Re: [PATCH] lock assertion macros for 2.5.30

[Bernd Eckenfels]

In article <E17d04X-0000eD-00@starship> you wrote:
> You would just have to break the patch up again when you submit it.  You
> might want create a patch that demonstrates its usage, by adding some
> asserts to core code and removing comments where the assert makes them
> redundant.

Yes, I defintely thing that those asserts are a good way of documenting
contracts. They can be used to document when a function expects a lock to be
held, and they will also be able to empirical test, if it is true.

It may even help static code analysers to find places where the assertion
macros are missing.

Greetings
Bernd