Skip to contentSkip to conversation

bug.h: Fix up powerpc build regression.

PATCH In review Paul Mundt 3 participants 6 messages Jun 18, 2012 – Jun 25, 2012 Generic include/asm header files

lore
Share & tools
lore

Paul Mundt

author
#0 lore raw

PATCH 1/1 bug.h: Fix up powerpc build regression.

The asm-generic/bug.h __ASSEMBLY__ guarding is completely bogus, which tripped up the powerpc build when the kernel.h include was added:

In file included from include/asm-generic/bug.h:5:0,
from arch/powerpc/include/asm/bug.h:127,
from arch/powerpc/kernel/head_64.S:31:
include/linux/kernel.h:44:0: warning: "ALIGN" redefined [enabled by default] include/linux/linkage.h:57:0: note: this is the location of the previous definition include/linux/sysinfo.h: Assembler messages:
include/linux/sysinfo.h:7: Error: Unrecognized opcode: struct' include/linux/sysinfo.h:8: Error: Unrecognized opcode: __kernel_long_t'

Moving the __ASSEMBLY__ guard up and stashing the kernel.h include under it fixes this up, as well as covering the case the original fix was attempting to handle.

Tested-by Stephen Rothwell <[email protected]>
Acked-by Arnd Bergmann <[email protected]>
Signed-off-by Paul Mundt <[email protected]>

---

 include/asm-generic/bug.h |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)
include/asm-generic/bug.h +5 -3
index 9f02005..506ec19 100644
@@ -2,7 +2,6 @@
 #define _ASM_GENERIC_BUG_H
 
 #include <linux/compiler.h>
-#include <linux/kernel.h>
 
 #ifdef CONFIG_BUG
 

@@ -32,6 +31,9 @@ struct bug_entry {
 
 #endif	/* CONFIG_GENERIC_BUG */
 
+#ifndef __ASSEMBLY__
+#include <linux/kernel.h>
+
 /*
  * Don't use BUG() or BUG_ON() unless there's really no way out; one
  * example might be detecting data structure corruption in the middle

@@ -61,7 +63,6 @@ struct bug_entry {
  * to provide better diagnostics.
  */
 #ifndef __WARN_TAINT
-#ifndef __ASSEMBLY__
 extern __printf(3, 4)
 void warn_slowpath_fmt(const char *file, const int line,
 		       const char *fmt, ...);

@@ -70,7 +71,6 @@ void warn_slowpath_fmt_taint(const char *file, const int line, unsigned taint,
 			     const char *fmt, ...);
 extern void warn_slowpath_null(const char *file, const int line);
 #define WANT_WARN_ON_SLOWPATH
-#endif
 #define __WARN()		warn_slowpath_null(__FILE__, __LINE__)
 #define __WARN_printf(arg...)	warn_slowpath_fmt(__FILE__, __LINE__, arg)
 #define __WARN_printf_taint(taint, arg...)				\

@@ -203,4 +203,6 @@ extern void warn_slowpath_null(const char *file, const int line);
 # define WARN_ON_SMP(x)			({0;})
 #endif
 
+#endif /* __ASSEMBLY__ */
+
 #endif
Tested-by on v1 1/1
#1 lore raw
Quoted 12 lines · Paul Mundt

The asm-generic/bug.h __ASSEMBLY__ guarding is completely bogus, which tripped up the powerpc build when the kernel.h include was added:

In file included from include/asm-generic/bug.h:5:0,
from arch/powerpc/include/asm/bug.h:127,
from arch/powerpc/kernel/head_64.S:31:
include/linux/kernel.h:44:0: warning: "ALIGN" redefined [enabled by default] include/linux/linkage.h:57:0: note: this is the location of the previous definition include/linux/sysinfo.h: Assembler messages:
include/linux/sysinfo.h:7: Error: Unrecognized opcode: struct' include/linux/sysinfo.h:8: Error: Unrecognized opcode: __kernel_long_t'

Quoted 6 lines

Moving the __ASSEMBLY__ guard up and stashing the kernel.h include under it fixes this up, as well as covering the case the original fix was attempting to handle.

Tested-by Stephen Rothwell <[email protected]>
Tested-by Vaidyanathan Srinivasan <[email protected]>

This build fix works fine for me for powerpc (pseries).

