cpuid leaf 0xb provides extended topology enumeration. This interface provides the 32-bit x2APIC id of the logical processor and it also provides a new mechanism to detect SMT and core siblings (which provides increased addressability). v2: Fix for !SMP build Signed-off-by: Suresh Siddha --- Index: tip/arch/x86/kernel/cpu/addon_cpuid_features.c =================================================================== --- tip.orig/arch/x86/kernel/cpu/addon_cpuid_features.c 2008-08-25 11:25:43.000000000 -0700 +++ tip/arch/x86/kernel/cpu/addon_cpuid_features.c 2008-08-25 12:06:46.000000000 -0700 @@ -7,6 +7,8 @@ #include #include +#include + struct cpuid_bit { u16 feature; u8 reg; @@ -48,6 +50,92 @@ } } +/* leaf 0xb SMT level */ +#define SMT_LEVEL 0 + +/* leaf 0xb sub-leaf types */ +#define INVALID_TYPE 0 +#define SMT_TYPE 1 +#define CORE_TYPE 2 + +#define LEAFB_SUBTYPE(ecx) (((ecx) >> 8) & 0xff) +#define BITS_SHIFT_NEXT_LEVEL(eax) ((eax) & 0x1f) +#define LEVEL_MAX_SIBLINGS(ebx) ((ebx) & 0xffff) + +/* + * Check for extended topology enumeration cpuid leaf 0xb and if it + * exists, use it for populating initial_apicid and cpu topology + * detection. + */ +void __cpuinit detect_extended_topology(struct cpuinfo_x86 *c) +{ +#ifdef CONFIG_SMP + unsigned int eax, ebx, ecx, edx, sub_index; + unsigned int ht_mask_width, core_plus_mask_width; + unsigned int core_select_mask, core_level_siblings; + + if (c->cpuid_level < 0xb) + return; + + cpuid_count(0xb, SMT_LEVEL, &eax, &ebx, &ecx, &edx); + + /* + * check if the cpuid leaf 0xb is actually implemented. + */ + if (ebx == 0 || (LEAFB_SUBTYPE(ecx) != SMT_TYPE)) + return; + + set_cpu_cap(c, X86_FEATURE_XTOPOLOGY); + + /* + * initial apic id, which also represents 32-bit extended x2apic id. + */ + c->initial_apicid = edx; + + /* + * Populate HT related information from sub-leaf level 0. + */ + core_level_siblings = smp_num_siblings = LEVEL_MAX_SIBLINGS(ebx); + core_plus_mask_width = ht_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + + sub_index = 1; + do { + cpuid_count(0xb, sub_index, &eax, &ebx, &ecx, &edx); + + /* + * Check for the Core type in the implemented sub leaves. + */ + if (LEAFB_SUBTYPE(ecx) == CORE_TYPE) { + core_level_siblings = LEVEL_MAX_SIBLINGS(ebx); + core_plus_mask_width = BITS_SHIFT_NEXT_LEVEL(eax); + break; + } + + sub_index++; + } while (LEAFB_SUBTYPE(ecx) != INVALID_TYPE); + + core_select_mask = (~(-1 << core_plus_mask_width)) >> ht_mask_width; + +#ifdef CONFIG_X86_32 + c->cpu_core_id = phys_pkg_id(c->initial_apicid, ht_mask_width) + & core_select_mask; + c->phys_proc_id = phys_pkg_id(c->initial_apicid, core_plus_mask_width); +#else + c->cpu_core_id = phys_pkg_id(ht_mask_width) & core_select_mask; + c->phys_proc_id = phys_pkg_id(core_plus_mask_width); +#endif + c->x86_max_cores = (core_level_siblings / smp_num_siblings); + + + printk(KERN_INFO "CPU: Physical Processor ID: %d\n", + c->phys_proc_id); + if (c->x86_max_cores > 1) + printk(KERN_INFO "CPU: Processor Core ID: %d\n", + c->cpu_core_id); + return; +#endif +} + #ifdef