Subject: Re: Why is using "inline" as a variable name a parse error for our compiler?
To: None <current-users@NetBSD.ORG>
From: der Mouse <mouse@Collatz.McRCIM.McGill.EDU>
Date: 07/19/1995 19:02:37
>> I, for one, strongly encourage compiling with -Werror, to the point
>> of having a strong desire to add it to the system Make rules for
>> 'standard' programs. If -Werror didn't exist, the Alpha port would
>> have been ... a lot harder...
> I am willing to put in the work to make the kernel compile with -Wall
> if people think that it is useful.
Yes! I've been toying with the idea of doing something similar myself,
specifically, making it compile with the gcc warning options I gave
before, possibly weakened if experience proves it necessary. (For
example, it may demand major work to make it clean wrt -Wcast-qual
-Wwrite-strings...proper const-poisoning is _not_ trivial. Even a few
of my own programs have resorted to a "deconst" trick, which strips the
const qualifier from a pointer in a way that doesn't trip -Wcast-qual.
Part of the problem, of course, is ANSI C's weird notion of pointer
compatability, which makes some pointer combinations clash when they
needn't...like casting "char **" to "const char **".)