This is the mail archive of the crossgcc@cygnus.com mailing list for the crossgcc project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Jens, What about changing your #define to #define SW_SRI (short)(0x00f0) I think this would have the desired effect. Good Luck, -------------------------------------------------------- Brett Carswell Ph: +61 7 3252 9722 Software Engineer Fax: +61 7 3253 1403 Tritronics (Australia) Email: brettc@tritro.com.au >-----Original Message----- >From: Jens Doege [SMTP:doege@ljk.atsugi.asahi-kasei.co.jp] >Sent: Thursday, June 19, 1997 9:35 PM >To: crossgcc@cygnus.com >Cc: doege@ljk.atsugi.asahi-kasei.co.jp >Subject: byte access in MC68k > >Hi again, > >I want to thank Doug Evans for his instant answer to my previous >question. I have another one: > >Is there a switch for gcc (68k) to force it not to use byte (char) >operations? >(I couldn't find something like this) > >All the variables I use are at least short. Nevertheless the compiler >detects that from some variables just the lower byte is used: > > if((ist & SW_SRI) != SW_SRI){...} > >In the ready code > > move.b -11(%a6),%d0 > and.b #128,%d0 > tst.b %d0 > >odd memory addresses are accessed which causes really big trouble for my >system. In the given case the preprocessor reduces > > #define SW_SRI 0x00f0 > >to just 0xf0. A precompilier switch to prevent that whould also be >possible. > >Thanks in advance > >Jens