Valgrind Messages - Revision history

Ethryn: add to category (Engine documents)

2011-03-17T00:29:58Z

add to category (Engine documents)

← Older revision		Revision as of 00:29, 17 March 2011
Line 168:		Line 168:

	In sprcal3d.cpp line 568. Memory allocated with a custom "new" can't be deallocated with a delete. The fix is to delete the custom part of new. This is crystal space issue [http://trac.crystalspace3d.org/trac/CS/ticket/315 315].		In sprcal3d.cpp line 568. Memory allocated with a custom "new" can't be deallocated with a delete. The fix is to delete the custom part of new. This is crystal space issue [http://trac.crystalspace3d.org/trac/CS/ticket/315 315].

			[[Category:Engine documents]]

Unkermouse: /* in libX11 */

2010-12-11T15:45:43Z

in libX11

← Older revision		Revision as of 15:45, 11 December 2010
Line 32:		Line 32:
	uginRequest>, CS::Memory::AllocatorMalloc, csArrayCapacityDefault> const&) (initapp.cpp:403)		uginRequest>, CS::Memory::AllocatorMalloc, csArrayCapacityDefault> const&) (initapp.cpp:403)
	</pre>		</pre>
	I don't know what this message total means, it is not in [[PlaneShift]] or [[~~Crystal Space~~]] and a variable unitalized passed to writev() shouldn't really be an error becasue it is writing memory to this location and not reading from it. It happens on starup.		I don't know what this message total means, it is not in [[PlaneShift]] or [[CrystalSpace]] and a variable unitalized passed to writev() shouldn't really be an error becasue it is writing memory to this location and not reading from it. It happens on starup.

	====nvidia====		====nvidia====

Josellis: /* in libX11 */ changed "PS" and "CS" to "PlaneShift" and "Crystal Space"

2007-04-28T11:00:20Z

in libX11: changed "PS" and "CS" to "PlaneShift" and "Crystal Space"

← Older revision		Revision as of 11:00, 28 April 2007
Line 32:		Line 32:
	uginRequest>, CS::Memory::AllocatorMalloc, csArrayCapacityDefault> const&) (initapp.cpp:403)		uginRequest>, CS::Memory::AllocatorMalloc, csArrayCapacityDefault> const&) (initapp.cpp:403)
	</pre>		</pre>
	I don't know what this message total means, it is not in [[PS]] or [[CS]] and a variable unitalized passed to writev() shouldn't really be an error becasue it is writing memory to this location and not reading from it. It happens on starup.		I don't know what this message total means, it is not in [[PlaneShift]] or [[Crystal Space]] and a variable unitalized passed to writev() shouldn't really be an error becasue it is writing memory to this location and not reading from it. It happens on starup.

