diff options
Diffstat (limited to 'tools')
31 files changed, 2450 insertions, 672 deletions
diff --git a/tools/lib/traceevent/.gitignore b/tools/lib/traceevent/.gitignore new file mode 100644 index 00000000000..35f56be5a4c --- /dev/null +++ b/tools/lib/traceevent/.gitignore @@ -0,0 +1 @@ +TRACEEVENT-CFLAGS diff --git a/tools/lib/traceevent/Makefile b/tools/lib/traceevent/Makefile index 46c2f6b7b12..14131cb0522 100644 --- a/tools/lib/traceevent/Makefile +++ b/tools/lib/traceevent/Makefile @@ -207,7 +207,7 @@ libtraceevent.so: $(PEVENT_LIB_OBJS) libtraceevent.a: $(PEVENT_LIB_OBJS) $(Q)$(do_build_static_lib) -$(PEVENT_LIB_OBJS): %.o: $(src)/%.c +$(PEVENT_LIB_OBJS): %.o: $(src)/%.c TRACEEVENT-CFLAGS $(Q)$(do_fpic_compile) define make_version.h @@ -272,6 +272,16 @@ ifneq ($(dep_includes),) include $(dep_includes) endif +### Detect environment changes +TRACK_CFLAGS = $(subst ','\'',$(CFLAGS)):$(ARCH):$(CROSS_COMPILE) + +TRACEEVENT-CFLAGS: force + @FLAGS='$(TRACK_CFLAGS)'; \ + if test x"$$FLAGS" != x"`cat TRACEEVENT-CFLAGS 2>/dev/null`" ; then \ + echo 1>&2 " * new build flags or cross compiler"; \ + echo "$$FLAGS" >TRACEEVENT-CFLAGS; \ + fi + tags: force $(RM) tags find . -name '*.[ch]' | xargs ctags --extra=+f --c-kinds=+px \ @@ -297,7 +307,7 @@ install: install_lib clean: $(RM) *.o *~ $(TARGETS) *.a *.so $(VERSION_FILES) .*.d - $(RM) tags TAGS + $(RM) TRACEEVENT-CFLAGS tags TAGS endif # skip-makefile diff --git a/tools/perf/Makefile b/tools/perf/Makefile index 75d74e5db8d..77f124fe57a 100644 --- a/tools/perf/Makefile +++ b/tools/perf/Makefile @@ -354,6 +354,7 @@ LIB_OBJS += $(OUTPUT)util/usage.o LIB_OBJS += $(OUTPUT)util/wrapper.o LIB_OBJS += $(OUTPUT)util/sigchain.o LIB_OBJS += $(OUTPUT)util/symbol.o +LIB_OBJS += $(OUTPUT)util/dso-test-data.o LIB_OBJS += $(OUTPUT)util/color.o LIB_OBJS += $(OUTPUT)util/pager.o LIB_OBJS += $(OUTPUT)util/header.o @@ -803,6 +804,9 @@ $(OUTPUT)ui/browsers/map.o: ui/browsers/map.c $(OUTPUT)PERF-CFLAGS $(OUTPUT)util/rbtree.o: ../../lib/rbtree.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -DETC_PERFCONFIG='"$(ETC_PERFCONFIG_SQ)"' $< +$(OUTPUT)util/parse-events.o: util/parse-events.c $(OUTPUT)PERF-CFLAGS + $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) -Wno-redundant-decls $< + $(OUTPUT)util/scripting-engines/trace-event-perl.o: util/scripting-engines/trace-event-perl.c $(OUTPUT)PERF-CFLAGS $(QUIET_CC)$(CC) -o $@ -c $(ALL_CFLAGS) $(PERL_EMBED_CCOPTS) -Wno-redundant-decls -Wno-strict-prototypes -Wno-unused-parameter -Wno-shadow $< diff --git a/tools/perf/builtin-test.c b/tools/perf/builtin-test.c index 5ce30305462..d909eb74a0e 100644 --- a/tools/perf/builtin-test.c +++ b/tools/perf/builtin-test.c @@ -1142,6 +1142,10 @@ static struct test { .func = test__perf_pmu, }, { + .desc = "Test dso data interface", + .func = dso__test_data, + }, + { .func = NULL, }, }; diff --git a/tools/perf/builtin-top.c b/tools/perf/builtin-top.c index e3cab5f088f..35e86c6df71 100644 --- a/tools/perf/builtin-top.c +++ b/tools/perf/builtin-top.c @@ -125,7 +125,7 @@ static int perf_top__parse_source(struct perf_top *top, struct hist_entry *he) /* * We can't annotate with just /proc/kallsyms */ - if (map->dso->symtab_type == SYMTAB__KALLSYMS) { + if (map->dso->symtab_type == DSO_BINARY_TYPE__KALLSYMS) { pr_err("Can't annotate %s: No vmlinux file was found in the " "path\n", sym->name); sleep(1); diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index 482f0517b61..413bd62eedb 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -978,8 +978,8 @@ static int hist_browser__dump(struct hist_browser *browser) fp = fopen(filename, "w"); if (fp == NULL) { char bf[64]; - strerror_r(errno, bf, sizeof(bf)); - ui_helpline__fpush("Couldn't write to %s: %s", filename, bf); + const char *err = strerror_r(errno, bf, sizeof(bf)); + ui_helpline__fpush("Couldn't write to %s: %s", filename, err); return -1; } diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index 8069dfb5ba7..3a282c0057d 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -426,7 +426,18 @@ int symbol__alloc_hist(struct symbol *sym) { struct annotation *notes = symbol__annotation(sym); const size_t size = symbol__size(sym); - size_t sizeof_sym_hist = (sizeof(struct sym_hist) + size * sizeof(u64)); + size_t sizeof_sym_hist; + + /* Check for overflow when calculating sizeof_sym_hist */ + if (size > (SIZE_MAX - sizeof(struct sym_hist)) / sizeof(u64)) + return -1; + + sizeof_sym_hist = (sizeof(struct sym_hist) + size * sizeof(u64)); + + /* Check for overflow in zalloc argument */ + if (sizeof_sym_hist > (SIZE_MAX - sizeof(*notes->src)) + / symbol_conf.nr_events) + return -1; notes->src = zalloc(sizeof(*notes->src) + symbol_conf.nr_events * sizeof_sym_hist); if (notes->src == NULL) @@ -777,7 +788,7 @@ fallback: free_filename = false; } - if (dso->symtab_type == SYMTAB__KALLSYMS) { + if (dso->symtab_type == DSO_BINARY_TYPE__KALLSYMS) { char bf[BUILD_ID_SIZE * 2 + 16] = " with build id "; char *build_id_msg = NULL; diff --git a/tools/perf/util/dso-test-data.c b/tools/perf/util/dso-test-data.c new file mode 100644 index 00000000000..541cdc72c7d --- /dev/null +++ b/tools/perf/util/dso-test-data.c @@ -0,0 +1,153 @@ +#include "util.h" + +#include <stdlib.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <string.h> + +#include "symbol.h" + +#define TEST_ASSERT_VAL(text, cond) \ +do { \ + if (!(cond)) { \ + pr_debug("FAILED %s:%d %s\n", __FILE__, __LINE__, text); \ + return -1; \ + } \ +} while (0) + +static char *test_file(int size) +{ + static char buf_templ[] = "/tmp/test-XXXXXX"; + char *templ = buf_templ; + int fd, i; + unsigned char *buf; + + fd = mkostemp(templ, O_CREAT|O_WRONLY|O_TRUNC); + + buf = malloc(size); + if (!buf) { + close(fd); + return NULL; + } + + for (i = 0; i < size; i++) + buf[i] = (unsigned char) ((int) i % 10); + + if (size != write(fd, buf, size)) + templ = NULL; + + close(fd); + return templ; +} + +#define TEST_FILE_SIZE (DSO__DATA_CACHE_SIZE * 20) + +struct test_data_offset { + off_t offset; + u8 data[10]; + int size; +}; + +struct test_data_offset offsets[] = { + /* Fill first cache page. */ + { + .offset = 10, + .data = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }, + .size = 10, + }, + /* Read first cache page. */ + { + .offset = 10, + .data = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }, + .size = 10, + }, + /* Fill cache boundary pages. */ + { + .offset = DSO__DATA_CACHE_SIZE - DSO__DATA_CACHE_SIZE % 10, + .data = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }, + .size = 10, + }, + /* Read cache boundary pages. */ + { + .offset = DSO__DATA_CACHE_SIZE - DSO__DATA_CACHE_SIZE % 10, + .data = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }, + .size = 10, + }, + /* Fill final cache page. */ + { + .offset = TEST_FILE_SIZE - 10, + .data = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }, + .size = 10, + }, + /* Read final cache page. */ + { + .offset = TEST_FILE_SIZE - 10, + .data = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 }, + .size = 10, + }, + /* Read final cache page. */ + { + .offset = TEST_FILE_SIZE - 3, + .data = { 7, 8, 9, 0, 0, 0, 0, 0, 0, 0 }, + .size = 3, + }, +}; + +int dso__test_data(void) +{ + struct machine machine; + struct dso *dso; + char *file = test_file(TEST_FILE_SIZE); + size_t i; + + TEST_ASSERT_VAL("No test file", file); + + memset(&machine, 0, sizeof(machine)); + + dso = dso__new((const char *)file); + + /* Basic 10 bytes tests. */ + for (i = 0; i < ARRAY_SIZE(offsets); i++) { + struct test_data_offset *data = &offsets[i]; + ssize_t