Oppolzer - Informatik / Blog

This is about the usability of the compiler listings, for example, when you need
the information there for dump reading etc. I have some suggestions about this
topic, but don't really know where to send them ... some kind of user forum for
the compilers would be nice ...

For a formal requirement to IBM, it is not important enough ...

Kind regards

Bernd



Am 01.07.2013 05:03, schrieb P.G.:
> On 2013-06-30, at 10:17, Bernd Oppolzer wrote:
>
>> The compilers (C, C++, PL/1, maybe COBOL, too) don't produce
>> correct ASSEMBLER syntax, but instead they provide as much information
>> as possible, using a somehow extended syntax which is not formally correct
>> ASSEMBLER. In this case, you have a symbolic name (#MX_TEMP1)
>> AND you have the offset and base reg, which is the resolution of the
>> symbolic name.
>> r13 = D
>> 152 = 0x098
>>
>> This may be a little confusing to "real" ASSEMBLER programmers ...
>>
> Indeed.  Sheesh!  They could have provided "as much information as
> possible" without doing violence to assembler syntax by reporting
> instead:
>
> 0000FA  4110  D098        000014 |                 LA        r1,#MX_TEMP1(,r13) 152
>
>
>> Am 30.06.2013 18:04, schrieb P.G.:
>>>
>>>   0000FA  4110  D098        000014 |                 LA       r1,#MX_TEMP1(,r13,152)
> -- gil
>