Xref: news-dnh.mv.net comp.os.msdos.djgpp:1628
Path: news-dnh.mv.net!mv!news.NH.Destek.Net!news2.net99.net!news.cais.net!ringer.cs.utsa.edu!swrinde!howland.reston.ans.net!Germany.EU.net!zib-berlin.de!solid.theo-physik.uni-kiel.de!news.informatik.uni-kiel.de!news.techfak.uni-kiel.de!news
From: Hartmut Schirmer <hsc AT techfak DOT uni-kiel DOT d400 DOT de>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: optimization.bug
Date: 18 Aug 1995 06:33:57 GMT
Organization: Technische Fakultaet
Lines: 33
References: <DDG7w2 DOT CIH AT jade DOT mv DOT net>
Nntp-Posting-Host: zora.techfak.uni-kiel.de
To: brendan AT rdt DOT monash DOT edu DOT au
Dj-Gateway: from newsgroup comp.os.msdos.djgpp

Brendan Simon <brendan AT rdt DOT monash DOT edu DOT au> wrote:
>
>
>I have some kind of wierd optimization bug in V2.  When reading an I/O status
>port on my A2D card, the BUSY bit seems to get stuck on after the first read
>of this port.  The code is :-
>
>short get_A2D(char channel)
>{
>	channel = channel % 8;
>	outportb(CONTROL_REG, channel);  /** select A2D channel to convert **/
>
>	while( (inportb(STATUS_REG) & BUSY_BIT) );	/** wait for not busy **/
>
>	outportb(START_CONVERSION, 0xff);						/** start A2D conversion **/
>
>	while( (inportb(STATUS_REG) & BUSY_BIT) );	/** wait for not busy **/
>
>	return( inportw(A2D_DATA) );
>}
>
>With any of the optimizations selected, the 2nd while loop never terminates
>as the busy bit always appears high.  With no optimization the code works
>as expected.
>
>Any ideas.  Is it a bug or is it a silly mistake somewhere???
>

Check the assembly output (gcc -S ...) and compare optimized and plain code.
Maybe the optimized version is just to quick and overrunning your A/D hardware.

Hartmut