According to Purify we're leaking lots of table column elements. I doubt there's anything special about test3 [W] MLK: Memory leak of 1904 bytes from 28 blocks allocated in NS_NewHTMLTableColElement(nsIHTMLContent * *,nsIAtom *) Distribution of leaked blocks Allocation location new(UINT) [new.cpp:23] NS_NewHTMLTableColElement(nsIHTMLContent * *,nsIAtom *) [nsHTMLTableColElement.cpp:97] if (nsnull == aInstancePtrResult) { return NS_ERROR_NULL_POINTER; } => nsIHTMLContent* it = new nsHTMLTableColElement(aTag); if (nsnull == it) { return NS_ERROR_OUT_OF_MEMORY; } MakeContentObject [nsHTMLContentSink.cpp:520] NS_CreateHTMLElement(nsIHTMLContent * *,nsString const&) [nsHTMLContentSink.cpp:838] nsTableFrame::EnsureColumns(nsIPresContext&) [nsTableFrame.cpp:737] nsTableFrame::SetInitialChildList(nsIPresContext&,nsIAtom *,nsIFrame *) [nsTableFrame.cpp:321] HTMLStyleSheetImpl::ConstructTableFrame(nsIPresContext *,nsIContent *,nsIFrame *,nsIStyleContext *,nsIFrame *&) [nsHTMLStyleSheet.cpp:1131] HTMLStyleSheetImpl::ConstructFrameByDisplayType(nsIPresContext *,nsStyleDisplay const*,nsIContent *,nsIFrame *,nsIStyleContext *,nsIFrame *&) [nsHTMLStyleSheet.cpp:1467] HTMLStyleSheetImpl::ConstructFrame(nsIPresContext *,nsIContent *,nsIFrame *,nsIFrame *&) [nsHTMLStyleSheet.cpp:1767] HTMLStyleSheetImpl::ContentAppended(nsIPresContext *,nsIContent *,int) [nsHTMLStyleSheet.cpp:1908]

removed a spurious GetContent that was the cause of a memory leak

who should verify this bug? could buster or troy verify?

I verified using purify