 1542	
	BusyIndicator.showWhile(Display, Runnable) line: 70	
	CompletionProposalPopup.incrementalComplete() line: 1536	
	ContentAssistant.showPossibleCompletions() line: 1659	
	CEditor$AdaptedSourceViewer.doOperation(int) line: 328	
	ContentAssistAction$1.run() line: 82	
	BusyIndicator.showWhile(Display, Runnable) line: 70	
	ContentAssistAction.run() line: 80	
	ContentAssistAction(Action).runWithEvent(Event) line: 498	
	ActionHandler.execute(Map) line: 185	
	LegacyHandlerWrapper.execute(ExecutionEvent) line: 109	
	Command.executeWithChecks(ExecutionEvent) line: 476	
	ParameterizedCommand.executeWithChecks(Object, Object) line: 508	
	HandlerService.executeCommand(ParameterizedCommand, Event) line: 169	
	WorkbenchKeyboard.executeCommand(Binding, Event) line: 468	
	WorkbenchKeyboard.press(List, Event) line: 786	
	WorkbenchKeyboard.processKeyEvent(List, Event) line: 885	
	WorkbenchKeyboard.filterKeySequenceBindings(Event) line: 567	
	WorkbenchKeyboard.access$3(WorkbenchKeyboard, Event) line: 508	
	WorkbenchKeyboard$KeyDownFilter.handleEvent(Event) line: 123	
	EventTable.sendEvent(Event) line: 84	
	Display.filterEvent(Event) line: 1253	
	StyledText(Widget).sendEvent(Event) line: 1052	
	StyledText(Widget).sendEvent(int, Event, boolean) line: 1077	
	StyledText(Widget).sendEvent(int, Event) line: 1062	
	StyledText(Widget).sendKeyEvent(int, int, int, int, Event) line: 1103	
	StyledText(Widget).sendKeyEvent(int, int, int, int) line: 1099	
	StyledText(Widget).wmChar(int, int, int) line: 1508	
	StyledText(Control).WM_CHAR(int, int) line: 4268	
	StyledText(Canvas).WM_CHAR(int, int) line: 345	
	StyledText(Control).windowProc(int, int, int, int) line: 4160	
	StyledText(Canvas).windowProc(int, int, int, int) line: 341	
	Display.windowProc(int, int, int, int) line: 4873	
	OS.DispatchMessageW(MSG) line: not available [native method]	
	OS.DispatchMessage(MSG) line: 2459	
	Display.readAndDispatch() line: 3655	
	Workbench.runEventLoop(Window$IExceptionHandler, Display) line: 2629	
	Workbench.runUI() line: 2593	
	Workbench.access$4(Workbench) line: 2427	
	Workbench$7.run() line: 670	
	Realm.runWithDefault(Realm, Runnable) line: 332	
	Workbench.createAndRunWorkbench(Display, WorkbenchAdvisor) line: 663	
	PlatformUI.createAndRunWorkbench(Display, WorkbenchAdvisor) line: 149	
	IDEApplication.start(IApplicationContext) line: 115	
	EclipseAppHandle.run(Object) line: 196	
	EclipseAppLauncher.runApplication(Object) line: 110	
	EclipseAppLauncher.start(Object) line: 79	
	EclipseStarter.run(Object) line: 369	
	EclipseStarter.run(String[], Runnable) line: 179	
	NativeMethodAccessorImpl.invoke0(Method, Object, Object[]) line: not available [native method]	
	NativeMethodAccessorImpl.invoke(Object, Object[]) line: 39	
	DelegatingMethodAccessorImpl.invoke(Object, Object[]) line: 25	
	Method.invoke(Object, Object...) line: 597	
	Main.invokeFramework(String[], URL[]) line: 619	
	Main.basicRun(String[]) line: 574	
	Main.run(String[]) line: 1407	
	Main.main(String[]) line: 1383
Comment 2 Markus Schorn CLA 2010-09-06 04:15:05 EDT 
Created attachment 178242 [details]
testcase + fix

Note, that there is something going wrong with the expansion of the macro. I guess that either the include file is not resolved or that the macro-definition is hidden in an inactive code branch. You probably have to adjust the configuration (include search path, defined symbols) of the indexer.

Without the macro expansion the construct is interpreted as the beginning of a K&R-style function definition. With that the parser runs into an infinite loop.
The provided patch fixes that.
Comment 4 Victor Yagusevich CLA 2010-09-06 09:48:58 EDT 
I've applied the attached patch but there is the exception when executing code completion for the following code (as usual !{cursor}! is the cursor position):

DBG_TAG_MODULE(AcMy)

int __AcMy_Init(P!{cursor}!AcMy self);

Exception stack:

eclipse.buildId=I20100608-0911
java.version=1.6.0_21
java.vendor=Sun Microsystems Inc.
BootLoader constants: OS=win32, ARCH=x86, WS=win32, NL=uk_UA
Framework arguments:  -product org.eclipse.platform.ide
Command-line arguments:  -product org.eclipse.platform.ide -data D:\Projects\Eclipse_plugins_helios/../runtime-EclipseApplication -dev file:D:/Projects/Eclipse_plugins_helios/.metadata/.plugins/org.eclipse.pde.core/Eclipse Application/dev.properties -os win32 -ws win32 -arch x86 -consoleLog


Error
Mon Sep 06 16:47:38 EEST 2010
Unhandled event loop exception

java.lang.OutOfMemoryError: Java heap space
at org.eclipse.cdt.internal.core.parser.scanner.CPreprocessor.nextToken(CPreprocessor.java:580)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.fetchToken(AbstractGNUSourceCodeParser.java:260)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.nextToken(AbstractGNUSourceCodeParser.java:284)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.lookaheadToken(AbstractGNUSourceCodeParser.java:294)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.LA(AbstractGNUSourceCodeParser.java:317)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.LT(AbstractGNUSourceCodeParser.java:444)
at org.eclipse.cdt.internal.core.dom.parser.c.GNUCSourceParser.createKnRCProblemDeclaration(GNUCSourceParser.java:1937)
at org.eclipse.cdt.internal.core.dom.parser.c.GNUCSourceParser.functionDeclarator(GNUCSourceParser.java:1535)
at org.eclipse.cdt.internal.core.dom.parser.c.GNUCSourceParser.declarator(GNUCSourceParser.java:1413)
at org.eclipse.cdt.internal.core.dom.parser.c.GNUCSourceParser.declarator(GNUCSourceParser.java:1348)
at org.eclipse.cdt.internal.core.dom.parser.c.GNUCSourceParser.initDeclarator(GNUCSourceParser.java:1307)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.declSpecifierSequence_initDeclarator(AbstractGNUSourceCodeParser.java:1616)
at org.eclipse.cdt.internal.core.dom.parser.c.GNUCSourceParser.simpleDeclaration(GNUCSourceParser.java:309)
at org.eclipse.cdt.internal.core.dom.parser.c.GNUCSourceParser.declaration(GNUCSourceParser.java:292)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.declarationList(AbstractGNUSourceCodeParser.java:1323)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.parseTranslationUnit(AbstractGNUSourceCodeParser.java:1261)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.translationUnit(AbstractGNUSourceCodeParser.java:1256)
at org.eclipse.cdt.internal.core.dom.parser.AbstractGNUSourceCodeParser.parse(AbstractGNUSourceCodeParser.java:645)
at org.eclipse.cdt.core.dom.parser.AbstractCLikeLanguage.getCompletionNode(AbstractCLikeLanguage.java:171)
at org.eclipse.cdt.internal.core.model.TranslationUnit.getCompletionNode(TranslationUnit.java:888)
at org.eclipse.cdt.internal.ui.text.contentassist.CContentAssistInvocationContext.getCompletionNode(CContentAssistInvocationContext.java:159)
at org.eclipse.cdt.internal.ui.text.contentassist.ParsingBasedProposalComputer.computeCompletionProposals(ParsingBasedProposalComputer.java:47)
at org.eclipse.cdt.internal.ui.text.contentassist.CompletionProposalComputerDescriptor.computeCompletionProposals(CompletionProposalComputerDescriptor.java:300)
at org.eclipse.cdt.internal.ui.text.contentassist.CompletionProposalCategory.computeCompletionProposals(CompletionProposalCategory.java:261)
at org.eclipse.cdt.internal.ui.text.contentassist.ContentAssistProcessor.collectProposals(ContentAssistProcessor.java:267)
at org.eclipse.cdt.internal.ui.text.contentassist.ContentAssistProcessor.computeCompletionProposals(ContentAssistProcessor.java:220)
at org.eclipse.jface.text.contentassist.ContentAssistant.computeCompletionProposals(ContentAssistant.java:1834)
at org.eclipse.jface.text.contentassist.CompletionProposalPopup.computeProposals(CompletionProposalPopup.java:556)
at org.eclipse.jface.text.contentassist.CompletionProposalPopup.access$16(CompletionProposalPopup.java:553)
at org.eclipse.jface.text.contentassist.CompletionProposalPopup$14.run(CompletionProposalPopup.java:1542)
at org.eclipse.swt.custom.BusyIndicator.showWhile(BusyIndicator.java:70)
at org.eclipse.jface.text.contentassist.CompletionProposalPopup.incrementalComplete(CompletionProposalPopup.java:1536)
Comment 5 Markus Schorn CLA 2010-09-06 10:14:10 EDT 
Created attachment 178267 [details]
fix + testcase

