1Date: Tue, 18 Sep 2001 00:38:37 -0500 (CDT)
2From: Chris Lattner <sabre@nondot.org>
3To: Vikram S. Adve <vadve@cs.uiuc.edu>
4Subject: Idea for a simple, useful link time optimization
5
6
7In C++ programs, exceptions suck, and here's why:
8
91. In virtually all function calls, you must assume that the function
10   throws an exception, unless it is defined as 'nothrow'.  This means
11   that every function call has to have code to invoke dtors on objects
12   locally if one is thrown by the function.  Most functions don't throw
13   exceptions, so this code is dead [with all the bad effects of dead
14   code, including icache pollution].
152. Declaring a function nothrow causes catch blocks to be added to every
16   call that isnot  provably nothrow.  This makes them very slow.
173. Extra extraneous exception edges reduce the opportunity for code
18   motion.
194. EH is typically implemented with large lookup tables.  Ours is going to
20   be much smaller (than the "standard" way of doing it) to start with,
21   but eliminating it entirely would be nice. :)
225. It is physically impossible to correctly put (accurate, correct)
23   exception specifications on generic, templated code.  But it is trivial
24   to analyze instantiations of said code.
256. Most large C++ programs throw few exceptions.  Most well designed
26   programs only throw exceptions in specific planned portions of the
27   code.
28
29Given our _planned_ model of handling exceptions, all of this would be
30pretty trivial to eliminate through some pretty simplistic interprocedural
31analysis.  The DCE factor alone could probably be pretty significant.  The
32extra code motion opportunities could also be exploited though...
33
34Additionally, this optimization can be implemented in a straight forward
35conservative manner, allowing libraries to be optimized or individual
36files even (if there are leaf functions visible in the translation unit
37that are called).
38
39I think it's a reasonable optimization that hasn't really been addressed
40(because assembly is way too low level for this), and could have decent
41payoffs... without being a overly complex optimization.
42
43After I wrote all of that, I found this page that is talking about
44basically the same thing I just wrote, except that it is translation unit
45at a time, tree based approach:
46http://www.ocston.org/~jls/ehopt.html
47
48but is very useful from "expected gain" and references perspective.  Note
49that their compiler is apparently unable to inline functions that use
50exceptions, so there numbers are pretty worthless... also our results
51would (hopefully) be better because it's interprocedural...
52
53What do you think?
54
55-Chris
56
57