From: Sam Ravnborg <sam@ravnborg.org>
To: sparclinux@vger.kernel.org
Subject: Re: [GIT PULL] unify sparc header files
Date: Mon, 07 Jul 2008 07:46:00 +0000 [thread overview]
Message-ID: <20080707074600.GA14300@uranus.ravnborg.org> (raw)
In-Reply-To: <20080628220235.GA3169@uranus.ravnborg.org>
On Sun, Jul 06, 2008 at 06:18:22PM -0700, David Miller wrote:
> From: Sam Ravnborg <sam@ravnborg.org>
> Date: Sun, 6 Jul 2008 22:06:37 +0200
>
> > I have now added the appended commit and the full series
> > can be pulled from:
> >
> > ssh://master.kernel.org/pub/scm/linux/kernel/git/sam/sparc2.git master
> >
> > I have tested that sparc64 continue to build and that we indeed
> > add the -m64 option to gcc when we build vmlinux.lds
> > iLet me know if this series is better.
>
> Something is still messed up, the whole build now complains about
> things like:
>
> init/initramfs.c: In function 'flush_window':
> init/initramfs.c:433: warning: pointer targets in passing argument 1 of 'flush_buffer' differ in signedness
> init/initramfs.c: In function 'unpack_to_rootfs':
> init/initramfs.c:474: warning: pointer targets in assignment differ in signedness
>
> So things are still busted with bi-arch compilers.
>
> It seems that adding "-m64" to the kbuild CPPFLAGS does not propagate
> "-m64" to the kbuild CFLAGS. So your "fix" changeset idea to remove
> "-m64" from the kbuild CFLAGS does not aparently work.
I actually checked it.
>
> Please build a bi-arch compiler that defaults to -m32 ;-) Or at least,
> look at the command line with some higher verbosity levels. You know
> how to do that right? :-)))
I checked that and it is present so we need to look somewhere else.
In top-level Makefile we do:
include arch/sparc64/Makefile
....
# disable pointer signed / unsigned warnings in gcc 4.0
KBUILD_CFLAGS += $(call cc-option,-Wno-pointer-sign,)
When we do the above check we use KBUILD_CFLAGS direct
and does not include KBUILD_CPPFLAGS - thus missing -m64.
So I think that your bi-arch sparc gcc only accepts
-Wno-pointer-sign if -m64 is specified and that this
option is missing with current patchset.
At least that is my best explanation so far.
I will upload a new patchset tonight where I am less smart
in setting -m64.
Note: On my cross compiled 64 bit sparc gcc I do not
have the -Wno-pointer-sign option so I cannot
see if -m64 matters or not.
Sam
next prev parent reply other threads:[~2008-07-07 7:46 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-06-28 22:02 [GIT PULL] unify sparc header files Sam Ravnborg
2008-06-28 22:29 ` Sam Ravnborg
2008-06-29 6:27 ` Sam Ravnborg
2008-06-30 19:49 ` Adrian Bunk
2008-07-03 7:50 ` David Miller
2008-07-03 8:12 ` David Miller
2008-07-03 10:01 ` David Miller
2008-07-03 13:45 ` Sam Ravnborg
2008-07-03 20:02 ` Adrian Bunk
2008-07-03 20:15 ` David Miller
2008-07-03 20:29 ` Sam Ravnborg
2008-07-03 20:29 ` Sam Ravnborg
2008-07-04 21:10 ` Sam Ravnborg
2008-07-05 20:55 ` Sam Ravnborg
2008-07-05 23:03 ` Sam Ravnborg
2008-07-06 1:56 ` David Miller
2008-07-06 1:57 ` David Miller
2008-07-06 20:06 ` Sam Ravnborg
2008-07-07 1:18 ` David Miller
2008-07-07 7:46 ` Sam Ravnborg [this message]
2008-07-07 7:54 ` David Miller
2008-07-07 20:54 ` Sam Ravnborg
2008-07-08 9:33 ` David Miller
2008-07-09 4:27 ` Sam Ravnborg
2008-07-15 5:58 ` David Miller
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=20080707074600.GA14300@uranus.ravnborg.org \
--to=sam@ravnborg.org \
--cc=sparclinux@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.