Re: Dillo_grows merged

On Sun, Apr 24, 2016 at 09:15:42PM +0200, Sebastian Geerken wrote:
> On Tue, Apr 19, 2016, eocene wrote:
> > 
> > Here's a segfault backtrace that I just got:
> > 
> > #0  0x080ab83d in dw::Textblock::sizeRequestReference (this=0x955e260, 
> >     index=-1224641713) at textblock.cc:423
> > #1  0x080bcd63 in dw::oof::OOFAwareWidget::OOFAwareWidgetIterator::numParts (
> >     this=0x9e446a8, sectionIndex=-1, numContentsInFlow=0)
> >     at oofawarewidget_iterator.cc:84
> 
> I do not understand what has actually happens (look at the code). Did
> you compile dillo with -O2 or -O0? Can you provide a test case how
> this is reproduced?
> 
> Perhaps this is related to another prolem: valgrind prints some
> warnings, which neither make sense to me.

As I recall, I was double-checking whether the textarea row attr default
was still 2 in html5 (https://www.w3.org/TR/html51/sec-forms.html#the-textarea-element)
searching for "rows ". But when I try this again, it doesn't crash.

FWIW, I compiled with -O0.

*tries valgrind*
I do get a number of "Conditional jump or move depends on uninitialised value(s)" when the page loads.

Searching the page doesn't cause any more, though.