Memory leak ... build from March 8, 1999 (about 6pm PST). Solaris 2.6. gcc 2.7.2.3 and Purify. MLK: 11200 bytes leaked in 100 blocks * This memory was allocated from: malloc [rtlib.o] __bUiLtIn_nEw [libgcc.a] __builtin_new [rtlib.o] CTokenRecycler::CreateTokenOfType(eHTMLTokenTypes,nsHTMLTag) [nsDTDUtils.cpp:375] nsHTMLTokenizer::ConsumeAttributes(unsigned short,CStartToken*,nsScanner&) [nsHTMLTokenizer.cpp:336] nsHTMLTokenizer::ConsumeStartTag(unsigned short,CToken*&,nsScanner&) [nsHTMLTokenizer.cpp:466] nsHTMLTokenizer::ConsumeTag(unsigned short,CToken*&,nsScanner&) [nsHTMLTokenizer.cpp:308] nsHTMLTokenizer::ConsumeToken(nsScanner&) [nsHTMLTokenizer.cpp:239] nsXMLTokenizer::ConsumeToken(nsScanner&) [nsXMLTokenizer.cpp:141] nsParser::Tokenize() [nsParser.cpp:1000] nsParser::ResumeParse(nsIDTD*) [nsParser.cpp:711] nsParser::OnDataAvailable(nsIURL*,nsIInputStream*,unsigned int) [nsParser.cpp:924] nsDocumentBindInfo::OnDataAvailable(nsIURL*,nsIInputStream*,unsigned int) [nsDocLoader.cpp:1694] stub_put_block(_NET_StreamClass*,const char*,int) [nsStubContext.cpp:647] net_read_file_chunk [mkfile.c:956] net_ProcessFile [mkfile.c:1327] NET_ProcessNet [mkgeturl.c:3371] NET_PollSockets [mkselect.c:298] nsNetlibService::NetPollSocketsCallback(nsITimer*,void*) [nsNetService.cpp:1217] TimerImpl::FireTimeout() [nsTimer.cpp:73] nsTimerExpired [nsTimer.cpp:188] g_timeout_dispatch [gmain.c:1147] g_main_dispatch [gmain.c:647] g_main_iterate [gmain.c:854] g_main_run [gmain.c:912] gtk_main [gtkmain.c:475] nsAppShell::Run() [nsAppShell.cpp:152] nsAppShellService::Run() [nsAppShellService.cpp:154] main [nsAppRunner.cpp:274] _start [crt1.o] * Block of 112 bytes (100 times); last block at 0x3ced88

This is a fairly major leak, happening well over a hundred times apparently and several K.

Upon looking at CWellFormedDTD::BuildModel(), is it possible to be leaking the tokens in there? Insure++ seems to think so in a log that I was given.

Improved tokenizer to remove leaks.

I'm still seeing this leak. Not sure if the magnitude is still the same. It seems that it is.

Fixed by healthy whackage of parsing engine by Purify.