WebKit Bugzilla
New
Browse
Search+
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED INVALID
35642
[Qt] Memory leak detected during widget launch, removal testing
https://bugs.webkit.org/show_bug.cgi?id=35642
Summary
[Qt] Memory leak detected during widget launch, removal testing
Tor Arne Vestbø
Reported
2010-03-03 03:28:41 PST
This bug report originated from issue QTBUG-6788
http://bugreports.qt.nokia.com/browse/QTBUG-6788
--- Description --- ===========<br/> Memory leak observed in the WRT and QWebkit Module. <p>Precondition:<br/> =============<br/> ---------</p> <p>Steps Leading to the problem:<br/> ============================</p> <p>1.Install a widget with the valid manifest file ( path to for the widget<br/> <span class="nobr"><a href="
http://waplabdc.nokia-boston.com/browser/users/amitha/widget/valid.wgz
">
http://waplabdc.nokia-boston.com/browser/users/amitha/widget/valid.wgz
<sup><img class="rendericon" src="
http://bugreports.qt.nokia.com/images/icons/linkext7.gif
" height="7" width="7" align="absmiddle" alt="" border="0"/></sup></a></span> )</p> <p>2.launch the widget with the valgrind options as follows:<br/> valgrind --log-file=/tmp/leaks.log --tool=memcheck --leak-check=yes<br/> --show-reachable=yes --num-callers=40 --track-fds=yes ./nokiaPhoneSimulator<br/> widgets</p> <p>3.Exit the widget</p> <p>(after this step you can find the log file in /tmp folder as .log extension)</p> <p>Expected Outcome:<br/> =================<br/> No memory leaks should be found.</p> <p>Actual Outcome:<br/> ===============<br/> ==12292== 41,424 (12,160 direct, 29,264 indirect) bytes in 44 blocks are<br/> definitely lost in loss record 436 of 581<br/> ==12292== at 0x4022B8E: realloc (vg_replace_malloc.c:429)<br/> ==12292== by 0x651BFD3: (within /usr/lib/libfontconfig.so.1.3.0)<br/> ==12292== by 0x650D3A0: (within /usr/lib/libfontconfig.so.1.3.0)<br/> ==12292== by 0x650E663: FcConfigSubstituteWithPat (in<br/> /usr/lib/libfontconfig.so.1.3.0)<br/> ==12292== by 0x650E8A6: FcConfigSubstitute (in<br/> /usr/lib/libfontconfig.so.1.3.0)<br/> ==12292== by 0x5C31F87: getFcPattern(QFontPrivate const*, int, QFontDef<br/> const&) (qfontdatabase_x11.cpp:1541)<br/> ==12292== by 0x5C322F9: loadFc(QFontPrivate const*, int, QFontDef const&)<br/> (qfontdatabase_x11.cpp:1691)<br/> ==12292== by 0x5C33D65: QFontDatabase::load(QFontPrivate const*, int)<br/> (qfontdatabase_x11.cpp:1927)<br/> ==12292== by 0x5C0E158: QFontPrivate::engineForScript(int) const<br/> (qfont.cpp:265)<br/> ==12292== by 0x5C25F02: QFontMetrics::ascent() const (qfontmetrics.cpp:300)<br/> ==12292== by 0x49EC8B4: WebCore::SimpleFontData::platformInit()<br/> (SimpleFontDataQt.cpp:45)<br/> ==12292== by 0x48493C6:<br/> WebCore::SimpleFontData::SimpleFontData(WebCore::FontPlatformData const&, bool,<br/> bool, WebCore::SVGFontData*) (SimpleFontData.cpp:101)<br/> ==12292== by 0x49EC3FF: WebCore::FontFallbackList::fontDataAt(WebCore::Font<br/> const*, unsigned) const (FontFallbackListQt.cpp:104)<br/> ==12292== by 0x483989B:<br/> WebCore::FontFallbackList::primaryFontData(WebCore::Font