Re: target/5621: ARM compiler aborts with asm "m" constraint

Re: target/5621: ARM compiler aborts with asm "m" constraint

[rearnsha]

Synopsis: ARM compiler aborts with asm "m" constraint

Responsible-Changed-From-To: unassigned->rearnsha
Responsible-Changed-By: rearnsha
Responsible-Changed-When: Tue Mar 26 10:08:22 2002
Responsible-Changed-Why:
    .
State-Changed-From-To: analyzed->closed
State-Changed-By: rearnsha
State-Changed-When: Tue Mar 26 10:08:22 2002
State-Changed-Why:
    It turns out that the compiler used to be able to do this,
    but that the patch was accidentally lost during the ARM/Thumb
    merge (I'd forgotten that I'd already fixed this once before :-)
    

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=5621

Re: target/5621: ARM compiler aborts with asm "m" constraint

[rearnsha]

Old Synopsis: assembler code
New Synopsis: ARM compiler aborts with asm "m" constraint

State-Changed-From-To: open->analyzed
State-Changed-By: rearnsha
State-Changed-When: Thu Feb  7 06:11:43 2002
State-Changed-Why:
    While the compiler shouldn't abort for this, you probably won't ever
    be able to make such an instruction work as you've expressed
    it.
    
    The best way of expressing what you want here, so that the
    compiler can handle it cleanly is
    
     asm volatile(
         "ldr %0,[%1]\n\t"
         :"=r"(i)
         :"r"(&x)
        );
    
    
    The compiler should be fixed to give an error in this case.

http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=5621