This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]

Re: debug_reload output


> > Ok, then how about having a global variable debug_reload_stream initialized
> > to stderr, instead of the parameter?  Does this sound like a better proposition?
> 
> This solves the practical problems, but why do you prefer adding a new global
> variable to adding a new function?  In general, adding global variables
> has a higher impact on subsequent maintenance than adding a new simple
> function.

'cause I get mails in shuffled order... :-)  As strange as it may seem, I get
my own mails back from the list later than some of the answers to them...

I agree that the two function thing is the best approach.  How about
debug_reload_to_stream(FILE *) as the new function?

					Christian


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]