site_archiver@lists.apple.com Delivered-To: darwin-dev@lists.apple.com User-agent: Mutt/1.4.2.1i Excellent. Apple is not a rogue entity, my code is now more compliant, *and* I now understand how the compiler was handling my old code. Thanks so much for the help! - Lowell On Sat, Nov 03, 2007 at 08:52:48AM -0700, Duane Murphy wrote:
--- At Sat, 3 Nov 2007 09:21:09 -0500, mothra wrote:
Thanks for the input - it appears that that solution does work. It is a bit distressing to me that Apple's gcc 4.0.1 is not consistent with either RedHat's gcc 4.1.* or their gcc 3.*, or Apple's gcc 3.* in this regard. Is Apple striking out on its own here? Or is this requirement for two lines of code to initialize such variables an existing standard that only Apple is currently enforcing?
More recent compilers have changed.
I used to do exactly what you did. Previous compilers would essentially treat the initialization as a immediate constant. That is the actual variable and memory would not be allocated, but just use the initializer value everywhere.
There must have been a determination that this was technically not allowed. The solution is to do as the standard says and specify the initializer separately. This has nothing to do with Apple, but rather the maturation of the compilers.
...Duane
_______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.appl... This email sent to site_archiver@lists.apple.com