[PATCH 0/3] fix recent parisc build breakages

[PATCH 0/3] fix recent parisc build breakages

[James Bottomley]

I was hoping that 3.5-rc1 would allow us to build perfectly.
Unfortunately, there were three compile breakages introduced in the last
few days of the merge window, so these are the three fixes

James Bottomley (3):
      fix missing TAINT_WARN problem
      fix compile break in use of lib/strncopy_from_user.c
      fix code to find libgcc

 arch/parisc/Makefile           |    3 ++-
 arch/parisc/include/asm/Kbuild |    1 +
 arch/parisc/include/asm/bug.h  |    2 ++
 3 files changed, 5 insertions(+), 1 deletions(-)


James

[PATCH 1/3] fix missing TAINT_WARN problem

[James Bottomley]

>From ad71a7070fd5912420c443b020c6cf796518f3e9 Mon Sep 17 00:00:00 2001
From: James Bottomley <James.Bottomley@HansenPartnership.com>
Date: Sun, 3 Jun 2012 01:49:32 -0700
Subject: fix missing TAINT_WARN problem

Al viro broke us with

commit edd63a2763bdae0daa4f0a4d4c5d61d1154352a5
Author: Al Viro <viro@zeniv.linux.org.uk>
Date:   Fri Apr 27 13:42:45 2012 -0400

    set_restore_sigmask() is never called without SIGPENDING (and never should be)

Although it's pretty much our fault since parisc's asm/bug.h uses
BUGWARN_TAINT but doesn't include the file that defines it.  Fix that.
---
 arch/parisc/include/asm/bug.h |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/arch/parisc/include/asm/bug.h b/arch/parisc/include/asm/bug.h
index 72cfdb0..62a3333 100644
--- a/arch/parisc/include/asm/bug.h
+++ b/arch/parisc/include/asm/bug.h
@@ -1,6 +1,8 @@
 #ifndef _PARISC_BUG_H
 #define _PARISC_BUG_H
 
+#include <linux/kernel.h>	/* for BUGFLAG_TAINT */
+
 /*
  * Tell the user there is some problem.
  * The offending file and line are encoded in the __bug_table section.
-- 
1.7.4.1

[PATCH 2/3] fix compile break in use of lib/strncopy_from_user.c

[James Bottomley]

Linus broke us with

commit 36126f8f2ed8168eb13aa0662b9b9585cba100a9
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Sat May 26 10:43:17 2012 -0700

    word-at-a-time: make the interfaces truly generic

By moving functions defined in strncopy_from_user.c into the asm-geneic
version word-at-a-time.h.  Spark and OpenRisc were fixed to use this, but
not parisc.  Fix by adding to generic-y in asm/Kbuild
---
 arch/parisc/include/asm/Kbuild |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/arch/parisc/include/asm/Kbuild b/arch/parisc/include/asm/Kbuild
index 19a434f..4383707 100644
--- a/arch/parisc/include/asm/Kbuild
+++ b/arch/parisc/include/asm/Kbuild
@@ -1,3 +1,4 @@
 include include/asm-generic/Kbuild.asm
 
 header-y += pdc.h
+generic-y += word-at-a-time.h
-- 
1.7.4.1

[PATCH 3/3] fix code to find libgcc

[James Bottomley]

Sam broke this with

commit 1f2bfbd00e466ff3489b2ca5cc75b1cccd14c123
Author: Sam Ravnborg <sam@ravnborg.org>
Date:   Sat May 5 10:18:41 2012 +0200

    kbuild: link of vmlinux moved to a script

But we should be deriving the location of libgcc in the same way as all
the other archs, so fix by adding a LIBGCC variable which is evaluated
in the makefile
---
 arch/parisc/Makefile |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)

diff --git a/arch/parisc/Makefile b/arch/parisc/Makefile
index dbc3850..5707f1a 100644
--- a/arch/parisc/Makefile
+++ b/arch/parisc/Makefile
@@ -21,6 +21,7 @@ KBUILD_DEFCONFIG := default_defconfig
 
 NM		= sh $(srctree)/arch/parisc/nm
 CHECKFLAGS	+= -D__hppa__=1
+LIBGCC		= $(shell $(CC) $(KBUILD_CFLAGS) -print-libgcc-file-name)
 
 MACHINE		:= $(shell uname -m)
 ifeq ($(MACHINE),parisc*)
@@ -79,7 +80,7 @@ kernel-y			:= mm/ kernel/ math-emu/
 kernel-$(CONFIG_HPUX)		+= hpux/
 
 core-y	+= $(addprefix arch/parisc/, $(kernel-y))
-libs-y	+= arch/parisc/lib/ `$(CC) -print-libgcc-file-name`
+libs-y	+= arch/parisc/lib/ $(LIBGCC)
 
 drivers-$(CONFIG_OPROFILE)		+= arch/parisc/oprofile/
 
-- 
1.7.4.1

Re: [PATCH 3/3] fix code to find libgcc

[Sam Ravnborg]

On Tue, Jun 05, 2012 at 01:54:09PM +0900, James Bottomley wrote:
> Sam broke this with
Soory!
> 
> commit 1f2bfbd00e466ff3489b2ca5cc75b1cccd14c123
> Author: Sam Ravnborg <sam@ravnborg.org>
> Date:   Sat May 5 10:18:41 2012 +0200
> 
>     kbuild: link of vmlinux moved to a script
> 
> But we should be deriving the location of libgcc in the same way as all
> the other archs, so fix by adding a LIBGCC variable which is evaluated
> in the makefile
It actually took me some time to realise why the formar approach was wrong.
It is still morning here..

Thanks for fixing this!

Acked-by: Sam Ravnborg <sam@ravnborg.org>

	Sam