From: Guenter Roeck <linux@roeck-us.net>
To: Arnd Bergmann <arnd@arndb.de>
Cc: stable <stable@vger.kernel.org>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
linuxppc-dev <linuxppc-dev@lists.ozlabs.org>
Subject: Re: Problems building ppc images in v4.14.y and v4.16.y using gcc 7.3.0 / 8.1.0 from kernel.org
Date: Tue, 5 Jun 2018 09:06:24 -0700 [thread overview]
Message-ID: <20180605160624.GA30373@roeck-us.net> (raw)
In-Reply-To: <CAK8P3a3CFhkjPCX8bgZbx7aYwrUVsKE5ct4ghnMxJLo5OSYqbA@mail.gmail.com>
On Tue, Jun 05, 2018 at 04:31:00PM +0200, Arnd Bergmann wrote:
> On Tue, Jun 5, 2018 at 3:52 PM, Guenter Roeck <linux@roeck-us.net> wrote:
> > Hi Arnd,
> >
> > when using the ppc64 compiler from kernel.org, I see the following problems
> > when trying to compile ppc:allnoconfig in v4.14.y or v4.16.y.
> >
> > gcc 7.3.0: Compilation of kernel.cpu.o hangs
> >
> > The problem goes away if I apply the following two patches (tested with
> > 4.16.y)
> >
> > 17a2f1ced028 cpu/hotplug: Merge cpuhp_bp_states and cpuhp_ap_states
> > fcb3029a8d89 cpu/hotplug: Fix unused function warning
>
> This is probably the same as
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84038
>
> I thought I had included the fix in my builds.
>
Guess not.
> > gcc 8.1.0: Compilation of kernel/cpu.o results in the following error
> >
> > powerpc64-linux-gcc: error: unrecognized command line option '-mno-spe'; did
> > you mean '-fno-see'?
> > powerpc64-linux-gcc: error: unrecognized command line option '-mspe=no'; did
> > you mean '-misel=no'?
> >
> > This problem is also seen with mainline.
>
> I've seen it, but couldn't figure out what the right fix is. I ended
> up commenting
> out those two lines in my private builds:
>
> --- a/arch/powerpc/Makefile
> +++ b/arch/powerpc/Makefile
> @@ -215,8 +215,8 @@ KBUILD_CFLAGS += $(call cc-option,-mno-vsx)
>
> # No SPE instruction when building kernel
> # (We use all available options to help semi-broken compilers)
> -KBUILD_CFLAGS += $(call cc-option,-mno-spe)
> -KBUILD_CFLAGS += $(call cc-option,-mspe=no)
> +#KBUILD_CFLAGS += $(call cc-option,-mno-spe)
> +#KBUILD_CFLAGS += $(call cc-option,-mspe=no)
>
> # Enable unit-at-a-time mode when possible. It shrinks the
> # kernel considerably.
>
> I think there were some changes in how cc-option gets evaluated, maybe
> those rely on something else to be enabled or disabled first?
>
I think it may have cached the flags from the other compiler version.
"make mrproper" prior to "make defconfig" took care of the issue.
However, that doesn't really help - I get lots of
error: 'sys_spu_create' alias between functions of incompatible types
error: 'strncpy' output truncated before terminating nul
if I try to use gcc 8.1.0.
Oh well. I'll try gcc 6.4.0 next.
Thanks,
Guenter
next prev parent reply other threads:[~2018-06-05 16:06 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <e11832b0-f071-b709-be34-417f2d3bc343@roeck-us.net>
2018-06-05 14:31 ` Problems building ppc images in v4.14.y and v4.16.y using gcc 7.3.0 / 8.1.0 from kernel.org Arnd Bergmann
2018-06-05 16:06 ` Guenter Roeck [this message]
2018-06-05 19:47 ` Arnd Bergmann
2018-06-06 6:44 ` Christophe LEROY
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20180605160624.GA30373@roeck-us.net \
--to=linux@roeck-us.net \
--cc=arnd@arndb.de \
--cc=gregkh@linuxfoundation.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=stable@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).