On Sun, 25 Dec 2016 21:15:19 -0500 Leo Famulari wrote: > On Sun, Dec 25, 2016 at 05:47:33PM -0800, Chris Marusich wrote: > > Danny Milosavljevic writes: > > > so I've finally found a 100% reproducible way to crash icecat. > > > Previously it has been sporadic and not reproducible. > > > > > > Try going to this site in a new icecat (guix master) instance: > > > > > > http://arstechnica.com/security/2016/12/this-low-cost-device-may-be-the-worlds-best-hope-against-account-takeovers/ > > > > > > Scroll down. It will crash. Every time. > > > > This doesn't crash for me. Note that I have applied the "fix" for > > the previously observed "crashiness", as described here - perhaps > > that is why it doesn't crash for me: > > > > https://lists.gnu.org/archive/html/help-guix/2016-11/msg00008.html > > Unfortunately for debugging purposes, this doesn't crash Icecat for me > either. I'm still using the Cairo rendering backend. > > It's difficult to use a site like this as a reproducer, since every > user will get totally different content via the advertisements. > > For the record, it also does not crash my Debian Firefox 45.5.1, which > uses uBlock Origin and Privacy Badger to block 1 and 9 domains, > respectively. > > If you can still reproduce the crash, I think it's worth it to save > the entire web page, ads included, and share that if you're satisfied > there's no sensitive data included. Hey, I'm also experiencing crashes with icecat since a few months. It just seems to happen randomly. I've found a note about that on the LFS website on the firefox page that says that using system cairo is causing a crash when it tries to do background rendering. So I've changed our icecat recipe to use the bundled version of cairo, and I don't get any crash anymore. Would something like this patch be acceptable?