[PATCH 1/6] modules: Fold percpu_modcopy into module.c linux-2.6.git

[PATCH 1/6] modules: Fold percpu_modcopy into module.c linux-2.6.git

[travis]

percpu_modcopy() is defined multiple times in arch files. However, the only
user is module.c. Put a static definition into module.c and remove
the definitions from the arch files.

Based on latest linux-2.6.git

Cc: Andi Kleen <ak@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: David Miller <davem@davemloft.net>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com>
Cc: linuxppc-dev@ozlabs.org
Cc: linux-ia64@vger.kernel.org

Signed-off-by: Mike Travis <travis@sgi.com>
---
linux-2.6.git:
  - added back in missing pieces from x86.git merge
---
 arch/ia64/kernel/module.c    |   11 -----------
 include/asm-ia64/percpu.h    |    5 -----
 include/asm-powerpc/percpu.h |    9 ---------
 include/asm-s390/percpu.h    |    9 ---------
 4 files changed, 34 deletions(-)

--- a/arch/ia64/kernel/module.c
+++ b/arch/ia64/kernel/module.c
@@ -940,14 +940,3 @@ module_arch_cleanup (struct module *mod)
 	if (mod->arch.core_unw_table)
 		unw_remove_unwind_table(mod->arch.core_unw_table);
 }
-
-#ifdef CONFIG_SMP
-void
-percpu_modcopy (void *pcpudst, const void *src, unsigned long size)
-{
-	unsigned int i;
-	for_each_possible_cpu(i) {
-		memcpy(pcpudst + per_cpu_offset(i), src, size);
-	}
-}
-#endif /* CONFIG_SMP */
--- a/include/asm-ia64/percpu.h
+++ b/include/asm-ia64/percpu.h
@@ -22,10 +22,6 @@
 #define DECLARE_PER_CPU(type, name)				\
 	extern PER_CPU_ATTRIBUTES __typeof__(type) per_cpu__##name
 
-/*
- * Pretty much a literal copy of asm-generic/percpu.h, except that percpu_modcopy() is an
- * external routine, to avoid include-hell.
- */
 #ifdef CONFIG_SMP
 
 extern unsigned long __per_cpu_offset[NR_CPUS];
@@ -38,7 +34,6 @@ DECLARE_PER_CPU(unsigned long, local_per
 #define __get_cpu_var(var) (*RELOC_HIDE(&per_cpu__##var, __ia64_per_cpu_var(local_per_cpu_offset)))
 #define __raw_get_cpu_var(var) (*RELOC_HIDE(&per_cpu__##var, __ia64_per_cpu_var(local_per_cpu_offset)))
 
-extern void percpu_modcopy(void *pcpudst, const void *src, unsigned long size);
 extern void setup_per_cpu_areas (void);
 extern void *per_cpu_init(void);
 
--- a/include/asm-powerpc/percpu.h
+++ b/include/asm-powerpc/percpu.h
@@ -21,15 +21,6 @@
 #define __get_cpu_var(var) (*RELOC_HIDE(&per_cpu__##var, __my_cpu_offset()))
 #define __raw_get_cpu_var(var) (*RELOC_HIDE(&per_cpu__##var, local_paca->data_offset))
 
-/* A macro to avoid #include hell... */
-#define percpu_modcopy(pcpudst, src, size)			\
-do {								\
-	unsigned int __i;					\
-	for_each_possible_cpu(__i)				\
-		memcpy((pcpudst)+__per_cpu_offset(__i),		\
-		       (src), (size));				\
-} while (0)
-
 extern void setup_per_cpu_areas(void);
 
 #else /* ! SMP */
--- a/include/asm-s390/percpu.h
+++ b/include/asm-s390/percpu.h
@@ -39,15 +39,6 @@ extern unsigned long __per_cpu_offset[NR
 #define per_cpu(var,cpu) __reloc_hide(var,__per_cpu_offset[cpu])
 #define per_cpu_offset(x) (__per_cpu_offset[x])
 
-/* A macro to avoid #include hell... */
-#define percpu_modcopy(pcpudst, src, size)			\
-do {								\
-	unsigned int __i;					\
-	for_each_possible_cpu(__i)				\
-		memcpy((pcpudst)+__per_cpu_offset[__i],		\
-		       (src), (size));				\
-} while (0)
-
 #else /* ! SMP */
 
