diff options
Diffstat (limited to '')
-rw-r--r-- | sys-utils/lscpu.c | 2545 |
1 files changed, 2545 insertions, 0 deletions
diff --git a/sys-utils/lscpu.c b/sys-utils/lscpu.c new file mode 100644 index 0000000..90c475a --- /dev/null +++ b/sys-utils/lscpu.c @@ -0,0 +1,2545 @@ +/* + * lscpu - CPU architecture information helper + * + * Copyright (C) 2008 Cai Qian <qcai@redhat.com> + * Copyright (C) 2008 Karel Zak <kzak@redhat.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it would be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. + */ + +#include <assert.h> +#include <ctype.h> +#include <dirent.h> +#include <errno.h> +#include <fcntl.h> +#include <getopt.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/utsname.h> +#include <unistd.h> +#include <stdarg.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/personality.h> + +#if (defined(__x86_64__) || defined(__i386__)) +# define INCLUDE_VMWARE_BDOOR +#endif + +#ifdef INCLUDE_VMWARE_BDOOR +# include <stdint.h> +# include <signal.h> +# include <strings.h> +# include <setjmp.h> +# ifdef HAVE_SYS_IO_H +# include <sys/io.h> +# endif +#endif + +#if defined(HAVE_LIBRTAS) +#include <librtas.h> +#endif + +#include <libsmartcols.h> + +#include "closestream.h" +#include "optutils.h" +#include "fileutils.h" + +#include "lscpu.h" + +#define CACHE_MAX 100 + +/* /sys paths */ +#define _PATH_SYS_SYSTEM "/sys/devices/system" +#define _PATH_SYS_HYP_FEATURES "/sys/hypervisor/properties/features" +#define _PATH_SYS_CPU _PATH_SYS_SYSTEM "/cpu" +#define _PATH_SYS_NODE _PATH_SYS_SYSTEM "/node" + +/* Xen Domain feature flag used for /sys/hypervisor/properties/features */ +#define XENFEAT_supervisor_mode_kernel 3 +#define XENFEAT_mmu_pt_update_preserve_ad 5 +#define XENFEAT_hvm_callback_vector 8 + +#define XEN_FEATURES_PV_MASK (1U << XENFEAT_mmu_pt_update_preserve_ad) +#define XEN_FEATURES_PVH_MASK ( (1U << XENFEAT_supervisor_mode_kernel) \ + | (1U << XENFEAT_hvm_callback_vector) ) + +static const char *virt_types[] = { + [VIRT_NONE] = N_("none"), + [VIRT_PARA] = N_("para"), + [VIRT_FULL] = N_("full"), + [VIRT_CONT] = N_("container"), +}; + +static const char *hv_vendors[] = { + [HYPER_NONE] = NULL, + [HYPER_XEN] = "Xen", + [HYPER_KVM] = "KVM", + [HYPER_MSHV] = "Microsoft", + [HYPER_VMWARE] = "VMware", + [HYPER_IBM] = "IBM", + [HYPER_VSERVER] = "Linux-VServer", + [HYPER_UML] = "User-mode Linux", + [HYPER_INNOTEK] = "Innotek GmbH", + [HYPER_HITACHI] = "Hitachi", + [HYPER_PARALLELS] = "Parallels", + [HYPER_VBOX] = "Oracle", + [HYPER_OS400] = "OS/400", + [HYPER_PHYP] = "pHyp", + [HYPER_SPAR] = "Unisys s-Par", + [HYPER_WSL] = "Windows Subsystem for Linux" +}; + +static const int hv_vendor_pci[] = { + [HYPER_NONE] = 0x0000, + [HYPER_XEN] = 0x5853, + [HYPER_KVM] = 0x0000, + [HYPER_MSHV] = 0x1414, + [HYPER_VMWARE] = 0x15ad, + [HYPER_VBOX] = 0x80ee, +}; + +static const int hv_graphics_pci[] = { + [HYPER_NONE] = 0x0000, + [HYPER_XEN] = 0x0001, + [HYPER_KVM] = 0x0000, + [HYPER_MSHV] = 0x5353, + [HYPER_VMWARE] = 0x0710, + [HYPER_VBOX] = 0xbeef, +}; + + +/* dispatching modes */ +static const char *disp_modes[] = { + [DISP_HORIZONTAL] = N_("horizontal"), + [DISP_VERTICAL] = N_("vertical") +}; + +static struct polarization_modes polar_modes[] = { + [POLAR_UNKNOWN] = {"U", "-"}, + [POLAR_VLOW] = {"VL", "vert-low"}, + [POLAR_VMEDIUM] = {"VM", "vert-medium"}, + [POLAR_VHIGH] = {"VH", "vert-high"}, + [POLAR_HORIZONTAL] = {"H", "horizontal"}, +}; + +static int maxcpus; /* size in bits of kernel cpu mask */ + +#define is_cpu_online(_d, _cpu) \ + ((_d) && (_d)->online ? \ + CPU_ISSET_S((_cpu), CPU_ALLOC_SIZE(maxcpus), (_d)->online) : 0) +#define is_cpu_present(_d, _cpu) \ + ((_d) && (_d)->present ? \ + CPU_ISSET_S((_cpu), CPU_ALLOC_SIZE(maxcpus), (_d)->present) : 0) + +#define real_cpu_num(_d, _i) ((_d)->idx2cpunum[(_i)]) + +/* + * IDs + */ +enum { + COL_CPU_CPU, + COL_CPU_CORE, + COL_CPU_SOCKET, + COL_CPU_NODE, + COL_CPU_BOOK, + COL_CPU_DRAWER, + COL_CPU_CACHE, + COL_CPU_POLARIZATION, + COL_CPU_ADDRESS, + COL_CPU_CONFIGURED, + COL_CPU_ONLINE, + COL_CPU_MAXMHZ, + COL_CPU_MINMHZ, +}; + +enum { + COL_CACHE_ALLSIZE, + COL_CACHE_LEVEL, + COL_CACHE_NAME, + COL_CACHE_ONESIZE, + COL_CACHE_TYPE, + COL_CACHE_WAYS, + COL_CACHE_ALLOCPOL, + COL_CACHE_WRITEPOL, + COL_CACHE_PHYLINE, + COL_CACHE_SETS, + COL_CACHE_COHERENCYSIZE +}; + + +/* column description + */ +struct lscpu_coldesc { + const char *name; + const char *help; + + int flags; + unsigned int is_abbr:1; /* name is abbreviation */ +}; + +static struct lscpu_coldesc coldescs_cpu[] = +{ + [COL_CPU_CPU] = { "CPU", N_("logical CPU number"), SCOLS_FL_RIGHT, 1 }, + [COL_CPU_CORE] = { "CORE", N_("logical core number"), SCOLS_FL_RIGHT }, + [COL_CPU_SOCKET] = { "SOCKET", N_("logical socket number"), SCOLS_FL_RIGHT }, + [COL_CPU_NODE] = { "NODE", N_("logical NUMA node number"), SCOLS_FL_RIGHT }, + [COL_CPU_BOOK] = { "BOOK", N_("logical book number"), SCOLS_FL_RIGHT }, + [COL_CPU_DRAWER] = { "DRAWER", N_("logical drawer number"), SCOLS_FL_RIGHT }, + [COL_CPU_CACHE] = { "CACHE", N_("shows how caches are shared between CPUs") }, + [COL_CPU_POLARIZATION] = { "POLARIZATION", N_("CPU dispatching mode on virtual hardware") }, + [COL_CPU_ADDRESS] = { "ADDRESS", N_("physical address of a CPU") }, + [COL_CPU_CONFIGURED] = { "CONFIGURED", N_("shows if the hypervisor has allocated the CPU") }, + [COL_CPU_ONLINE] = { "ONLINE", N_("shows if Linux currently makes use of the CPU"), SCOLS_FL_RIGHT }, + [COL_CPU_MAXMHZ] = { "MAXMHZ", N_("shows the maximum MHz of the CPU"), SCOLS_FL_RIGHT }, + [COL_CPU_MINMHZ] = { "MINMHZ", N_("shows the minimum MHz of the CPU"), SCOLS_FL_RIGHT } +}; + +static struct lscpu_coldesc coldescs_cache[] = +{ + [COL_CACHE_ALLSIZE] = { "ALL-SIZE", N_("size of all system caches"), SCOLS_FL_RIGHT }, + [COL_CACHE_LEVEL] = { "LEVEL", N_("cache level"), SCOLS_FL_RIGHT }, + [COL_CACHE_NAME] = { "NAME", N_("cache name") }, + [COL_CACHE_ONESIZE] = { "ONE-SIZE", N_("size of one cache"), SCOLS_FL_RIGHT }, + [COL_CACHE_TYPE] = { "TYPE", N_("cache type") }, + [COL_CACHE_WAYS] = { "WAYS", N_("ways of associativity"), SCOLS_FL_RIGHT }, + [COL_CACHE_ALLOCPOL] = { "ALLOC-POLICY", N_("allocation policy") }, + [COL_CACHE_WRITEPOL] = { "WRITE-POLICY", N_("write policy") }, + [COL_CACHE_PHYLINE] = { "PHY-LINE", N_("number of physical cache line per cache t"), SCOLS_FL_RIGHT }, + [COL_CACHE_SETS] = { "SETS", N_("number of sets in the cache; set lines has the same cache index"), SCOLS_FL_RIGHT }, + [COL_CACHE_COHERENCYSIZE] = { "COHERENCY-SIZE", N_("minimum amount of data in bytes transferred from memory to cache"), SCOLS_FL_RIGHT } +}; + + +static int get_cache_full_size(struct lscpu_desc *desc, struct cpu_cache *ca, uint64_t *res); + +static int +cpu_column_name_to_id(const char *name, size_t namesz) +{ + size_t i; + + for (i = 0; i < ARRAY_SIZE(coldescs_cpu); i++) { + const char *cn = coldescs_cpu[i].name; + + if (!strncasecmp(name, cn, namesz) && !*(cn + namesz)) + return i; + } + warnx(_("unknown column: %s"), name); + return -1; +} + +static int +cache_column_name_to_id(const char *name, size_t namesz) +{ + size_t i; + + for (i = 0; i < ARRAY_SIZE(coldescs_cache); i++) { + const char *cn = coldescs_cache[i].name; + + if (!strncasecmp(name, cn, namesz) && !*(cn + namesz)) + return i; + } + warnx(_("unknown column: %s"), name); + return -1; +} + +/* Lookup a pattern and get the value from cpuinfo. + * Format is: + * + * "<pattern> : <key>" + */ +static int +lookup(char *line, char *pattern, char **value) +{ + char *p, *v; + int len = strlen(pattern); + + /* don't re-fill already found tags, first one wins */ + if (!*line || *value) + return 0; + + /* pattern */ + if (strncmp(line, pattern, len) != 0) + return 0; + + /* white spaces */ + for (p = line + len; isspace(*p); p++); + + /* separator */ + if (*p != ':') + return 0; + + /* white spaces */ + for (++p; isspace(*p); p++); + + /* value */ + if (!*p) + return 0; + v = p; + + /* end of value */ + len = strlen(line) - 1; + for (p = line + len; isspace(*(p-1)); p--); + *p = '\0'; + + *value = xstrdup(v); + return 1; +} + +/* Parse extra cache lines contained within /proc/cpuinfo but which are not + * part of the cache topology information within the sysfs filesystem. + * This is true for all shared caches on e.g. s390. When there are layers of + * hypervisors in between it is not knows which CPUs share which caches. + * Therefore information about shared caches is only available in + * /proc/cpuinfo. + * Format is: + * "cache<nr> : level=<lvl> type=<type> scope=<scope> size=<size> line_size=<lsz> associativity=<as>" + */ +static int +lookup_cache(char *line, struct lscpu_desc *desc) +{ + struct cpu_cache *cache; + long long size; + char *p, type; + int level, line_size, associativity; + + /* Make sure line starts with "cache<nr> :" */ + if (strncmp(line, "cache", 5) != 0) + return 0; + for (p = line + 5; isdigit(*p); p++); + for (; isspace(*p); p++); + if (*p != ':') + return 0; + + p = strstr(line, "scope=") + 6; + /* Skip private caches, also present in sysfs */ + if (!p || strncmp(p, "Private", 7) == 0) + return 0; + p = strstr(line, "level="); + if (!p || sscanf(p, "level=%d", &level) != 1) + return 0; + p = strstr(line, "type=") + 5; + if (!p || !