From 0a3aee0da4402aa19b66e458038533c896fb80c6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 18 Dec 2010 16:28:55 +0100 Subject: x86: Use this_cpu_ops to optimize code Go through x86 code and replace __get_cpu_var and get_cpu_var instances that refer to a scalar and are not used for address determinations. Cc: Yinghai Lu Cc: Ingo Molnar Acked-by: Tejun Heo Acked-by: "H. Peter Anvin" Signed-off-by: Christoph Lameter Signed-off-by: Tejun Heo --- arch/x86/kernel/smpboot.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86/kernel/smpboot.c') diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 083e99d1b7df..ff4e5a113a5b 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -1377,7 +1377,7 @@ void play_dead_common(void) mb(); /* Ack it */ - __get_cpu_var(cpu_state) = CPU_DEAD; + __this_cpu_write(cpu_state, CPU_DEAD); /* * With physical CPU hotplug, we should halt the cpu -- cgit From 7b543a5334ff4ea2e3ad3b777fc23cdb8072a988 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 18 Dec 2010 16:30:05 +0100 Subject: x86: Replace uses of current_cpu_data with this_cpu ops Replace all uses of current_cpu_data with this_cpu operations on the per cpu structure cpu_info. The scala accesses are replaced with the matching this_cpu ops which results in smaller and more efficient code. In the long run, it might be a good idea to remove cpu_data() macro too and use per_cpu macro directly. tj: updated description Cc: Yinghai Lu Cc: Ingo Molnar Acked-by: H. Peter Anvin Acked-by: Tejun Heo Signed-off-by: Christoph Lameter Signed-off-by: Tejun Heo --- arch/x86/kernel/smpboot.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) (limited to 'arch/x86/kernel/smpboot.c') diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index ff4e5a113a5b..0720071086d1 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -430,7 +430,7 @@ void __cpuinit set_cpu_sibling_map(int cpu) cpumask_set_cpu(cpu, c->llc_shared_map); - if (current_cpu_data.x86_max_cores == 1) { + if (__this_cpu_read(cpu_info.x86_max_cores) == 1) { cpumask_copy(cpu_core_mask(cpu), cpu_sibling_mask(cpu)); c->booted_cores = 1; return; @@ -1094,7 +1094,7 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) preempt_disable(); smp_cpu_index_default(); - current_cpu_data = boot_cpu_data; + memcpy(__this_cpu_ptr(&cpu_info), &boot_cpu_data, sizeof(cpu_info)); cpumask_copy(cpu_callin_mask, cpumask_of(0)); mb(); /* @@ -1397,11 +1397,11 @@ static inline void mwait_play_dead(void) int i; void *mwait_ptr; - if (!cpu_has(¤t_cpu_data, X86_FEATURE_MWAIT)) + if (!cpu_has(__this_cpu_ptr(&cpu_info), X86_FEATURE_MWAIT)) return; - if (!cpu_has(¤t_cpu_data, X86_FEATURE_CLFLSH)) + if (!cpu_has(__this_cpu_ptr(&cpu_info), X86_FEATURE_CLFLSH)) return; - if (current_cpu_data.cpuid_level < CPUID_MWAIT_LEAF) + if (__this_cpu_read(cpu_info.cpuid_level) < CPUID_MWAIT_LEAF) return; eax = CPUID_MWAIT_LEAF; @@ -1452,7 +1452,7 @@ static inline void mwait_play_dead(void) static inline void hlt_play_dead(void) { - if (current_cpu_data.x86 >= 4) + if (__this_cpu_read(cpu_info.x86) >= 4) wbinvd(); while (1) { -- cgit