It seems that 'aEventStatus' can have an uninitialized value still when coming out of the call to mInner.HandleDOMEvent() in nsHTMLButtonElement::HandleDOMEvent(). This happens when submitting via the HTML button. The stack trace is: nsHTMLButtonElement::HandleDOMEvent() nsHTMLButtonControlFrame::MouseClicked() nsHTMLButtonControlFrame::HandleEvent() PresShell::HandleEvent() Build from March 28, 1999. Solaris 2.6, gcc 2.7.2.3, Purify.

Eric V, if you don't want this, please reassign it to Eric Pollmann.

What does the 'UMR:' mean in the summary? uninitialized memory reference? (just guessing). If this is a crash, the severity should be major or critical. targetting m4.

I have not worked with any of the html4 content model code. I have not idea.

Marking M5

I'm arbitrarily declaring M8 my UMR/MLK milestone. :)

I still haven't got Purify working on Solaris, and won't receive Purify for NT for a few weeks. Marking these M10

UMR = Uninitialized Memory Read. This looks like a simple lack of initialization in nsGenericElement::HandleDOMEvent().

Working on a fix...

Kevin fixed this on 9/13. nsButtonControlFrame::MouseClicked now contains this line: nsEventStatus status = nsEventStatus_eIgnore; Before any of the calls to HandleDOMEvent. This is a code level change and is not verifiable by running the program (except for less random incorrect event behaviour.) :)

If it isn't verifiable by running the app... I'm going to just go on ahead and mark it as verified. bruce, if you see any further problems in this area, please reopen. thanks!