*p) + return 0; + type = 0; + if (strncmp(p, "Data", 4) == 0) + type = 'd'; + else if (strncmp(p, "Instruction", 11) == 0) + type = 'i'; + else if (strncmp(p, "Unified", 7) == 0) + type = 'u'; + p = strstr(line, "size="); + if (!p || sscanf(p, "size=%lld", &size) != 1) + return 0; + + p = strstr(line, "line_size="); + if (!p || sscanf(p, "line_size=%u", &line_size) != 1) + return 0; + + p = strstr(line, "associativity="); + if (!p || sscanf(p, "associativity=%u", &associativity) != 1) + return 0; + + desc->necaches++; + desc->ecaches = xrealloc(desc->ecaches, + desc->necaches * sizeof(struct cpu_cache)); + cache = &desc->ecaches[desc->necaches - 1]; + memset(cache, 0 , sizeof(*cache)); + + if (type == 'i' || type == 'd') + xasprintf(&cache->name, "L%d%c", level, type); + else + xasprintf(&cache->name, "L%d", level); + + cache->level = level; + cache->size = size * 1024; + cache->ways_of_associativity = associativity; + cache->coherency_line_size = line_size; + /* Number of sets for s390. For safety, just check divide by zero */ + cache->number_of_sets = line_size ? (cache->size / line_size): 0; + cache->number_of_sets = associativity ? (cache->number_of_sets / associativity) : 0; + + cache->type = type == 'i' ? xstrdup("Instruction") : + type == 'd' ? xstrdup("Data") : + type == 'u' ? xstrdup("Unified") : NULL; + return 1; +} + +/* Don't init the mode for platforms where we are not able to + * detect that CPU supports 64-bit mode. + */ +static int +init_mode(struct lscpu_modifier *mod) +{ + int m = 0; + + if (mod->system == SYSTEM_SNAPSHOT) + /* reading info from any /{sys,proc} dump, don't mix it with + * information about our real CPU */ + return 0; + +#if defined(__alpha__) || defined(__ia64__) + m |= MODE_64BIT; /* 64bit platforms only */ +#endif + /* platforms with 64bit flag in /proc/cpuinfo, define + * 32bit default here */ +#if defined(__i386__) || defined(__x86_64__) || \ + defined(__s390x__) || defined(__s390__) || defined(__sparc_v9__) + m |= MODE_32BIT; +#endif + +#if defined(__aarch64__) + { + /* personality() is the most reliable way (since 4.7) + * to determine aarch32 support */ + int pers = personality(PER_LINUX32); + if (pers != -1) { + personality(pers); + m |= MODE_32BIT; + } + m |= MODE_64BIT; + } +#endif + return m; +} + +#if defined(HAVE_LIBRTAS) +#define PROCESSOR_MODULE_INFO 43 +static int strbe16toh(const char *buf, int offset) +{ + return (buf[offset] << 8) + buf[offset+1]; +} + +static void read_physical_info_powerpc(struct lscpu_desc *desc) +{ + char buf[BUFSIZ]; + int rc, len, ntypes; + + desc->physsockets = desc->physchips = desc->physcoresperchip = 0; + + rc = rtas_get_sysparm(PROCESSOR_MODULE_INFO, sizeof(buf), buf); + if (rc < 0) + return; + + len = strbe16toh(buf, 0); + if (len < 8) + return; + + ntypes = strbe16toh(buf, 2); + if (!ntypes) + return; + + desc->physsockets = strbe16toh(buf, 4); + desc->physchips = strbe16toh(buf, 6); + desc->physcoresperchip = strbe16toh(buf, 8); +} +#else +static void read_physical_info_powerpc( + struct lscpu_desc *desc __attribute__((__unused__))) +{ +} +#endif + +static int cmp_vulnerability_name(const void *a0, const void *b0) +{ + const struct cpu_vulnerability *a = (const struct cpu_vulnerability *) a0, + *b = (const struct cpu_vulnerability *) b0; + return strcmp(a->name, b->name); +} + +static void read_vulnerabilities(struct lscpu_desc *desc) +{ + struct dirent *d; + DIR *dir = ul_path_opendir(desc->syscpu, "vulnerabilities"); + int n = 0; + + if (!dir) + return; + + desc->nvuls = n = 0; + + while (xreaddir(dir)) + n++; + if (!n) + return; + + rewinddir(dir); + desc->vuls = xcalloc(n, sizeof(struct cpu_vulnerability)); + + while (desc->nvuls < n && (d = xreaddir(dir))) { + char *str, *p; + struct cpu_vulnerability *vu; + +#ifdef _DIRENT_HAVE_D_TYPE + if (d->d_type == DT_DIR || d->d_type == DT_UNKNOWN) + continue; +#endif + if (ul_path_readf_string(desc->syscpu, &str, + "vulnerabilities/%s", d->d_name) <= 0) + continue; + + vu = &desc->vuls[desc->nvuls++]; + + /* Name */ + vu->name = xstrdup(d->d_name); + *vu->name = toupper(*vu->name); + strrep(vu->name, '_', ' '); + + /* Description */ + vu->text = str; + p = (char *) startswith(vu->text, "Mitigation"); + if (p) { + *p = ';'; + strrem(vu->text, ':'); + } + } + closedir(dir); + + qsort(desc->vuls, desc->nvuls, + sizeof(struct cpu_vulnerability), cmp_vulnerability_name); +} + + + + +static void +read_basicinfo(struct lscpu_desc *desc, struct lscpu_modifier *mod) +{ + FILE *fp; + char buf[BUFSIZ]; + struct utsname utsbuf; + size_t setsize; + cpu_set_t *cpuset = NULL; + + /* architecture */ + if (uname(&utsbuf) == -1) + err(EXIT_FAILURE, _("error: uname failed")); + + fp = ul_path_fopen(desc->procfs, "r", "cpuinfo"); + if (!fp) + err(EXIT_FAILURE, _("cannot open %s"), "/proc/cpuinfo"); + desc->arch = xstrdup(utsbuf.machine); + + /* details */ + while (fgets(buf, sizeof(buf), fp) != NULL) { + if (lookup(buf, "vendor", &desc->vendor)) ; + else if (lookup(buf, "vendor_id", &desc->vendor)) ; + else if (lookup(buf, "CPU implementer", &desc->vendor)) ; /* ARM and aarch64 */ + else if (lookup(buf, "family", &desc->family)) ; + else if (lookup(buf, "cpu family", &desc->family)) ; + else if (lookup(buf, "model", &desc->model)) ; + else if (lookup(buf, "CPU part", &desc->model)) ; /* ARM and aarch64 */ + else if (lookup(buf, "cpu model", &desc->model)) ; /* mips */ + else if (lookup(buf, "model name", &desc->modelname)) ; + else if (lookup(buf, "stepping", &desc->stepping)) ; + else if (lookup(buf, "CPU variant", &desc->stepping)) ; /* aarch64 */ + else if (lookup(buf, "cpu MHz", &desc->mhz)) ; + else if (lookup(buf, "cpu MHz dynamic", &desc->dynamic_mhz)) ; /* s390 */ + else if (lookup(buf, "cpu MHz static", &desc->static_mhz)) ; /* s390 */ + else if (lookup(buf, "flags", &desc->flags)) ; /* x86 */ + else if (lookup(buf, "features", &desc->flags)) ; /* s390 */ + else if (lookup(buf, "Features", &desc->flags)) ; /* aarch64 */ + else if (lookup(buf, "ASEs implemented", &desc->flags)) ; /* mips */ + else if (lookup(buf, "type", &desc->flags)) ; /* sparc64 */ + else if (lookup(buf, "bogomips", &desc->bogomips)) ; + else if (lookup(buf, "BogoMIPS", &desc->bogomips)) ; /* aarch64 */ + else if (lookup(buf, "bogomips per cpu", &desc->bogomips)) ; /* s390 */ + else if (lookup(buf, "cpu", &desc->cpu)) ; + else if (lookup(buf, "revision", &desc->revision)) ; + else if (lookup(buf, "CPU revision", &desc->revision)) ; /* aarch64 */ + else if (lookup(buf, "max thread id", &desc->mtid)) ; /* s390 */ + else if (lookup(buf, "address sizes", &desc->addrsz)) ; /* x86 */ + else if (lookup_cache(buf, desc)) ; + else + continue; + } + + desc->mode = init_mode(mod); + + if (desc->flags) { + snprintf(buf, sizeof(buf), " %s ", desc->flags); + if (strstr(buf, " svm ")) + desc->virtflag = xstrdup("svm"); + else if (strstr(buf, " vmx ")) + desc->virtflag = xstrdup("vmx"); + if (strstr(buf, " lm ")) + desc->mode |= MODE_32BIT | MODE_64BIT; /* x86_64 */ + if (strstr(buf, " zarch ")) + desc->mode |= MODE_32BIT | MODE_64BIT; /* s390x */ + if (strstr(buf, " sun4v ") || strstr(buf, " sun4u ")) + desc->mode |= MODE_32BIT | MODE_64BIT; /* sparc64 */ + } + + if (desc->arch && mod->system != SYSTEM_SNAPSHOT) { + if (strcmp(desc->arch, "ppc64") == 0) + desc->mode |= MODE_32BIT | MODE_64BIT; + else if (strcmp(desc->arch, "ppc") == 0) + desc->mode |= MODE_32BIT; + } + + fclose(fp); + + if (ul_path_read_s32(desc->syscpu, &maxcpus, "kernel_max") == 0) + /* note that kernel_max is maximum index [NR_CPUS-1] */ + maxcpus += 1; + + else if (mod->system == SYSTEM_LIVE) + /* the root is '/' so we are working with data from the current kernel */ + maxcpus = get_max_number_of_cpus(); + + if (maxcpus <= 0) + /* error or we are reading some /sys snapshot instead of the + * real /sys, let's use any crazy number... */ + maxcpus = 2048; + + setsize = CPU_ALLOC_SIZE(maxcpus); + + if (ul_path_readf_cpulist(desc->syscpu, &cpuset, maxcpus, "possible") == 0) { + int num, idx; + + desc->ncpuspos = CPU_COUNT_S(setsize, cpuset); + desc->idx2cpunum = xcalloc(desc->ncpuspos, sizeof(int)); + + for (num = 0, idx = 0; num < maxcpus; num++) { + if (CPU_ISSET_S(num, setsize, cpuset)) + desc->idx2cpunum[idx++] = num; + } + cpuset_free(cpuset); + cpuset = NULL; + } else + err(EXIT_FAILURE, _("failed to determine number of CPUs: %s"), + _PATH_SYS_CPU "/possible"); + + + /* get mask for present CPUs */ + if (ul_path_readf_cpulist(desc->syscpu, &desc->present, maxcpus, "present") == 0) + desc->ncpus = CPU_COUNT_S(setsize, desc->present); + + /* get mask for online CPUs */ + if (ul_path_readf_cpulist(desc->syscpu, &desc->online, maxcpus, "online") == 0) + desc->nthreads = CPU_COUNT_S(setsize, desc->online); + + /* get dispatching mode */ + if (ul_path_read_s32(desc->syscpu, &desc->dispatching, "dispatching") != 0) + desc->dispatching = -1; + + /* get cpufreq boost mode */ + if (ul_path_read_s32(desc->syscpu, &desc->freqboost, "cpufreq/boost") != 0) + desc->freqboost = -1; + + if (mod->system == SYSTEM_LIVE) + read_physical_info_powerpc(desc); + + if ((fp = ul_path_fopen(desc->procfs, "r", "sysinfo"))) { + while (fgets(buf, sizeof(buf), fp) != NULL) { + if (lookup(buf, "Type", &desc->machinetype)) + break; + } + fclose(fp); + } + + /* vulnerabilities */ + if (ul_path_access(desc->syscpu, F_OK, "vulnerabilities") == 0) + read_vulnerabilities(desc); +} + +static int +has_pci_device(struct lscpu_desc *desc, unsigned int vendor, unsigned int device) +{ + FILE *f; + unsigned int num, fn, ven, dev; + int res = 1; + + f = ul_path_fopen(desc->procfs, "r", "bus/pci/devices"); + if (!f) + return 0; + + /* for more details about bus/pci/devices format see + * drivers/pci/proc.c in linux kernel + */ + while(fscanf(f, "%02x%02x\t%04x%04x\t%*[^\n]", + &num, &fn, &ven, &dev) == 4) { + + if (ven == vendor && dev == device) + goto found; + } + + res = 0; +found: + fclose(f); + return res; +} + +#if defined(__x86_64__) || defined(__i386__) + +/* + * This CPUID leaf returns the information about the hypervisor. + * EAX : maximum input value for CPUID supported by the hypervisor. + * EBX, ECX, EDX : Hypervisor vendor ID signature. E.g. VMwareVMware. + */ +#define HYPERVISOR_INFO_LEAF 0x40000000 + +static inline void +cpuid(unsigned int op, unsigned int *eax, unsigned int *ebx, + unsigned int *ecx, unsigned int *edx) +{ + __asm__( +#if defined(__PIC__) && defined(__i386__) + /* x86 PIC cannot clobber ebx -- gcc bitches */ + "xchg %%ebx, %%esi;" + "cpuid;" + "xchg %%esi, %%ebx;" + : "=S" (*ebx), +#else + "cpuid;" + : "=b" (*ebx), +#endif + "=a" (*eax), + "=c" (*ecx), + "=d" (*edx) + : "1" (op), "c"(0)); +} + +static void +read_hypervisor_cpuid(struct lscpu_desc *desc) +{ + unsigned int eax = 0, ebx = 0, ecx = 0, edx = 0; + char hyper_vendor_id[13]; + + memset(hyper_vendor_id, 0, sizeof(hyper_vendor_id)); + + cpuid(HYPERVISOR_INFO_LEAF, &eax, &ebx, &ecx, &edx); + memcpy(hyper_vendor_id + 0, &ebx, 4); + memcpy(hyper_vendor_id + 4, &ecx, 4); + memcpy(hyper_vendor_id + 8, &edx, 4); + hyper_vendor_id[12] = '\0'; + + if (!hyper_vendor_id[0]) + return; + + if (!strncmp("XenVMMXenVMM", hyper_vendor_id, 12)) + desc->hyper = HYPER_XEN; + else if (!strncmp("KVMKVMKVM", hyper_vendor_id, 9)) + desc->hyper = HYPER_KVM; + else if (!strncmp("Microsoft Hv", hyper_vendor_id, 12)) + desc->hyper = HYPER_MSHV; + else if (!strncmp("VMwareVMware", hyper_vendor_id, 12)) + desc->hyper = HYPER_VMWARE; + else if (!strncmp("UnisysSpar64", hyper_vendor_id, 12)) + desc->hyper = HYPER_SPAR; +} + +#else /* ! (__x86_64__ || __i386__) */ +static void +read_hypervisor_cpuid(struct lscpu_desc *desc __attribute__((__unused__))) +{ +} +#endif + +static int is_devtree_compatible(struct lscpu_desc *desc, const char *str) +{ + FILE *fd = ul_path_fopen(desc->procfs, "r", "device-tree/compatible"); + + if (fd) { + char buf[256]; + size_t i, len; + + memset(buf, 0, sizeof(buf)); + len = fread(buf, 1, sizeof(buf) - 1, fd); + fclose(fd); + + for (i = 0; i < len;) { + if (!strcmp(&buf[i], str)) + return 1; + i += strlen(&buf[i]); + i++; + } + } + + return 0; +} + +static int +read_hypervisor_powerpc(struct lscpu_desc *desc) +{ + assert(!desc->hyper); + + /* IBM iSeries: legacy, para-virtualized on top of OS/400 */ + if (ul_path_access(desc->procfs, F_OK, "iSeries") == 0) { + desc->hyper = HYPER_OS400; + desc->virtype = VIRT_PARA; + + /* PowerNV (POWER Non-Virtualized, bare-metal) */ + } else if (is_devtree_compatible(desc, "ibm,powernv")) { + desc->hyper = HYPER_NONE; + desc->virtype = VIRT_NONE; + + /* PowerVM (IBM's proprietary hypervisor, aka pHyp) */ + } else if (ul_path_access(desc->procfs, F_OK, "device-tree/ibm,partition-name") == 0 + && ul_path_access(desc->procfs, F_OK, "device-tree/hmc-managed?") == 0 + && ul_path_access(desc->procfs, F_OK, "device-tree/chosen/qemu,graphic-width") != 0) { + + FILE *fd; + desc->hyper = HYPER_PHYP; + desc->virtype = VIRT_PARA; + + fd = ul_path_fopen(desc->procfs, "r", "device-tree/ibm,partition-name"); + if (fd) { + char buf[256]; + if (fscanf(fd, "%255s", buf) == 1 && !strcmp(buf, "full")) + desc->virtype = VIRT_NONE; + fclose(fd); + } + + /* Qemu */ + } else if (is_devtree_compatible(desc, "qemu,pseries")) { + desc->hyper = HYPER_KVM; + desc->virtype = VIRT_PARA; + } + return desc->hyper; +} + +#ifdef INCLUDE_VMWARE_BDOOR + +#define VMWARE_BDOOR_MAGIC 0x564D5868 +#define VMWARE_BDOOR_PORT 0x5658 +#define VMWARE_BDOOR_CMD_GETVERSION 10 + +static UL_ASAN_BLACKLIST +void vmware_bdoor(uint32_t *eax, uint32_t *ebx, uint32_t *ecx, uint32_t *edx) +{ + __asm__( +#if defined(__PIC__) && defined(__i386__) + /* x86 PIC cannot clobber ebx -- gcc bitches */ + "xchg %%ebx, %%esi;" + "inl (%%dx), %%eax;" + "xchg %%esi, %%ebx;" + : "=S" (*ebx), +#else + "inl (%%dx), %%eax;" + : "=b" (*ebx), +#endif + "=a" (*eax), + "=c" (*ecx), + "=d" (*edx) + : "0" (VMWARE_BDOOR_MAGIC), + "1" (VMWARE_BDOOR_CMD_GETVERSION), + "2" (VMWARE_BDOOR_PORT), + "3" (0) + : "memory"); +} + +static jmp_buf segv_handler_env; + +static void +segv_handler(__attribute__((__unused__)) int sig, + __attribute__((__unused__)) siginfo_t *info, + __attribute__((__unused__)) void *ignored) +{ + siglongjmp(segv_handler_env, 1); +} + +static int +is_vmware_platform(void) +{ + uint32_t eax, ebx, ecx, edx; + struct sigaction act, oact; + + /* + * FIXME: Not reliable for non-root users. Note it works as expected if + * vmware_bdoor() is not optimized for PIE, but then it fails to build + * on 32bit x86 systems. See lscpu git log for more details (commit + * 7845b91dbc7690064a2be6df690e4aaba728fb04). kzak [3-Nov-2016] + */ + if (getuid() != 0) + return 0; + + /* + * The assembly routine for vmware detection works + * fine under vmware, even if ran as regular user. But + * on real HW or under other hypervisors, it segfaults (which is + * expected). So we temporarily install SIGSEGV handler to catch + * the signal. All this magic is needed because lscpu + * isn't supposed to require root privileges. + */ + if (sigsetjmp(segv_handler_env, 1)) + return 0; + + memset(&act, 0, sizeof(act)); + act.sa_sigaction = segv_handler; + act.sa_flags = SA_SIGINFO; + + if (sigaction(SIGSEGV, &act, &oact)) + err(EXIT_FAILURE, _("cannot set signal handler")); + + vmware_bdoor(&eax, &ebx, &ecx, &edx); + + if (sigaction(SIGSEGV, &oact, NULL)) + err(EXIT_FAILURE, _("cannot restore signal handler")); + + return eax != (uint32_t)-1 && ebx == VMWARE_BDOOR_MAGIC; +} + +#else /* ! INCLUDE_VMWARE_BDOOR */ + +static int +is_vmware_platform(void) +{ + return 0; +} + +#endif /* INCLUDE_VMWARE_BDOOR */ + +static void +read_hypervisor(struct lscpu_desc *desc, struct lscpu_modifier *mod) +{ + FILE *fd; + + /* We have to detect WSL first. is_vmware_platform() crashes on Windows 10. */ + + if ((fd = ul_path_fopen(desc->procfs, "r", "sys/kernel/osrelease"))) { + char buf[256]; + + if (fgets(buf, sizeof(buf), fd) != NULL) { + if (strstr(buf, "Microsoft")) { + desc->hyper = HYPER_WSL; + desc->virtype = VIRT_CONT; + } + } + fclose(fd); + if (desc->virtype) + return; + } + + if (mod->system != SYSTEM_SNAPSHOT) { + read_hypervisor_cpuid(desc); + if (!desc->hyper) + desc->hyper = read_hypervisor_dmi(); + if (!desc->hyper && is_vmware_platform()) + desc->hyper = HYPER_VMWARE; + } + + if (desc->hyper) { + desc->virtype = VIRT_FULL; + + if (desc->hyper == HYPER_XEN) { + uint32_t features; + + fd = ul_prefix_fopen(desc->prefix, "r", _PATH_SYS_HYP_FEATURES); + + if (fd && fscanf(fd, "%x", &features) == 1) { + /* Xen PV domain */ + if (features & XEN_FEATURES_PV_MASK) + desc->virtype = VIRT_PARA; + /* Xen PVH domain */ + else if ((features & XEN_FEATURES_PVH_MASK) + == XEN_FEATURES_PVH_MASK) + desc->virtype = VIRT_PARA; + } + if (fd) + fclose(fd); + } + } else if (read_hypervisor_powerpc(desc) > 0) { + /* read_hypervisor_powerpc() sets all necessary stuff to @desc */ + ; + /* Xen para-virt or dom0 */ + } else if (ul_path_access(desc->procfs, F_OK, "xen") == 0) { + int dom0 = 0; + + fd = ul_path_fopen(desc->procfs, "r", "xen/capabilities"); + if (fd) { + char buf[256]; + + if (fscanf(fd, "%255s", buf) == 1 && + !strcmp(buf, "control_d")) + dom0 = 1; + fclose(fd); + } + desc->virtype = dom0 ? VIRT_NONE : VIRT_PARA; + desc->hyper = HYPER_XEN; + + /* Xen full-virt on non-x86_64 */ + } else if (has_pci_device(desc, hv_vendor_pci[HYPER_XEN], hv_graphics_pci[HYPER_XEN])) { + desc->hyper = HYPER_XEN; + desc->virtype = VIRT_FULL; + } else if (has_pci_device(desc, hv_vendor_pci[HYPER_VMWARE], hv_graphics_pci[HYPER_VMWARE])) { + desc->hyper = HYPER_VMWARE; + desc->virtype = VIRT_FULL; + } else if (has_pci_device(desc, hv_vendor_pci[HYPER_VBOX], hv_graphics_pci[HYPER_VBOX])) { + desc->hyper = HYPER_VBOX; + desc->virtype = VIRT_FULL; + + /* IBM PR/SM */ + } else if ((fd = ul_path_fopen(desc->procfs, "r", "sysinfo"))) { + char buf[BUFSIZ]; + + desc->hyper = HYPER_IBM; + desc->hypervisor = "PR/SM"; + desc->virtype = VIRT_FULL; + while (fgets(buf, sizeof(buf), fd) != NULL) { + char *str, *p; + + if (!strstr(buf, "Control Program:")) + continue; + if (!strstr(buf, "KVM")) + desc->hyper = HYPER_IBM; + else + desc->hyper = HYPER_KVM; + p = strchr(buf, ':'); + if (!p) + continue; + xasprintf(&str, "%s", p + 1); + + /* remove leading, trailing and repeating whitespace */ + while (*str == ' ') + str++; + desc->hypervisor = str; + str += strlen(str) - 1; + while ((*str == '\n') || (*str == ' ')) + *(str--) = '\0'; + while ((str = strstr(desc->hypervisor, " "))) + memmove(str, str + 1, strlen(str)); + break; + } + fclose(fd); + } + + /* OpenVZ/Virtuozzo - /proc/vz dir should exist + * /proc/bc should not */ + else if (ul_path_access(desc->procfs, F_OK, "vz") == 0 && + ul_path_access(desc->procfs, F_OK, "bc") != 0) { + desc->hyper = HYPER_PARALLELS; + desc->virtype = VIRT_CONT; + + /* IBM */ + } else if (desc->vendor && + (strcmp(desc->vendor, "PowerVM Lx86") == 0 || + strcmp(desc->vendor, "IBM/S390") == 0)) { + desc->hyper = HYPER_IBM; + desc->virtype = VIRT_FULL; + + /* User-mode-linux */ + } else if (desc->modelname && strstr(desc->modelname, "UML")) { + desc->hyper = HYPER_UML; + desc->virtype = VIRT_PARA; + + /* Linux-VServer */ + } else if ((fd = ul_path_fopen(desc->procfs, "r", "self/status"))) { + char buf[BUFSIZ]; + char *val = NULL; + + while (fgets(buf, sizeof(buf), fd) != NULL) { + if (lookup(buf, "VxID", &val)) + break; + } + fclose(fd); + + if (val) { + char *org = val; + + while (isdigit(*val)) + ++val; + if (!*val) { + desc->hyper = HYPER_VSERVER; + desc->virtype = VIRT_CONT; + } + free(org); + } + } +} + +/* add @set to the @ary, unnecessary set is deallocated. */ +static int add_cpuset_to_array(cpu_set_t **ary, int *items, cpu_set_t *set) +{ + int i; + size_t setsize = CPU_ALLOC_SIZE(maxcpus); + + if (!ary) + return -1; + + for (i = 0; i < *items; i++) { + if (CPU_EQUAL_S(setsize, set, ary[i])) + break; + } + if (i == *items) { + ary[*items] = set; + ++*items; + return 0; + } + CPU_FREE(set); + return 1; +} + +static void +read_topology(struct lscpu_desc *desc, int idx) +{ + cpu_set_t *thread_siblings, *core_siblings; + cpu_set_t *book_siblings, *drawer_siblings; + int coreid, socketid, bookid, drawerid; + int i, num = real_cpu_num(desc, idx); + + if (ul_path_accessf(desc->syscpu, F_OK, "cpu%d/topology/thread_siblings", num) != 0) + return; + + ul_path_readf_cpuset(desc->syscpu, &thread_siblings, maxcpus, + "cpu%d/topology/thread_siblings", num); + ul_path_readf_cpuset(desc->syscpu, &core_siblings, maxcpus, + "cpu%d/topology/core_siblings", num); + ul_path_readf_cpuset(desc->syscpu, &book_siblings, maxcpus, + "cpu%d/topology/book_siblings", num); + ul_path_readf_cpuset(desc->syscpu, &drawer_siblings, maxcpus, + "cpu%d/topology/drawer_siblings", num); + + if (ul_path_readf_s32(desc->syscpu, &coreid, "cpu%d/topology/core_id", num) != 0) + coreid = -1; + + if (ul_path_readf_s32(desc->syscpu, &socketid, "cpu%d/topology/physical_package_id", num) != 0) + socketid = -1; + + if (ul_path_readf_s32(desc->syscpu, &bookid, "cpu%d/topology/book_id", num) != 0) + bookid = -1; + + if (ul_path_readf_s32(desc->syscpu, &drawerid, "cpu%d/topology/drawer_id", num) != 0) + drawerid = -1; + + if (!desc->coremaps) { + int ndrawers, nbooks, nsockets, ncores, nthreads; + size_t setsize = CPU_ALLOC_SIZE(maxcpus); + + /* threads within one core */ + nthreads = CPU_COUNT_S(setsize, thread_siblings); + if (!nthreads) + nthreads = 1; + + /* cores within one socket */ + ncores = CPU_COUNT_S(setsize, core_siblings) / nthreads; + if (!ncores) + ncores = 1; + + /* number of sockets within one book. Because of odd / + * non-present cpu maps and to keep calculation easy we make + * sure that nsockets and nbooks is at least 1. + */ + nsockets = desc->ncpus / nthreads / ncores; + if (!nsockets) + nsockets = 1; + + /* number of books */ + nbooks = desc->ncpus / nthreads / ncores / nsockets; + if (!nbooks) + nbooks = 1; + + /* number of drawers */ + ndrawers = desc->ncpus / nbooks / nthreads / ncores / nsockets; + if (!ndrawers) + ndrawers = 1; + + /* all threads, see also read_basicinfo() + * -- fallback for kernels without + * /sys/devices/system/cpu/online. + */ + if (!desc->nthreads) + desc->nthreads = ndrawers * nbooks * nsockets * ncores * nthreads; + + /* For each map we make sure that it can have up to ncpuspos + * entries. This is because we cannot reliably calculate the + * number of cores, sockets and books on all architectures. + * E.g. completely virtualized architectures like s390 may + * have multiple sockets of different sizes. + */ + desc->coremaps = xcalloc(desc->ncpuspos, sizeof(cpu_set_t *)); + desc->socketmaps = xcalloc(desc->ncpuspos, sizeof(cpu_set_t *)); + desc->coreids = xcalloc(desc->ncpuspos, sizeof(*desc->coreids)); + desc->socketids = xcalloc(desc->ncpuspos, sizeof(*desc->socketids)); + + for (i = 0; i < desc->ncpuspos; i++) + desc->coreids[i] = desc->socketids[i] = -1; + + if (book_siblings) { + desc->bookmaps = xcalloc(desc->ncpuspos, sizeof(cpu_set_t *)); + desc->bookids = xcalloc(desc->ncpuspos, sizeof(*desc->bookids)); + for (i = 0; i < desc->ncpuspos; i++) + desc->bookids[i] = -1; + } + if (drawer_siblings) { + desc->drawermaps = xcalloc(desc->ncpuspos, sizeof(cpu_set_t *)); + desc->drawerids = xcalloc(desc->ncpuspos, sizeof(*desc->drawerids)); + for (i = 0; i < desc->ncpuspos; i++) + desc->drawerids[i] = -1; + } + } + + add_cpuset_to_array(desc->socketmaps, &desc->nsockets, core_siblings); + desc->coreids[idx] = coreid; + add_cpuset_to_array(desc->coremaps, &desc->ncores, thread_siblings); + desc->socketids[idx] = socketid; + + if (book_siblings && desc->bookmaps && desc->bookids) { + add_cpuset_to_array(desc->bookmaps, &desc->nbooks, book_siblings); + desc->bookids[idx] = bookid; + } + if (drawer_siblings && desc->drawermaps && desc->drawerids) { + add_cpuset_to_array(desc->drawermaps, &desc->ndrawers, drawer_siblings); + desc->drawerids[idx] = drawerid; + } +} + +static void +read_polarization(struct lscpu_desc *desc, int idx) +{ + char mode[64]; + int num = real_cpu_num(desc, idx); + + if (desc->dispatching < 0) + return; + if (ul_path_accessf(desc->syscpu, F_OK, "cpu%d/polarization", num) != 0) + return; + if (!desc->polarization) + desc->polarization = xcalloc(desc->ncpuspos, sizeof(int)); + + ul_path_readf_buffer(desc->syscpu, mode, sizeof(mode), "cpu%d/polarization", num); + + if (strncmp(mode, "vertical:low", sizeof(mode)) == 0) + desc->polarization[idx] = POLAR_VLOW; + else if (strncmp(mode, "vertical:medium", sizeof(mode)) == 0) + desc->polarization[idx] = POLAR_VMEDIUM; + else if (strncmp(mode, "vertical:high", sizeof(mode)) == 0) + desc->polarization[idx] = POLAR_VHIGH; + else if (strncmp(mode, "horizontal", sizeof(mode)) == 0) + desc->polarization[idx] = POLAR_HORIZONTAL; + else + desc->polarization[idx] = POLAR_UNKNOWN; +} + +static void +read_address(struct lscpu_desc *desc, int idx) +{ + int num = real_cpu_num(desc, idx); + + if (ul_path_accessf(desc->syscpu, F_OK, "cpu%d/address", num) != 0) + return; + if (!desc->addresses) + desc->addresses = xcalloc(desc->ncpuspos, sizeof(int)); + ul_path_readf_s32(desc->syscpu, &desc->addresses[idx], "cpu%d/address", num); +} + +static void +read_configured(struct lscpu_desc *desc, int idx) +{ + int num = real_cpu_num(desc, idx); + + if (ul_path_accessf(desc->syscpu, F_OK, "cpu%d/configure", num) != 0) + return; + if (!desc->configured) + desc->configured = xcalloc(desc->ncpuspos, sizeof(int)); + ul_path_readf_s32(desc->syscpu, &desc->configured[idx], "cpu%d/configure", num); +} + +/* Read overall maximum frequency of cpu */ +static char * +cpu_max_mhz(struct lscpu_desc *desc, char *buf, size_t bufsz) +{ + int i; + float cpu_freq = 0.0; + size_t setsize = CPU_ALLOC_SIZE(maxcpus); + + if (desc->present) { + for (i = 0; i < desc->ncpuspos; i++) { + if (CPU_ISSET_S(real_cpu_num(desc, i), setsize, desc->present) + && desc->maxmhz[i]) { + float freq = atof(desc->maxmhz[i]); + + if (freq > cpu_freq) + cpu_freq = freq; + } + } + } + snprintf(buf, bufsz, "%.4f", cpu_freq); + return buf; +} + +/* Read overall minimum frequency of cpu */ +static char * +cpu_min_mhz(struct lscpu_desc *desc, char *buf, size_t bufsz) +{ + int i; + float cpu_freq = -1.0; + size_t setsize = CPU_ALLOC_SIZE(maxcpus); + + if (desc->present) { + for (i = 0; i < desc->ncpuspos; i++) { + if (CPU_ISSET_S(real_cpu_num(desc, i), setsize, desc->present) + && desc->minmhz[i]) { + float freq = atof(desc->minmhz[i]); + + if (cpu_freq < 0.0 || freq < cpu_freq) + cpu_freq = freq; + } + } + } + snprintf(buf, bufsz, "%.4f", cpu_freq); + return buf; +} + + +static void +read_max_mhz(struct lscpu_desc *desc, int idx) +{ + int num = real_cpu_num(desc, idx); + int mhz; + + if (ul_path_readf_s32(desc->syscpu, &mhz, "cpu%d/cpufreq/cpuinfo_max_freq", num) != 0) + return; + if (!desc->maxmhz) + desc->maxmhz = xcalloc(desc->ncpuspos, sizeof(char *)); + xasprintf(&desc->maxmhz[idx], "%.4f", (float) mhz / 1000); +} + +static void +read_min_mhz(struct lscpu_desc *desc, int idx) +{ + int num = real_cpu_num(desc, idx); + int mhz; + + if (ul_path_readf_s32(desc->syscpu, &mhz, "cpu%d/cpufreq/cpuinfo_min_freq", num) != 0) + return; + if (!desc->minmhz) + desc->minmhz = xcalloc(desc->ncpuspos, sizeof(char *)); + xasprintf(&desc->minmhz[idx], "%.4f", (float) mhz / 1000); +} + +static int +cachecmp(const void *a, const void *b) +{ + struct cpu_cache *c1 = (struct cpu_cache *) a; + struct cpu_cache *c2 = (struct cpu_cache *) b; + + return strcmp(c2->name, c1->name); +} + +static void +read_cache(struct lscpu_desc *desc, int idx) +{ + char buf[256]; + int i; + int num = real_cpu_num(desc, idx); + + if (!desc->ncaches) { + while (ul_path_accessf(desc->syscpu, F_OK, + "cpu%d/cache/index%d", + num, desc->ncaches) == 0) + desc->ncaches++; + + if (!desc->ncaches) + return; + desc->caches = xcalloc(desc->ncaches, sizeof(*desc->caches)); + } + for (i = 0; i < desc->ncaches; i++) { + struct cpu_cache *ca = &desc->caches[i]; + cpu_set_t *map; + + if (ul_path_accessf(desc->syscpu, F_OK, + "cpu%d/cache/index%d", num, i) != 0) + continue; + if (!ca->name) { + int type = 0; + + /* cache type */ + if (ul_path_readf_string(desc->syscpu, &ca->type, + "cpu%d/cache/index%d/type", num, i) > 0) { + if (!strcmp(ca->type, "Data")) + type = 'd'; + else if (!strcmp(ca->type, "Instruction")) + type = 'i'; + } + + /* cache level */ + ul_path_readf_s32(desc->syscpu, &ca->level, + "cpu%d/cache/index%d/level", num, i); + if (type) + snprintf(buf, sizeof(buf), "L%d%c", ca->level, type); + else + snprintf(buf, sizeof(buf), "L%d", ca->level); + + ca->name = xstrdup(buf); + + ul_path_readf_u32(desc->syscpu, &ca->ways_of_associativity, + "cpu%d/cache/index%d/ways_of_associativity", num, i); + ul_path_readf_u32(desc->syscpu, &ca->physical_line_partition, + "cpu%d/cache/index%d/physical_line_partition", num, i); + ul_path_readf_u32(desc->syscpu, &ca->number_of_sets, + "cpu%d/cache/index%d/number_of_sets", num, i); + ul_path_readf_u32(desc->syscpu, &ca->coherency_line_size, + "cpu%d/cache/index%d/coherency_line_size", num, i); + + ul_path_readf_string(desc->syscpu, &ca->allocation_policy, + "cpu%d/cache/index%d/allocation_policy", num, i); + ul_path_readf_string(desc->syscpu, &ca->write_policy, + "cpu%d/cache/index%d/write_policy", num, i); + + /* cache size */ + if (ul_path_readf_buffer(desc->syscpu, buf, sizeof(buf), + "cpu%d/cache/index%d/size", num, i) > 0) + parse_size(buf, &ca->size, NULL); + else + ca->size = 0; + } + + /* information about how CPUs share different caches */ + ul_path_readf_cpuset(desc->syscpu, &map, maxcpus, + "cpu%d/cache/index%d/shared_cpu_map", num, i); + + if (!ca->sharedmaps) + ca->sharedmaps = xcalloc(desc->ncpuspos, sizeof(cpu_set_t *)); + add_cpuset_to_array(ca->sharedmaps, &ca->nsharedmaps, map); + } +} + +static inline int is_node_dirent(struct dirent *d) +{ + return + d && +#ifdef _DIRENT_HAVE_D_TYPE + (d->d_type == DT_DIR || d->d_type == DT_UNKNOWN) && +#endif + strncmp(d->d_name, "node", 4) == 0 && + isdigit_string(d->d_name + 4); +} + +static int +nodecmp(const void *ap, const void *bp) +{ + int *a = (int *) ap, *b = (int *) bp; + return *a - *b; +} + +static void +read_nodes(struct lscpu_desc *desc) +{ + int i = 0; + DIR *dir; + struct dirent *d; + struct path_cxt *sysnode; + + desc->nnodes = 0; + + sysnode = ul_new_path(_PATH_SYS_NODE); + if (!sysnode) + err(EXIT_FAILURE, _("failed to initialize %s handler"), _PATH_SYS_NODE); + ul_path_set_prefix(sysnode, desc->prefix); + + dir = ul_path_opendir(sysnode, NULL); + if (!dir) + goto done; + + while ((d = readdir(dir))) { + if (is_node_dirent(d)) + desc->nnodes++; + } + + if (!desc->nnodes) { + closedir(dir); + goto done; + } + + desc->nodemaps = xcalloc(desc->nnodes, sizeof(cpu_set_t *)); + desc->idx2nodenum = xmalloc(desc->nnodes * sizeof(int)); + + rewinddir(dir); + while ((d = readdir(dir)) && i < desc->nnodes) { + if (is_node_dirent(d)) + desc->idx2nodenum[i++] = strtol_or_err(((d->d_name) + 4), + _("Failed to extract the node number")); + } + closedir(dir); + qsort(desc->idx2nodenum, desc->nnodes, sizeof(int), nodecmp); + + /* information about how nodes share different CPUs */ + for (i = 0; i < desc->nnodes; i++) + ul_path_readf_cpuset(sysnode, &desc->nodemaps[i], maxcpus, + "node%d/cpumap", desc->idx2nodenum[i]); +done: + ul_unref_path(sysnode); +} + +static char * +get_cell_data(struct lscpu_desc *desc, int idx, int col, + struct lscpu_modifier *mod, + char *buf, size_t bufsz) +{ + size_t setsize = CPU_ALLOC_SIZE(maxcpus); + size_t i; + int cpu = real_cpu_num(desc, idx); + + *buf = '\0'; + + switch (col) { + case COL_CPU_CPU: + snprintf(buf, bufsz, "%d", cpu); + break; + case COL_CPU_CORE: + if (mod->physical) { + if (desc->coreids[idx] == -1) + snprintf(buf, bufsz, "-"); + else + snprintf(buf, bufsz, "%d", desc->coreids[idx]); + } else { + if (cpuset_ary_isset(cpu, desc->coremaps, + desc->ncores, setsize, &i) == 0) + snprintf(buf, bufsz, "%zu", i); + } + break; + case COL_CPU_SOCKET: + if (mod->physical) { + if (desc->socketids[idx] == -1) + snprintf(buf, bufsz, "-"); + else + snprintf(buf, bufsz, "%d", desc->socketids[idx]); + } else { + if (cpuset_ary_isset(cpu, desc->socketmaps, + desc->nsockets, setsize, &i) == 0) + snprintf(buf, bufsz, "%zu", i); + } + break; + case COL_CPU_NODE: + if (cpuset_ary_isset(cpu, desc->nodemaps, + desc->nnodes, setsize, &i) == 0) + snprintf(buf, bufsz, "%d", desc->idx2nodenum[i]); + break; + case COL_CPU_DRAWER: + if (!desc->drawerids || !desc->drawermaps) + break; + if (mod->physical) { + if (desc->drawerids[idx] == -1) + snprintf(buf, bufsz, "-"); + else + snprintf(buf, bufsz, "%d", desc->drawerids[idx]); + } else { + if (cpuset_ary_isset(cpu, desc->drawermaps, + desc->ndrawers, setsize, &i) == 0) + snprintf(buf, bufsz, "%zu", i); + } + break; + case COL_CPU_BOOK: + if (!desc->bookids || !desc->bookmaps) + break; + if (mod->physical) { + if (desc->bookids[idx] == -1) + snprintf(buf, bufsz, "-"); + else + snprintf(buf, bufsz, "%d", desc->bookids[idx]); + } else { + if (cpuset_ary_isset(cpu, desc->bookmaps, + desc->nbooks, setsize, &i) == 0) + snprintf(buf, bufsz, "%zu", i); + } + break; + case COL_CPU_CACHE: + { + char *p = buf; + size_t sz = bufsz; + int j; + + for (j = desc->ncaches - 1; j >= 0; j--) { + struct cpu_cache *ca = &desc->caches[j]; + + if (cpuset_ary_isset(cpu, ca->sharedmaps, + ca->nsharedmaps, setsize, &i) == 0) { + int x = snprintf(p, sz, "%zu", i); + if (x < 0 || (size_t) x >= sz) + return NULL; + p += x; + sz -= x; + } + if (j != 0) { + if (sz < 2) + return NULL; + *p++ = mod->compat ? ',' : ':'; + *p = '\0'; + sz--; + } + } + break; + } + case COL_CPU_POLARIZATION: + if (desc->polarization) { + int x = desc->polarization[idx]; + + snprintf(buf, bufsz, "%s", + mod->mode == OUTPUT_PARSABLE ? + polar_modes[x].parsable : + polar_modes[x].readable); + } + break; + case COL_CPU_ADDRESS: + if (desc->addresses) + snprintf(buf, bufsz, "%d", desc->addresses[idx]); + break; + case COL_CPU_CONFIGURED: + if (!desc->configured) + break; + if (mod->mode == OUTPUT_PARSABLE) + snprintf(buf, bufsz, "%s", + desc->configured[idx] ? _("Y") : _("N")); + else + snprintf(buf, bufsz, "%s", + desc->configured[idx] ? _("yes") : _("no")); + break; + case COL_CPU_ONLINE: + if (!desc->online) + break; + if (mod->mode == OUTPUT_PARSABLE) + snprintf(buf, bufsz, "%s", + is_cpu_online(desc, cpu) ? _("Y") : _("N")); + else + snprintf(buf, bufsz, "%s", + is_cpu_online(desc, cpu) ? _("yes") : _("no")); + break; + case COL_CPU_MAXMHZ: + if (desc->maxmhz && desc->maxmhz[idx]) + xstrncpy(buf, desc->maxmhz[idx], bufsz); + break; + case COL_CPU_MINMHZ: + if (desc->minmhz && desc->minmhz[idx]) + xstrncpy(buf, desc->minmhz[idx], bufsz); + break; + } + return buf; +} + +static char * +get_cell_header(struct lscpu_desc *desc, int col, + struct lscpu_modifier *mod, + char *buf, size_t bufsz) +{ + *buf = '\0'; + + if (col == COL_CPU_CACHE) { + char *p = buf; + size_t sz = bufsz; + int i; + + for (i = desc->ncaches - 1; i >= 0; i--) { + int x = snprintf(p, sz, "%s", desc->caches[i].name); + if (x < 0 || (size_t) x >= sz) + return NULL; + sz -= x; + p += x; + if (i > 0) { + if (sz < 2) + return NULL; + *p++ = mod->compat ? ',' : ':'; + *p = '\0'; + sz--; + } + } + if (desc->ncaches) + return buf; + } + snprintf(buf, bufsz, "%s", coldescs_cpu[col].name); + return buf; +} + +/* + * [-C] backend + */ +static void +print_caches_readable(struct lscpu_desc *desc, int cols[], int ncols, + struct lscpu_modifier *mod) +{ + struct libscols_table *table; + struct cpu_cache *cachesrc; + int i, end, j, shared_allsize; + + scols_init_debug(0); + + table = scols_new_table(); + if (!table) + err(EXIT_FAILURE, _("failed to allocate output table")); + if (mod->json) { + scols_table_enable_json(table, 1); + scols_table_set_name(table, "caches"); + } + + for (i = 0; i < ncols; i++) { + struct lscpu_coldesc *cd = &coldescs_cache[cols[i]]; + if (!scols_table_new_column(table, cd->name, 0, cd->flags)) + err(EXIT_FAILURE, _("failed to allocate output column")); + } + + for (j = 0; j < 2; j++) { + /* First check the caches from /sys/devices */ + if (j == 0) { + cachesrc = desc->caches; + end = desc->ncaches - 1; + shared_allsize = 0; + } else { + /* Check shared caches from /proc/cpuinfo s390 */ + cachesrc = desc->ecaches; + end = desc->necaches - 1; + /* Dont use get_cache_full_size */ + shared_allsize = 1; + } + + for (i = end; i >= 0; i--) { + struct libscols_line *line; + struct cpu_cache *ca = &cachesrc[i]; + int c; + + line = scols_table_new_line(table, NULL); + if (!line) + err(EXIT_FAILURE, _("failed to allocate output line")); + + for (c = 0; c < ncols; c++) { + char *data = NULL; + + switch (cols[c]) { + case COL_CACHE_NAME: + if (ca->name) + data = xstrdup(ca->name); + break; + case COL_CACHE_ONESIZE: + if (!ca->size) + break; + if (mod->bytes) + xasprintf(&data, "%" PRIu64, ca->size); + else + data = size_to_human_string(SIZE_SUFFIX_1LETTER, ca->size); + break; + case COL_CACHE_ALLSIZE: + { + uint64_t sz = 0; + if (shared_allsize) + break; + if (get_cache_full_size(desc, ca, &sz) != 0) + break; + if (mod->bytes) + xasprintf(&data, "%" PRIu64, sz); + else + data = size_to_human_string(SIZE_SUFFIX_1LETTER, sz); + break; + } + case COL_CACHE_WAYS: + if (ca->ways_of_associativity) + xasprintf(&data, "%u", ca->ways_of_associativity); + break; + case COL_CACHE_TYPE: + if (ca->type) + data = xstrdup(ca->type); + break; + case COL_CACHE_LEVEL: + if (ca->level) + xasprintf(&data, "%d", ca->level); + break; + case COL_CACHE_ALLOCPOL: + if (ca->allocation_policy) + data = xstrdup(ca->allocation_policy); + break; + case COL_CACHE_WRITEPOL: + if (ca->write_policy) + data = xstrdup(ca->write_policy); + break; + case COL_CACHE_PHYLINE: + if (ca->physical_line_partition) + xasprintf(&data, "%u", ca->physical_line_partition); + break; + case COL_CACHE_SETS: + if (ca->number_of_sets) + xasprintf(&data, "%u", ca->number_of_sets); + break; + case COL_CACHE_COHERENCYSIZE: + if (ca->coherency_line_size) + xasprintf(&data, "%u", ca->coherency_line_size); + break; + } + + if (data && scols_line_refer_data(line, c, data)) + err(EXIT_FAILURE, _("failed to add output data")); + } + } + + } + + scols_print_table(table); + scols_unref_table(table); +} + +/* + * [-p] backend, we support two parsable formats: + * + * 1) "compatible" -- this format is compatible with the original lscpu(1) + * output and it contains fixed set of the columns. The CACHE columns are at + * the end of the line and the CACHE is not printed if the number of the caches + * is zero. The CACHE columns are separated by two commas, for example: + * + * $ lscpu --parse + * # CPU,Core,Socket,Node,,L1d,L1i,L2 + * 0,0,0,0,,0,0,0 + * 1,1,0,0,,1,1,0 + * + * 2) "user defined output" -- this format prints always all columns without + * special prefix for CACHE column. If there are not CACHEs then the column is + * empty and the header "Cache" is printed rather than a real name of the cache. + * The CACHE columns are separated by ':'. + * + * $ lscpu --parse=CPU,CORE,SOCKET,NODE,CACHE + * # CPU,Core,Socket,Node,L1d:L1i:L2 + * 0,0,0,0,0:0:0 + * 1,1,0,0,1:1:0 + */ +static void +print_cpus_parsable(struct lscpu_desc *desc, int cols[], int ncols, + struct lscpu_modifier *mod) +{ + char buf[BUFSIZ], *data; + int i; + + /* + * Header + */ + printf(_( + "# The following is the parsable format, which can be fed to other\n" + "# programs. Each different item in every column has an unique ID\n" + "# starting from zero.\n")); + + fputs("# ", stdout); + for (i = 0; i < ncols; i++) { + int col = cols[i]; + + if (col == COL_CPU_CACHE) { + if (mod->compat && !desc->ncaches) + continue; + if (mod->compat && i != 0) + putchar(','); + } + if (i > 0) + putchar(','); + + data = get_cell_header(desc, col, mod, buf, sizeof(buf)); + + if (data && * data && col != COL_CPU_CACHE && + !coldescs_cpu[col].is_abbr) { + /* + * For normal column names use mixed case (e.g. "Socket") + */ + char *p = data + 1; + + while (p && *p != '\0') { + *p = tolower((unsigned int) *p); + p++; + } + } + fputs(data && *data ? data : "", stdout); + } + putchar('\n'); + + /* + * Data + */ + for (i = 0; i < desc->ncpuspos; i++) { + int c; + int cpu = real_cpu_num(desc, i); + + if (desc->online) { + if (!mod->offline && !is_cpu_online(desc, cpu)) + continue; + if (!mod->online && is_cpu_online(desc, cpu)) + continue; + } + if (desc->present && !is_cpu_present(desc, cpu)) + continue; + for (c = 0; c < ncols; c++) { + if (mod->compat && cols[c] == COL_CPU_CACHE) { + if (!desc->ncaches) + continue; + if (c > 0) + putchar(','); + } + if (c > 0) + putchar(','); + + data = get_cell_data(desc, i, cols[c], mod, + buf, sizeof(buf)); + fputs(data && *data ? data : "", stdout); + *buf = '\0'; + } + putchar('\n'); + } +} + +/* + * [-e] backend + */ +static void +print_cpus_readable(struct lscpu_desc *desc, int cols[], int ncols, + struct lscpu_modifier *mod) +{ + int i; + char buf[BUFSIZ]; + const char *data; + struct libscols_table *table; + + scols_init_debug(0); + + table = scols_new_table(); + if (!table) + err(EXIT_FAILURE, _("failed to allocate output table")); + if (mod->json) { + scols_table_enable_json(table, 1); + scols_table_set_name(table, "cpus"); + } + + for (i = 0; i < ncols; i++) { + data = get_cell_header(desc, cols[i], mod, buf, sizeof(buf)); + if (!scols_table_new_column(table, data, 0, coldescs_cpu[cols[i]].flags)) + err(EXIT_FAILURE, _("failed to allocate output column")); + } + + for (i = 0; i < desc->ncpuspos; i++) { + int c; + struct libscols_line *line; + int cpu = real_cpu_num(desc, i); + + if (desc->online) { + if (!mod->offline && !is_cpu_online(desc, cpu)) + continue; + if (!mod->online && is_cpu_online(desc, cpu)) + continue; + } + if (desc->present && !is_cpu_present(desc, cpu)) + continue; + + line = scols_table_new_line(table, NULL); + if (!line) + err(EXIT_FAILURE, _("failed to allocate output line")); + + for (c = 0; c < ncols; c++) { + data = get_cell_data(desc, i, cols[c], mod, + buf, sizeof(buf)); + if (!data || !*data) + data = "-"; + if (scols_line_set_data(line, c, data)) + err(EXIT_FAILURE, _("failed to add output data")); + } + } + + scols_print_table(table); + scols_unref_table(table); +} + + +static void __attribute__ ((__format__(printf, 3, 4))) + add_summary_sprint(struct libscols_table *tb, + const char *txt, + const char *fmt, + ...) +{ + struct libscols_line *ln = scols_table_new_line(tb, NULL); + char *data; + va_list args; + + if (!ln) + err(EXIT_FAILURE, _("failed to allocate output line")); + + /* description column */ + if (txt && scols_line_set_data(ln, 0, txt)) + err(EXIT_FAILURE, _("failed to add output data")); + + /* data column */ + va_start(args, fmt); + xvasprintf(&data, fmt, args); + va_end(args); + + if (data && scols_line_refer_data(ln, 1, data)) + err(EXIT_FAILURE, _("failed to add output data")); +} + +#define add_summary_n(tb, txt, num) add_summary_sprint(tb, txt, "%d", num) +#define add_summary_s(tb, txt, str) add_summary_sprint(tb, txt, "%s", str) + +static void +print_cpuset(struct libscols_table *tb, + const char *key, cpu_set_t *set, int hex) +{ + size_t setsize = CPU_ALLOC_SIZE(maxcpus); + size_t setbuflen = 7 * maxcpus; + char setbuf[setbuflen], *p; + + if (hex) { + p = cpumask_create(setbuf, setbuflen, set, setsize); + add_summary_s(tb, key, p); + } else { + p = cpulist_create(setbuf, setbuflen, set, setsize); + add_summary_s(tb, key, p); + } +} + +static int get_cache_full_size(struct lscpu_desc *desc, + struct cpu_cache *ca, uint64_t *res) +{ + size_t setsize = CPU_ALLOC_SIZE(maxcpus); + int i, nshares = 0; + + /* Count number of CPUs which shares the cache */ + for (i = 0; i < desc->ncpuspos; i++) { + int cpu = real_cpu_num(desc, i); + + if (desc->present && !is_cpu_present(desc, cpu)) + continue; + if (CPU_ISSET_S(cpu, setsize, ca->sharedmaps[0])) + nshares++; + } + + /* Correction for CPU threads */ + if (desc->nthreads > desc->ncores) + nshares /= (desc->nthreads / desc->ncores); + if (nshares < 1) + nshares = 1; + + *res = (desc->ncores / nshares) * ca->size; + return 0; +} + +/* + * default output + */ +static void +print_summary(struct lscpu_desc *desc, struct lscpu_modifier *mod) +{ + char buf[BUFSIZ]; + int i = 0; + size_t setsize = CPU_ALLOC_SIZE(maxcpus); + struct libscols_table *tb; + + scols_init_debug(0); + + tb = scols_new_table(); + if (!tb) + err(EXIT_FAILURE, _("failed to allocate output table")); + + scols_table_enable_noheadings(tb, 1); + if (mod->json) { + scols_table_enable_json(tb, 1); + scols_table_set_name(tb, "lscpu"); + } + + if (scols_table_new_column(tb, "field", 0, 0) == NULL || + scols_table_new_column(tb, "data", 0, SCOLS_FL_NOEXTREMES | SCOLS_FL_WRAP) == NULL) + err(EXIT_FAILURE, _("failed to initialize output column")); + + add_summary_s(tb, _("Architecture:"), desc->arch); + if (desc->mode) { + char *p = buf; + + if (desc->mode & MODE_32BIT) { + strcpy(p, "32-bit, "); + p += 8; + } + if (desc->mode & MODE_64BIT) { + strcpy(p, "64-bit, "); + p += 8; + } + *(p - 2) = '\0'; + add_summary_s(tb, _("CPU op-mode(s):"), buf); + } +#if !defined(WORDS_BIGENDIAN) + add_summary_s(tb, _("Byte Order:"), "Little Endian"); +#else + add_summary_s(tb, _("Byte Order:"), "Big Endian"); +#endif + + if (desc->addrsz) + add_summary_s(tb, _("Address sizes:"), desc->addrsz); + + add_summary_n(tb, _("CPU(s):"), desc->ncpus); + + if (desc->online) + print_cpuset(tb, mod->hex ? _("On-line CPU(s) mask:") : + _("On-line CPU(s) list:"), + desc->online, mod->hex); + + if (desc->online && CPU_COUNT_S(setsize, desc->online) != desc->ncpus) { + cpu_set_t *set; + + /* Linux kernel provides cpuset of off-line CPUs that contains + * all configured CPUs (see /sys/devices/system/cpu/offline), + * but want to print real (present in system) off-line CPUs only. + */ + set = cpuset_alloc(maxcpus, NULL, NULL); + if (!set) + err(EXIT_FAILURE, _("failed to callocate cpu set")); + CPU_ZERO_S(setsize, set); + for (i = 0; i < desc->ncpuspos; i++) { + int cpu = real_cpu_num(desc, i); + if (!is_cpu_online(desc, cpu) && is_cpu_present(desc, cpu)) + CPU_SET_S(cpu, setsize, set); + } + print_cpuset(tb, mod->hex ? _("Off-line CPU(s) mask:") : + _("Off-line CPU(s) list:"), + set, mod->hex); + cpuset_free(set); + } + + if (desc->nsockets) { + int threads_per_core, cores_per_socket, sockets_per_book; + int books_per_drawer, drawers; + FILE *fd; + + threads_per_core = cores_per_socket = sockets_per_book = 0; + books_per_drawer = drawers = 0; + /* s390 detects its cpu topology via /proc/sysinfo, if present. + * Using simply the cpu topology masks in sysfs will not give + * usable results since everything is virtualized. E.g. + * virtual core 0 may have only 1 cpu, but virtual core 2 may + * five cpus. + * If the cpu topology is not exported (e.g. 2nd level guest) + * fall back to old calculation scheme. + */ + if ((fd = ul_path_fopen(desc->procfs, "r", "sysinfo"))) { + int t0, t1; + + while (fd && fgets(buf, sizeof(buf), fd) != NULL) { + if (sscanf(buf, "CPU Topology SW:%d%d%d%d%d%d", + &t0, &t1, &drawers, &books_per_drawer, + &sockets_per_book, + &cores_per_socket) == 6) + break; + } + if (fd) + fclose(fd); + } + if (desc->mtid) + threads_per_core = atoi(desc->mtid) + 1; + add_summary_n(tb, _("Thread(s) per core:"), + threads_per_core ?: desc->nthreads / desc->ncores); + add_summary_n(tb, _("Core(s) per socket:"), + cores_per_socket ?: desc->ncores / desc->nsockets); + if (desc->nbooks) { + add_summary_n(tb, _("Socket(s) per book:"), + sockets_per_book ?: desc->nsockets / desc->nbooks); + if (desc->ndrawers) { + add_summary_n(tb, _("Book(s) per drawer:"), + books_per_drawer ?: desc->nbooks / desc->ndrawers); + add_summary_n(tb, _("Drawer(s):"), drawers ?: desc->ndrawers); + } else { + add_summary_n(tb, _("Book(s):"), books_per_drawer ?: desc->nbooks); + } + } else { + add_summary_n(tb, _("Socket(s):"), sockets_per_book ?: desc->nsockets); + } + } + if (desc->nnodes) + add_summary_n(tb, _("NUMA node(s):"), desc->nnodes); + if (desc->vendor) + add_summary_s(tb, _("Vendor ID:"), desc->vendor); + if (desc->machinetype) + add_summary_s(tb, _("Machine type:"), desc->machinetype); + if (desc->family) + add_summary_s(tb, _("CPU family:"), desc->family); + if (desc->model || desc->revision) + add_summary_s(tb, _("Model:"), desc->revision ? desc->revision : desc->model); + if (desc->modelname || desc->cpu) + add_summary_s(tb, _("Model name:"), desc->cpu ? desc->cpu : desc->modelname); + if (desc->stepping) + add_summary_s(tb, _("Stepping:"), desc->stepping); + if (desc->freqboost >= 0) + add_summary_s(tb, _("Frequency boost:"), desc->freqboost ? + _("enabled") : _("disabled")); + if (desc->mhz) + add_summary_s(tb, _("CPU MHz:"), desc->mhz); + if (desc->dynamic_mhz) + add_summary_s(tb, _("CPU dynamic MHz:"), desc->dynamic_mhz); + if (desc->static_mhz) + add_summary_s(tb, _("CPU static MHz:"), desc->static_mhz); + if (desc->maxmhz) + add_summary_s(tb, _("CPU max MHz:"), cpu_max_mhz(desc, buf, sizeof(buf))); + if (desc->minmhz) + add_summary_s(tb, _("CPU min MHz:"), cpu_min_mhz(desc, buf, sizeof(buf))); + if (desc->bogomips) + add_summary_s(tb, _("BogoMIPS:"), desc->bogomips); + if (desc->virtflag) { + if (!strcmp(desc->virtflag, "svm")) + add_summary_s(tb, _("Virtualization:"), "AMD-V"); + else if (!strcmp(desc->virtflag, "vmx")) + add_summary_s(tb, _("Virtualization:"), "VT-x"); + } + if (desc->hypervisor) + add_summary_s(tb, _("Hypervisor:"), desc->hypervisor); + if (desc->hyper) { + add_summary_s(tb, _("Hypervisor vendor:"), hv_vendors[desc->hyper]); + add_summary_s(tb, _("Virtualization type:"), _(virt_types[desc->virtype])); + } + if (desc->dispatching >= 0) + add_summary_s(tb, _("Dispatching mode:"), _(disp_modes[desc->dispatching])); + if (desc->ncaches) { + for (i = desc->ncaches - 1; i >= 0; i--) { + uint64_t sz = 0; + char *tmp; + struct cpu_cache *ca = &desc->caches[i]; + + if (ca->size == 0) + continue; + if (get_cache_full_size(desc, ca, &sz) != 0 || sz == 0) + continue; + if (mod->bytes) + xasprintf(&tmp, "%" PRIu64, sz); + else + tmp = size_to_human_string( + SIZE_SUFFIX_3LETTER | SIZE_SUFFIX_SPACE, + sz); + snprintf(buf, sizeof(buf), _("%s cache:"), ca->name); + add_summary_s(tb, buf, tmp); + free(tmp); + } + } + if (desc->necaches) { + for (i = desc->necaches - 1; i >= 0; i--) { + char *tmp; + struct cpu_cache *ca = &desc->ecaches[i]; + + if (ca->size == 0) + continue; + if (mod->bytes) + xasprintf(&tmp, "%" PRIu64, ca->size); + else + tmp = size_to_human_string( + SIZE_SUFFIX_3LETTER | SIZE_SUFFIX_SPACE, + ca->size); + snprintf(buf, sizeof(buf), _("%s cache:"), ca->name); + add_summary_s(tb, buf, tmp); + free(tmp); + } + } + + for (i = 0; i < desc->nnodes; i++) { + snprintf(buf, sizeof(buf), _("NUMA node%d CPU(s):"), desc->idx2nodenum[i]); + print_cpuset(tb, buf, desc->nodemaps[i], mod->hex); + } + + if (desc->physsockets) { + add_summary_n(tb, _("Physical sockets:"), desc->physsockets); + add_summary_n(tb, _("Physical chips:"), desc->physchips); + add_summary_n(tb, _("Physical cores/chip:"), desc->physcoresperchip); + } + + if (desc->vuls) { + for (i = 0; i < desc->nvuls; i++) { + snprintf(buf, sizeof(buf), ("Vulnerability %s:"), desc->vuls[i].name); + add_summary_s(tb, buf, desc->vuls[i].text); + } + } + + if (desc->flags) + add_summary_s(tb, _("Flags:"), desc->flags); + + scols_print_table(tb); + scols_unref_table(tb); +} + +static void __attribute__((__noreturn__)) usage(void) +{ + FILE *out = stdout; + size_t i; + + fputs(USAGE_HEADER, out); + fprintf(out, _(" %s [options]\n"), program_invocation_short_name); + + fputs(USAGE_SEPARATOR, out); + fputs(_("Display information about the CPU architecture.\n"), out); + + fputs(USAGE_OPTIONS, out); + fputs(_(" -a, --all print both online and offline CPUs (default for -e)\n"), out); + fputs(_(" -b, --online print online CPUs only (default for -p)\n"), out); + fputs(_(" -B, --bytes print sizes in bytes rather than in human readable format\n"), out); + fputs(_(" -C, --caches[=<list>] info about caches in extended readable format\n"), out); + fputs(_(" -c, --offline print offline CPUs only\n"), out); + fputs(_(" -J, --json use JSON for default or extended format\n"), out); + fputs(_(" -e, --extended[=<list>] print out an extended readable format\n"), out); + fputs(_(" -p, --parse[=<list>] print out a parsable format\n"), out); + fputs(_(" -s, --sysroot <dir> use specified directory as system root\n"), out); + fputs(_(" -x, --hex print hexadecimal masks rather than lists of CPUs\n"), out); + fputs(_(" -y, --physical print physical instead of logical IDs\n"), out); + fputs(_(" --output-all print all available columns for -e, -p or -C\n"), out); + fputs(USAGE_SEPARATOR, out); + printf(USAGE_HELP_OPTIONS(25)); + + fputs(_("\nAvailable output columns for -e or -p:\n"), out); + for (i = 0; i < ARRAY_SIZE(coldescs_cpu); i++) + fprintf(out, " %13s %s\n", coldescs_cpu[i].name, _(coldescs_cpu[i].help)); + + fputs(_("\nAvailable output columns for -C:\n"), out); + for (i = 0; i < ARRAY_SIZE(coldescs_cache); i++) + fprintf(out, " %13s %s\n", coldescs_cache[i].name, _(coldescs_cache[i].help)); + + printf(USAGE_MAN_TAIL("lscpu(1)")); + + exit(EXIT_SUCCESS); +} + +int main(int argc, char *argv[]) +{ + struct lscpu_modifier _mod = { .mode = OUTPUT_SUMMARY }, *mod = &_mod; + struct lscpu_desc _desc = { .flags = NULL }, *desc = &_desc; + int c, i, all = 0; + int columns[ARRAY_SIZE(coldescs_cpu)], ncolumns = 0; + int cpu_modifier_specified = 0; + size_t setsize; + + enum { + OPT_OUTPUT_ALL = CHAR_MAX + 1, + }; + static const struct option longopts[] = { + { "all", no_argument, NULL, 'a' }, + { "online", no_argument, NULL, 'b' }, + { "bytes", no_argument, NULL, 'B' }, + { "caches", optional_argument, NULL, 'C' }, + { "offline", no_argument, NULL, 'c' }, + { "help", no_argument, NULL, 'h' }, + { "extended", optional_argument, NULL, 'e' }, + { "json", no_argument, NULL, 'J' }, + { "parse", optional_argument, NULL, 'p' }, + { "sysroot", required_argument, NULL, 's' }, + { "physical", no_argument, NULL, 'y' }, + { "hex", no_argument, NULL, 'x' }, + { "version", no_argument, NULL, 'V' }, + { "output-all", no_argument, NULL, OPT_OUTPUT_ALL }, + { NULL, 0, NULL, 0 } + }; + + static const ul_excl_t excl[] = { /* rows and cols in ASCII order */ + { 'C','e','p' }, + { 'a','b','c' }, + { 0 } + }; + int excl_st[ARRAY_SIZE(excl)] = UL_EXCL_STATUS_INIT; + + setlocale(LC_ALL, ""); + bindtextdomain(PACKAGE, LOCALEDIR); + textdomain(PACKAGE); + close_stdout_atexit(); + + while ((c = getopt_long(argc, argv, "aBbC::ce::hJp::s:xyV", longopts, NULL)) != -1) { + + err_exclusive_options(c, longopts, excl, excl_st); + + switch (c) { + case 'a': + mod->online = mod->offline = 1; + cpu_modifier_specified = 1; + break; + case 'B': + mod->bytes = 1; + break; + case 'b': + mod->online = 1; + cpu_modifier_specified = 1; + break; + case 'c': + mod->offline = 1; + cpu_modifier_specified = 1; + break; + case 'C': + if (optarg) { + if (*optarg == '=') + optarg++; + ncolumns = string_to_idarray(optarg, + columns, ARRAY_SIZE(columns), + cache_column_name_to_id); + if (ncolumns < 0) + return EXIT_FAILURE; + } + mod->mode = OUTPUT_CACHES; + break; + case 'J': + mod->json = 1; + break; + case 'p': + case 'e': + if (optarg) { + if (*optarg == '=') + optarg++; + ncolumns = string_to_idarray(optarg, + columns, ARRAY_SIZE(columns), + cpu_column_name_to_id); + if (ncolumns < 0) + return EXIT_FAILURE; + } + mod->mode = c == 'p' ? OUTPUT_PARSABLE : OUTPUT_READABLE; + break; + case 's': + desc->prefix = optarg; + mod->system = SYSTEM_SNAPSHOT; + break; + case 'x': + mod->hex = 1; + break; + case 'y': + mod->physical = 1; + break; + case OPT_OUTPUT_ALL: + all = 1; + break; + + case 'h': + usage(); + case 'V': + print_version(EXIT_SUCCESS); + default: + errtryhelp(EXIT_FAILURE); + } + } + + if (all && ncolumns == 0) { + size_t sz, maxsz = mod->mode == OUTPUT_CACHES ? + ARRAY_SIZE(coldescs_cache) : + ARRAY_SIZE(coldescs_cpu); + + for (sz = 0; sz < maxsz; sz++) + columns[ncolumns++] = sz; + } + + if (cpu_modifier_specified && mod->mode == OUTPUT_SUMMARY) { + fprintf(stderr, + _("%s: options --all, --online and --offline may only " + "be used with options --extended or --parse.\n"), + program_invocation_short_name); + return EXIT_FAILURE; + } + + if (argc != optind) { + warnx(_("bad usage")); + errtryhelp(EXIT_FAILURE); + } + + /* set default cpu display mode if none was specified */ + if (!mod->online && !mod->offline) { + mod->online = 1; + mod->offline = mod->mode == OUTPUT_READABLE ? 1 : 0; + } + + ul_path_init_debug(); + + /* /sys/devices/system/cpu */ + desc->syscpu = ul_new_path(_PATH_SYS_CPU); + if (!desc->syscpu) + err(EXIT_FAILURE, _("failed to initialize CPUs sysfs handler")); + if (desc->prefix) + ul_path_set_prefix(desc->syscpu, desc->prefix); + + /* /proc */ + desc->procfs = ul_new_path("/proc"); + if (!desc->procfs) + err(EXIT_FAILURE, _("failed to initialize procfs handler")); + if (desc->prefix) + ul_path_set_prefix(desc->procfs, desc->prefix); + + read_basicinfo(desc, mod); + + setsize = CPU_ALLOC_SIZE(maxcpus); + + for (i = 0; i < desc->ncpuspos; i++) { + /* only consider present CPUs */ + if (desc->present && + !CPU_ISSET_S(real_cpu_num(desc, i), setsize, desc->present)) + continue; + read_topology(desc, i); + read_cache(desc, i); + read_polarization(desc, i); + read_address(desc, i); + read_configured(desc, i); + read_max_mhz(desc, i); + read_min_mhz(desc, i); + } + + if (desc->caches) + qsort(desc->caches, desc->ncaches, + sizeof(struct cpu_cache), cachecmp); + + if (desc->ecaches) + qsort(desc->ecaches, desc->necaches, + sizeof(struct cpu_cache), cachecmp); + + read_nodes(desc); + read_hypervisor(desc, mod); + arm_cpu_decode(desc); + + switch(mod->mode) { + case OUTPUT_SUMMARY: + print_summary(desc, mod); + break; + case OUTPUT_CACHES: + if (!ncolumns) { + columns[ncolumns++] = COL_CACHE_NAME; + columns[ncolumns++] = COL_CACHE_ONESIZE; + columns[ncolumns++] = COL_CACHE_ALLSIZE; + columns[ncolumns++] = COL_CACHE_WAYS; + columns[ncolumns++] = COL_CACHE_TYPE; + columns[ncolumns++] = COL_CACHE_LEVEL; + columns[ncolumns++] = COL_CACHE_SETS; + columns[ncolumns++] = COL_CACHE_PHYLINE; + columns[ncolumns++] = COL_CACHE_COHERENCYSIZE; + } + print_caches_readable(desc, columns, ncolumns, mod); + break; + case OUTPUT_PARSABLE: + if (!ncolumns) { + columns[ncolumns++] = COL_CPU_CPU; + columns[ncolumns++] = COL_CPU_CORE; + columns[ncolumns++] = COL_CPU_SOCKET; + columns[ncolumns++] = COL_CPU_NODE; + columns[ncolumns++] = COL_CPU_CACHE; + mod->compat = 1; + } + print_cpus_parsable(desc, columns, ncolumns, mod); + break; + case OUTPUT_READABLE: + if (!ncolumns) { + /* No list was given. Just print whatever is there. */ + columns[ncolumns++] = COL_CPU_CPU; + if (desc->nodemaps) + columns[ncolumns++] = COL_CPU_NODE; + if (desc->drawermaps) + columns[ncolumns++] = COL_CPU_DRAWER; + if (desc->bookmaps) + columns[ncolumns++] = COL_CPU_BOOK; + if (desc->socketmaps) + columns[ncolumns++] = COL_CPU_SOCKET; + if (desc->coremaps) + columns[ncolumns++] = COL_CPU_CORE; + if (desc->caches) + columns[ncolumns++] = COL_CPU_CACHE; + if (desc->online) + columns[ncolumns++] = COL_CPU_ONLINE; + if (desc->configured) + columns[ncolumns++] = COL_CPU_CONFIGURED; + if (desc->polarization) + columns[ncolumns++] = COL_CPU_POLARIZATION; + if (desc->addresses) + columns[ncolumns++] = COL_CPU_ADDRESS; + if (desc->maxmhz) + columns[ncolumns++] = COL_CPU_MAXMHZ; + if (desc->minmhz) + columns[ncolumns++] = COL_CPU_MINMHZ; + } + print_cpus_readable(desc, columns, ncolumns, mod); + break; + } + + ul_unref_path(desc->syscpu); + ul_unref_path(desc->procfs); + return EXIT_SUCCESS; +} |