Date: Tue, 07 Aug 2001 20:29:52 +0300
From: "Eli Zaretskii" <eliz AT is DOT elta DOT co DOT il>
Sender: halo1 AT zahav DOT net DOT il
To: "Andrew Cottrell" <acottrel AT ihug DOT com DOT au>
Message-Id: <9003-Tue07Aug2001202952+0300-eliz@is.elta.co.il>
X-Mailer: Emacs 20.6 (via feedmail 8.3.emacs20_6 I) and Blat ver 1.8.9
CC: djgpp-workers AT delorie DOT com
In-reply-to: <012401c11f40$19d063e0$0a02a8c0@acceleron> (acottrel@ihug.com.au)
Subject: Re: Fw: Windows 2000 /dev/null permission query
References:  <012401c11f40$19d063e0$0a02a8c0 AT acceleron>
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

> From: "Andrew Cottrell" <acottrel AT ihug DOT com DOT au>
> Date: Tue, 7 Aug 2001 21:43:20 +1000
>
> The call to 71A6 returns the following data if passed the file handle from
> 6C00h or 3Dh. I do not have any code to decipher the results, except the
> length of the file which is correct.
>
> DJGPP_204 D:\dj204\contrib\touch>test
> -rw-r--r--   1 AC       root            4 Aug 25  2001 new.txt
> file_attrib = 0x0020
> creation_time_h = 0x20626972
> creation_time_l = 0x6E6F6974
> access_time_h =  0x206C5F65
> access_time_l =  0x32783020
> write_time_h =  0xD303230
> write_time_l =  0x33323330
> hd_serial =  0x84950A0A
> retval_l =  0x0004
> retval_h =  0x0000
> links =  0x0001
> uniq_id_high =  0x80000
> uniq_id_low =  0x88E0
>
> Hope this helps.

To see if these values are okay, I need to know the file's times as
reported by some independent program.  Can someone suggest a way to
print the 3 file times up to the last bit of accuracy using some Win32
application?

> Now for the next hard questions: What is the consensus as to the way
> to head on fixing the LFN issues on Win2K?  Do we fix on a case by
> case basis for each of the file I/O functions that fail?  Test each
> of the LFN API function calls and then make a descision on what to
> do next?

We should at least test each LFN function called by the library.  We
should do that for both regular files and devices.  We should also try
functions, whether from the LFN API or not, which accept a handle,
since we already know they might be broken for handles created by
716C.

Since IOCTL is broken on W2K, we probably should decide how to fix
that (see my other mail), and then begin fixing it.  The first step
would be to prepare a list of IOCTL subfunctions called by the library
and rgeir callers; the next step is to see whether all the IOCTL
subfunctions are broken or only some.