svn commit: r257203 - head/sys/arm/arm

[Bruce Evans]

On Sun, 27 Oct 2013, Ian Lepore wrote:

> On Mon, 2013-10-28 at 04:42 +1100, Bruce Evans wrote:
>> On Sun, 27 Oct 2013, Ian Lepore wrote:
>>
>>> Log:
>>>  Eliminate a compiler warning about extraneous parens.
>>
>> Wow, what flags give these warnings?
>
> --- busdma_machdep.o ---
> /local/build/staging/freebsd/dp10/src/sys/arm/arm/busdma_machdep.c:811:24: warning: equality comparison with extraneous parentheses [-Wparentheses-equality]
>        if ((map->pagesneeded == 0)) {
>             ~~~~~~~~~~~~~~~~~^~~~

Ahh.  I guess what I really want is -Wredundant-parentheses
-Wredundant-braces ...  The parser can easily tell what is redundant
than what is bogus since bogusness depends on style and too many
suboptions would be required to specify styles
(-Wparentheses[-operator][-context].

Oops, the parser easily can't do this right because macro parameters
must usually have redundant parentheses and C parsers are specified
to act as if on the output of the preprocessor so they can't easily
either have a special case to ignore these redundant parentheses or
know that they aren't in the unpreprocessed source so they should be
ignored.

> /local/build/staging/freebsd/dp10/src/sys/arm/arm/busdma_machdep.c:811:24: note: remove extraneous parentheses around the comparison to silence this warning
>        if ((map->pagesneeded == 0)) {
>            ~                 ^   ~
> /local/build/staging/freebsd/dp10/src/sys/arm/arm/busdma_machdep.c:811:24: note: use '=' to turn this equality comparison into an assignment
>        if ((map->pagesneeded == 0)) {
>                              ^~
>                              =
>
> That's what the compiler had to say about it.  I guess in somebody's
> mind if it's a probable error to have done
> ...

This expression could be the result of a macro

#define	ISZERO(x)	((x) == 0)

which for if (ISZERO(map->pagesneeded)) expands to almost the above.  I
guess clang warns excessively about this.

> The warning doesn't bother me as much as the two useless notes that
> follow it (which are annoying when combined with my editor's feature of
> parsing compiler output and jumping to the next error point in the
> code).  I've notice clang is particularly chatty, with things like
> suggesting what you might have meant when you spell a variable name
> wrong.  Before long it's going to be like RPG and COBOL where it's
> willing to assume what you meant and try to generate runnable code in
> the face of almost any error.

The verbose prettyprinted output can be annoying.  If clang would print
its man page then I might know its options so I could turn it off :-).

> I probably date myself with references to RPG and COBOL, but it does
> bring back fond memories of slapping a boilerplate header on every
> interoffice memo issued by management and running it through the cobol
> compiler, and grading them on whether their memos generated runnable
> code or so many errors the compiler gave up.

I started late enough to not touch these or spend more than 2 years on
accounting-related software.

Bruce