--Vaidy

#2 lore raw

PATCH 1/1 bug.h: Fix up x86 build regression

In reply to Paul Mundt
Quoted 12 lines · Paul Mundt

The asm-generic/bug.h __ASSEMBLY__ guarding is completely bogus, which tripped up the powerpc build when the kernel.h include was added:

In file included from include/asm-generic/bug.h:5:0,
from arch/powerpc/include/asm/bug.h:127,
from arch/powerpc/kernel/head_64.S:31:
include/linux/kernel.h:44:0: warning: "ALIGN" redefined [enabled by default] include/linux/linkage.h:57:0: note: this is the location of the previous definition include/linux/sysinfo.h: Assembler messages:
include/linux/sysinfo.h:7: Error: Unrecognized opcode: struct' include/linux/sysinfo.h:8: Error: Unrecognized opcode: __kernel_long_t'

Quoted 7 lines

Moving the __ASSEMBLY__ guard up and stashing the kernel.h include under it fixes this up, as well as covering the case the original fix was attempting to handle.

Tested-by Stephen Rothwell <[email protected]>
Acked-by Arnd Bergmann <[email protected]>
Signed-off-by Paul Mundt <[email protected]>

This commit broke most x86 !CONFIG_BUG builds in -rc4:

In file included from /home/mingo/tip/arch/x86/include/asm/cpumask.h:4:0,

                 from /home/mingo/tip/arch/x86/include/asm/msr.h:18,
                 from /home/mingo/tip/arch/x86/include/asm/processor.h:20,
                 from /home/mingo/tip/arch/x86/include/asm/atomic.h:6,
                 from include/linux/atomic.h:4,
                 from include/linux/crypto.h:20,
                 from arch/x86/kernel/asm-offsets.c:8:
include/linux/cpumask.h: In function ‘cpumask_check’:
include/linux/cpumask.h:108:2: error: implicit declaration of function ‘WARN_ON_ONCE’ [-Werror=implicit-function-declaration]
In file included from include/linux/thread_info.h:54:0,
                 from include/linux/preempt.h:9,
                 from include/linux/spinlock.h:50,
                 from include/linux/mmzone.h:7,
                 from include/linux/gfp.h:4,
                 from include/linux/slab.h:12,
                 from include/linux/crypto.h:24,
                 from arch/x86/kernel/asm-offsets.c:8:
/home/mingo/tip/arch/x86/include/asm/thread_info.h: In function ‘set_restore_sigmask’:
/home/mingo/tip/arch/x86/include/asm/thread_info.h:251:2: error: implicit declaration of function ‘WARN_ON’ [-Werror=implicit-function-declaration]
In file included from /home/mingo/tip/arch/x86/include/asm/apic.h:12:0,
                 from /home/mingo/tip/arch/x86/include/asm/smp.h:13,
                 from include/linux/smp.h:38,
                 from include/linux/topology.h:33,
                 from include/linux/gfp.h:7,
                 from include/linux/slab.h:12,
                 from include/linux/crypto.h:24,
                 from arch/x86/kernel/asm-offsets.c:8:
/home/mingo/tip/arch/x86/include/asm/fixmap.h: In function ‘virt_to_fix’:
/home/mingo/tip/arch/x86/include/asm/fixmap.h:219:2: error: implicit declaration of function ‘BUG_ON’ [-Werror=implicit-function-declaration]
In file included from include/linux/slab.h:185:0,
                 from include/linux/crypto.h:24,
                 from arch/x86/kernel/asm-offsets.c:8:

include/linux/slub_def.h: In function ‘kmalloc_index’:
include/linux/slub_def.h:192:2: error: implicit declaration of function ‘BUG’ [-Werror=implicit-function-declaration]
cc1: some warnings being treated as errors

The commit mis-placed the !ASSEMBLY block, mixing it with the CONFIG_BUG section.

Hiding an #include <kernel.h> deep inside a generic header is in pretty bad taste anyway, it creates kconfig dependent breakages and makes it harder to untangle header files.

The patch below fixes most of the bad effects for me on x86, but I have not tested the PowerPC cross-build yet.

Signed-off-by Ingo Molnar <[email protected]>
include/asm-generic/bug.h +5 -5
index 506ec19..25565b7 100644
@@ -3,6 +3,10 @@
 
 #include <linux/compiler.h>
 
+#ifndef __ASSEMBLY__
+#include <linux/kernel.h>
+#endif /* __ASSEMBLY__ */
+
 #ifdef CONFIG_BUG
 
 #ifdef CONFIG_GENERIC_BUG

@@ -31,9 +35,6 @@ struct bug_entry {
 
 #endif	/* CONFIG_GENERIC_BUG */
 
-#ifndef __ASSEMBLY__
-#include <linux/kernel.h>
-
 /*
  * Don't use BUG() or BUG_ON() unless there's really no way out; one
  * example might be detecting data structure corruption in the middle

