ARM patches applied

Richard Earnshaw rearnsha@arm.com
Thu Feb 25 04:16:00 GMT 1999


> I think it should be OK.  I can't think of anything in particular that's 
> likely to break and as you mentioned yesterday I'd like to move the Linux ABI 
> towards full ATPCS conformance.  

Hmm, ATPCS isn't the same as APCS (which is what I mentioned yesterday).  
ATPCS compliance is a fairly major chunk of work on the to-do list.

> 
> >That might help to tidy up some of the internal code (though it some ways 
> >it obfuscates what is happening); but it still misses my point, which is 
> >that the user option -mshort-load-bytes is an inaccurate description of 
> >what really happens and why we do it.
> 
> Agreed.  I get a fairly regular stream of requests from people who want to use 
> it to suppress ldrh/strh on StrongARM machines where the bus doesn't support 
> 16-bit accesses (Acorn RiscPCs usually).

This is precisely the confusion I want to avoid, and which renaming the 
option should achieve.

R.



More information about the Gcc-patches mailing list