Re: [PATCH -next] treewide: Remove stringification from __alias macro definition

[Ard Biesheuvel]

(+ Russell)

On Thu, 22 Oct 2020 at 18:20, Joe Perches <joe@xxxxxxxxxxx> wrote:
>
> On Thu, 2020-10-22 at 09:33 +0200, Peter Zijlstra wrote:
> > On Wed, Oct 21, 2020 at 11:58:25AM -0700, Joe Perches wrote:
> > > Like the __section macro, the __alias macro uses
> > > macro # stringification to create quotes around
> > > the section name used in the __attribute__.
> > >
> > > Remove the stringification and add quotes or a
> > > stringification to the uses instead.
> >
> > There's a complete lack of rationale for this change.
>
> I'll eventually post V2.
> I'm waiting to see if there are more comments.
>
> As I wrote in reply to Ard:
>
> https://lore.kernel.org/lkml/1cecfbfc853b2e71a96ab58661037c28a2f9280e.camel@xxxxxxxxxxx/
>
> Using quotes in __section caused/causes differences
> between clang and gcc.
>
> https://lkml.org/lkml/2020/9/29/2187
>
> Using common styles for details like this is good.
>

This patch is now causing problems in the ARM tree, because some new
uses of __alias() have been queued (for KASAN), and since this is a
non-backwards compatible change, we have to choose between breaking
the maintainer's tree or breaking -next (given that the change has
been pulled in there now)

I am still not convinced we need this change, as I don't see how the
concerns regarding __section apply to __alias. But if we do, can we
please use the same approach, i.e., revert the current patch, and
queue it again after v5.11-rc1 with all new occurrences covered as
well?