Re: [PATCH 4/8] Implement target_emit_ops

[Antoine Tremblay <antoine.tremblay@ericsson.com>]

On 02/08/2016 12:30 PM, Yao Qi wrote:
> Antoine Tremblay <antoine.tremblay@ericsson.com> writes:
>
>>    I'm wondering is there a reason we don't use gcc to generate the
>> native bytecode like done on x86 ?
>>
>> x86 has a macro like so :
>>
>> #define EMIT_ASM(NAME, INSNS)						
>>    do									
>>      {									
>>        extern unsigned char start_ ## NAME, end_ ## NAME;		
>>        add_insns (&start_ ## NAME, &end_ ## NAME - &start_ ## NAME);	
>>        __asm__ ("jmp end_" #NAME "\n"					
>> 	       "\t" "start_" #NAME ":"					
>> 	       "\t" INSNS "\n"						
>> 	       "\t" "end_" #NAME ":");					
>>      } while (0)
>>
>
> The reason is about the code readability.  We find the way we are using
> in aarch64 is more readable than the way for x86.
>

OK.

>    p += emit_stp (p, x0, x1, sp, preindex_memory_operand (-2 * 16));
>    p += emit_str (p, lr, sp, offset_memory_operand (3 * 8));
>    p += emit_str (p, fp, sp, offset_memory_operand (2 * 8));
>    p += emit_add (p, fp, sp, immediate_operand (2 * 8));
>
> vs.
>
>    EMIT_ASM (amd64_prologue,
>              "pushq %rbp\n\t"
>              "movq %rsp,%rbp\n\t"
>              "sub $0x20,%rsp\n\t"
>              "movq %rdi,-8(%rbp)\n\t"
>              "movq %rsi,-16(%rbp)");
>
>> Could the same be done for armv7 or there's a reason that prevents it ?
>>
>
> No reason prevents it, but the code should be readable.
>

OK thanks, It is readable that way but I think I'll go with EMIT_ASM... 
given that this should save me quite a bit of time considering thumb/arm 
versions and that there's not much emit_ functions done at the moment.

It should still be readable.

Regards,
Antoine