Quote:Original post by ukdeveloper
As for Torvalds' views, I hardly think his views of C++ are valid given the appalling nature of some of the C code that got posted earlier in the thread. Anyone using goto ought to be pointed to and laughed at, then damned to Hell. If you're using that, then you've most likely got some kind of design flaw. That's advice I picked up from GDNet itself, by the way, and because of that I have never once used it in any serious projects.

Quote:Original post by eedok
If something was completely useless they would have removed it from the language,

Quote:
it's not so much as goto is a horrible disease that plagues whatever touches it, as much as it's goto is very easily misused.

Quote:
If you think there's unnecessary goto's in some of his code, refactor it to get rid of the goto, test it, then submit it as a patch.

void release_task(struct task_struct * p) 136 { 137         struct task_struct *leader; 138         int zap_leader; 139 repeat: 140         atomic_dec(&p->user->processes);     [... blabla ...] 173         p = leader; 174         if (unlikely(zap_leader)) 175                 goto repeat; 176 }

Quote:
I agree with the best programmer is the one who gets things done, I generally use the best offerings for whatever task I have at hand, and the ones that are finished are usually much better than ones that don't work.

Quote:Original post by SiCrane
Let's put it another way: I consider myself to be an expert C++ programmer. The thing is, I need to be an expert C++ programmer in order to write relatively bug-free code. I've been using the language for something like 15 years, and I still have WTF moments when my code behaves in a manner completely different than I intended. I'm probably only a decent Python, Visual Basic and C# programmer, but I don't need to be any better than that to get my programs to do what I want in those languages.

Quote:Original post by crusadingknight
That was the only abuse I could see (could replace it with a do-while loop, which would be about the same). The rest of the goto uses are common practices in kernel code in the name of keeping code size down, decreasing cache footprint, error/stack unwinding, etc.;

Quote:Original post by Yann L

Quote:Original post by crusadingknight
That was the only abuse I could see (could replace it with a do-while loop, which would be about the same). The rest of the goto uses are common practices in kernel code in the name of keeping code size down, decreasing cache footprint, error/stack unwinding, etc.;

Ah, the "but but but - it's kernel code" excuse [wink]

Of course this doesn't hold up. As I said before, kernel code is nothing special - it's just code. Sure, you have to be careful about a few specific things, but this is just as valid for device drivers and the like. Any advantage of using gotos over cleaner alternatives would be absolutely minimal in almost all cases (and even counterproductive). We're not working on the 6502 anymore, where every single byte counted. The kernel I'm currently running is about 1.6MB, compressed. I've written entire 3D engines that were smaller than that (in C++ :). And about using gotos for performance - now that would probably be the textbook example of completely futile micro-optimization with exactly zero benefit.

Torvalds is just using gotos for convenience, because it's his (very bad) coding style. Nothing more.