<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META content="text/html; charset=iso-8859-1" http-equiv=Content-Type>
<META content="MSHTML 5.00.3315.2870" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2>Hi all,<BR>from my application on coldfire targets
I have evidence that gcc3.2.1 is<BR>affected by a bug similar to
PR8309 (if I compile my code with optimization level>
0).</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>I obtain correct code (<FONT face=Arial
size=2></FONT>and all work fine) if I use the option -O4
-fno-force-mem. (O4 or O1, O2 O3)</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>I am trying to investigate the
problem:</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>from a C code as<BR></FONT></DIV>
<DIV><FONT face=Arial size=2>short I<BR>for(I=0; I<xx
;I++)<BR>{<BR>......<BR>.........<BR>}</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>but located inside a function that use many
dinamic variables,<BR>I obtain from gcc an assembler code as:</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>move. w d0,
0x10(a6)
(I=0 ; I is declared as short<BR>and 0x10(a6) is
his place)<BR>...</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>moveq.l
#1,d5
(I++)<BR>add.l
d5,0x10(a6) (!!!!! : correct code is
add.w d5,0x10(a6) )</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2></FONT> </DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>Best regards</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2>Fabrizio Pirovano</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=Arial size=2><BR></FONT> </DIV></BODY></HTML>