Thanks, this patch adresses the second issue.
Comment 6 Markus Schorn CLA 2010-09-06 10:16:47 EDT 
Fixed in 7.0.1 and 8.0 > 20100906.
Comment 9 John Liu CLA 2010-09-07 14:48:40 EDT 
(In reply to comment #5)
> Created an attachment (id=178267) [details]
> fix + testcase
> Thanks, this patch adresses the second issue.

Hi, Markus:

For the test case testIncompleteKnrFunction_Bug324384b, the LR parser returns a proposal of CImplicitTypedef with a name "__builtin_va_list", so that it is failed.

The parser return a completion node with an empty ASTName, which is bound to   org.eclipse.cdt.internal.core.dom.parser.c.CImplicitTypedef later, so proposal is returned as CImplicitTypedef.

I am wondering if it is correct too?
Comment 10 Markus Schorn CLA 2010-09-08 02:17:21 EDT 
(In reply to comment #9)
> (In reply to comment #5)
> > Created an attachment (id=178267) [details] [details]
> > fix + testcase
> > Thanks, this patch adresses the second issue.
> Hi, Markus:
> For the test case testIncompleteKnrFunction_Bug324384b, the LR parser returns a
> proposal of CImplicitTypedef with a name "__builtin_va_list", so that it is
> failed.
> The parser return a completion node with an empty ASTName, which is bound to  
> org.eclipse.cdt.internal.core.dom.parser.c.CImplicitTypedef later, so proposal
> is returned as CImplicitTypedef.
> I am wondering if it is correct too?

Because the example contains invalid syntax, that's certainly ok. I'll slightly change the test-case such that it passes for the LR-parser, also.
Comment 11 Markus Schorn CLA 2010-09-08 06:01:26 EDT 
(In reply to comment #10)
> ...
> Because the example contains invalid syntax, that's certainly ok. I'll 
> slightly change the test-case such that it passes for the LR-parser, also.

I have adjusted the testcase such that there is a prefix for the completion. With that the proposal __builtin.. is wrong. Nevertheless with the LR parser it is computed. I assume that the LR parser erroneously returns a completion name without the prefix.
Comment 13 Victor Yagusevich CLA 2010-09-08 08:53:39 EDT 
I've applied the second patch two and noticed more bugs. I created a source file with only one line:

#include "my.h"

Bug I: Auto-completion writes additional symbols '__' before show proposals.
#include !{cursor}!"my.h" ==> #include __!{proposals list}!"my.h"

Bug II: Auto-completion writes characters 'de' and doesn't show proposals
#inclu!{cursor}!de "my.h" ==> #include !{cursor}!de "my.h"
Comment 14 John Liu CLA 2010-09-08 09:48:03 EDT 
(In reply to comment #11)
> (In reply to comment #10)
> > ...
> > Because the example contains invalid syntax, that's certainly ok. I'll 
> > slightly change the test-case such that it passes for the LR-parser, also.
> I have adjusted the testcase such that there is a prefix for the completion.
> With that the proposal __builtin.. is wrong. Nevertheless with the LR parser it
> is computed. I assume that the LR parser erroneously returns a completion name
> without the prefix.

Thanks, Markus. The updated test case is passed by LR parser in my workspace testing.
Comment 15 Markus Schorn CLA 2010-09-08 10:42:42 EDT 
(In reply to comment #13)
> I've applied the second patch two and noticed more bugs. I created a source
> file with only one line:
> #include "my.h"
> Bug I: Auto-completion writes additional symbols '__' before show proposals.
> #include !{cursor}!"my.h" ==> #include __!{proposals list}!"my.h"
> Bug II: Auto-completion writes characters 'de' and doesn't show proposals
> #inclu!{cursor}!de "my.h" ==> #include !{cursor}!de "my.h"

These issues are unreated, please create a separate bug report for that.
Comment 16 Victor Yagusevich CLA 2010-09-08 10:56:34 EDT 
(In reply to comment #15)
> (In reply to comment #13)
> These issues are unreated, please create a separate bug report for that.

OK. I've created bug 324754.