 #define __get_cpu_var(var) __reloc_hide(var,0)

-- 

[PATCH 2/6] percpu: Change Kconfig to HAVE_SETUP_PER_CPU_AREA linux-2.6.git

[travis]

Change:
	config ARCH_SETS_UP_PER_CPU_AREA
to:
	config HAVE_SETUP_PER_CPU_AREA

Based on latest linux-2.6.git

Cc: Andi Kleen <ak@suse.de>
Cc: Tony Luck <tony.luck@intel.com>
Cc: David Miller <davem@davemloft.net>
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Geert Uytterhoeven <Geert.Uytterhoeven@sonycom.com>
Cc: linuxppc-dev@ozlabs.org
Cc: linux-ia64@vger.kernel.org

Signed-off-by: Mike Travis <travis@sgi.com>
---
linux-2.6.git:
  - added back in missing pieces from x86.git merge

The change to using "select xxx" as suggested by Sam
requires an addition to a non-existant file (arch/Kconfig)
so I went back to using "config xxx" to introduce the flag.

---
 arch/ia64/Kconfig    |    2 +-
 arch/powerpc/Kconfig |    2 +-
 arch/sparc64/Kconfig |    2 +-
 init/main.c          |    2 ++
 4 files changed, 5 insertions(+), 3 deletions(-)

--- a/arch/ia64/Kconfig
+++ b/arch/ia64/Kconfig
@@ -80,7 +80,7 @@ config GENERIC_TIME_VSYSCALL
 	bool
 	default y
 
-config ARCH_SETS_UP_PER_CPU_AREA
+config HAVE_SETUP_PER_CPU_AREA
 	def_bool y
 
 config DMI
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -42,7 +42,7 @@ config GENERIC_HARDIRQS
 	bool
 	default y
 
-config ARCH_SETS_UP_PER_CPU_AREA
+config HAVE_SETUP_PER_CPU_AREA
 	def_bool PPC64
 
 config IRQ_PER_CPU
--- a/arch/sparc64/Kconfig
+++ b/arch/sparc64/Kconfig
@@ -66,7 +66,7 @@ config AUDIT_ARCH
 	bool
 	default y
 
-config ARCH_SETS_UP_PER_CPU_AREA
+config HAVE_SETUP_PER_CPU_AREA
 	def_bool y
 
 config ARCH_NO_VIRT_TO_BUS
--- a/init/main.c
+++ b/init/main.c
@@ -380,6 +380,8 @@ static void __init setup_per_cpu_areas(v
 
 	/* Copy section for each CPU (we discard the original) */
 	size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE);
+	printk(KERN_INFO
+	    "PERCPU: Allocating %lu bytes of per cpu data (main)\n", size);
 	ptr = alloc_bootmem_pages(size * nr_possible_cpus);
 
 	for_each_possible_cpu(i) {

-- 

Re: [PATCH 2/6] percpu: Change Kconfig to HAVE_SETUP_PER_CPU_AREA

[Ingo Molnar]

* travis@sgi.com <travis@sgi.com> wrote:

> Change:
> 	config ARCH_SETS_UP_PER_CPU_AREA
> to:
> 	config HAVE_SETUP_PER_CPU_AREA

undocumented change:

>  config ARCH_NO_VIRT_TO_BUS
> --- a/init/main.c
> +++ b/init/main.c
> @@ -380,6 +380,8 @@ static void __init setup_per_cpu_areas(v
>  
>  	/* Copy section for each CPU (we discard the original) */
>  	size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE);
> +	printk(KERN_INFO
> +	    "PERCPU: Allocating %lu bytes of per cpu data (main)\n", size);
>  	ptr = alloc_bootmem_pages(size * nr_possible_cpus);

but looks fine to me.

	Ingo

Re: [PATCH 2/6] percpu: Change Kconfig to HAVE_SETUP_PER_CPU_AREA

[Mike Travis]

Ingo Molnar wrote:
> * travis@sgi.com <travis@sgi.com> wrote:
> 
>> Change:
>> 	config ARCH_SETS_UP_PER_CPU_AREA
>> to:
>> 	config HAVE_SETUP_PER_CPU_AREA
> 
> undocumented change:
> 
>>  config ARCH_NO_VIRT_TO_BUS
>> --- a/init/main.c
>> +++ b/init/main.c
>> @@ -380,6 +380,8 @@ static void __init setup_per_cpu_areas(v
>>  
>>  	/* Copy section for each CPU (we discard the original) */
>>  	size = ALIGN(PERCPU_ENOUGH_ROOM, PAGE_SIZE);
>> +	printk(KERN_INFO
>> +	    "PERCPU: Allocating %lu bytes of per cpu data (main)\n", size);
>>  	ptr = alloc_bootmem_pages(size * nr_possible_cpus);
> 
> but looks fine to me.
> 
> 	Ingo

