Date:	Sun, 20 Sep 1998 12:17:23 +0300 (IDT)
From:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To:	bowman <bowman AT montana DOT com>
cc:	djgpp AT delorie DOT com
Subject:	Re: getmntent problem
In-Reply-To:	<36049D63.D9365D09@montana.com>
Message-ID:	<Pine.SUN.3.91.980920120811.406X-100000@is>
MIME-Version:	1.0

On Sun, 20 Sep 1998, bowman wrote:

> I am having a problem with getmntent, specifically when running in a DOS
> box under Win95. If there isn't a diskette in the A:/ drive, the flag to
> skip the B: drive is not set. When the 'biosdisk' call is made on the
> non-existent B: drive, it will hang in '_dpmi_int'. While 'getmntent'
> does make several calls to biosdisk before moving on, I don't see a loop
> in '_dpmi_int'. 

The flag cannot be set, because if the drive is empty, its DOS mapping 
(either A: or B:) cannot be known.

However, `getmntent' works for me on Windows 95 on a single-floppy 
system, so maybe it is some other reason that haunts you.

> Could this possibly be something in the Win95 dpmi?

Not likely.  Seems more like a BIOS setup problem.

> I can't duplicate the situation on my DOS box, since it does have a 5
> 1/4 B: drive. For the time being, I'm just linking a version of mntent.c
> that unconditionally skips drive 2, but this isn't very elegant.

Maybe you could suggest a change to mntent.c that works around this 
problem, and also works on two-floppy systems.  Right now, I would like 
to understand why does it work for me but not for you.

- Raw text -