Re: ongoing rump udf build failure - and strict-aliasing warnings generally

[Simon Burge <simonb%NetBSD.org@localhost>]

Daniel Carosone wrote:

> On Mon, Jul 21, 2008 at 01:02:46PM +1000, Simon Burge wrote:
> > Daniel Carosone wrote:
> > > Therefore, surely it seems therefore that one should *also* not
> > > squelch it with blanket -Wno-* options either.  So, do we have a brave
> > > volunteer who knows what he's doing to squelch them properly?
> >
> > I think the short answer is "use a temporary variable of the right type"
> > and not to throw casts in just to stop the warnings.
> 
> I assume gcc is then still free to optimise away the temporary
> variable, now that you've used the extra semantics to avoid telling
> puns?

I would assume so too, but can't say for sure.

> > Just to build an amd64 GENERIC kernel without -fno-strict-aliasing give
> > 167 type-punned warnings, so that brave volunteer needs to understand
> > the code in question and also have a bit of time up their sleeves as
> > well.
> 
> Certainly - but across how many files?  Even just capturing the
> current set of affected files (with explicit per-file CFLAGS) will
> help avoid new ones creeping in under the umbrella, no matter how slow
> the progress at cleaning those up might be.

There's 167 warnings across 85 files, 61 of those files are in sys/dev.
The next largest chunk are under dist/ipf with 10 files offending.

Simon.