1. The single signon code should not be statically referenced by the form code; instead it should use the service manager or the repository to come up with an interface that the layout form code can use. 2. It crashes today (test5 in our samples triggers the crash) under purify. [E] IPR: Invalid pointer read in strlen {1 occurrence} Reading 1 byte from 0xdeadbeef (1 byte at 0xdeadbeef illegal) Address 0xdeadbeef is argument #1 of strlen Address 0xdeadbeef points into invalid memory Thread ID: 0xe4 Error location strlen [strlen.asm:54] si_StrippedURL [singsign.c:303] si_GetURL [singsign.c:402] si_GetUser [singsign.c:547] SI_RestoreSignonData [singsign.c:2062] nsNetlibService::SI_RestoreSignonData(char *,char *,char * *) [nsNetService.cpp:597] nsTextControlFrame::PostCreateWidget(nsIPresContext *,int&,int&) [nsTextControlFrame.cpp:363] nsFormControlFrame::Reflow(nsIPresContext&,nsHTMLReflowMetrics&,nsHTMLReflowStat e const&,UINT&) [nsFormControlFrame.cpp:293] nsFileControlFrame::Reflow(nsIPresContext&,nsHTMLReflowMetrics&,nsHTMLReflowStat e const&,UINT&) [nsFileControlFrame.cpp:244] nsInlineReflow::ReflowFrame(nsHTMLReflowMetrics&,UINT&) [nsInlineReflow.cpp:433]

Kipp, how can I reproduce this?

Now that this bug has been assigned to me, let me say that I don't understand it. The report says that the single signon code should not be referenced statically but should use the service manager. But that is exactly what it is doing. So I am going to resolve this report as being invalid. If there is indeed a crash and you can provide information as to how I can reproduce it, then reopen it and I'll investigate.

Marking Verified/Invalid