	====nvidia====		====nvidia====

Brandon rioja: /* crystal space */

2007-04-14T22:10:10Z

crystal space

← Older revision		Revision as of 22:10, 14 April 2007
Line 123:		Line 123:
	<res2k> besides, freefnt2 is not the only font server		<res2k> besides, freefnt2 is not the only font server
	</pre>		</pre>
			===in cal3d===
			<pre>
			==6802== Mismatched free() / delete / delete []
			==6802== at 0x4004D31: operator delete(void*) (vg_replace_malloc.c:244)
			==6802== by 0x7BBF1A4: CalCoreMaterial::~CalCoreMaterial() (corematerial.h:40)
			==6802== by 0x7BB8635: cal3d::RefCounted::decRef() (refcounted.h:69)
			==6802== by 0x7BB8658: cal3d::explicitDecRef(cal3d::RefCounted*) (refcounted.h:99)
			==6802== by 0x7BB86F7: cal3d::RefPtr<CalCoreMaterial>::~RefPtr() (refptr.h:32)
			==6802== by 0x7BC88E0: void std::_Destroy<cal3d::RefPtr<CalCoreMaterial> >(cal3d::RefPtr<CalCoreMaterial>*) (stl_construct
			.h:107)
			==6802== by 0x7BC89B0: void std::__destroy_aux<cal3d::RefPtr<CalCoreMaterial>>(cal3d::RefPtr<CalCoreMaterial>, cal3d::Re
			fPtr<CalCoreMaterial>*, __false_type) (stl_construct.h:122)
			==6802== by 0x7BC89EF: void std::_Destroy<cal3d::RefPtr<CalCoreMaterial>>(cal3d::RefPtr<CalCoreMaterial>, cal3d::RefPtr<
			CalCoreMaterial>*) (stl_construct.h:155)
			==6802== by 0x7BC8A19: void std::_Destroy<cal3d::RefPtr<CalCoreMaterial>*, cal3d::RefPtr<CalCoreMaterial> >(cal3d::RefPtr<
			CalCoreMaterial>, cal3d::RefPtr<CalCoreMaterial>, std::allocator<cal3d::RefPtr<CalCoreMaterial> >) (stl_construct.h:182)
			==6802== by 0x7BCBE8E: std::vector<cal3d::RefPtr<CalCoreMaterial>, std::allocator<cal3d::RefPtr<CalCoreMaterial> > >::~vec
			tor() (stl_vector.h:272)
			==6802== by 0x7BC20BA: CalCoreModel::~CalCoreModel() (coremodel.cpp:58)
			==6802== by 0x7AC8696: CS::Plugin::SprCal3d::csSpriteCal3DMeshObjectFactory::~csSpriteCal3DMeshObjectFactory() (sprcal3d.c
			pp:192)
			==6802== Address 0x40A6630 is 0 bytes inside a block of size 48 alloc'd
			==6802== at 0x4005400: malloc (vg_replace_malloc.c:149)
			==6802== by 0x7AE69C0: operator new(unsigned, CS::AllocPlatform const&) (platform_new.cpp:45)
			==6802== by 0x7AC62E2: CS::Plugin::SprCal3d::csSpriteCal3DMeshObjectFactory::AddCoreMaterial(iMaterialWrapper*) (sprcal3d.
			cpp:568)
			==6802== by 0x7A8D6F3: CS::Plugins::SprCal3dLoader::csSpriteCal3DFactoryLoader::LoadMaterialTag(iSpriteCal3DFactoryState*,
			iDocumentNode, iLoaderContext, char const, char const) (sprcal3dldr.cpp:485)
			==6802== by 0x7A8F89C: CS::Plugins::SprCal3dLoader::csSpriteCal3DFactoryLoader::Parse(iDocumentNode, iStreamSource, iLoa
			derContext, iBase) (sprcal3dldr.cpp:332)
			==6802== by 0x4A432DF: csLoader::LoadMeshObjectFactory(iLoaderContext, iMeshFactoryWrapper, iMeshFactoryWrapper*, iDocum
			entNode, csReversibleTransform, iStreamSource*) (csloader.cpp:2121)
			==6802== by 0x4A4C3DB: csLoader::Load(iDocumentNode, iBase&, iRegion, bool, bool, iStreamSource, char const*, iMissing
			LoaderData*) (csloader.cpp:776)
			==6802== by 0x4A3B88C: csLoader::Load(iDataBuffer, char const, iBase&, iRegion, bool, bool, iStreamSource*, char const
			, iMissingLoaderData) (csloader.cpp:682)
			==6802== by 0x4A3BD75: csLoader::Load(char const, iBase&, iRegion, bool, bool, iStreamSource, char const*, iMissingLoa
			derData*) (csloader.cpp:742)
			==6802== by 0x804E66A: ClientCacheManager::LoadNewFactory(char const*) (clientcachemanager.cpp:91)
			==6802== by 0x8068E44: psEngine::PreloadModels() (psengine.cpp:1360)
			==6802== by 0x805DE26: psEngine::HandleEvent(iEvent&) (psengine.cpp:712)
			==6802==
			</pre>

			In sprcal3d.cpp line 568. Memory allocated with a custom "new" can't be deallocated with a delete. The fix is to delete the custom part of new. This is crystal space issue [http://trac.crystalspace3d.org/trac/CS/ticket/315 315].

Brandon rioja: /* crystal space */

2007-04-14T18:58:44Z

crystal space

@@ Line 106: / Line 106: @@
      by 0x80CD3EE: psEntityLabels::CreateLabelOfObject(GEMClientObject*) (entitylabels.cpp:300)
 </pre>
-This is a bug in crystal space's code, (unfixed at the time of the writing.) glfontcache.cpp line 447 csGLFontCache::CopyGlyphData(). The for loop reads a line ahead at the end of the loop, and assigns it to dest value on the next iteration.  The problem is when it reaches the end of the array, the memory reads ahead to a non allocated area in memory.
+This is a bug in crystal space's code in glfontcache.cpp line 447 csGLFontCache::CopyGlyphData(). Crystal space ticket [http://trac.crystalspace3d.org/trac/CS/ticket/314 314].