size; + u8 buf[10]; + + memset(buf, 0, 10); + size = dso__data_read_offset(dso, &machine, data->offset, + buf, 10); + + TEST_ASSERT_VAL("Wrong size", size == data->size); + TEST_ASSERT_VAL("Wrong data", !memcmp(buf, data->data, 10)); + } + + /* Read cross multiple cache pages. */ + { + ssize_t size; + int c; + u8 *buf; + + buf = malloc(TEST_FILE_SIZE); + TEST_ASSERT_VAL("ENOMEM\n", buf); + + /* First iteration to fill caches, second one to read them. */ + for (c = 0; c < 2; c++) { + memset(buf, 0, TEST_FILE_SIZE); + size = dso__data_read_offset(dso, &machine, 10, + buf, TEST_FILE_SIZE); + + TEST_ASSERT_VAL("Wrong size", + size == (TEST_FILE_SIZE - 10)); + + for (i = 0; i < (size_t)size; i++) + TEST_ASSERT_VAL("Wrong data", + buf[i] == (i % 10)); + } + + free(buf); + } + + dso__delete(dso); + unlink(file); + return 0; +} diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c index f74e9560350..3edfd348381 100644 --- a/tools/perf/util/evlist.c +++ b/tools/perf/util/evlist.c @@ -214,7 +214,7 @@ int perf_evlist__add_tracepoints(struct perf_evlist *evlist, attrs[i].type = PERF_TYPE_TRACEPOINT; attrs[i].config = err; attrs[i].sample_type = (PERF_SAMPLE_RAW | PERF_SAMPLE_TIME | - PERF_SAMPLE_CPU); + PERF_SAMPLE_CPU | PERF_SAMPLE_PERIOD); attrs[i].sample_period = 1; } diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index 5a47aba4675..3a6d2044333 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -1212,6 +1212,12 @@ static void print_event_desc(struct perf_header *ph, int fd, FILE *fp) attr.exclude_user, attr.exclude_kernel); + fprintf(fp, ", excl_host = %d, excl_guest = %d", + attr.exclude_host, + attr.exclude_guest); + + fprintf(fp, ", precise_ip = %d", attr.precise_ip); + if (nr) fprintf(fp, ", id = {"); diff --git a/tools/perf/util/hist.c b/tools/perf/util/hist.c index 514e2a4b367..f247ef2789a 100644 --- a/tools/perf/util/hist.c +++ b/tools/perf/util/hist.c @@ -708,7 +708,7 @@ static size_t callchain__fprintf_graph(FILE *fp, struct rb_root *root, bool printed = false; struct rb_node *node; int i = 0; - int ret; + int ret = 0; /* * If have one single callchain root, don't bother printing @@ -747,8 +747,11 @@ static size_t callchain__fprintf_graph(FILE *fp, struct rb_root *root, root = &cnode->rb_root; } - return __callchain__fprintf_graph(fp, root, total_samples, + ret += __callchain__fprintf_graph(fp, root, total_samples, 1, 1, left_margin); + ret += fprintf(fp, "\n"); + + return ret; } static size_t __callchain__fprintf_flat(FILE *fp, diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index a1f4e366914..cc33486ad9e 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -7,6 +7,8 @@ #include <stdio.h> #include <unistd.h> #include "map.h" +#include "thread.h" +#include "strlist.h" const char *map_type__name[MAP__NR_TYPES] = { [MAP__FUNCTION] = "Functions", @@ -585,7 +587,21 @@ int machine__init(struct machine *self, const char *root_dir, pid_t pid) self->kmaps.machine = self; self->pid = pid; self->root_dir = strdup(root_dir); - return self->root_dir == NULL ? -ENOMEM : 0; + if (self->root_dir == NULL) + return -ENOMEM; + + if (pid != HOST_KERNEL_ID) { + struct thread *thread = machine__findnew_thread(self, pid); + char comm[64]; + + if (thread == NULL) + return -ENOMEM; + + snprintf(comm, sizeof(comm), "[guest/%d]", pid); + thread__set_comm(thread, comm); + } + + return 0; } static void dsos__delete(struct list_head *self) @@ -680,7 +696,15 @@ struct machine *machines__findnew(struct rb_root *self, pid_t pid) (symbol_conf.guestmount)) { sprintf(path, "%s/%d", symbol_conf.guestmount, pid); if (access(path, R_OK)) { - pr_err("Can't access file %s\n", path); + static struct strlist *seen; + + if (!seen) + seen = strlist__new(true, NULL); + + if (!strlist__has_entry(seen, path)) { + pr_err("Can't access file %s\n", path); + strlist__add(seen, path); + } machine = NULL; goto out; } @@ -714,3 +738,16 @@ char *machine__mmap_name(struct machine *self, char *bf, size_t size) return bf; } + +void machines__set_id_hdr_size(struct rb_root *machines, u16 id_hdr_size) +{ + struct rb_node *node; + struct machine *machine; + + for (node = rb_first(machines); node; node = rb_next(node)) { + machine = rb_entry(node, struct machine, rb_node); + machine->id_hdr_size = id_hdr_size; + } + + return; +} diff --git a/tools/perf/util/map.h b/tools/perf/util/map.h index c14c665d9a2..03a1e9b08b2 100644 --- a/tools/perf/util/map.h +++ b/tools/perf/util/map.h @@ -151,6 +151,7 @@ struct machine *machines__add(struct rb_root *self, pid_t pid, struct machine *machines__find_host(struct rb_root *self); struct machine *machines__find(struct rb_root *self, pid_t pid); struct machine *machines__findnew(struct rb_root *self, pid_t pid); +void machines__set_id_hdr_size(struct rb_root *self, u16 id_hdr_size); char *machine__mmap_name(struct machine *self, char *bf, size_t size); int machine__init(struct machine *self, const char *root_dir, pid_t pid); void machine__exit(struct machine *self); diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index 1aa721d7c10..74a5af4d33e 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -377,6 +377,7 @@ static int add_tracepoint(struct list_head **list, int *idx, attr.sample_type |= PERF_SAMPLE_RAW; attr.sample_type |= PERF_SAMPLE_TIME; attr.sample_type |= PERF_SAMPLE_CPU; + attr.sample_type |= PERF_SAMPLE_PERIOD; attr.sample_period = 1; snprintf(name, MAX_NAME_LEN, "%s:%s", sys_name, evt_name); @@ -489,6 +490,7 @@ int parse_events_add_breakpoint(struct list_head **list, int *idx, attr.bp_len = HW_BREAKPOINT_LEN_4; attr.type = PERF_TYPE_BREAKPOINT; + attr.sample_period = 1; return add_event(list, idx, &attr, NULL); } diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 8e485592ca2..8e4f0755d2a 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -87,6 +87,7 @@ void perf_session__update_sample_type(struct perf_session *self) self->sample_id_all = perf_evlist__sample_id_all(self->evlist); self->id_hdr_size = perf_evlist__id_hdr_size(self->evlist); self->host_machine.id_hdr_size = self->id_hdr_size; + machines__set_id_hdr_size(&self->machines, self->id_hdr_size); } int perf_session__create_kernel_maps(struct perf_session *self) @@ -918,7 +919,9 @@ static struct machine * { const u8 cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK; - if (cpumode == PERF_RECORD_MISC_GUEST_KERNEL && perf_guest) { + if (perf_guest && + ((cpumode == PERF_RECORD_MISC_GUEST_KERNEL) || + (cpumode == PERF_RECORD_MISC_GUEST_USER))) { u32 pid; if (event->header.type == PERF_RECORD_MMAP) diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 50958bbeb26..fdad4eeeb42 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -29,6 +29,7 @@ #define NT_GNU_BUILD_ID 3 #endif +static void dso_cache__free(struct rb_root *root); static bool dso__build_id_equal(const struct dso *dso, u8 *build_id); static int elf_read_build_id(Elf *elf, void *bf, size_t size); static void dsos__add(struct list_head *head, struct dso *dso); @@ -48,6 +49,31 @@ struct symbol_conf symbol_conf = { .symfs = "", }; +static enum dso_binary_type binary_type_symtab[] = { + DSO_BINARY_TYPE__KALLSYMS, + DSO_BINARY_TYPE__GUEST_KALLSYMS, + DSO_BINARY_TYPE__JAVA_JIT, + DSO_BINARY_TYPE__DEBUGLINK, + DSO_BINARY_TYPE__BUILD_ID_CACHE, + DSO_BINARY_TYPE__FEDORA_DEBUGINFO, + DSO_BINARY_TYPE__UBUNTU_DEBUGINFO, + DSO_BINARY_TYPE__BUILDID_DEBUGINFO, + DSO_BINARY_TYPE__SYSTEM_PATH_DSO, + DSO_BINARY_TYPE__GUEST_KMODULE, + DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE, + DSO_BINARY_TYPE__NOT_FOUND, +}; + +#define DSO_BINARY_TYPE__SYMTAB_CNT