@@ -108,6 +109,7 @@ extern void warn_slowpath_null(const char *file, const int line);
 })
 
 #else /* !CONFIG_BUG */
+
 #ifndef HAVE_ARCH_BUG
 #define BUG() do {} while(0)
 #endif

@@ -203,6 +205,4 @@ extern void warn_slowpath_null(const char *file, const int line);
 # define WARN_ON_SMP(x)			({0;})
 #endif
 
-#endif /* __ASSEMBLY__ */
-
 #endif
#3 lore raw

PATCH v2 1/1 bug.h: Fix up x86 build regression

Quoted 64 lines · Ingo Molnar
Quoted 12 lines · Paul Mundt

The asm-generic/bug.h __ASSEMBLY__ guarding is completely bogus, which tripped up the powerpc build when the kernel.h include was added:

In file included from include/asm-generic/bug.h:5:0,
from arch/powerpc/include/asm/bug.h:127,
from arch/powerpc/kernel/head_64.S:31:
include/linux/kernel.h:44:0: warning: "ALIGN" redefined [enabled by default] include/linux/linkage.h:57:0: note: this is the location of the previous definition include/linux/sysinfo.h: Assembler messages:
include/linux/sysinfo.h:7: Error: Unrecognized opcode: struct' include/linux/sysinfo.h:8: Error: Unrecognized opcode: __kernel_long_t'

Quoted 7 lines

Moving the __ASSEMBLY__ guard up and stashing the kernel.h include under it fixes this up, as well as covering the case the original fix was attempting to handle.

Tested-by Stephen Rothwell <[email protected]>
Acked-by Arnd Bergmann <[email protected]>
Signed-off-by Paul Mundt <[email protected]>

This commit broke most x86 !CONFIG_BUG builds in -rc4:

In file included from /home/mingo/tip/arch/x86/include/asm/cpumask.h:4:0,

                 from /home/mingo/tip/arch/x86/include/asm/msr.h:18,
                 from /home/mingo/tip/arch/x86/include/asm/processor.h:20,
                 from /home/mingo/tip/arch/x86/include/asm/atomic.h:6,
                 from include/linux/atomic.h:4,
                 from include/linux/crypto.h:20,
                 from arch/x86/kernel/asm-offsets.c:8:
include/linux/cpumask.h: In function ‘cpumask_check’:
include/linux/cpumask.h:108:2: error: implicit declaration of function ‘WARN_ON_ONCE’ [-Werror=implicit-function-declaration]
In file included from include/linux/thread_info.h:54:0,
                 from include/linux/preempt.h:9,
                 from include/linux/spinlock.h:50,
                 from include/linux/mmzone.h:7,
                 from include/linux/gfp.h:4,
                 from include/linux/slab.h:12,
                 from include/linux/crypto.h:24,
                 from arch/x86/kernel/asm-offsets.c:8:
/home/mingo/tip/arch/x86/include/asm/thread_info.h: In function ‘set_restore_sigmask’:
/home/mingo/tip/arch/x86/include/asm/thread_info.h:251:2: error: implicit declaration of function ‘WARN_ON’ [-Werror=implicit-function-declaration]
In file included from /home/mingo/tip/arch/x86/include/asm/apic.h:12:0,
                 from /home/mingo/tip/arch/x86/include/asm/smp.h:13,
                 from include/linux/smp.h:38,
                 from include/linux/topology.h:33,
                 from include/linux/gfp.h:7,
                 from include/linux/slab.h:12,
                 from include/linux/crypto.h:24,
                 from arch/x86/kernel/asm-offsets.c:8:
/home/mingo/tip/arch/x86/include/asm/fixmap.h: In function ‘virt_to_fix’:
/home/mingo/tip/arch/x86/include/asm/fixmap.h:219:2: error: implicit declaration of function ‘BUG_ON’ [-Werror=implicit-function-declaration]
In file included from include/linux/slab.h:185:0,
                 from include/linux/crypto.h:24,
                 from arch/x86/kernel/asm-offsets.c:8:

include/linux/slub_def.h: In function ‘kmalloc_index’:
include/linux/slub_def.h:192:2: error: implicit declaration of function ‘BUG’ [-Werror=implicit-function-declaration]
cc1: some warnings being treated as errors

The commit mis-placed the !ASSEMBLY block, mixing it with the CONFIG_BUG section.

Quoted 6 lines

Hiding an #include <kernel.h> deep inside a generic header is in pretty bad taste anyway, it creates kconfig dependent breakages and makes it harder to untangle header files.

