Re: Inline assembly syntax

[Stan Shebs <shebs at apple dot com>]

Ian Lance Taylor wrote:

> These lead us in the direction of
>
> asm
>  {
>    add r0, foo, bar    // register r0 = variable foo + variable bar
>    add r0, "r1", r2    // register r0 = variable r1 + register r2
>    {                   // instructions in block may not be rearranged
>      cli               // disable interrupts
>      add 0(r0), 1      // * (register r0) += 1
>      sti               // enable interuppts
>    }
>    svc 21 __clobber__(r0,r1,r2) // supervisor call
>    __noschedule__      // scheduling barrier
>    addi r0, r1, constant __constraint__(i) // specify constraint
>  }
Coool... This is rather more ambitious than what I was proposing, but
I'm not going to complain! This is both compatible with CW syntax,
and addresses many of its limitations.

Incidentally, as I was thinking about the assembly grokking problem,
it occurred that we already have a rich mine of info about operand
constraints - the machine descriptions themselves. Not always
expressed as per-instruction, but for at least the one-instruction
output strings, one could make up a table of instructions/constraint
letters/types that would be pretty accurate, and not need special
per-target hackery.

Stan