Re: indirect threading for bytecode interpreter

[Tom Tromey]

>>>>> "Dan" == Dan Nicolaescu <address@hidden> writes:

Dan> Making the GC generational might help.  The first GC after startup does
Dan> 200K mark_object calls, so if most of those don't need to be scanned for
Dan> each GC, then that might help.

Yeah.

Whatever happened to the Boehm GC branch?  That seems like the simplest
way to implement this feature.

Dan> If you have a testcase that shows GC problems, maybe posting the test
Dan> case would enticemore people look into it...
     
Here's a script I've been running on the GDB sources:

http://sourceware.org/ml/gdb-patches/2009-08/msg00390.html

I've profiled (using oprofile) Emacs running this a few times; typical
results show mark_object at 10-12% of runtime.  It is probably pretty
easy to reproduce this with a variety of elisp programs.

Tom