Brandon rioja: /* crystal space */

2007-04-14T18:37:55Z

crystal space

← Older revision		Revision as of 18:37, 14 April 2007
Line 106:		Line 106:
	by 0x80CD3EE: psEntityLabels::CreateLabelOfObject(GEMClientObject*) (entitylabels.cpp:300)		by 0x80CD3EE: psEntityLabels::CreateLabelOfObject(GEMClientObject*) (entitylabels.cpp:300)
	</pre>		</pre>
	This is a bug in crystal space's code, (unfixed at the time of the writing.) glfontcache.cpp line 447 csGLFontCache::CopyGlyphData(). The for loop reads a line ahead at the end of the loop, and assigns it to dest value on the next iteration. The problem is when it reaches the end of the array, the memory reads ahead to a non allocated area in memory~~. [[brandon_rioja]] suspectes that this can be the cause of [[Mononoke]]'s [http://pojebana.ovecka.be/~social/bt_onjoin crash]~~.		This is a bug in crystal space's code, (unfixed at the time of the writing.) glfontcache.cpp line 447 csGLFontCache::CopyGlyphData(). The for loop reads a line ahead at the end of the loop, and assigns it to dest value on the next iteration. The problem is when it reaches the end of the array, the memory reads ahead to a non allocated area in memory.

Brandon rioja at 18:33, 14 April 2007

2007-04-14T18:33:50Z

← Older revision		Revision as of 18:33, 14 April 2007
Line 69:		Line 69:
	</pre>		</pre>
	There are some errors in nvidia. Not much can be done here...		There are some errors in nvidia. Not much can be done here...
			===crystal space===
			<pre>
			Invalid read of size 1
			at 0x462227E: csGLFontCache::CopyGlyphData(iFont*, unsigned, unsigned, csBitmapMetrics const&, csRect const&, iDa
			taBuffer, iDataBuffer) (glfontcache.cpp:447)
			by 0x4622EC8: csGLFontCache::InternalCacheGlyph(csFontCache::KnownFont*, unsigned, unsigned) (glfontcache.cpp:336
			)
			by 0x4665B8E: csFontCache::CacheGlyphUnsafe(csFontCache::KnownFont*, unsigned, unsigned) (fontcache.cpp:423)
			by 0x462161F: csGLFontCache::WriteString(iFont, int, int, int, int, void const, bool, unsigned) (glfontcache.cp
			p:696)
			by 0x466AAE8: csGraphics2D::Write(iFont, int, int, int, int, char const, unsigned) (graph2d.cpp:827)
			by 0x837FA24: psEffectObjText::DrawTextElement(psEffectTextElement const&) (pseffectobjtext.cpp:305)
			by 0x83807D1: psEffectObjText::SetText(csArray<psEffectTextElement, csArrayElementHandler<psEffectTextElement>, C
			S::Memory::AllocatorMalloc, csArrayCapacityDefault> const&) (pseffectobjtext.cpp:121)
			by 0x8380A88: psEffectObjText::SetText(int, ...) (pseffectobjtext.cpp:205)
			by 0x80CD0EF: psEntityLabels::SetObjectText(GEMClientObject*) (entitylabels.cpp:272)
			by 0x80CD3EE: psEntityLabels::CreateLabelOfObject(GEMClientObject*) (entitylabels.cpp:300)
			by 0x80CD526: psEntityLabels::OnObjectArrived(GEMClientObject*) (entitylabels.cpp:332)
			by 0x808EDEE: GEMClientActor::GEMClientActor(psCelClient*, psPersistActor&) (pscelclient.cpp:1046)
			Address 0x56EA769 is 0 bytes after a block of size 497 alloc'd
			at 0x4005835: operator new[](unsigned) (vg_replace_malloc.c:195)
			by 0x44B739A: CS::Plugin::FreeFont2::csFreeType2Font::GetGlyphBitmap(unsigned, csBitmapMetrics&) (freefnt2.cpp:44
			0)
			by 0x4508BAF: CS::Plugin::FontPlex::csFontPlexer::GetGlyphBitmap(unsigned, csBitmapMetrics&) (fontplex.cpp:461)
			by 0x46226E2: csGLFontCache::InternalCacheGlyph(csFontCache::KnownFont*, unsigned, unsigned) (glfontcache.cpp:229
			)
			by 0x4665B8E: csFontCache::CacheGlyphUnsafe(csFontCache::KnownFont*, unsigned, unsigned) (fontcache.cpp:423)
			by 0x462161F: csGLFontCache::WriteString(iFont, int, int, int, int, void const, bool, unsigned) (glfontcache.cp
			p:696)
			by 0x466AAE8: csGraphics2D::Write(iFont, int, int, int, int, char const, unsigned) (graph2d.cpp:827)
			by 0x837FA24: psEffectObjText::DrawTextElement(psEffectTextElement const&) (pseffectobjtext.cpp:305)
			by 0x83807D1: psEffectObjText::SetText(csArray<psEffectTextElement, csArrayElementHandler<psEffectTextElement>, C
			S::Memory::AllocatorMalloc, csArrayCapacityDefault> const&) (pseffectobjtext.cpp:121)
			by 0x8380A88: psEffectObjText::SetText(int, ...) (pseffectobjtext.cpp:205)
			by 0x80CD0EF: psEntityLabels::SetObjectText(GEMClientObject*) (entitylabels.cpp:272)
			by 0x80CD3EE: psEntityLabels::CreateLabelOfObject(GEMClientObject*) (entitylabels.cpp:300)
			</pre>
			This is a bug in crystal space's code, (unfixed at the time of the writing.) glfontcache.cpp line 447 csGLFontCache::CopyGlyphData(). The for loop reads a line ahead at the end of the loop, and assigns it to dest value on the next iteration. The problem is when it reaches the end of the array, the memory reads ahead to a non allocated area in memory. [[brandon_rioja]] suspectes that this can be the cause of [[Mononoke]]'s [http://pojebana.ovecka.be/~social/bt_onjoin crash].

