[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH gnumach 1/3] Use Mach native error codes instead of POSIX err
From: |
Guillem Jover |
Subject: |
Re: [PATCH gnumach 1/3] Use Mach native error codes instead of POSIX errno E* ones |
Date: |
Tue, 6 Sep 2011 02:39:08 +0200 |
User-agent: |
Mutt/1.5.21 (2010-09-15) |
On Tue, 2011-09-06 at 01:04:02 +0200, Thomas Schwinge wrote:
> On Thu, 1 Sep 2011 20:24:56 +0200, Guillem Jover <guillem@hadrons.org> wrote:
> > * device/errno.h: Remove file.
> > [...]
>
> > diff --git a/i386/i386at/kd_mouse.c b/i386/i386at/kd_mouse.c
> > index 640209c..770a3fe 100644
> > --- a/i386/i386at/kd_mouse.c
> > +++ b/i386/i386at/kd_mouse.c
> > @@ -179,10 +179,10 @@ mouseopen(dev, flags)
> > #ifdef MACH_KERNEL
> > #else /* MACH_KERNEL */
> > if (flags & FWRITE)
> > - return(ENODEV);
> > + return (D_NO_SUCH_DEVICE);
>
> This has actually been !MACH_KERNEL usage of ENODEV.
> > #ifndef MACH_KERNEL
> > if (tp->t_state & TS_XCLUDE && u.u_uid != 0)
> > - return(EBUSY);
> > + return (D_ALREADY_OPEN);
> > #endif /* MACH_KERNEL */
>
> This has actually been !MACH_KERNEL usage of EBUSY.
Hmmm, right, I doubted about those ones, but cannot remember why I
ended up switching them. Checking now it seems !MACH_KERNEL is in
use when combined with a BSD kernel?
> Should we get rid of the !MACH_KERNEL code?
Probably, I doubt it has not bit rot during this time, and it's dead
code for us anyway. Otherwise I can send a patch reverting those
hunks.
regards,
guillem