Re: [RFA] MIPS16 FP manual call/return fixes

[Jan Kratochvil <jan.kratochvil@redhat.com>]

On Mon, 14 May 2012 11:11:14 +0200, Maciej W. Rozycki wrote:
[ Mostly unrelated to this patch:  ]

> strictly speaking we can keep the structure flat if 
> we decide that complicating it is not worth the saving.

Technically yes but it is more readable to know this field is valid/used only
with this subclass.  And it is better not just to rely on unstructured
comments content.  And to give the field name its real meaning ("func_addr")
and not just a generic name of overload ("related").  A good example of these
overloaded generic fields without subclassing is main_type which is such as
mess I still have problems to fully understand it.


>  Actually, I relied on the lack of compilation errors so far; it seems 
> that while we do have -Wall -Werror, we have -Wno-unused as well which 
> implies -Wno-unused-variable and has defeated my assumptions.  Any 
> particular reason why we disable this warning?

Because there are now many such unused-variable cases needing to be fixed,
there were recent threads about it (see subject /unused/) by Sergio.
It looks still not all the cases are fixed to enable the warnings.


> > >  	      b->type = bp_gnu_ifunc_resolver;
> > 
> > Empty line before a comment according to GDB Coding Standards.
> > 
> > > +	      /* Remember the resolver's address for use by the return
[...]
>  I believe this requirement applies to function and variable/type/etc. 
> definitions only and I haven't seen this style applied inline in many 
> places (including the original comment above, actually).  Are you sure?

I find it more readable with the empty line even in this case.
You are right there are too many of such cases without empty line in GDB.
Therefore I do not mind, check it in either way.


Thanks,
Jan