Re: + uapi-linux-consth-prefer-iso-friendly-__typeof__.patch added to mm-nonmm-unstable branch

[Petr Vorel <pvorel@suse.cz>]

> Hi Andrew,



> On Wed, Apr 12, 2023 at 2:34 PM Petr Vorel <pvorel@suse.cz> wrote:

> > > On Wed, Apr 12, 2023 at 9:24 AM Andrew Morton <akpm@linux-foundation.org> wrote:


> > > > The patch titled
> > > >      Subject: uapi/linux/const.h: prefer ISO-friendly __typeof__
> > > > has been added to the -mm mm-nonmm-unstable branch.  Its filename is
> > > >      uapi-linux-consth-prefer-iso-friendly-__typeof__.patch

> > > > This patch will shortly appear at
> > > >      https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/uapi-linux-consth-prefer-iso-friendly-__typeof__.patch

> > > > This patch will later appear in the mm-nonmm-unstable branch at
> > > >     git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm

> > > > Before you just go and hit "reply", please:
> > > >    a) Consider who else should be cc'ed
> > > >    b) Prefer to cc a suitable mailing list as well
> > > >    c) Ideally: find the original patch on the mailing list and do a
> > > >       reply-to-all to that, adding suitable additional cc's

> > > > *** Remember to use Documentation/process/submit-checklist.rst when testing your code ***

> > > > The -mm tree is included into linux-next via the mm-everything
> > > > branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
> > > > and is updated there every 2-3 working days

> > > > ------------------------------------------------------
> > > > From: Kevin Brodsky <kevin.brodsky@arm.com>
> > > > Subject: uapi/linux/const.h: prefer ISO-friendly __typeof__
> > > > Date: Tue, 11 Apr 2023 10:27:47 +0100

> > > > typeof is (still) a GNU extension, which means that it cannot be used when
> > > > building ISO C (e.g.  -std=c99).  It should therefore be avoided in uapi
> > > > headers in favour of the ISO-friendly __typeof__.

> > > > Unfortunately this issue could not be detected by
> > > > CONFIG_UAPI_HEADER_TEST=y as the __ALIGN_KERNEL() macro is not expanded in
> > > > any uapi header.

> > > > Link: https://lkml.kernel.org/r/20230411092747.3759032-1-kevin.brodsky@arm.com
> > > > Fixes: d6fc9fcbaa65 ("kbuild: compile-test exported headers to ensure they are self-contained")


> > > This tag is wrong.

> > > The correct one is:

> > > Fixes: a85cbe6159ff ("uapi: move constants from <linux/kernel.h> to
> > > <linux/const.h>")

> > I'm sorry if I misinterpreted the history. The reason why I think it was
> > d6fc9fcbaa65 (which introduced -std=c90) is because there was already some
> > typeof() uses before this commit (uapi/linux/kernel.h and
> > uapi/linux/netfilter/x_tables.h).



> Sorry, I wrongly read the code.


> Fixes: a85cbe6159ff ("uapi: move constants from <linux/kernel.h> to
> <linux/const.h>")

> was also wrong.


> Before a85cbe6159ff, 'typeof' existed in include/uapi/linux/kernel.h.
> So, it was exported to userspace.


> Kevin suspects 607ca46e97a1 ("UAPI: (Scripted) Disintegrate
> include/linux"), but I think the issue goes back more.




> I think the following is the first commit that exported it.

> Fixes: a79ff731a1b2 ("netfilter: xtables: make XT_ALIGN() usable in
> exported headers by exporting __ALIGN_KERNEL()")

Yes, that's the first commit which adds typeof().

I pointed out d6fc9fcbaa65, because that brought -std=c90. I thought that commit
introduced expectation of C without non-gnu extensions for headers. But if this
expectation was "since ever" then obviously a79ff731a1b2 is the faulty commit.

Anyway, I'm going to send patch, which replaces typeof() in
netfilter/x_tables.h.

Kind regards,
Petr