Re: lockdep problem conversion semaphore->mutex (dev->sem)

["Remy Bohmer" <linux@bohmer.net>]

Hello Ingo,

> no, you are wrong. If you want to do complex locking, you can still do
> it: take a look at the dev->sem conversion patches from Peter which
> correctly do this. Lockdep has all the facilities for that.
> (you just dont know about them)

Ok.

> the general policy message here is: do not implement complex locking. It
> hurts. It's hard to maintain. It's easy to mess up and leads to bugs.

Agree

> Lockdep just makes that plain obvious.
> Your mail and your frustration shows this general concept in happy

Please, do not see it as frustration, because it isn't...
I just want to understand it better.

> action: judging from your comments you have little clue about dev->sem
> locking and its implications and you'd happily go along and pollute the
> kernel with complex and hard to maintain nested locking constructs.

Now, you assume that i would _like_ complex locking. This is not true.

I just want to understand what was so wrong with dev->sem, I even read
in the discussions before about dev->sem, that it still was on the old
semaphore interface to get around lockdep issues, and _that_ is wrong.
That is bug-hiding from either the code or the tool. I just wanted to
understand if this was a lockdep bug, or a real code bug.

> Lockdep prevents you from doing it mindlessly, it _forces_ you to first
> understand the data structures, their locking and their relationship
> with each other. Then you can implement complexity, if you still want
> it.
>
> That, Sir, is a Good Thing (tm).

Completely agree.


Remy