Re: [RFC][PATCH 1/4] sched: Force the address order of each sched class descriptor

[Steven Rostedt]

On Fri, 20 Dec 2019 13:44:05 +0300
Kirill Tkhai <ktkhai@xxxxxxxxxxxxx> wrote:

> My opinion is to better make some less beautiful thing in a single synchronous place,
> than to distribute the redundancy over all the code (especially, when it is asynchronous).

I very much subscribe to this notion. The TRACE_EVENT() and NMI code
does the same. Keep all other use cases simple by making it complex in
one locality.

-- Steve