www.digitalmars.com         C & C++   DMDScript  

D.gnu - (mac only) "Invalid mnemonic"?

reply Carlos Santander <csantander619 gmail.com> writes:
I tested this with DMD 0.125 and gdc 0.11 on linux (gcc 3.4.2) and it 
didn't give an error. However on mac (gdc 0.11, gcc 3.4.3) I had an error:

//---------------------------------------
void Α()
{}

void main()
{
	Α();
}
//---------------------------------------

I get "/var/tmp//ccismieQ.s:29:Invalid mnemonic '?FZv'". I also tried 
with this characters: αινσϊρΡΑΙΝΣΪ. It doesn't matter if I start the 
function with a non-accented letter. The weird thing is that sometimes I 
can use these characters on identifiers. Sometimes it works, sometimes 
it doesn't. There doesn't seem to be a pattern.

-- 
Carlos Santander Bernal
May 29 2005
next sibling parent David Friedman <d3rdclsmail_a_ _t_earthlink_d_._t_net> writes:
Carlos Santander wrote:
 I tested this with DMD 0.125 and gdc 0.11 on linux (gcc 3.4.2) and it 
 didn't give an error. However on mac (gdc 0.11, gcc 3.4.3) I had an error:
 
 //---------------------------------------
 void Α()
 {}
 
 void main()
 {
     Α();
 }
 //---------------------------------------
 
 I get "/var/tmp//ccismieQ.s:29:Invalid mnemonic '?FZv'". I also tried 
 with this characters: αινσϊρΡΑΙΝΣΪ. It doesn't matter if I start the 
 function with a non-accented letter. The weird thing is that sometimes I 
 can use these characters on identifiers. Sometimes it works, sometimes 
 it doesn't. There doesn't seem to be a pattern.
 

The assembler doesn't like those characters. This problem seems to have been fixed in Tiger (tried it just now). If you are sticking with Panther, you might still be able use the assembler from the latest Darwin sources/binaries. David
May 29 2005
prev sibling parent =?ISO-8859-1?Q?Anders_F_Bj=F6rklund?= <afb algonet.se> writes:
Carlos Santander wrote:

 I tested this with DMD 0.125 and gdc 0.11 on linux (gcc 3.4.2) and it 
 didn't give an error. However on mac (gdc 0.11, gcc 3.4.3) I had an error:

I reported this as D.gnu/983, but then Walter said that he had "fixed" the D name mangling somehow ? (DMD 0.116: http://www.digitalmars.com/d/changelog.html#new0116) Maybe he didn't change all possibilities, or something. --anders
May 30 2005