Brandon rioja at 16:13, 14 April 2007

2007-04-14T16:13:43Z

@@ Line 1: / Line 1: @@
 This page is a dump of all the valgrind messages I see:
+==Linux==
-==In psclient==
+===In psclient===
-===in libX11===
+====in libX11====
 <pre>
 Syscall param writev(vector[...]) points to uninitialised byte(s)
@@ Line 33: / Line 33: @@
 </pre>
 I don't know what this message total means, it is not in [[PS]] or [[CS]] and a variable unitalized passed to writev() shouldn't really be an error becasue it is writing memory to this location and not reading from it. It happens on starup.

Brandon rioja at 15:54, 14 April 2007

2007-04-14T15:54:23Z

New page

This page is a dump of all the valgrind messages I see:

==In psclient==
===in libX11===
<pre>
Syscall param writev(vector[...]) points to uninitialised byte(s)
at 0xD31B78: writev (in /lib/libc-2.5.so)
by 0x15098D: (within /usr/lib/libX11.so.6.2.0)
by 0x15077E: _X11TransWritev (in /usr/lib/libX11.so.6.2.0)
by 0x156468: _XSend (in /usr/lib/libX11.so.6.2.0)
by 0x14738A: XQueryExtension (in /usr/lib/libX11.so.6.2.0)
by 0x13BCAA: XInitExtension (in /usr/lib/libX11.so.6.2.0)
by 0x254AD4: XRenderFindDisplay (in /usr/lib/libXrender.so.1.3.0)
by 0x25589C: XRenderQueryExtension (in /usr/lib/libXrender.so.1.3.0)
by 0x550DC1: _XcursorGetDisplayInfo (in /usr/lib/libXcursor.so.1.0.2)
by 0x5514EC: XcursorSupportsARGB (in /usr/lib/libXcursor.so.1.0.2)
by 0x553C1B: XcursorNoticeCreateBitmap (in /usr/lib/libXcursor.so.1.0.2)
by 0x12FB24: _XNoticeCreateBitmap (in /usr/lib/libX11.so.6.2.0)
Address 0x443D95D is 165 bytes inside a block of size 16,384 alloc'd
at 0x400473F: calloc (vg_replace_malloc.c:279)
by 0x141406: XOpenDisplay (in /usr/lib/libX11.so.6.2.0)
by 0x46C32C0: csXWindow::Initialize(iObjectRegistry*) (xwindow.cpp:139)
by 0x8129DCB: csPluginManager::LoadPlugin(char const*, bool) (plugmgr.cpp:203)
by 0x4610110: csPtr<iXWindow> csLoadPlugin<iXWindow>(iPluginManager*, char const*) (plugin.h:152)
by 0x460EB29: csGraphics2DGLX::Initialize(iObjectRegistry*) (glx2d.cpp:107)
by 0x8129DCB: csPluginManager::LoadPlugin(char const*, bool) (plugmgr.cpp:203)
by 0x4575E20: csPtr<iGraphics2D> csLoadPlugin<iGraphics2D>(iPluginManager*, char const*) (plugin.h:152)
by 0x45578B8: csGLGraphics3D::Initialize(iObjectRegistry*) (gl_render3d.cpp:3513)
by 0x8129DCB: csPluginManager::LoadPlugin(char const*, bool) (plugmgr.cpp:203)
by 0x8139F71: csPluginLoader::LoadPlugins() (plugldr.cpp:357)
by 0x8104ADE: csInitializer::RequestPlugins(iObjectRegistry*, csArray<csPluginRequest, csArrayElementHandler<csPl
uginRequest>, CS::Memory::AllocatorMalloc, csArrayCapacityDefault> const&) (initapp.cpp:403)
</pre>
I don't know what this message total means, it is not in [[PS]] or [[CS]] and a variable unitalized passed to writev() shouldn't really be an error becasue it is writing memory to this location and not reading from it. It happens on starup.