A few weeks ago, bug 940 was added. A segfault. The submitter gets it with Tiny Core. I don't see it with my non-TC setup, so I asked Colin about it. He suspected that it had to do with TC's dearth of fonts. http://www.fltk.org/str.php?L2319 was added yesterday, and is about someone with dillo-2.2 getting font crashes. fltk::setfont() calling fontopen(), which gets NULL from XftFontOpen() (as the manpage says can happen). Since this is a fltk2 thing, I think this will fall to us to fix. As for whether we simply want to stick something like if (!xftfont) return; after the fontopen() call, or we want to try harder in some way, I don't know yet.
I wrote:
A few weeks ago, bug 940 was added. A segfault. The submitter gets it with Tiny Core. I don't see it with my non-TC setup, so I asked Colin about it. He suspected that it had to do with TC's dearth of fonts.
Just remembered Colin reproduced the crash with xft disabled. So this might still be a fltk font problem, but if so, not the same exact one.
On Thu, Feb 18, 2010 at 04:20:40PM +0000, corvid wrote:
A few weeks ago, bug 940 was added. A segfault. The submitter gets it with Tiny Core. I don't see it with my non-TC setup, so I asked Colin about it. He suspected that it had to do with TC's dearth of fonts.
http://www.fltk.org/str.php?L2319 was added yesterday, and is about someone with dillo-2.2 getting font crashes. fltk::setfont() calling fontopen(), which gets NULL from XftFontOpen() (as the manpage says can happen).
Since this is a fltk2 thing, I think this will fall to us to fix.
As for whether we simply want to stick something like if (!xftfont) return; after the fontopen() call, or we want to try harder in some way, I don't know yet.
I don't understand all the details especially whether it is ok to return without assigning anything to current. But it might be good to suggest the fix on the fltk list.
participants (2)
-
corvid＠lavabit.com -
Johannes.Hofmann＠gmx.de