sizeof(binary_type_symtab) + +static enum dso_binary_type binary_type_data[] = { + DSO_BINARY_TYPE__BUILD_ID_CACHE, + DSO_BINARY_TYPE__SYSTEM_PATH_DSO, + DSO_BINARY_TYPE__NOT_FOUND, +}; + +#define DSO_BINARY_TYPE__DATA_CNT sizeof(binary_type_data) + int dso__name_len(const struct dso *dso) { if (!dso) @@ -318,7 +344,9 @@ struct dso *dso__new(const char *name) dso__set_short_name(dso, dso->name); for (i = 0; i < MAP__NR_TYPES; ++i) dso->symbols[i] = dso->symbol_names[i] = RB_ROOT; - dso->symtab_type = SYMTAB__NOT_FOUND; + dso->cache = RB_ROOT; + dso->symtab_type = DSO_BINARY_TYPE__NOT_FOUND; + dso->data_type = DSO_BINARY_TYPE__NOT_FOUND; dso->loaded = 0; dso->sorted_by_name = 0; dso->has_build_id = 0; @@ -352,6 +380,7 @@ void dso__delete(struct dso *dso) free((char *)dso->short_name); if (dso->lname_alloc) free(dso->long_name); + dso_cache__free(&dso->cache); free(dso); } @@ -806,9 +835,9 @@ int dso__load_kallsyms(struct dso *dso, const char *filename, symbols__fixup_end(&dso->symbols[map->type]); if (dso->kernel == DSO_TYPE_GUEST_KERNEL) - dso->symtab_type = SYMTAB__GUEST_KALLSYMS; + dso->symtab_type = DSO_BINARY_TYPE__GUEST_KALLSYMS; else - dso->symtab_type = SYMTAB__KALLSYMS; + dso->symtab_type = DSO_BINARY_TYPE__KALLSYMS; return dso__split_kallsyms(dso, map, filter); } @@ -1660,32 +1689,110 @@ out: char dso__symtab_origin(const struct dso *dso) { static const char origin[] = { - [SYMTAB__KALLSYMS] = 'k', - [SYMTAB__JAVA_JIT] = 'j', - [SYMTAB__DEBUGLINK] = 'l', - [SYMTAB__BUILD_ID_CACHE] = 'B', - [SYMTAB__FEDORA_DEBUGINFO] = 'f', - [SYMTAB__UBUNTU_DEBUGINFO] = 'u', - [SYMTAB__BUILDID_DEBUGINFO] = 'b', - [SYMTAB__SYSTEM_PATH_DSO] = 'd', - [SYMTAB__SYSTEM_PATH_KMODULE] = 'K', - [SYMTAB__GUEST_KALLSYMS] = 'g', - [SYMTAB__GUEST_KMODULE] = 'G', + [DSO_BINARY_TYPE__KALLSYMS] = 'k', + [DSO_BINARY_TYPE__JAVA_JIT] = 'j', + [DSO_BINARY_TYPE__DEBUGLINK] = 'l', + [DSO_BINARY_TYPE__BUILD_ID_CACHE] = 'B', + [DSO_BINARY_TYPE__FEDORA_DEBUGINFO] = 'f', + [DSO_BINARY_TYPE__UBUNTU_DEBUGINFO] = 'u', + [DSO_BINARY_TYPE__BUILDID_DEBUGINFO] = 'b', + [DSO_BINARY_TYPE__SYSTEM_PATH_DSO] = 'd', + [DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE] = 'K', + [DSO_BINARY_TYPE__GUEST_KALLSYMS] = 'g', + [DSO_BINARY_TYPE__GUEST_KMODULE] = 'G', }; - if (dso == NULL || dso->symtab_type == SYMTAB__NOT_FOUND) + if (dso == NULL || dso->symtab_type == DSO_BINARY_TYPE__NOT_FOUND) return '!'; return origin[dso->symtab_type]; } +int dso__binary_type_file(struct dso *dso, enum dso_binary_type type, + char *root_dir, char *file, size_t size) +{ + char build_id_hex[BUILD_ID_SIZE * 2 + 1]; + int ret = 0; + + switch (type) { + case DSO_BINARY_TYPE__DEBUGLINK: { + char *debuglink; + + strncpy(file, dso->long_name, size); + debuglink = file + dso->long_name_len; + while (debuglink != file && *debuglink != '/') + debuglink--; + if (*debuglink == '/') + debuglink++; + filename__read_debuglink(dso->long_name, debuglink, + size - (debuglink - file)); + } + break; + case DSO_BINARY_TYPE__BUILD_ID_CACHE: + /* skip the locally configured cache if a symfs is given */ + if (symbol_conf.symfs[0] || + (dso__build_id_filename(dso, file, size) == NULL)) + ret = -1; + break; + + case DSO_BINARY_TYPE__FEDORA_DEBUGINFO: + snprintf(file, size, "%s/usr/lib/debug%s.debug", + symbol_conf.symfs, dso->long_name); + break; + + case DSO_BINARY_TYPE__UBUNTU_DEBUGINFO: + snprintf(file, size, "%s/usr/lib/debug%s", + symbol_conf.symfs, dso->long_name); + break; + + case DSO_BINARY_TYPE__BUILDID_DEBUGINFO: + if (!dso->has_build_id) { + ret = -1; + break; + } + + build_id__sprintf(dso->build_id, + sizeof(dso->build_id), + build_id_hex); + snprintf(file, size, + "%s/usr/lib/debug/.build-id/%.2s/%s.debug", + symbol_conf.symfs, build_id_hex, build_id_hex + 2); + break; + + case DSO_BINARY_TYPE__SYSTEM_PATH_DSO: + snprintf(file, size, "%s%s", + symbol_conf.symfs, dso->long_name); + break; + + case DSO_BINARY_TYPE__GUEST_KMODULE: + snprintf(file, size, "%s%s%s", symbol_conf.symfs, + root_dir, dso->long_name); + break; + + case DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE: + snprintf(file, size, "%s%s", symbol_conf.symfs, + dso->long_name); + break; + + default: + case DSO_BINARY_TYPE__KALLSYMS: + case DSO_BINARY_TYPE__GUEST_KALLSYMS: + case DSO_BINARY_TYPE__JAVA_JIT: + case DSO_BINARY_TYPE__NOT_FOUND: + ret = -1; + break; + } + + return ret; +} + int dso__load(struct dso *dso, struct map *map, symbol_filter_t filter) { - int size = PATH_MAX; char *name; int ret = -1; int fd; + u_int i; struct machine *machine; - const char *root_dir; + char *root_dir = (char *) ""; int want_symtab; dso__set_loaded(dso, map->type); @@ -1700,7 +1807,7 @@ int dso__load(struct dso *dso, struct map *map, symbol_filter_t filter) else machine = NULL; - name = malloc(size); + name = malloc(PATH_MAX); if (!name) return -1; @@ -1719,81 +1826,27 @@ int dso__load(struct dso *dso, struct map *map, symbol_filter_t filter) } ret = dso__load_perf_map(dso, map, filter); - dso->symtab_type = ret > 0 ? SYMTAB__JAVA_JIT : - SYMTAB__NOT_FOUND; + dso->symtab_type = ret > 0 ? DSO_BINARY_TYPE__JAVA_JIT : + DSO_BINARY_TYPE__NOT_FOUND; return ret; } + if (machine) + root_dir = machine->root_dir; + /* Iterate over candidate debug images. * On the first pass, only load images if they have a full symtab. * Failing that, do a second pass where we accept .dynsym also */ want_symtab = 1; restart: - for (dso->symtab_type = SYMTAB__DEBUGLINK; - dso->symtab_type != SYMTAB__NOT_FOUND; - dso->symtab_type++) { - switch (dso->symtab_type) { - case SYMTAB__DEBUGLINK: { - char *debuglink; - strncpy(name, dso->long_name, size); - debuglink = name + dso->long_name_len; - while (debuglink != name && *debuglink != '/') - debuglink--; - if (*debuglink == '/') - debuglink++; - filename__read_debuglink(dso->long_name, debuglink, - size - (debuglink - name)); - } - break; - case SYMTAB__BUILD_ID_CACHE: - /* skip the locally configured cache if a symfs is given */ - if (symbol_conf.symfs[0] || - (dso__build_id_filename(dso, name, size) == NULL)) { - continue; - } - break; - case SYMTAB__FEDORA_DEBUGINFO: - snprintf(name, size, "%s/usr/lib/debug%s.debug", - symbol_conf.symfs, dso->long_name); - break; - case SYMTAB__UBUNTU_DEBUGINFO: - snprintf(name, size, "%s/usr/lib/debug%s", - symbol_conf.symfs, dso->long_name); - break; - case SYMTAB__BUILDID_DEBUGINFO: { - char build_id_hex[BUILD_ID_SIZE * 2 + 1]; + for (i = 0; i < DSO_BINARY_TYPE__SYMTAB_CNT; i++) { - if (!dso->has_build_id) - continue; + dso->symtab_type = binary_type_symtab[i]; - build_id__sprintf(dso->build_id, - sizeof(dso->build_id), - build_id_hex); - snprintf(name, size, - "%s/usr/lib/debug/.build-id/%.2s/%s.debug", - symbol_conf.symfs, build_id_hex, build_id_hex + 2); - } - break; - case SYMTAB__SYSTEM_PATH_DSO: - snprintf(name, size, "%s%s", - symbol_conf.symfs, dso->long_name); - break; - case SYMTAB__GUEST_KMODULE: - if (map->groups && machine) - root_dir = machine->root_dir; - else - root_dir = ""; - snprintf(name, size, "%s%s%s", symbol_conf.symfs, - root_dir, dso->long_name); - break; - - case SYMTAB__SYSTEM_PATH_KMODULE: - snprintf(name, size, "%s%s", symbol_conf.symfs, - dso->long_name); - break; - default:; - } + if (dso__binary_type_file(dso, dso->symtab_type, + root_dir, name, PATH_MAX)) + continue; /* Name is now the name of the next image to try */ fd = open(name, O_RDONLY); @@ -2010,9 +2063,9 @@ struct map *machine__new_module(struct machine *machine, u64 start, return NULL; if (machine__is_host(machine)) - dso->symtab_type = SYMTAB__SYSTEM_PATH_KMODULE; + dso->symtab_type = DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE; else - dso->symtab_type = SYMTAB__GUEST_KMODULE; + dso->symtab_type = DSO_BINARY_TYPE__GUEST_KMODULE; map_groups__insert(&machine->kmaps, map); return map; } @@ -2564,8 +2617,15 @@ int machine__create_kernel_maps(struct machine *machine) __machine__create_kernel_maps(machine, kernel) < 0) return -1; - if (symbol_conf.use_modules && machine__create_modules(machine) < 0) - pr_debug("Problems creating module maps, continuing anyway...\n"); + if (symbol_conf.use_modules && machine__create_modules(machine) < 0) { + if (machine__is_host(machine)) + pr_debug("Problems creating module maps, " + "continuing anyway...\n"); + else + pr_debug("Problems creating module maps for guest %d, " + "continuing anyway...\n", machine->pid); + } + /* * Now that we have all the maps created, just set the ->end of them: */ @@ -2905,3 +2965,218 @@ struct map *dso__new_map(const char *name) return map; } + +static int open_dso(struct dso *dso, struct machine *machine) +{ + char *root_dir = (char *) ""; + char *name; + int fd; + + name = malloc(PATH_MAX); + if (!name) + return -ENOMEM; + + if (machine) + root_dir = machine->root_dir; + + if (dso__binary_type_file(dso, dso->data_type, + root_dir, name, PATH_MAX)) { + free(name); + return -EINVAL; + } + + fd = open(name, O_RDONLY); + free(name); + return fd; +} + +int dso__data_fd(struct dso *dso, struct machine *machine) +{ + int i = 0; + + if (dso->data_type != DSO_BINARY_TYPE__NOT_FOUND) + return open_dso(dso, machine); + + do { + int fd; + + dso->data_type = binary_type_data[i++]; + + fd = open_dso(dso, machine); + if (fd >= 0) + return fd; + + } while (dso->data_type != DSO_BINARY_TYPE__NOT_FOUND); + + return -EINVAL; +} + +static void +dso_cache__free(struct rb_root *root) +{ + struct rb_node *next = rb_first(root); + + while (next) { + struct dso_cache *cache; + + cache = rb_entry(next, struct dso_cache, rb_node); + next = rb_next(&cache->rb_node); + rb_erase(&cache->rb_node, root); + free(cache); + } +} + +static struct dso_cache* +dso_cache__find(struct rb_root *root, u64 offset) +{ + struct rb_node **p = &root->rb_node; + struct rb_node *parent = NULL; + struct dso_cache *cache; + + while (*p != NULL) { + u64 end; + + parent = *p; + cache = rb_entry(parent, struct dso_cache, rb_node); + end = cache->offset + DSO__DATA_CACHE_SIZE; + + if (offset < cache->offset) + p = &(*p)->rb_left; + else if (offset >= end) + p = &(*p)->rb_right; + else + return cache; + } + return NULL; +} + +static void +dso_cache__insert(struct rb_root *root, struct dso_cache *new) +{ + struct rb_node **p = &root->rb_node; + struct rb_node *parent = NULL; + struct dso_cache *cache; + u64 offset = new->offset; + + while (*p != NULL) { + u64 end; + + parent = *p; + cache = rb_entry(parent, struct dso_cache, rb_node); + end = cache->offset + DSO__DATA_CACHE_SIZE; + + if (offset < cache->offset) + p = &(*p)->rb_left; + else if (offset >= end) + p = &(*p)->rb_right; + } + + rb_link_node(&new->rb_node, parent, p); + rb_insert_color(&new->rb_node, root); +} + +static ssize_t +dso_cache__memcpy(struct dso_cache *cache, u64 offset, + u8 *data, u64 size) +{ + u64 cache_offset = offset - cache->offset; + u64 cache_size = min(cache->size - cache_offset, size); + + memcpy(data, cache->data + cache_offset, cache_size); + return cache_size; +} + +static ssize_t +dso_cache__read(struct dso *dso, struct machine *machine, + u64 offset, u8 *data, ssize_t size) +{ + struct dso_cache *cache; + ssize_t ret; + int fd; + + fd = dso__data_fd(dso, machine); + if (fd < 0) + return -1; + + do { + u64 cache_offset; + + ret = -ENOMEM; + + cache = zalloc(sizeof(*cache) + DSO__DATA_CACHE_SIZE); + if (!cache) + break; + + cache_offset = offset & DSO__DATA_CACHE_MASK; + ret = -EINVAL; + + if (-1 == lseek(fd, cache_offset, SEEK_SET)) + break; + + ret = read(fd, cache->data, DSO__DATA_CACHE_SIZE); + if (ret <= 0) + break; + + cache->offset = cache_offset; + cache->size = ret; + dso_cache__insert(&dso->cache, cache); + + ret = dso_cache__memcpy(cache, offset, data, size); + + } while (0); + + if (ret <= 0) + free(cache); + + close(fd); + return ret; +} + +static ssize_t dso_cache_read(struct dso *dso, struct machine *machine, + u64 offset, u8 *data, ssize_t size) +{ + struct dso_cache *cache; + + cache = dso_cache__find(&dso->cache, offset); + if (cache) + return dso_cache__memcpy(cache, offset, data, size); + else + return dso_cache__read(dso, machine, offset, data, size); +} + +ssize_t dso__data_read_offset(struct dso *dso, struct machine *machine, + u64 offset, u8 *data, ssize_t size) +{ + ssize_t r = 0; + u8 *p = data; + + do { + ssize_t ret; + + ret = dso_cache_read(dso, machine, offset, p, size); + if (ret < 0) + return ret; + + /* Reached EOF, return what we have. */ + if (!ret) + break; + + BUG_ON(ret > size); + + r += ret; + p += ret; + offset += ret; + size -= ret; + + } while (size); + + return r; +} + +ssize_t dso__data_read_addr(struct dso *dso, struct map *map, + struct machine *machine, u64 addr, + u8 *data, ssize_t size) +{ + u64 offset = map->map_ip(map, addr); + return dso__data_read_offset(dso, machine, offset, data, size); +} diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index a884b99017f..1fe733a1e21 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -155,6 +155,21 @@ struct addr_location { s32 cpu; }; +enum dso_binary_type { + DSO_BINARY_TYPE__KALLSYMS = 0, + DSO_BINARY_TYPE__GUEST_KALLSYMS, + DSO_BINARY_TYPE__JAVA_JIT, + DSO_BINARY_TYPE__DEBUGLINK, + DSO_BINARY_TYPE__BUILD_ID_CACHE, + DSO_BINARY_TYPE__FEDORA_DEBUGINFO, + DSO_BINARY_TYPE__UBUNTU_DEBUGINFO, + DSO_BINARY_TYPE__BUILDID_DEBUGINFO, + DSO_BINARY_TYPE__SYSTEM_PATH_DSO, + DSO_BINARY_TYPE__GUEST_KMODULE, + DSO_BINARY_TYPE__SYSTEM_PATH_KMODULE, + DSO_BINARY_TYPE__NOT_FOUND, +}; + enum dso_kernel_type { DSO_TYPE_USER = 0, DSO_TYPE_KERNEL, @@ -167,19 +182,31 @@ enum dso_swap_type { DSO_SWAP__YES, }; +#define DSO__DATA_CACHE_SIZE 4096 +#define DSO__DATA_CACHE_MASK ~(DSO__DATA_CACHE_SIZE - 1) + +struct dso_cache { + struct rb_node rb_node; + u64 offset; + u64 size; + char data[0]; +}; + struct dso { struct list_head node; struct rb_root symbols[MAP__NR_TYPES]; struct rb_root symbol_names[MAP__NR_TYPES]; + struct rb_root cache; enum dso_kernel_type kernel; enum dso_swap_type needs_swap; + enum dso_binary_type symtab_type; + enum dso_binary_type data_type; u8 adjust_symbols:1; u8 has_build_id:1; u8 hit:1; u8 annotate_warned:1; u8 sname_alloc:1; u8 lname_alloc:1; - unsigned char symtab_type; u8 sorted_by_name; u8 loaded; u8 build_id[BUILD_ID_SIZE]; @@ -253,21 +280,6 @@ size_t dso__fprintf_symbols_by_name(struct dso *dso, enum map_type type, FILE *fp); size_t dso__fprintf(struct dso *dso, enum map_type type, FILE *fp); -enum symtab_type { - SYMTAB__KALLSYMS = 0, - SYMTAB__GUEST_KALLSYMS, - SYMTAB__JAVA_JIT, - SYMTAB__DEBUGLINK, - SYMTAB__BUILD_ID_CACHE, - SYMTAB__FEDORA_DEBUGINFO, - SYMTAB__UBUNTU_DEBUGINFO, - SYMTAB__BUILDID_DEBUGINFO, - SYMTAB__SYSTEM_PATH_DSO, - SYMTAB__GUEST_KMODULE, - SYMTAB__SYSTEM_PATH_KMODULE, - SYMTAB__NOT_FOUND, -}; - char dso__symtab_origin(const struct dso *dso); void dso__set_long_name(struct dso *dso, char *name); void dso__set_build_id(struct dso *dso, void *build_id); @@ -304,4 +316,14 @@ bool symbol_type__is_a(char symbol_type, enum map_type map_type); size_t machine__fprintf_vmlinux_path(struct machine *machine, FILE *fp); +int dso__binary_type_file(struct dso *dso, enum dso_binary_type type, + char *root_dir, char *file, size_t size); + +int dso__data_fd(struct dso *dso, struct machine *machine); +ssize_t dso__data_read_offset(struct dso *dso, struct machine *machine, + u64 offset, u8 *data, ssize_t size); +ssize_t dso__data_read_addr(struct dso *dso, struct map *map, + struct machine *machine, u64 addr, + u8 *data, ssize_t size); +int dso__test_data(void); #endif /* __PERF_SYMBOL */ diff --git a/tools/perf/util/target.c b/tools/perf/util/target.c index 1064d5b148a..3f59c496e64 100644 --- a/tools/perf/util/target.c +++ b/tools/perf/util/target.c @@ -110,8 +110,17 @@ int perf_target__strerror(struct perf_target *target, int errnum, int idx; const char *msg; + BUG_ON(buflen > 0); + if (errnum >= 0) { - strerror_r(errnum, buf, buflen); + const char *err = strerror_r(errnum, buf, buflen); + + if (err != buf) { + size_t len = strlen(err); + char *c = mempcpy(buf, err, min(buflen - 1, len)); + *c = '\0'; + } + return 0; } diff --git a/tools/power/x86/turbostat/Makefile b/tools/power/x86/turbostat/Makefile index fd8e1f1297a..f8564955419 100644 --- a/tools/power/x86/turbostat/Makefile +++ b/tools/power/x86/turbostat/Makefile @@ -1,4 +1,5 @@ turbostat : turbostat.c +CFLAGS += -Wall clean : rm -f turbostat diff --git a/tools/power/x86/turbostat/turbostat.8 b/tools/power/x86/turbostat/turbostat.8 index adf175f6149..74e44507dfe 100644 --- a/tools/power/x86/turbostat/turbostat.8 +++ b/tools/power/x86/turbostat/turbostat.8 @@ -27,7 +27,11 @@ supports an "invariant" TSC, plus the APERF and MPERF MSRs. on processors that additionally support C-state residency counters. .SS Options -The \fB-s\fP option prints only a 1-line summary for each sample interval. +The \fB-s\fP option limits output to a 1-line system summary for each interval. +.PP +The \fB-c\fP option limits output to the 1st thread in each core. +.PP +The \fB-p\fP option limits output to the 1st thread in each package. .PP The \fB-v\fP option increases verbosity. .PP @@ -65,19 +69,19 @@ Subsequent rows show per-CPU statistics. .nf [root@x980]# ./turbostat cor CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6 - 0.60 1.63 3.38 2.91 0.00 96.49 0.00 76.64 - 0 0 0.59 1.62 3.38 4.51 0.00 94.90 0.00 76.64 - 0 6 1.13 1.64 3.38 3.97 0.00 94.90 0.00 76.64 - 1 2 0.08 1.62 3.38 0.07 0.00 99.85 0.00 76.64 - 1 8 0.03 1.62 3.38 0.12 0.00 99.85 0.00 76.64 - 2 4 0.01 1.62 3.38 0.06 0.00 99.93 0.00 76.64 - 2 10 0.04 1.62 3.38 0.02 0.00 99.93 0.00 76.64 - 8 1 2.85 1.62 3.38 11.71 0.00 85.44 0.00 76.64 - 8 7 1.98 1.62 3.38 12.58 0.00 85.44 0.00 76.64 - 9 3 0.36 1.62 3.38 0.71 0.00 98.93 0.00 76.64 - 9 9 0.09 1.62 3.38 0.98 0.00 98.93 0.00 76.64 - 10 5 0.03 1.62 3.38 0.09 0.00 99.87 0.00 76.64 - 10 11 0.07 1.62 3.38 0.06 0.00 99.87 0.00 76.64 + 0.09 1.62 3.38 1.83 0.32 97.76 1.26 83.61 + 0 0 0.15 1.62 3.38 10.23 0.05 89.56 1.26 83.61 + 0 6 0.05 1.62 3.38 10.34 + 1 2 0.03 1.62 3.38 0.07 0.05 99.86 + 1 8 0.03 1.62 3.38 0.06 + 2 4 0.21 1.62 3.38 0.10 1.49 98.21 + 2 10 0.02 1.62 3.38 0.29 + 8 1 0.04 1.62 3.38 0.04 0.08 99.84 + 8 7 0.01 1.62 3.38 0.06 + 9 3 0.53 1.62 3.38 0.10 0.20 99.17 + 9 9 0.02 1.62 3.38 0.60 + 10 5 0.01 1.62 3.38 0.02 0.04 99.92 + 10 11 0.02 1.62 3.38 0.02 .fi .SH SUMMARY EXAMPLE The "-s" option prints the column headers just once, @@ -86,9 +90,10 @@ and then the one line system summary for each sample interval. .nf [root@x980]# ./turbostat -s %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6 - 0.61 1.89 3.38 5.95 0.00 93.44 0.00 66.33 - 0.52 1.62 3.38 6.83 0.00 92.65 0.00 61.11 - 0.62 1.92 3.38 5.47 0.00 93.91 0.00 67.31 + 0.23 1.67 3.38 2.00 0.30 97.47 1.07 82.12 + 0.10 1.62 3.38 1.87 2.25 95.77 12.02 72.60 + 0.20 1.64 3.38 1.98 0.11 97.72 0.30 83.36 + 0.11 1.70 3.38 1.86 1.81 96.22 9.71 74.90 .fi .SH VERBOSE EXAMPLE The "-v" option adds verbosity to the output: @@ -120,30 +125,28 @@ until ^C while the other CPUs are mostly idle: [root@x980 lenb]# ./turbostat cat /dev/zero > /dev/null ^C cor CPU %c0 GHz TSC %c1 %c3 %c6 %pc3 %pc6 - 8.63 3.64 3.38 14.46 0.49 76.42 0.00 0.00 - 0 0 0.34 3.36 3.38 99.66 0.00 0.00 0.00 0.00 - 0 6 99.96 3.64 3.38 0.04 0.00 0.00 0.00 0.00 - 1 2 0.14 3.50 3.38 1.75 2.04 96.07 0.00 0.00 - 1 8 0.38 3.57 3.38 1.51 2.04 96.07 0.00 0.00 - 2 4 0.01 2.65 3.38 0.06 0.00 99.93 0.00 0.00 - 2 10 0.03 2.12 3.38 0.04 0.00 99.93 0.00 0.00 - 8 1 0.91 3.59 3.38 35.27 0.92 62.90 0.00 0.00 - 8 7 1.61 3.63 3.38 34.57 0.92 62.90 0.00 0.00 - 9 3 0.04 3.38 3.38 0.20 0.00 99.76 0.00 0.00 - 9 9 0.04 3.29 3.38 0.20 0.00 99.76 0.00 0.00 - 10 5 0.03 3.08 3.38 0.12 0.00 99.85 0.00 0.00 - 10 11 0.05 3.07 3.38 0.10 0.00 99.85 0.00 0.00 -4.907015 sec - + 8.86 3.61 3.38 15.06 31.19 44.89 0.00 0.00 + 0 0 1.46 3.22 3.38 16.84 29.48 52.22 0.00 0.00 + 0 6 0.21 3.06 3.38 18.09 + 1 2 0.53 3.33 3.38 2.80 46.40 50.27 + 1 8 0.89 3.47 3.38 2.44 + 2 4 1.36 3.43 3.38 9.04 23.71 65.89 + 2 10 0.18 2.86 3.38 10.22 + 8 1 0.04 2.87 3.38 99.96 0.01 0.00 + 8 7 99.72 3.63 3.38 0.27 + 9 3 0.31 3.21 3.38 7.64 56.55 35.50 + 9 9 0.08 2.95 3.38 7.88 + 10 5 1.42 3.43 3.38 2.14 30.99 65.44 + 10 11 0.16 2.88 3.38 3.40 .fi -Above the cycle soaker drives cpu6 up 3.6 Ghz turbo limit +Above the cycle soaker drives cpu7 up its 3.6 Ghz turbo limit while the other processors are generally in various states of idle. -Note that cpu0 is an HT sibling sharing core0 -with cpu6, and thus it is unable to get to an idle state -deeper than c1 while cpu6 is busy. +Note that cpu1 and cpu7 are HT siblings within core8. +As cpu7 is very busy, it prevents its sibling, cpu1, +from entering a c-state deeper than c1. -Note that turbostat reports average GHz of 3.64, while +Note that turbostat reports average GHz of 3.63, while the arithmetic average of the GHz column above is lower. This is a weighted average, where the weight is %c0. ie. it is the total number of un-halted cycles elapsed per time divided by the number of CPUs. diff --git a/tools/power/x86/turbostat/turbostat.c b/tools/power/x86/turbostat/turbostat.c index 16de7ad4850..861d7719020 100644 --- a/tools/power/x86/turbostat/turbostat.c +++ b/tools/power/x86/turbostat/turbostat.c @@ -67,92 +67,119 @@ double bclk; unsigned int show_pkg; unsigned int show_core; unsigned int show_cpu; +unsigned int show_pkg_only; +unsigned int show_core_only; +char *output_buffer, *outp; int aperf_mperf_unstable; int backwards_count; char *progname; -int num_cpus; -cpu_set_t *cpu_present_set, *cpu_mask; -size_t cpu_present_setsize, cpu_mask_size; - -struct counters { - unsigned long long tsc; /* per thread */ - unsigned long long aperf; /* per thread */ - unsigned long long mperf; /* per thread */ - unsigned long long c1; /* per thread (calculated) */ - unsigned long long c3; /* per core */ - unsigned long long c6; /* per core */ - unsigned long long c7; /* per core */ - unsigned long long pc2; /* per package */ - unsigned long long pc3; /* per package */ - unsigned long long pc6; /* per package */ - unsigned long long pc7; /* per package */ - unsigned long long extra_msr; /* per thread */ - int pkg; - int core; - int cpu; - struct counters *next; -}; - -struct counters *cnt_even; -struct counters *cnt_odd; -struct counters *cnt_delta; -struct counters *cnt_average; -struct timeval tv_even; -struct timeval tv_odd; -struct timeval tv_delta; - -int mark_cpu_present(int pkg, int core, int cpu) +cpu_set_t *cpu_present_set, *cpu_affinity_set; +size_t cpu_present_setsize, cpu_affinity_setsize; + +struct thread_data { + unsigned long long tsc; + unsigned long long aperf; + unsigned long long mperf; + unsigned long long c1; /* derived */ + unsigned long long extra_msr; + unsigned int cpu_id; + unsigned int flags; +#define CPU_IS_FIRST_THREAD_IN_CORE 0x2 +#define CPU_IS_FIRST_CORE_IN_PACKAGE 0x4 +} *thread_even, *thread_odd; + +struct core_data { + unsigned long long c3; + unsigned long long c6; + unsigned long long c7; + unsigned int core_id; +} *core_even, *core_odd; + +struct pkg_data { + unsigned long long pc2; + unsigned long long pc3; + unsigned long long pc6; + unsigned long long pc7; + unsigned int package_id; +} *package_even, *package_odd; + +#define ODD_COUNTERS thread_odd, core_odd, package_odd +#define EVEN_COUNTERS thread_even, core_even, package_even + +#define GET_THREAD(thread_base, thread_no, core_no, pkg_no) \ + (thread_base + (pkg_no) * topo.num_cores_per_pkg * \ + topo.num_threads_per_core + \ + (core_no) * topo.num_threads_per_core + (thread_no)) +#define GET_CORE(core_base, core_no, pkg_no) \ + (core_base + (pkg_no) * topo.num_cores_per_pkg + (core_no)) +#define GET_PKG(pkg_base, pkg_no) (pkg_base + pkg_no) + +struct system_summary { + struct thread_data threads; + struct core_data cores; + struct pkg_data packages; +} sum, average; + + +struct topo_params { + int num_packages; + int num_cpus; + int num_cores; + int max_cpu_num; + int num_cores_per_pkg; + int num_threads_per_core; +} topo; + +struct timeval tv_even, tv_odd, tv_delta; + +void setup_all_buffers(void); + +int cpu_is_not_present(int cpu) { - CPU_SET_S(cpu, cpu_present_setsize, cpu_present_set); - return 0; + return !CPU_ISSET_S(cpu, cpu_present_setsize, cpu_present_set); } - /* - * cpu_mask_init(ncpus) - * - * allocate and clear cpu_mask - * set cpu_mask_size + * run func(thread, core, package) in topology order + * skip non-present cpus */ -void cpu_mask_init(int ncpus) + +int for_all_cpus(int (func)(struct thread_data *, struct core_data *, struct pkg_data *), + struct thread_data *thread_base, struct core_data *core_base, struct pkg_data *pkg_base) { - cpu_mask = CPU_ALLOC(ncpus); - if (cpu_mask == NULL) { - perror("CPU_ALLOC"); - exit(3); - } - cpu_mask_size = CPU_ALLOC_SIZE(ncpus); - CPU_ZERO_S(cpu_mask_size, cpu_mask); + int retval, pkg_no, core_no, thread_no; - /* - * Allocate and initialize cpu_present_set - */ - cpu_present_set = CPU_ALLOC(ncpus); - if (cpu_present_set == NULL) { - perror("CPU_ALLOC"); - exit(3); - } - cpu_present_setsize = CPU_ALLOC_SIZE(ncpus); - CPU_ZERO_S(cpu_present_setsize, cpu_present_set); - for_all_cpus(mark_cpu_present); -} + for (pkg_no = 0; pkg_no < topo.num_packages; ++pkg_no) { + for (core_no = 0; core_no < topo.num_cores_per_pkg; ++core_no) { + for (thread_no = 0; thread_no < + topo.num_threads_per_core; ++thread_no) { + struct thread_data *t; + struct core_data *c; + struct pkg_data *p; -void cpu_mask_uninit() -{ - CPU_FREE(cpu_mask); - cpu_mask = NULL; - cpu_mask_size = 0; - CPU_FREE(cpu_present_set); - cpu_present_set = NULL; - cpu_present_setsize = 0; + t = GET_THREAD(thread_base, thread_no, core_no, pkg_no); + + if (cpu_is_not_present(t->cpu_id)) + continue; + + c = GET_CORE(core_base, core_no, pkg_no); + p = GET_PKG(pkg_base, pkg_no); + + retval = func(t, c, p); + if (retval) + return retval; + } + } + } + return 0; } int cpu_migrate(int cpu) { - CPU_ZERO_S(cpu_mask_size, cpu_mask); - CPU_SET_S(cpu, cpu_mask_size, cpu_mask); - if (sched_setaffinity(0, cpu_mask_size, cpu_mask) == -1) + CPU_ZERO_S(cpu_affinity_setsize, cpu_affinity_set); + CPU_SET_S(cpu, cpu_affinity_setsize, cpu_affinity_set); + if (sched_setaffinity(0, cpu_affinity_setsize, cpu_affinity_set) == -1) return -1; else return 0; @@ -181,67 +208,72 @@ int get_msr(int cpu, off_t offset, unsigned long long *msr) void print_header(void) { if (show_pkg) - fprintf(stderr, "pk"); + outp += sprintf(outp, "pk"); if (show_pkg) - fprintf(stderr, " "); + outp += sprintf(outp, " "); if (show_core) - fprintf(stderr, "cor"); + outp += sprintf(outp, "cor"); if (show_cpu) - fprintf(stderr, " CPU"); + outp += sprintf(outp, " CPU"); if (show_pkg || show_core || show_cpu) - fprintf(stderr, " "); + outp += sprintf(outp, " "); if (do_nhm_cstates) - fprintf(stderr, " %%c0"); + outp += sprintf(outp, " %%c0"); if (has_aperf) - fprintf(stderr, " GHz"); - fprintf(stderr, " TSC"); + outp += sprintf(outp, " GHz"); + outp += sprintf(outp, " TSC"); if (do_nhm_cstates) - fprintf(stderr, " %%c1"); + outp += sprintf(outp, " %%c1"); if (do_nhm_cstates) - fprintf(stderr, " %%c3"); + outp += sprintf(outp, " %%c3"); if (do_nhm_cstates) - fprintf(stderr, " %%c6"); + outp += sprintf(outp, " %%c6"); if (do_snb_cstates) - fprintf(stderr, " %%c7"); + outp += sprintf(outp, " %%c7"); if (do_snb_cstates) - fprintf(stderr, " %%pc2"); + outp += sprintf(outp, " %%pc2"); if (do_nhm_cstates) - fprintf(stderr, " %%pc3"); + outp += sprintf(outp, " %%pc3"); if (do_nhm_cstates) - fprintf(stderr, " %%pc6"); + outp += sprintf(outp, " %%pc6"); if (do_snb_cstates) - fprintf(stderr, " %%pc7"); + outp += sprintf(outp, " %%pc7"); if (extra_msr_offset) - fprintf(stderr, " MSR 0x%x ", extra_msr_offset); + outp += sprintf(outp, " MSR 0x%x ", extra_msr_offset); - putc('\n', stderr); + outp += sprintf(outp, "\n"); } -void dump_cnt(struct counters *cnt) +int dump_counters(struct thread_data *t, struct core_data *c, + struct pkg_data *p) { - if (!cnt) - return; - if (cnt->pkg) fprintf(stderr, "package: %d ", cnt->pkg); - if (cnt->core) fprintf(stderr, "core:: %d ", cnt->core); - if (cnt->cpu) fprintf(stderr, "CPU: %d ", cnt->cpu); - if (cnt->tsc) fprintf(stderr, "TSC: %016llX\n", cnt->tsc); - if (cnt->c3) fprintf(stderr, "c3: %016llX\n", cnt->c3); - if (cnt->c6) fprintf(stderr, "c6: %016llX\n", cnt->c6); - if (cnt->c7) fprintf(stderr, "c7: %016llX\n", cnt->c7); - if (cnt->aperf) fprintf(stderr, "aperf: %016llX\n", cnt->aperf); - if (cnt->pc2) fprintf(stderr, "pc2: %016llX\n", cnt->pc2); - if (cnt->pc3) fprintf(stderr, "pc3: %016llX\n", cnt->pc3); - if (cnt->pc6) fprintf(stderr, "pc6: %016llX\n", cnt->pc6); - if (cnt->pc7) fprintf(stderr, "pc7: %016llX\n", cnt->pc7); - if (cnt->extra_msr) fprintf(stderr, "msr0x%x: %016llX\n", extra_msr_offset, cnt->extra_msr); -} + fprintf(stderr, "t %p, c %p, p %p\n", t, c, p); + + if (t) { + fprintf(stderr, "CPU: %d flags 0x%x\n", t->cpu_id, t->flags); + fprintf(stderr, "TSC: %016llX\n", t->tsc); + fprintf(stderr, "aperf: %016llX\n", t->aperf); + fprintf(stderr, "mperf: %016llX\n", t->mperf); + fprintf(stderr, "c1: %016llX\n", t->c1); + fprintf(stderr, "msr0x%x: %016llX\n", + extra_msr_offset, t->extra_msr); + } -void dump_list(struct counters *cnt) -{ - printf("dump_list 0x%p\n", cnt); + if (c) { + fprintf(stderr, "core: %d\n", c->core_id); + fprintf(stderr, "c3: %016llX\n", c->c3); + fprintf(stderr, "c6: %016llX\n", c->c6); + fprintf(stderr, "c7: %016llX\n", c->c7); + } - for (; cnt; cnt = cnt->next) - dump_cnt(cnt); + if (p) { + fprintf(stderr, "package: %d\n", p->package_id); + fprintf(stderr, "pc2: %016llX\n", p->pc2); + fprintf(stderr, "pc3: %016llX\n", p->pc3); + fprintf(stderr, "pc6: %016llX\n", p->pc6); + fprintf(stderr, "pc7: %016llX\n", p->pc7); + } + return 0; } /* @@ -253,321 +285,389 @@ void dump_list(struct counters *cnt) * TSC: "TSC" 3 columns %3.2 * percentage " %pc3" %6.2 */ -void print_cnt(struct counters *p) +int format_counters(struct thread_data *t, struct core_data *c, + struct pkg_data *p) { double interval_float; + /* if showing only 1st thread in core and this isn't one, bail out */ + if (show_core_only && !(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) + return 0; + + /* if showing only 1st thread in pkg and this isn't one, bail out */ + if (show_pkg_only && !(t->flags & CPU_IS_FIRST_CORE_IN_PACKAGE)) + return 0; + interval_float = tv_delta.tv_sec + tv_delta.tv_usec/1000000.0; - /* topology columns, print blanks on 1st (average) line */ - if (p == cnt_average) { + /* topo columns, print blanks on 1st (average) line */ + if (t == &average.threads) { if (show_pkg) - fprintf(stderr, " "); + outp += sprintf(outp, " "); if (show_pkg && show_core) - fprintf(stderr, " "); + outp += sprintf(outp, " "); if (show_core) - fprintf(stderr, " "); + outp += sprintf(outp, " "); if (show_cpu) - fprintf(stderr, " " " "); + outp += sprintf(outp, " " " "); } else { - if (show_pkg) - fprintf(stderr, "%2d", p->pkg); + if (show_pkg) { + if (p) + outp += sprintf(outp, "%2d", p->package_id); + else + outp += sprintf(outp, " "); + } if (show_pkg && show_core) - fprintf(stderr, " "); - if (show_core) - fprintf(stderr, "%3d", p->core); + outp += sprintf(outp, " "); + if (show_core) { + if (c) + outp += sprintf(outp, "%3d", c->core_id); + else + outp += sprintf(outp, " "); + } if (show_cpu) - fprintf(stderr, " %3d", p->cpu); + outp += sprintf(outp, " %3d", t->cpu_id); } /* %c0 */ if (do_nhm_cstates) { if (show_pkg || show_core || show_cpu) - fprintf(stderr, " "); + outp += sprintf(outp, " "); if (!skip_c0) - fprintf(stderr, "%6.2f", 100.0 * p->mperf/p->tsc); + outp += sprintf(outp, "%6.2f", 100.0 * t->mperf/t->tsc); else - fprintf(stderr, " ****"); + outp += sprintf(outp, " ****"); } /* GHz */ if (has_aperf) { if (!aperf_mperf_unstable) { - fprintf(stderr, " %3.2f", - 1.0 * p->tsc / units * p->aperf / - p->mperf / interval_float); + outp += sprintf(outp, " %3.2f", + 1.0 * t->tsc / units * t->aperf / + t->mperf / interval_float); } else { - if (p->aperf > p->tsc || p->mperf > p->tsc) { - fprintf(stderr, " ***"); + if (t->aperf > t->tsc || t->mperf > t->tsc) { + outp += sprintf(outp, " ***"); } else { - fprintf(stderr, "%3.1f*", - 1.0 * p->tsc / - units * p->aperf / - p->mperf / interval_float); + outp += sprintf(outp, "%3.1f*", + 1.0 * t->tsc / + units * t->aperf / + t->mperf / interval_float); } } } /* TSC */ - fprintf(stderr, "%5.2f", 1.0 * p->tsc/units/interval_float); + outp += sprintf(outp, "%5.2f", 1.0 * t->tsc/units/interval_float); if (do_nhm_cstates) { if (!skip_c1) - fprintf(stderr, " %6.2f", 100.0 * p->c1/p->tsc); + outp += sprintf(outp, " %6.2f", 100.0 * t->c1/t->tsc); else - fprintf(stderr, " ****"); + outp += sprintf(outp, " ****"); } + + /* print per-core data only for 1st thread in core */ + if (!(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) + goto done; + if (do_nhm_cstates) - fprintf(stderr, " %6.2f", 100.0 * p->c3/p->tsc); + outp += sprintf(outp, " %6.2f", 100.0 * c->c3/t->tsc); if (do_nhm_cstates) - fprintf(stderr, " %6.2f", 100.0 * p->c6/p->tsc); + outp += sprintf(outp, " %6.2f", 100.0 * c->c6/t->tsc); if (do_snb_cstates) - fprintf(stderr, " %6.2f", 100.0 * p->c7/p->tsc); + outp += sprintf(outp, " %6.2f", 100.0 * c->c7/t->tsc); + + /* print per-package data only for 1st core in package */ + if (!(t->flags & CPU_IS_FIRST_CORE_IN_PACKAGE)) + goto done; + if (do_snb_cstates) - fprintf(stderr, " %6.2f", 100.0 * p->pc2/p->tsc); + outp += sprintf(outp, " %6.2f", 100.0 * p->pc2/t->tsc); if (do_nhm_cstates) - fprintf(stderr, " %6.2f", 100.0 * p->pc3/p->tsc); + outp += sprintf(outp, " %6.2f", 100.0 * p->pc3/t->tsc); if (do_nhm_cstates) - fprintf(stderr, " %6.2f", 100.0 * p->pc6/p->tsc); + outp += sprintf(outp, " %6.2f", 100.0 * p->pc6/t->tsc); if (do_snb_cstates) - fprintf(stderr, " %6.2f", 100.0 * p->pc7/p->tsc); + outp += sprintf(outp, " %6.2f", 100.0 * p->pc7/t->tsc); +done: if (extra_msr_offset) - fprintf(stderr, " 0x%016llx", p->extra_msr); - putc('\n', stderr); + outp += sprintf(outp, " 0x%016llx", t->extra_msr); + outp += sprintf(outp, "\n"); + + return 0; } -void print_counters(struct counters *counters) +void flush_stdout() +{ + fputs(output_buffer, stdout); + outp = output_buffer; +} +void flush_stderr() +{ + fputs(output_buffer, stderr); + outp = output_buffer; +} +void format_all_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) { - struct counters *cnt; static int printed; - if (!printed || !summary_only) print_header(); - if (num_cpus > 1) - print_cnt(cnt_average); + if (topo.num_cpus > 1) + format_counters(&average.threads, &average.cores, + &average.packages); printed = 1; if (summary_only) return; - for (cnt = counters; cnt != NULL; cnt = cnt->next) - print_cnt(cnt); - + for_all_cpus(format_counters, t, c, p); } -#define SUBTRACT_COUNTER(after, before, delta) (delta = (after - before), (before > after)) +void +delta_package(struct pkg_data *new, struct pkg_data *old) +{ + old->pc2 = new->pc2 - old->pc2; + old->pc3 = new->pc3 - old->pc3; + old->pc6 = new->pc6 - old->pc6; + old->pc7 = new->pc7 - old->pc7; +} -int compute_delta(struct counters *after, - struct counters *before, struct counters *delta) +void +delta_core(struct core_data *new, struct core_data *old) { - int errors = 0; - int perf_err = 0; + old->c3 = new->c3 - old->c3; + old->c6 = new->c6 - old->c6; + old->c7 = new->c7 - old->c7; +} - skip_c0 = skip_c1 = 0; +/* + * old = new - old + */ +void +delta_thread(struct thread_data *new, struct thread_data *old, + struct core_data *core_delta) +{ + old->tsc = new->tsc - old->tsc; + + /* check for TSC < 1 Mcycles over interval */ + if (old->tsc < (1000 * 1000)) { + fprintf(stderr, "Insanely slow TSC rate, TSC stops in idle?\n"); + fprintf(stderr, "You can disable all c-states by booting with \"idle=poll\"\n"); + fprintf(stderr, "or just the deep ones with \"processor.max_cstate=1\"\n"); + exit(-3); + } - for ( ; after && before && delta; - after = after->next, before = before->next, delta = delta->next) { - if (before->cpu != after->cpu) { - printf("cpu configuration changed: %d != %d\n", - before->cpu, after->cpu); - return -1; - } + old->c1 = new->c1 - old->c1; - if (SUBTRACT_COUNTER(after->tsc, before->tsc, delta->tsc)) { - fprintf(stderr, "cpu%d TSC went backwards %llX to %llX\n", - before->cpu, before->tsc, after->tsc); - errors++; - } - /* check for TSC < 1 Mcycles over interval */ - if (delta->tsc < (1000 * 1000)) { - fprintf(stderr, "Insanely slow TSC rate," - " TSC stops in idle?\n"); - fprintf(stderr, "You can disable all c-states" - " by booting with \"idle=poll\"\n"); - fprintf(stderr, "or just the deep ones with" - " \"processor.max_cstate=1\"\n"); - exit(-3); - } - if (SUBTRACT_COUNTER(after->c3, before->c3, delta->c3)) { - fprintf(stderr, "cpu%d c3 counter went backwards %llX to %llX\n", - before->cpu, before->c3, after->c3); - errors++; - } - if (SUBTRACT_COUNTER(after->c6, before->c6, delta->c6)) { - fprintf(stderr, "cpu%d c6 counter went backwards %llX to %llX\n", - before->cpu, before->c6, after->c6); - errors++; - } - if (SUBTRACT_COUNTER(after->c7, before->c7, delta->c7)) { - fprintf(stderr, "cpu%d c7 counter went backwards %llX to %llX\n", - before->cpu, before->c7, after->c7); - errors++; - } - if (SUBTRACT_COUNTER(after->pc2, before->pc2, delta->pc2)) { - fprintf(stderr, "cpu%d pc2 counter went backwards %llX to %llX\n", - before->cpu, before->pc2, after->pc2); - errors++; - } - if (SUBTRACT_COUNTER(after->pc3, before->pc3, delta->pc3)) { - fprintf(stderr, "cpu%d pc3 counter went backwards %llX to %llX\n", - before->cpu, before->pc3, after->pc3); - errors++; - } - if (SUBTRACT_COUNTER(after->pc6, before->pc6, delta->pc6)) { - fprintf(stderr, "cpu%d pc6 counter went backwards %llX to %llX\n", - before->cpu, before->pc6, after->pc6); - errors++; - } - if (SUBTRACT_COUNTER(after->pc7, before->pc7, delta->pc7)) { - fprintf(stderr, "cpu%d