This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: g++ compile-time regression
- To: dewar at gnat dot com
- Subject: Re: g++ compile-time regression
- From: Joe Buck <jbuck at synopsys dot COM>
- Date: Wed, 28 Feb 2001 16:56:25 -0800 (PST)
- Cc: carlo at alinoe dot com, gcc at gcc dot gnu dot org
> <<3.1 is ten times slower than 2.95.1 for a 'hello world' program :).
> >>
>
> Well hello world is hardly an interesting example, so this data point alone
> is not something to get too worried about.
I suspect that the slowdown is caused mainly by the much more complicated
libstdc++-v3 ... it takes longer to parse those headers. This complexity
is unavoidable if we want to be standards conformant: in the old library,
ostream wasn't a template.
Still, anything that lets us parse those headers faster would be a win.