This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Less radical proposal (was Re: Radical proposal: skip 3.4)
- From: Chris Lattner <sabre at nondot dot org>
- To: Daniel Berlin <dberlin at dberlin dot org>
- Cc: Nathanael Nerode <neroden at twcny dot rr dot com>, <gcc at gcc dot gnu dot org>
- Date: Sat, 10 Jan 2004 17:04:14 -0600 (CST)
- Subject: Re: Less radical proposal (was Re: Radical proposal: skip 3.4)
Daniel Berlin wrote:
> One of the things i love about LLVM is bugpoint.
> It can tell you what pass is causing a bug, etc, just by handing it a
> piece of code.
If you guys are interested in bugpoint, what it does and how it works,
there are fairly decent dox here:
http://llvm.cs.uiuc.edu/docs/CommandGuide/bugpoint.html
Actually the nicest thing about it isn't narrowing down the problem to a
particular pass, the nice thing is that it can _reduce testcases_ for
crashes or miscompilations. Though the miscompilation reducer is
(currently) limited to narrowing it down to the function being
miscompiled, it is still incredibly useful for dealing with large
programs. For optimizer crashes, it often narrows it down a testcase to
the relevant fragment of the CFG and the few instructions necessary to
reproduce the problem.
-Chris
--
http://llvm.cs.uiuc.edu/
http://www.nondot.org/~sabre/Projects/