Update: Just managed to reproduce it with `emacs -Q` I just had to scroll for almost 10 minutes for it to kick in. So ignore my init.el red herring. //Wasim > On 17 Aug 2020, at 15:10, Wasim Abbas wrote: > > So I spend some more time on this today. Its painful to keep scrolling forever but I think I have found a pattern that works reliably. > > Open emacs with a specific header file from command line. /usr/local/Cellar/emacs-head/HEAD-86fef6a_1/Emacs.app/Contents/MacOS/Emacs /personal/roar_engine/core/rhi/rorrhi_types.hpp > M-x split-window-horizontally and start scrolling between the two windows interchangeably. > > I did a binary search through my packages but it was inconclusive. It appeared with single package, but different single packages at different times. So I started commenting out my whole init.el file. > > (setq user-full-name "Wasim Abbas") > (setq user-mail-address "abbas.wasim@gmail.com ") > > ;; load my sexp from elisp dir > (add-to-list 'load-path (expand-file-name "elisp" user-emacs-directory)) > ;; add custom theme folder to load themes from ~/.emacs.d/themes > (add-to-list 'custom-theme-load-path (expand-file-name "themes" user-emacs-directory)) > > This is the only snippet that I can reproduce it with. This doesn’t make much sense. I am not using any themes later on. The whole init.el is empty after these lines. I have reproduced it with this a few times. > > Although I have still not managed to reproduce it with `-Q` option yet. > > Also at one point I got the following segfault. > > > > > //Wasim > >> On 10 Aug 2020, at 20:11, Alan Third > wrote: >> >> On Mon, Aug 10, 2020 at 12:40:31PM +0100, Wasim Abbas wrote: >>> I thought since drawing buffers and rendering is such a core thing >>> to emacs I wasn’t sure it will be anything else (like a package) but >>> looks like it could be? Would you now think it's something to do >>> with any package? >> >> It's unlikely to be a package directly causing it, but it may be doing >> something which causes a bug in Emacs itself to show. >> >>> Is the standard disable each package one by one and see if it can be >>> narrowed down to a single package the way to go forward now? >> >> The usual way would be a binary search type strategy. Comment out half >> your config, then see if you still have the problem, if not try the >> other half. Then you half what remains, and repeat until you've found >> the cause. >> >> The problem is that it may be a combination of things that are causing >> it, but at least it may give us a chance to get a minimal recipe to >> replicate the bug. >> -- >> Alan Third >