Re: [RFC PATCH 8/9] block: genhd: export-GPL generic disk device type

[Greg Kroah-Hartman]

On Fri, Apr 10, 2020 at 11:44:44AM -0400, Steven Rostedt wrote:
> On Fri, 10 Apr 2020 08:33:57 +0200
> Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> wrote:
> 
> > I understand your need here, however we do not export things for
> > modules, when there are no in-kernel module users, sorry.
> 
> This "we don't cater to out-of-tree modules" even when they are GPL seems
> to always baffle me. Especially since we have a high bar of accepting out
> of tree modules especially if they duplicate some functionality of an
> existing infrastructure of the kernel. I like choice, and coming from
> someone that spent over a decade working on code that has been out of tree,
> I'm a little sympathetic to the cause ;-)

We can't do anything for out-of-tree modules as they suddenly become
"higher priority" than in-tree code if you have to not do specific
changes or extra work for them.  Which is not fair at all to the in-tree
code developers at all.

With drivers/staging/ we removed the barrier for accepting any license
compliant driver, so that solved the huge majority of these issues.

> I guess we should be open to allowing LTTng modules in the kernel as well,
> even though it is yet another tracing framework. It's not like its going
> away. And perhaps by doing so, ftrace and perf could start taking advantage
> of anything that LTTng brings.

That is up to you all, as you are the one preventing this from being
merged in the tree, not me :)

Again, don't make us do _more_ work for out-of-tree modules than we do
for in-tree modules, that's just crazy to expect.

thanks,

greg k-h