Re: [for-linus][PATCH 4/5] timers: Add timer_shutdown_sync() and timer_shutdown() to be called before freeing timers

[Steven Rostedt]

On Tue, 08 Nov 2022 00:45:55 +0100
Thomas Gleixner <tglx@xxxxxxxxxxxxx> wrote:

> > Sure, but I'm traveling this week and may not get to it until Friday.  
> 
> That's fine as I think this whole approach is wrong to begin with.
> 
> We are not doing a tree wide change at rc4 just to scratch an itch which
> is there for ages and has been debated for months.
> 
> Q: Did you try a merge of the result against -next?
> A: Definitely not.

OK, Yep, I was only working with Linus's tree, not next, something I
overlooked.

> 
> The proper approach is to provide all the infrastructure, i.e. the new
> interfaces and merge them during the next merge window. Right before rc1
> provide a script or a script converted tree to Linus which switches the
> tree over to the new world order.

Well, actually this was my original plan, but Linus asked for the
treewide script to be done in this -rc release, and that part was only
done on his request.

  https://lore.kernel.org/all/CAHk-=whKE5UL+AuCC2wK8oq8D_ueSO_T7-9Acx4POouqVi8ZHg@xxxxxxxxxxxxxx/


> 
> How long have you been around doing kernel development and how much fuzz
> do you make about even trivial changes to ftrace?

This is my first "treewide" change I'm doing, so I'm going to make
mistakes. Hence why I started this all off with RFC.

I'll rebase this on next and break this patch up into more comprehensible bits.

-- Steve