[PATCH] : Demangle PR libstdc++/12600

Carlo Wood carlo@alinoe.com
Tue Oct 14 17:21:00 GMT 2003


On Tue, Oct 14, 2003 at 12:04:24AM -0500, Benjamin Kosnik wrote:
> This is fine, but if this fixes a bug you should put in a testcase. 

Common sense would *definitely* not add a test case for this.

I understand why we need to have rules, but we're all more-than-average
intelligent people who understand the WHY of a rule, and as such know
when they do not apply no?

I will add a test case now, but under protest: it is 10,000,000 times
more likely that there is a *similar* bug in the demangler code
that is not caught by this test case then that this particular way
of breakage will ever occur again.  If that ever happens then
I can only think of one way: Someone accidently downloaded this
patch, then accidently reverse applied it, using patch -R and
then accidently committed it at a moment that the cvs report-back-
to-the-mailinglist was broken so that I didn't see it.

;)  :p

-- 
Carlo Wood <carlo@alinoe.com>



More information about the Libstdc++ mailing list