Re: C-h f now permanently loads ~2MB

[Dave Love]

Richard Stallman <address@hidden> writes:

> My Emacs sessions usually end with "Memory exhausted", so it does
> matter for me.

I expected the Boehm GC work I did to help with that.  It avoids the
GC markbit and I suspect it manages freelists/pages better, though I
didn't get to the stage of measuring.  It may also have been possible
to have it use a mmapped arena to avoid competing with normal malloc
for address space.

I think you could also allow dlmalloc to use mmapped storage for Lisp
objects in a dumped Emacs, possibly at the cost of not being able to
re-dump it, but I don't know how much that might help.