Re: [PATCH] rcu: Is it safe to enter an RCU read-side critical section?

[Ingo Molnar <mingo@kernel.org>]

* Christoph Lameter <cl@linux.com> wrote:

> On Mon, 9 Sep 2013, Peter Zijlstra wrote:
> 
> > > Slander. Certainly validation is good. Its just that PREEMPT kernels 
> > > are not in use
> >
> > Complete bullshit, its part of the mainline kernel, lots of people run 
> > them -- including me, and any patch is supposed to keep it working.
> 
> Nonsense. There is no main line distro that supports full preempt. Its 
> an academic exercise.

You are dead wrong on multiple levels.

Firstly, here's a "popularity list" of preempt kernel config options, 
distilled from kernel configs sent to lkml, from a time span of a couple 
of months:

     73 CONFIG_PREEMPT_NONE=y
     81 CONFIG_PREEMPT_VOLUNTARY=y
    135 CONFIG_PREEMPT=y

CONFIG_PREEMPT=y is in fact more popular than the other two modes of 
preemption, amongst kernel testers who post configs to lkml.

Secondly, even if, hypotethically, in an alternate universe, 
CONFIG_PREEMPT=y was used only rarely, the preempt debug checks are still 
very important for automated testing: they often catch bugs that are 
relevant on !PREEMPT as well. (such as accidental unlocked access) The 
last thing we want to do is to reduce the 'reach' of debug checks.

Distros typically go for the lowest overhead preemption option with server 
loads in mind - still even they enjoy the bug fixes generated by the 
preempt checks.

So my NAK stands: you are still in denial, you should stop the silly 
arguing and you should stop wasting maintainer time. You need to address 
PeterZ's review feedback and fix the bugs in your patches, ASAP.

Thanks,

	Ingo