linux-kernel - Re: [PATCH] sh: add missing EXPORT_SYMBOL() for _

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAMuHMdUvSVs7vQ7GDhHx4Rp-DJZuSU2xtPVtX86+p9GptHfBhQ@mail.gmail.com>
Date:   Mon, 3 Aug 2020 12:01:29 +0200
From:   Geert Uytterhoeven <geert@...ux-m68k.org>
To:     Guenter Roeck <linux@...ck-us.net>
Cc:     Rich Felker <dalias@...c.org>,
        Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>,
        Yoshinori Sato <ysato@...rs.sourceforge.jp>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Linux-sh list <linux-sh@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] sh: add missing EXPORT_SYMBOL() for __delay

On Thu, Jul 23, 2020 at 1:53 AM Guenter Roeck <linux@...ck-us.net> wrote:
> On 7/22/20 3:52 PM, Rich Felker wrote:
> > On Tue, Jul 21, 2020 at 07:38:40PM -0700, Guenter Roeck wrote:
> >> On Thu, Dec 12, 2019 at 11:38:43AM +0900, Kuninori Morimoto wrote:
> >>> From: Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>
> >>>
> >>> __delay() is used from kernel module.
> >>> We need EXPORT_SYMBOL(), otherwise we will get compile error.
> >>>
> >>> ERROR: "__delay" [drivers/net/phy/mdio-cavium.ko] undefined!
> >>>
> >>> Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@...esas.com>
> >>
> >> I must admit that this patch completely baffles me. __delay was
> >> already exported, only elsewhere in the file. With this patch
> >> in place, it is exported twice, and all sh builds in -next fail
> >> with
> >>
> >> In file included from include/linux/linkage.h:7,
> >>                  from arch/sh/include/asm/bug.h:5,
> >>                  from include/linux/bug.h:5,
> >>                  from include/linux/thread_info.h:12,
> >>                  from include/asm-generic/current.h:5,
> >>                  from ./arch/sh/include/generated/asm/current.h:1,
> >>                  from include/linux/sched.h:12,
> >>                  from arch/sh/lib/delay.c:8:
> >> include/linux/export.h:67:36: error: redefinition of '__ksymtab___delay'
> >>
> >> Guenter
> >>
> >>> ---
> >>>  arch/sh/lib/delay.c | 1 +
> >>>  1 file changed, 1 insertion(+)
> >>>
> >>> diff --git a/arch/sh/lib/delay.c b/arch/sh/lib/delay.c
> >>> index dad8e6a..540e670 100644
> >>> --- a/arch/sh/lib/delay.c
> >>> +++ b/arch/sh/lib/delay.c
> >>> @@ -29,6 +29,7 @@ void __delay(unsigned long loops)
> >>>             : "0" (loops)
> >>>             : "t");
> >>>  }
> >>> +EXPORT_SYMBOL(__delay);
> >>>
> >>>  inline void __const_udelay(unsigned long xloops)
> >>>  {
> >
> > I presently have a revert of this commit in queue for next. If it's
> > sufficiently breaking (and especially if there are other regressions
> > that need to be fixed, see the pmd_free thing) I could try to get it
> > in for 5.8 still but that's getting a bit late.
> >
>
> The patch in mainline is ok. It appears that it has been applied
> again in -next.
>
> "git log --oneline v5.7.. arch/sh/lib/delay.c" on top of next-20200721
> reports:
>
> ee0e4f15dfd4 (origin/akpm) sh: add missing EXPORT_SYMBOL() for __delay
> d1f56f318d23 sh: add missing EXPORT_SYMBOL() for __delay
>
> Maybe it just needs to be dropped from the akpm tree in -next ?

IMHO all of them should be dropped/reverted.
__delay is an internal implementation detail, not to be used by drivers.

See also include/asm-generic/delay.h:

    /* Undefined functions to get compile-time errors */
    ...
    extern void __delay(unsigned long loops);

Cfr. '[PATCH] Revert "sh: add missing EXPORT_SYMBOL() for __delay"'
https://lore.kernel.org/lkml/20200608080636.27862-1-geert+renesas@glider.be/

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@...ux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds