1=========================================== 2How CPU topology info is exported via sysfs 3=========================================== 4 5Export CPU topology info via sysfs. Items (attributes) are similar 6to /proc/cpuinfo output of some architectures. They reside in 7/sys/devices/system/cpu/cpuX/topology/: 8 9physical_package_id: 10 11 physical package id of cpuX. Typically corresponds to a physical 12 socket number, but the actual value is architecture and platform 13 dependent. 14 15die_id: 16 17 the CPU die ID of cpuX. Typically it is the hardware platform's 18 identifier (rather than the kernel's). The actual value is 19 architecture and platform dependent. 20 21core_id: 22 23 the CPU core ID of cpuX. Typically it is the hardware platform's 24 identifier (rather than the kernel's). The actual value is 25 architecture and platform dependent. 26 27book_id: 28 29 the book ID of cpuX. Typically it is the hardware platform's 30 identifier (rather than the kernel's). The actual value is 31 architecture and platform dependent. 32 33drawer_id: 34 35 the drawer ID of cpuX. Typically it is the hardware platform's 36 identifier (rather than the kernel's). The actual value is 37 architecture and platform dependent. 38 39core_cpus: 40 41 internal kernel map of CPUs within the same core. 42 (deprecated name: "thread_siblings") 43 44core_cpus_list: 45 46 human-readable list of CPUs within the same core. 47 (deprecated name: "thread_siblings_list"); 48 49package_cpus: 50 51 internal kernel map of the CPUs sharing the same physical_package_id. 52 (deprecated name: "core_siblings") 53 54package_cpus_list: 55 56 human-readable list of CPUs sharing the same physical_package_id. 57 (deprecated name: "core_siblings_list") 58 59die_cpus: 60 61 internal kernel map of CPUs within the same die. 62 63die_cpus_list: 64 65 human-readable list of CPUs within the same die. 66 67book_siblings: 68 69 internal kernel map of cpuX's hardware threads within the same 70 book_id. 71 72book_siblings_list: 73 74 human-readable list of cpuX's hardware threads within the same 75 book_id. 76 77drawer_siblings: 78 79 internal kernel map of cpuX's hardware threads within the same 80 drawer_id. 81 82drawer_siblings_list: 83 84 human-readable list of cpuX's hardware threads within the same 85 drawer_id. 86 87Architecture-neutral, drivers/base/topology.c, exports these attributes. 88However, the book and drawer related sysfs files will only be created if 89CONFIG_SCHED_BOOK and CONFIG_SCHED_DRAWER are selected, respectively. 90 91CONFIG_SCHED_BOOK and CONFIG_SCHED_DRAWER are currently only used on s390, 92where they reflect the cpu and cache hierarchy. 93 94For an architecture to support this feature, it must define some of 95these macros in include/asm-XXX/topology.h:: 96 97 #define topology_physical_package_id(cpu) 98 #define topology_die_id(cpu) 99 #define topology_core_id(cpu) 100 #define topology_book_id(cpu) 101 #define topology_drawer_id(cpu) 102 #define topology_sibling_cpumask(cpu) 103 #define topology_core_cpumask(cpu) 104 #define topology_die_cpumask(cpu) 105 #define topology_book_cpumask(cpu) 106 #define topology_drawer_cpumask(cpu) 107 108The type of ``**_id macros`` is int. 109The type of ``**_cpumask macros`` is ``(const) struct cpumask *``. The latter 110correspond with appropriate ``**_siblings`` sysfs attributes (except for 111topology_sibling_cpumask() which corresponds with thread_siblings). 112 113To be consistent on all architectures, include/linux/topology.h 114provides default definitions for any of the above macros that are 115not defined by include/asm-XXX/topology.h: 116 1171) topology_physical_package_id: -1 1182) topology_die_id: -1 1193) topology_core_id: 0 1204) topology_sibling_cpumask: just the given CPU 1215) topology_core_cpumask: just the given CPU 1226) topology_die_cpumask: just the given CPU 123 124For architectures that don't support books (CONFIG_SCHED_BOOK) there are no 125default definitions for topology_book_id() and topology_book_cpumask(). 126For architectures that don't support drawers (CONFIG_SCHED_DRAWER) there are 127no default definitions for topology_drawer_id() and topology_drawer_cpumask(). 128 129Additionally, CPU topology information is provided under 130/sys/devices/system/cpu and includes these files. The internal 131source for the output is in brackets ("[]"). 132 133 =========== ========================================================== 134 kernel_max: the maximum CPU index allowed by the kernel configuration. 135 [NR_CPUS-1] 136 137 offline: CPUs that are not online because they have been 138 HOTPLUGGED off (see cpu-hotplug.txt) or exceed the limit 139 of CPUs allowed by the kernel configuration (kernel_max 140 above). [~cpu_online_mask + cpus >= NR_CPUS] 141 142 online: CPUs that are online and being scheduled [cpu_online_mask] 143 144 possible: CPUs that have been allocated resources and can be 145 brought online if they are present. [cpu_possible_mask] 146 147 present: CPUs that have been identified as being present in the 148 system. [cpu_present_mask] 149 =========== ========================================================== 150 151The format for the above output is compatible with cpulist_parse() 152[see <linux/cpumask.h>]. Some examples follow. 153 154In this example, there are 64 CPUs in the system but cpus 32-63 exceed 155the kernel max which is limited to 0..31 by the NR_CPUS config option 156being 32. Note also that CPUs 2 and 4-31 are not online but could be 157brought online as they are both present and possible:: 158 159 kernel_max: 31 160 offline: 2,4-31,32-63 161 online: 0-1,3 162 possible: 0-31 163 present: 0-31 164 165In this example, the NR_CPUS config option is 128, but the kernel was 166started with possible_cpus=144. There are 4 CPUs in the system and cpu2 167was manually taken offline (and is the only CPU that can be brought 168online.):: 169 170 kernel_max: 127 171 offline: 2,4-127,128-143 172 online: 0-1,3 173 possible: 0-127 174 present: 0-3 175 176See cpu-hotplug.txt for the possible_cpus=NUM kernel start parameter 177as well as more information on the various cpumasks. 178