const*) const<br/> (FontFallbackList.h:69)<br/> ==12292== by 0x49EC4B3:<br/> WebCore::FontFallbackList::determinePitch(WebCore::Font const*) const<br/> (FontFallbackListQt.cpp:68)<br/> ==12292== by 0x483919D:<br/> WebCore::FontFallbackList::isFixedPitch(WebCore::Font const*) const<br/> (FontFallbackList.h:50)<br/> ==12292== by 0x4839232: WebCore::Font::isFixedPitch() const (Font.cpp:160)<br/> ==12292== by 0x48AF1CB:<br/> WebCore::RenderBlock::findNextLineBreak(WebCore::BidiResolver<WebCore::InlineIterator,<br/> WebCore::BidiRun>&, bool, bool&, bool&, WebCore::EClear*) (bidi.cpp:1777)<br/> ==12292== by 0x48B3884: WebCore::RenderBlock::layoutInlineChildren(bool,<br/> int&, int&) (bidi.cpp:945)<br/> ==12292== by 0x48EFBD3: WebCore::RenderBlock::layoutBlock(bool)<br/> (RenderBlock.cpp:779)<br/> ==12292== by 0x48CB2C9: WebCore::RenderBlock::layout() (RenderBlock.cpp:698)<br/> ==12292== by 0x48EE5BF: WebCore::RenderBlock::layoutBlockChildren(bool,<br/> int&) (RenderBlock.cpp:1373)<br/> ==12292== by 0x48EFBF5: WebCore::RenderBlock::layoutBlock(bool)<br/> (RenderBlock.cpp:781)<br/> ==12292== by 0x48CB2C9: WebCore::RenderBlock::layout() (RenderBlock.cpp:698)<br/> ==12292== by 0x48EE5BF: WebCore::RenderBlock::layoutBlockChildren(bool,<br/> int&) (RenderBlock.cpp:1373)<br/> ==12292== by 0x48EFBF5: WebCore::RenderBlock::layoutBlock(bool)<br/> (RenderBlock.cpp:781)<br/> ==12292== by 0x48CB2C9: WebCore::RenderBlock::layout() (RenderBlock.cpp:698)<br/> ==12292== by 0x498D574: WebCore::RenderView::layout() (RenderView.cpp:122)<br/> ==12292== by 0x4811464: WebCore::FrameView::layout(bool) (FrameView.cpp:610)<br/> ==12292== by 0x4811CB6:<br/> WebCore::FrameView::layoutTimerFired(WebCore::Timer<WebCore::FrameView>*)<br/> (FrameView.cpp:924)<br/> ==12292== by 0x480A54C: WebCore::Timer<WebCore::FrameView>::fired()<br/> (Timer.h:98)<br/> ==12292== by 0x4885052: WebCore::ThreadTimers::fireTimers(double,<br/> WTF::Vector<WebCore::TimerBase*, 0> const&) (ThreadTimers.cpp:111)<br/> ==12292== by 0x4885B46: WebCore::ThreadTimers::sharedTimerFiredInternal()<br/> (ThreadTimers.cpp:141)<br/> ==12292== by 0x4885BC0: WebCore::ThreadTimers::sharedTimerFired()<br/> (ThreadTimers.cpp:122)<br/> ==12292== by 0x49F8001: WebCore::SharedTimerQt::timerEvent(QTimerEvent*)<br/> (SharedTimerQt.cpp:105)<br/> ==12292== by 0x6A7A809: QObject::event(QEvent*) (qobject.cpp:1073)<br/> ==12292== by 0x599F406: QApplicationPrivate::notify_helper(QObject*,<br/> QEvent*) (qapplication.cpp:4057)<br/> ==12292== by 0x599F77F: QApplication::notify(QObject*, QEvent*)<br/> (qapplication.cpp:3604)<br/> ==12292== by 0x6A63387: QCoreApplication::notifyInternal(QObject*, QEvent*)<br/> (qcoreapplication.cpp:610)<br/> ==12292== by 0x8071FF7: QCoreApplication::sendEvent(QObject*, QEvent*)<br/> (qcoreapplication.h:213)</p>
Attachments
Add attachment
proposed patch, testcase, etc.
Andreas Kling
Comment 1
2010-03-17 04:54:25 PDT
Not a leak, but rather a known issue with Fontconfig. See
https://bugs.freedesktop.org/show_bug.cgi?id=18301
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug