I upgraded yesterday and took that opportunity to track down which plugin it was. I tested it with 1.30 and 1.31 before I went to sleep. I searched for the issue today and saw another person had an issue with markdown, tried to reproduce the stall, but today it was working fine. I'll add a stack trace if it reoccurs. Figured I'd still add what I could above since the next person to have the issue might use Process Hacker, double click on geany, then double click on the thread with the highest cpu usage, and create a stack frame trace. But as you said, it is likely an issue with webkit/libfontconfig of which you only have so much control of.