From: sandmann AT clio DOT rice DOT edu (Charles Sandmann)
Message-Id: <10206101828.AA23165@clio.rice.edu>
Subject: Re: Win2K bad errno; proposed patch
To: djgpp-workers AT delorie DOT com
Date: Mon, 10 Jun 2002 13:28:42 -0500 (CDT)
In-Reply-To: <3D046476.D20A30B6@phekda.freeserve.co.uk> from "Richard Dawe" at Jun 10, 2002 09:33:58 AM
X-Mailer: ELM [version 2.5 PL2]
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> > If file name has illegal character, dos error is 123 0x7b.  This seems to
> > fix it.  Comments?  If not I'll commit it.
> 
> Does that error code mean anything else? Is it used by anthing other than
> Windows 2000/XP?

As Andrew noted (and others earlier in the original discussions) it seems
these error codes as currently exist in DJGPP were for Netware.  Given the
number of users today on Win2K and WinXP - compared to those using Netware,
I think the Win2K/XP definition should win.  It's not even clear to me how
to get this error code on Netware (how common) - much less who might be
able to see if replacing EFAULT with ENOENT in that situation is OK.  
According to RBIL, that code for Netware4 is "invalid character or
bad file-system name" - which is exactly what we are seeing (so Win2K
is consistent) - and on an open we want that to be ENOENT.
It's quite possible that ENOENT is just as good a translation for any
example of this error code vs EFAULT (and it's more consistent with
pre-Win2K behavior).