Sorry, I should have noted this.  The primary reason I put this in, is
that if the HAVE_SETUP_PER_CPU_AREA is not set when it should be, then
the incorrect (generic) setup_per_cpu_areas() is used and weird things
happen later on.  The above line documents that PERCPU has been allocated
by init/main.c version of this function in the startup messages.
(Since it's a static function, there is no "duplicate label" error in
the linker.)

Thanks,
Mike

[PATCH 4/6] ia64: Use generic percpu linux-2.6.git

[travis]

ia64 has a special processor specific mapping that can be used to locate the
offset for the current per cpu area.

Based on latest linux-2.6.git

Cc: Tony Luck <tony.luck@intel.com>
Cc: linux-ia64@vger.kernel.org

Signed-off-by: Mike Travis <travis@sgi.com>
---
linux-2.6.git:
  - added back in missing pieces from x86.git merge
---
 include/asm-ia64/percpu.h |   24 +++++++-----------------
 include/linux/percpu.h    |    4 ----
 2 files changed, 7 insertions(+), 21 deletions(-)

--- a/include/asm-ia64/percpu.h
+++ b/include/asm-ia64/percpu.h
@@ -19,29 +19,14 @@
 # define PER_CPU_ATTRIBUTES	__attribute__((__model__ (__small__)))
 #endif
 
-#define DECLARE_PER_CPU(type, name)				\
-	extern PER_CPU_ATTRIBUTES __typeof__(type) per_cpu__##name
-
 #ifdef CONFIG_SMP
 
-extern unsigned long __per_cpu_offset[NR_CPUS];
-#define per_cpu_offset(x) (__per_cpu_offset[x])
-
-/* Equal to __per_cpu_offset[smp_processor_id()], but faster to access: */
-DECLARE_PER_CPU(unsigned long, local_per_cpu_offset);
-
-#define per_cpu(var, cpu)  (*RELOC_HIDE(&per_cpu__##var, __per_cpu_offset[cpu]))
-#define __get_cpu_var(var) (*RELOC_HIDE(&per_cpu__##var, __ia64_per_cpu_var(local_per_cpu_offset)))
-#define __raw_get_cpu_var(var) (*RELOC_HIDE(&per_cpu__##var, __ia64_per_cpu_var(local_per_cpu_offset)))
+#define __my_cpu_offset	__ia64_per_cpu_var(local_per_cpu_offset)
 
-extern void setup_per_cpu_areas (void);
 extern void *per_cpu_init(void);
 
 #else /* ! SMP */
 
-#define per_cpu(var, cpu)			(*((void)(cpu), &per_cpu__##var))
-#define __get_cpu_var(var)			per_cpu__##var
-#define __raw_get_cpu_var(var)			per_cpu__##var
 #define per_cpu_init()				(__phys_per_cpu_start)
 
 #endif	/* SMP */
@@ -52,7 +37,12 @@ extern void *per_cpu_init(void);
  * On the positive side, using __ia64_per_cpu_var() instead of __get_cpu_var() is slightly
  * more efficient.
  */
-#define __ia64_per_cpu_var(var)	(per_cpu__##var)
+#define __ia64_per_cpu_var(var)	per_cpu__##var
+
+#include <asm-generic/percpu.h>
+
+/* Equal to __per_cpu_offset[smp_processor_id()], but faster to access: */
+DECLARE_PER_CPU(unsigned long, local_per_cpu_offset);
 
 #endif /* !__ASSEMBLY__ */
 
--- a/include/linux/percpu.h
+++ b/include/linux/percpu.h
@@ -9,10 +9,6 @@
 
 #include <asm/percpu.h>
 
-#ifndef PER_CPU_ATTRIBUTES
-#define PER_CPU_ATTRIBUTES
-#endif
-
 #ifdef CONFIG_SMP
 #define DEFINE_PER_CPU(type, name)					\
 	__attribute__((__section__(".data.percpu")))			\

--