Re: cvs commit: src/usr.bin/tail Makefile tail.c

[ Available lists | Index of cvs-all | Month of Jul 2002 | Week of 15 Jul 2002 | Raw email | View thread | Wrap long lines | Reply | Tag ]
From
Bruce Evans <bde@zeta.org.au>
Date
15 Jul 2002 12:30:28
Subject
Re: cvs commit: src/usr.bin/tail Makefile tail.c
Message-ID
20020716052659.G41739-100000@gamplex.bde.org

In reply to

[ Hide this part ]
On Mon, 15 Jul 2002, Alfred Perlstein wrote:

> * Bruce Evans <bde@zeta.org.au> [020715 03:27] wrote:
> > On Mon, 15 Jul 2002, Alfred Perlstein wrote:
> >
> > > Er, both size_t and off_t are pretty much the same one i386 and alpha...
> >
> > Actually, the are quite different.
>
> BUHHH...
>
> size_t is 64 bits on alpha (versus 32 on i386), isn't it?

Yes.

> > > Also, casting enomem of line 257 of reverse.c doesn't fix this warning
> > > which makes no sense...
> > >
> > > Index: reverse.c
> > > - warnx("warning: %qd bytes discarded", enomem);
> > > + warnx("warning: %qd bytes discarded", (quad_t)enomem);
> >
> > This makes sense. %qd is just an alias for %lld, so it unsuitable for
> > printing anything except long longs. Neither off_t nor quad_t is long
> > long on alphas.
>
> That's just obnoxious. :)
>
> Any chance you can suggest how to clean this up? The fact that we
> can't compile tail(1) without warnings bugs me.

Just cast off_t's to intmax_t and print them using %jd. The correct way
to print most other foo_t's is not so clear. %j[du...] is overkill for
most of them, but might become necessary.

I won't clean up anything related to quad_t's since I want them to go away :-).

Bruce


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe cvs-all" in the body of the message



Elapsed time: 0.479 seconds