CONFIG_X86_PAT void __cpuinit validate_pat_support(struct cpuinfo_x86 *c) { Index: tip/arch/x86/kernel/cpu/common_64.c =================================================================== --- tip.orig/arch/x86/kernel/cpu/common_64.c 2008-08-25 11:25:43.000000000 -0700 +++ tip/arch/x86/kernel/cpu/common_64.c 2008-08-25 11:25:55.000000000 -0700 @@ -128,6 +128,9 @@ u32 eax, ebx, ecx, edx; int index_msb, core_bits; + if (cpu_has(c, X86_FEATURE_XTOPOLOGY)) + return; + cpuid(1, &eax, &ebx, &ecx, &edx); Index: tip/arch/x86/kernel/cpu/intel_64.c =================================================================== --- tip.orig/arch/x86/kernel/cpu/intel_64.c 2008-08-25 11:25:43.000000000 -0700 +++ tip/arch/x86/kernel/cpu/intel_64.c 2008-08-25 11:25:55.000000000 -0700 @@ -80,7 +80,10 @@ if (c->x86 == 6) set_cpu_cap(c, X86_FEATURE_REP_GOOD); set_cpu_cap(c, X86_FEATURE_LFENCE_RDTSC); - c->x86_max_cores = intel_num_cpu_cores(c); + + detect_extended_topology(c); + if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) + c->x86_max_cores = intel_num_cpu_cores(c); srat_detect_node(); } Index: tip/include/asm-x86/cpufeature.h =================================================================== --- tip.orig/include/asm-x86/cpufeature.h 2008-08-25 11:25:43.000000000 -0700 +++ tip/include/asm-x86/cpufeature.h 2008-08-25 11:25:55.000000000 -0700 @@ -81,6 +81,7 @@ #define X86_FEATURE_LFENCE_RDTSC (3*32+18) /* Lfence synchronizes RDTSC */ #define X86_FEATURE_11AP (3*32+19) /* Bad local APIC aka 11AP */ #define X86_FEATURE_NOPL (3*32+20) /* The NOPL (0F 1F) instructions */ +#define X86_FEATURE_XTOPOLOGY (3*32+21) /* cpu topology enum extensions */ /* Intel-defined CPU features, CPUID level 0x00000001 (ecx), word 4 */ #define X86_FEATURE_XMM3 (4*32+ 0) /* Streaming SIMD Extensions-3 */ Index: tip/include/asm-x86/processor.h =================================================================== --- tip.orig/include/asm-x86/processor.h 2008-08-25 11:25:43.000000000 -0700 +++ tip/include/asm-x86/processor.h 2008-08-25 11:25:55.000000000 -0700 @@ -166,6 +166,7 @@ extern unsigned int init_intel_cacheinfo(struct cpuinfo_x86 *c); extern unsigned short num_cache_leaves; +extern void detect_extended_topology(struct cpuinfo_x86 *c); #if defined(CONFIG_X86_HT) || defined(CONFIG_X86_64) extern void detect_ht(struct cpuinfo_x86 *c); #else Index: tip/arch/x86/kernel/cpu/intel.c =================================================================== --- tip.orig/arch/x86/kernel/cpu/intel.c 2008-08-25 11:25:43.000000000 -0700 +++ tip/arch/x86/kernel/cpu/intel.c 2008-08-25 11:25:55.000000000 -0700 @@ -176,9 +176,16 @@ if (p) strcpy(c->x86_model_id, p); - c->x86_max_cores = num_cpu_cores(c); + detect_extended_topology(c); - detect_ht(c); + if (!cpu_has(c, X86_FEATURE_XTOPOLOGY)) { + /* + * let's use the legacy cpuid vector 0x1 and 0x4 for topology + * detection. + */ + c->x86_max_cores = num_cpu_cores(c); + detect_ht(c); + } /* Work around errata */ Intel_errata_workarounds(c); -- -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/