RE: [PATCH] Bump gc-cons-threshold to 4MB.
I actually have a pending patch that redoes the determination of GC to be a
% of total memory usage. This sums up the total amount as computed during
garbage collection and then GC's when a specified percent of this has been
consed -- currently I have it set at 40%. This should make GC'ing always be
no more than a constant "amortized" percent of total time since GC time is
proportional (more or less) to memory usage.
I'll post this soon.
ben
-----Original Message-----
From: xemacs-patches-bounces(a)xemacs.org
[mailto:xemacs-patches-bounces＠xemacs.org] On Behalf Of Aidan Kehoe
Sent: Sunday, February 27, 2005 1:57 PM
To: xemacs-patches(a)xemacs.org
Subject: [PATCH] Bump gc-cons-threshold to 4MB.
It's been 500k since 1996. Computers ship with significantly
more RAM now. As discussed about a month ago.
XEmacs Trunk source patch:
Diff command: cvs -q diff -u
Files affected: src/alloc.c
Index: src/alloc.c
===================================================================
RCS file: /pack/xemacscvs/XEmacs/xemacs/src/alloc.c,v
retrieving revision 1.106
diff -u -u -r1.106 alloc.c
--- src/alloc.c 2005/02/03 16:14:04 1.106
+++ src/alloc.c 2005/02/27 18:49:10
＠＠ -5340,7 +5340,7 ＠＠
funcall_alloca_count = 0;
#if 1
- gc_cons_threshold = 500000; /* XEmacs change */
+ gc_cons_threshold = 4194304; /* XEmacs change */
#else
gc_cons_threshold = 15000; /* debugging */
#endif
--
"I, for instance, am gung-ho about open source because my
family is being held hostage in Rob Malda's basement. But who
fact-checks me, or Enderle, when we say something in public?
No-one!" -- Danny O'Brien