The patch below fixes most of the bad effects for me on x86, but I have not tested the PowerPC cross-build yet.

I did a PowerPC cross-build and it was not quite there yet, but the second version below does the trick and fixes x86 and powerpc as well.

Linus, if there are no objections to the patch then I can push this fix to you via tip:core/urgent (where it's queued up currently), or you could apply it via email as well.

Thanks,

Ingo

------------->

Quoted 1 line

From 7b37393f13d511fa076d4527e35e86cc45d974c5 Mon Sep 17 00:00:00 2001

From: Ingo Molnar <[email protected]>
Date: Mon, 25 Jun 2012 11:42:15 +0200
Subject: [PATCH] bug.h: Fix x86 !CONFIG_BUG build regression
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Commit 2603efa31a03 ("bug.h: Fix up powerpc build regression.") broke most x86 !CONFIG_BUG builds in -rc4:

include/linux/cpumask.h: In function ‘cpumask_check’:
include/linux/cpumask.h:108:2: error: implicit declaration of function ‘WARN_ON_ONCE’ [...]

The commit mis-placed the !ASSEMBLY block, mixing it with the CONFIG_BUG section.

Hiding an #include <kernel.h> deep inside a generic header is in pretty bad taste anyway, it creates kconfig dependent breakages and makes it harder to untangle header files.

Signed-off-by Ingo Molnar <[email protected]>
Cc Linus Torvalds <[email protected]>
Cc Stephen Rothwell <[email protected]>
Cc Randy Dunlap <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Paul Mundt <[email protected]>
Link http://lkml.kernel.org/r/[email protected] open here
Signed-off-by Ingo Molnar <[email protected]>
---
 include/asm-generic/bug.h |   14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)
include/asm-generic/bug.h +9 -5
index 506ec19..53f3eb1 100644
@@ -3,6 +3,10 @@
 
 #include <linux/compiler.h>
 
+#ifndef __ASSEMBLY__
+# include <linux/kernel.h>
+#endif
+
 #ifdef CONFIG_BUG
 
 #ifdef CONFIG_GENERIC_BUG

@@ -31,9 +35,6 @@ struct bug_entry {
 
 #endif	/* CONFIG_GENERIC_BUG */
 
-#ifndef __ASSEMBLY__
-#include <linux/kernel.h>
-
 /*
  * Don't use BUG() or BUG_ON() unless there's really no way out; one
  * example might be detecting data structure corruption in the middle

@@ -63,6 +64,8 @@ struct bug_entry {
  * to provide better diagnostics.
  */
 #ifndef __WARN_TAINT
+
+#ifndef __ASSEMBLY__
 extern __printf(3, 4)
 void warn_slowpath_fmt(const char *file, const int line,
 		       const char *fmt, ...);

@@ -70,6 +73,8 @@ extern __printf(4, 5)
 void warn_slowpath_fmt_taint(const char *file, const int line, unsigned taint,
 			     const char *fmt, ...);
 extern void warn_slowpath_null(const char *file, const int line);
+#endif /* __ASSEMBLY__ */
+
 #define WANT_WARN_ON_SLOWPATH
 #define __WARN()		warn_slowpath_null(__FILE__, __LINE__)
 #define __WARN_printf(arg...)	warn_slowpath_fmt(__FILE__, __LINE__, arg)

@@ -108,6 +113,7 @@ extern void warn_slowpath_null(const char *file, const int line);
 })
 
 #else /* !CONFIG_BUG */
+
 #ifndef HAVE_ARCH_BUG
 #define BUG() do {} while(0)
 #endif

@@ -203,6 +209,4 @@ extern void warn_slowpath_null(const char *file, const int line);
 # define WARN_ON_SMP(x)			({0;})
 #endif
 
-#endif /* __ASSEMBLY__ */
-
 #endif
#4 lore raw

[tip:core/urgent] bug.h: Fix x86 !CONFIG_BUG build regression

In reply to Ingo Molnar
Commit-ID:  7b37393f13d511fa076d4527e35e86cc45d974c5
Gitweb:     http://git.kernel.org/tip/7b37393f13d511fa076d4527e35e86cc45d974c5
Author:     Ingo Molnar <[email protected]>
AuthorDate: Mon, 25 Jun 2012 11:42:15 +0200
Committer:  Ingo Molnar <[email protected]>
CommitDate: Mon, 25 Jun 2012 12:48:29 +0200

bug.h: Fix x86 !CONFIG_BUG build regression

Commit 2603efa31a03 ("bug.h: Fix up powerpc build regression.") broke most x86 !CONFIG_BUG builds in -rc4:

include/linux/cpumask.h: In function ‘cpumask_check’:
include/linux/cpumask.h:108:2: error: implicit declaration of function ‘WARN_ON_ONCE’ [...]

