Re: OBSOLETE STUFF
ba0421ch@soback.kornet.nm.kr
Wed, 08 Jul 1998 18:42:02 -0700
dalecki wrote:
>
> On Tue, 7 Jul 1998, Rogier Wolff wrote:
>
> > dalecki wrote:
> > > > understand how to judge an OS based on such aesthetic qualities, I don't
> > > > mind catering to their little obsessions if it doesn't take much
> > > > effort. In this particular case the code has gone through enough
> > > > testing that I'm reasonably certain it won't cause a problem, which is
> > > > why I submitted a bug to Linus to disable it by default.
> > >
> > > ;-))) Ted ted. That was exactly my point. After some time, like years in
> > > this case, the code should be in a state where the debugging code should
> > > be just obsolete and removed. It was just a remainder what I did. And btw.
> > > is't not my obsession it's my hobby :-))))))).
> >
> > I hold the opinion that unless it is really hindering your "movements"
> > in the code, debugging code should remain in place. At some point in
> > the future, you're going to get a problem, where enabeling the
> > debugging code leads you to the problem quickly, while having to redo
> > the debugging stuff takes a while.
> >
> > Linux also has a lot of stuff that checks for stuff that "usually
> > doesn't go wrong". For example pushing data on skbuffs check if that
> > space is available. You could argue that Linux has been debugged
> > enough that this is no longer neccesary, but I tell you that it is
> > much nicer to get a reproducable kernel panic while developing a
> > driver than it is to have clients wine about unstable software.
> >
> > And I bet I'd have forgotten to turn that debugging on before starting
> > developing, as well as I'd think stuff was stable before shipping, so
> > the field would see "unstable behaviour" instead of the reliable
> > panic. So Linux indeed has some debugging code, that remains enabled
> > even in production kernels. This means that the cause for problems is
> > more easily traced. This is a feature.
> >
> > Roger.
>
> Please see the difference between debugging/ and assertion/ code. If there
> is such: Please make it clear what's for which purpose. BTW. If I'm in a
> bad mood I always look at the anal debugging code in the ftape driver to
> get a lought :-).
>
> Marcin
> =========================================================================
> In real life: System Programmer at AIS AXON GmbH
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.rutgers.edu
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu