Newsgroups: comp.os.msdos.djgpp
From: manni AT hotbot DOT com (Manni Heumann)
Subject: Re: string class
References: <Pine DOT SUN DOT 3 DOT 91 DOT 981126094447 DOT 14160D-100000 AT is> <19981127 DOT 7330233 AT eev6 DOT eev>
X-Newsreader: News Xpress 2.01
Date: Fri, 27 Nov 1998 09:18:36 GMT
NNTP-Posting-Host: dhcp33-242.uni-bielefeld.de
Message-ID: <365e6e6b.0@news.uni-bielefeld.de>
X-Trace: 27 Nov 1998 10:18:35 +0200, dhcp33-242.uni-bielefeld.de
Lines: 37
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

In article <19981127 DOT 7330233 AT eev6 DOT eev>, Michael Schuster <schuster AT eev DOT e-technik DOT uni-erlangen DOT de> wrote:

>doesn't gxx does this inline - expanding.
>Also djgpp version 2.7.xx worked without -O.
>Additionally does it mean, when declaring something as inline and=20
>don't use the -O option this prog won't compile???
>
>Just thinking....
>
>Gruesse=20
>Michi
>
>
>

=20?

Anyway, good thinking!
I thought using the inline keyword / including the function code in the 
definition, was a kind of suggestion to the compiler. Whether the function was 
then actually inlined is another question. But inlined or not, the function 
should be defined and declared that way.
Is DJGPP forgetting functions that were supposed to be inlined, but weren't 
because of the missing commandline option?

Would be good to no, no matter whether it concerns the string class or my own 
classes, wouldn't it?


-----------------------------------------------------------
Manni Heumann

Bielefeld, Germany

Spammers use reply-adress,
all others: mheumann AT post DOT uni-bielefeld DOT de
-----------------------------------------------------------