percpu changes for v2.6.40

Hello, Linus.

Please pull from the following branch.

git://git.kernel.org/pub/scm/linux/kernel/git/tj/percpu.git for-2.6.40

The only interesting one is Eric's extra NOP removal from percpu_cmpxchg16b_double(). Other commits shouldn't cause noticeable behavior difference. The extra merge commit is to pull in non-critical fix patches which were waiting for further fixes during 2.6.39 cycle (only the merge commit itself is new).

When pulled into the current mainline, the following conflict occurs.

#ifdef CONFIG_SMP
#define CMPXCHG16B_EMU_CALL "call this_cpu_cmpxchg16b_emu\n\t" P6_NOP3 #else
#define CMPXCHG16B_EMU_CALL "call this_cpu_cmpxchg16b_emu\n\t" P6_NOP2 #endif
#define percpu_cmpxchg16b_double(pcp1, o1, o2, n1, n2) \

  ({									\
	  char __ret;							\
	  typeof(o1) __o1 = o1;						\
	  typeof(o1) __n1 = n1;						\
	  typeof(o2) __o2 = o2;						\
	  typeof(o2) __n2 = n2;						\
	  typeof(o2) __dummy;						\
  <<<<<<< HEAD
	  alternative_io("call this_cpu_cmpxchg16b_emu\n\t" ASM_NOP4,	\
  =======
	  alternative_io(CMPXCHG16B_EMU_CALL,				\

			 "cmpxchg16b " __percpu_prefix "(%%rsi)\n\tsetz %0\n\t",	\
			 X86_FEATURE_CX16,				\
			 ASM_OUTPUT2("=a"(__ret), "=d"(__dummy)),		\
			 "S" (&pcp1), "b"(__n1), "c"(__n2),		\
			 "a"(__o1), "d"(__o2) : "memory");		\
	  __ret;								\

})

This is caused between b1e7734f02 (x86, percpu: Use ASM_NOP4 instead of hardcoding P6_NOP4) and 5f55924dea (percpu: Avoid extra NOP in percpu_cmpxchg16b_double). The former updates P6_NOP4 used in percpu_cmpxchg16b_double() to ASM_NOP4 while the latter makes the whole thing into CMPXCHG16B_EMU_CALL macro which changes depending on CONFIG_SMP.

Updating CMPXCHG16B_EMU_CALL macro to use ASM_NOP* and using it resolves the conflict, which looks like the following.

#ifdef CONFIG_SMP
#define CMPXCHG16B_EMU_CALL "call this_cpu_cmpxchg16b_emu\n\t" ASM_NOP3 #else
#define CMPXCHG16B_EMU_CALL "call this_cpu_cmpxchg16b_emu\n\t" ASM_NOP2 #endif
#define percpu_cmpxchg16b_double(pcp1, o1, o2, n1, n2) \

  ({									\
	  char __ret;							\
	  typeof(o1) __o1 = o1;						\
	  typeof(o1) __n1 = n1;						\
	  typeof(o2) __o2 = o2;						\
	  typeof(o2) __n2 = n2;						\
	  typeof(o2) __dummy;						\
	  alternative_io(CMPXCHG16B_EMU_CALL,				\
			 "cmpxchg16b " __percpu_prefix "(%%rsi)\n\tsetz %0\n\t",	\
			 X86_FEATURE_CX16,				\
			 ASM_OUTPUT2("=a"(__ret), "=d"(__dummy)),		\
			 "S" (&pcp1), "b"(__n1), "c"(__n2),		\
			 "a"(__o1), "d"(__o2) : "memory");		\
	  __ret;								\

})

Just in case, the merged tree is available in the following git branch.

git://git.kernel.org/pub/scm/linux/kernel/git/tj/percpu.git merged-2.6.40

Thanks.

Eric Dumazet (1):
percpu: Avoid extra NOP in percpu_cmpxchg16b_double

Mike Frysinger (2):
percpu: Cast away printk format warning
percpu: Unify input section names

Tejun Heo (2):
percpu: Always align percpu output section to PAGE_SIZE Merge branch 'fixes-2.6.39' into for-2.6.40

 arch/alpha/kernel/vmlinux.lds.S    |    2 +-
 arch/arm/kernel/vmlinux.lds.S      |    2 +-
 arch/blackfin/kernel/vmlinux.lds.S |    2 +-
 arch/cris/kernel/vmlinux.lds.S     |    2 +-
 arch/frv/kernel/vmlinux.lds.S      |    2 +-
 arch/m32r/kernel/vmlinux.lds.S     |    2 +-
 arch/mips/kernel/vmlinux.lds.S     |    2 +-
 arch/mn10300/kernel/vmlinux.lds.S  |    2 +-
 arch/parisc/kernel/vmlinux.lds.S   |    2 +-
 arch/powerpc/kernel/vmlinux.lds.S  |    2 +-
 arch/s390/kernel/vmlinux.lds.S     |    2 +-
 arch/sh/kernel/vmlinux.lds.S       |    2 +-
 arch/sparc/kernel/vmlinux.lds.S    |    2 +-
 arch/tile/kernel/vmlinux.lds.S     |    2 +-
 arch/um/include/asm/common.lds.S   |    2 +-
 arch/x86/include/asm/percpu.h      |    7 +++-
 arch/x86/kernel/vmlinux.lds.S      |    2 +-
 arch/xtensa/kernel/vmlinux.lds.S   |    2 +-
 include/asm-generic/vmlinux.lds.h  |   61 +++++++++++++++++++-----------------
 kernel/workqueue.c                 |    4 +--
 mm/percpu.c                        |    6 ++-
 21 files changed, 60 insertions(+), 52 deletions(-)

--
tejun