The commit mis-placed the !ASSEMBLY block, mixing it with the CONFIG_BUG section.

Hiding an #include <kernel.h> deep inside a generic header is in pretty bad taste anyway, it creates kconfig dependent breakages and makes it harder to untangle header files.

Signed-off-by Ingo Molnar <[email protected]>
Cc Linus Torvalds <[email protected]>
Cc Stephen Rothwell <[email protected]>
Cc Randy Dunlap <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Paul Mundt <[email protected]>
Link http://lkml.kernel.org/r/[email protected] open here
Signed-off-by Ingo Molnar <[email protected]>
---
 include/asm-generic/bug.h |   14 +++++++++-----
 1 files changed, 9 insertions(+), 5 deletions(-)
include/asm-generic/bug.h +9 -5
index 506ec19..53f3eb1 100644
@@ -3,6 +3,10 @@
 
 #include <linux/compiler.h>
 
+#ifndef __ASSEMBLY__
+# include <linux/kernel.h>
+#endif
+
 #ifdef CONFIG_BUG
 
 #ifdef CONFIG_GENERIC_BUG

@@ -31,9 +35,6 @@ struct bug_entry {
 
 #endif	/* CONFIG_GENERIC_BUG */
 
-#ifndef __ASSEMBLY__
-#include <linux/kernel.h>
-
 /*
  * Don't use BUG() or BUG_ON() unless there's really no way out; one
  * example might be detecting data structure corruption in the middle

@@ -63,6 +64,8 @@ struct bug_entry {
  * to provide better diagnostics.
  */
 #ifndef __WARN_TAINT
+
+#ifndef __ASSEMBLY__
 extern __printf(3, 4)
 void warn_slowpath_fmt(const char *file, const int line,
 		       const char *fmt, ...);

@@ -70,6 +73,8 @@ extern __printf(4, 5)
 void warn_slowpath_fmt_taint(const char *file, const int line, unsigned taint,
 			     const char *fmt, ...);
 extern void warn_slowpath_null(const char *file, const int line);
+#endif /* __ASSEMBLY__ */
+
 #define WANT_WARN_ON_SLOWPATH
 #define __WARN()		warn_slowpath_null(__FILE__, __LINE__)
 #define __WARN_printf(arg...)	warn_slowpath_fmt(__FILE__, __LINE__, arg)

@@ -108,6 +113,7 @@ extern void warn_slowpath_null(const char *file, const int line);
 })
 
 #else /* !CONFIG_BUG */
+
 #ifndef HAVE_ARCH_BUG
 #define BUG() do {} while(0)
 #endif

@@ -203,6 +209,4 @@ extern void warn_slowpath_null(const char *file, const int line);
 # define WARN_ON_SMP(x)			({0;})
 #endif
 
-#endif /* __ASSEMBLY__ */
-
 #endif
author
#5 lore raw

Re: [PATCH] bug.h: Fix up x86 build regression

In reply to Ingo Molnar
Quoted 19 lines · Ingo Molnar
Quoted 16 lines · Paul Mundt

The asm-generic/bug.h __ASSEMBLY__ guarding is completely bogus, which tripped up the powerpc build when the kernel.h include was added:

In file included from include/asm-generic/bug.h:5:0,
from arch/powerpc/include/asm/bug.h:127,
from arch/powerpc/kernel/head_64.S:31:
include/linux/kernel.h:44:0: warning: "ALIGN" redefined [enabled by default] include/linux/linkage.h:57:0: note: this is the location of the previous definition include/linux/sysinfo.h: Assembler messages:
include/linux/sysinfo.h:7: Error: Unrecognized opcode: struct' include/linux/sysinfo.h:8: Error: Unrecognized opcode: __kernel_long_t'

Moving the __ASSEMBLY__ guard up and stashing the kernel.h include under it fixes this up, as well as covering the case the original fix was attempting to handle.

Quoted 6 lines
Quoted 3 lines
Tested-by Stephen Rothwell <[email protected]>
Acked-by Arnd Bergmann <[email protected]>
Signed-off-by Paul Mundt <[email protected]>

This commit broke most x86 !CONFIG_BUG builds in -rc4:

Yes, that was reported already and I've sent out a fix:

https://lkml.org/lkml/2012/6/25/98