[Info-vax] decrementing & for loops in C

[Bill Gunshannon]

In article <gk5gsk$q46$1 at usenet01.boi.hp.com>,
	"Ed Vogel" <edward.vogel at hp_stopping_spam.com> writes:
> 
> "Bill Gunshannon" <billg999 at cs.uofs.edu> wrote in message 
> news:6smia4F73tecU6 at mid.individual.net...
>> In article <gk534r$677$1 at south.jnrs.ja.net>,
>> "Richard Brodie" <R.Brodie at rl.ac.uk> writes:
>>>
>>> <johnwallace4 at yahoo.co.uk> wrote in message news:09411eb9-a739-4e87-9278-
>>>
>>>> Some kinder compilers will (optionally) offer a warning when an
>>>> assignment rather than a logical expression is used in this context.
>>>
>>> /warn=enable=questcode is a reasonable balance between passing
>>> probable errors silently.
>>
>> The original code is not "questionable", it is 100% correct.
> 
> This depends upon what you mean by "correct".  Sure, it's
> valid syntax, but it may not do what the user intended.

The day we can truly trust the computer to understand what we meant
rather than what we said I will be very afraid.  :-)

> 
> The purpose of /warn=enable=questcode is to detect such
> cases.

I prefer relying on code reviews by humans.

> 
>>And,
>> as I stated elsewhere, a really good optimizing compiler, after
>> examining the assembler generated by that piece of code would find
>> that it was always successful and would optimize the loop out
>> completely.
> 
> I believe the HP C compiler for OpenVMS detects this and will
> generate no code for the loop.
> 

Knowing the skill level of the compiler writers there I would have
been very surprised if it didn't.

bill

-- 
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg999 at cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>