Subject: Re: Option to make cpp(1) not accept named pipes or devices as
To: Eric Christopher <>
From: Jim Wise <>
List: tech-toolchain
Date: 11/30/2004 01:48:34
Hash: SHA1

On Mon, 29 Nov 2004, Eric Christopher wrote:

>> I'd like to commit this sometime tonight or tomorrow, and then feed back 
>> to FSF.
>> Any suggested improvements before I do so?
>The idea for the patch looks good, though I'm not sure it even needs to
>be an option IMO, but others may care more about that.
>Couple of comments:
>1) Whitespace. For the diff for the FSF you'll want to use unified (as I
>mentioned before) and probably use -w since you appear to have a lot of
>whitespace changes. Also, you'll only need to send patches for non-
>generated files at that point.

Agreed, will do.

>2) Word of warning: Mainline has changed a bit. cpp is now in it's own
>top level directory.

Okay -- I'll see how easy it is to drag the changes across, and will do 
so if I can.

>3) Extra bits:
>+   The `-isystem-cxx' command line option adds its argument to the list
>+of C++ system headers, similar to `-isystem' for C headers.
>You seem to have some various bits for -isystem-cxx in the patch...

This seems to be an import glitch with our sources, actually -- our 
in-cvs .texi has this text and our in-cvs .info doesn't, so when I 
re-ran makeinfo, in it came.

>Otherwise it looks good to me for FSF - not that it's an area I can approve
>mind you :)

Cool, thanks for the help!

I'll try to get this sent over some time Wednesday.

- -- 
				Jim Wise
Version: GnuPG v1.2.6 (NetBSD)