> You said earlier that setting gc-cons-threshold to a big value stopped the problem. What happens if you set it small, say to 1/10th or 1/100 the default value?

I tried that, but that did not really seem to have any effect. :-(

I also tried to come up with some reproduction case creating a huge undo list, small gc-cons-threshold, and then just running gc/modify undo list a lot, but that did not cause any crash, either.

Jorgen

On Mon, Mar 14, 2016 at 9:28 PM Phillip Lord <phillip.lord@russet.org.uk> wrote:
Jorgen Schäfer <jorgen.schaefer@gmail.com> writes:
> Sorry to kill the good news – it took a while longer, but my Emacs just
> crashed with a segfault, so I'm afraid emacs-25 is still affected by this
> problem.

Unsurprising. There is not a huge amount of work happening in master.

You said earlier that setting gc-cons-threshold to a big value stopped
the problem. What happens if you set it small, say to 1/10th or 1/100
the default value?

If we can provoke the seqfault much faster then it should be possible to
bisect it.

Phil