pc7 counter went backwards %llX to %llX\n", - before->cpu, before->pc7, after->pc7); - errors++; - } + if ((new->aperf > old->aperf) && (new->mperf > old->mperf)) { + old->aperf = new->aperf - old->aperf; + old->mperf = new->mperf - old->mperf; + } else { - perf_err = SUBTRACT_COUNTER(after->aperf, before->aperf, delta->aperf); - if (perf_err) { - fprintf(stderr, "cpu%d aperf counter went backwards %llX to %llX\n", - before->cpu, before->aperf, after->aperf); - } - perf_err |= SUBTRACT_COUNTER(after->mperf, before->mperf, delta->mperf); - if (perf_err) { - fprintf(stderr, "cpu%d mperf counter went backwards %llX to %llX\n", - before->cpu, before->mperf, after->mperf); - } - if (perf_err) { - if (!aperf_mperf_unstable) { - fprintf(stderr, "%s: APERF or MPERF went backwards *\n", progname); - fprintf(stderr, "* Frequency results do not cover entire interval *\n"); - fprintf(stderr, "* fix this by running Linux-2.6.30 or later *\n"); + if (!aperf_mperf_unstable) { + fprintf(stderr, "%s: APERF or MPERF went backwards *\n", progname); + fprintf(stderr, "* Frequency results do not cover entire interval *\n"); + fprintf(stderr, "* fix this by running Linux-2.6.30 or later *\n"); - aperf_mperf_unstable = 1; - } - /* - * mperf delta is likely a huge "positive" number - * can not use it for calculating c0 time - */ - skip_c0 = 1; - skip_c1 = 1; + aperf_mperf_unstable = 1; } - /* - * As mperf and tsc collection are not atomic, - * it is possible for mperf's non-halted cycles - * to exceed TSC's all cycles: show c1 = 0% in that case. + * mperf delta is likely a huge "positive" number + * can not use it for calculating c0 time */ - if (delta->mperf > delta->tsc) - delta->c1 = 0; - else /* normal case, derive c1 */ - delta->c1 = delta->tsc - delta->mperf - - delta->c3 - delta->c6 - delta->c7; + skip_c0 = 1; + skip_c1 = 1; + } - if (delta->mperf == 0) - delta->mperf = 1; /* divide by 0 protection */ - /* - * for "extra msr", just copy the latest w/o subtracting - */ - delta->extra_msr = after->extra_msr; - if (errors) { - fprintf(stderr, "ERROR cpu%d before:\n", before->cpu); - dump_cnt(before); - fprintf(stderr, "ERROR cpu%d after:\n", before->cpu); - dump_cnt(after); - errors = 0; - } + /* + * As counter collection is not atomic, + * it is possible for mperf's non-halted cycles + idle states + * to exceed TSC's all cycles: show c1 = 0% in that case. + */ + if ((old->mperf + core_delta->c3 + core_delta->c6 + core_delta->c7) > old->tsc) + old->c1 = 0; + else { + /* normal case, derive c1 */ + old->c1 = old->tsc - old->mperf - core_delta->c3 + - core_delta->c6 - core_delta->c7; } + + if (old->mperf == 0) { + if (verbose > 1) fprintf(stderr, "cpu%d MPERF 0!\n", old->cpu_id); + old->mperf = 1; /* divide by 0 protection */ + } + + /* + * for "extra msr", just copy the latest w/o subtracting + */ + old->extra_msr = new->extra_msr; +} + +int delta_cpu(struct thread_data *t, struct core_data *c, + struct pkg_data *p, struct thread_data *t2, + struct core_data *c2, struct pkg_data *p2) +{ + /* calculate core delta only for 1st thread in core */ + if (t->flags & CPU_IS_FIRST_THREAD_IN_CORE) + delta_core(c, c2); + + /* always calculate thread delta */ + delta_thread(t, t2, c2); /* c2 is core delta */ + + /* calculate package delta only for 1st core in package */ + if (t->flags & CPU_IS_FIRST_CORE_IN_PACKAGE) + delta_package(p, p2); + return 0; } -void compute_average(struct counters *delta, struct counters *avg) +void clear_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) +{ + t->tsc = 0; + t->aperf = 0; + t->mperf = 0; + t->c1 = 0; + + /* tells format_counters to dump all fields from this set */ + t->flags = CPU_IS_FIRST_THREAD_IN_CORE | CPU_IS_FIRST_CORE_IN_PACKAGE; + + c->c3 = 0; + c->c6 = 0; + c->c7 = 0; + + p->pc2 = 0; + p->pc3 = 0; + p->pc6 = 0; + p->pc7 = 0; +} +int sum_counters(struct thread_data *t, struct core_data *c, + struct pkg_data *p) { - struct counters *sum; + average.threads.tsc += t->tsc; + average.threads.aperf += t->aperf; + average.threads.mperf += t->mperf; + average.threads.c1 += t->c1; - sum = calloc(1, sizeof(struct counters)); - if (sum == NULL) { - perror("calloc sum"); - exit(1); - } + /* sum per-core values only for 1st thread in core */ + if (!(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) + return 0; - for (; delta; delta = delta->next) { - sum->tsc += delta->tsc; - sum->c1 += delta->c1; - sum->c3 += delta->c3; - sum->c6 += delta->c6; - sum->c7 += delta->c7; - sum->aperf += delta->aperf; - sum->mperf += delta->mperf; - sum->pc2 += delta->pc2; - sum->pc3 += delta->pc3; - sum->pc6 += delta->pc6; - sum->pc7 += delta->pc7; - } - avg->tsc = sum->tsc/num_cpus; - avg->c1 = sum->c1/num_cpus; - avg->c3 = sum->c3/num_cpus; - avg->c6 = sum->c6/num_cpus; - avg->c7 = sum->c7/num_cpus; - avg->aperf = sum->aperf/num_cpus; - avg->mperf = sum->mperf/num_cpus; - avg->pc2 = sum->pc2/num_cpus; - avg->pc3 = sum->pc3/num_cpus; - avg->pc6 = sum->pc6/num_cpus; - avg->pc7 = sum->pc7/num_cpus; - - free(sum); + average.cores.c3 += c->c3; + average.cores.c6 += c->c6; + average.cores.c7 += c->c7; + + /* sum per-pkg values only for 1st core in pkg */ + if (!(t->flags & CPU_IS_FIRST_CORE_IN_PACKAGE)) + return 0; + + average.packages.pc2 += p->pc2; + average.packages.pc3 += p->pc3; + average.packages.pc6 += p->pc6; + average.packages.pc7 += p->pc7; + + return 0; +} +/* + * sum the counters for all cpus in the system + * compute the weighted average + */ +void compute_average(struct thread_data *t, struct core_data *c, + struct pkg_data *p) +{ + clear_counters(&average.threads, &average.cores, &average.packages); + + for_all_cpus(sum_counters, t, c, p); + + average.threads.tsc /= topo.num_cpus; + average.threads.aperf /= topo.num_cpus; + average.threads.mperf /= topo.num_cpus; + average.threads.c1 /= topo.num_cpus; + + average.cores.c3 /= topo.num_cores; + average.cores.c6 /= topo.num_cores; + average.cores.c7 /= topo.num_cores; + + average.packages.pc2 /= topo.num_packages; + average.packages.pc3 /= topo.num_packages; + average.packages.pc6 /= topo.num_packages; + average.packages.pc7 /= topo.num_packages; } -int get_counters(struct counters *cnt) +static unsigned long long rdtsc(void) { - for ( ; cnt; cnt = cnt->next) { + unsigned int low, high; - if (cpu_migrate(cnt->cpu)) - return -1; + asm volatile("rdtsc" : "=a" (low), "=d" (high)); - if (get_msr(cnt->cpu, MSR_TSC, &cnt->tsc)) - return -1; + return low | ((unsigned long long)high) << 32; +} - if (has_aperf) { - if (get_msr(cnt->cpu, MSR_APERF, &cnt->aperf)) - return -1; - if (get_msr(cnt->cpu, MSR_MPERF, &cnt->mperf)) - return -1; - } - if (do_nhm_cstates) { - if (get_msr(cnt->cpu, MSR_CORE_C3_RESIDENCY, &cnt->c3)) - return -1; - if (get_msr(cnt->cpu, MSR_CORE_C6_RESIDENCY, &cnt->c6)) - return -1; - } +/* + * get_counters(...) + * migrate to cpu + * acquire and record local counters for that cpu + */ +int get_counters(struct thread_data *t, struct core_data *c, struct pkg_data *p) +{ + int cpu = t->cpu_id; - if (do_snb_cstates) - if (get_msr(cnt->cpu, MSR_CORE_C7_RESIDENCY, &cnt->c7)) - return -1; + if (cpu_migrate(cpu)) + return -1; - if (do_nhm_cstates) { - if (get_msr(cnt->cpu, MSR_PKG_C3_RESIDENCY, &cnt->pc3)) - return -1; - if (get_msr(cnt->cpu, MSR_PKG_C6_RESIDENCY, &cnt->pc6)) - return -1; - } - if (do_snb_cstates) { - if (get_msr(cnt->cpu, MSR_PKG_C2_RESIDENCY, &cnt->pc2)) - return -1; - if (get_msr(cnt->cpu, MSR_PKG_C7_RESIDENCY, &cnt->pc7)) - return -1; - } - if (extra_msr_offset) - if (get_msr(cnt->cpu, extra_msr_offset, &cnt->extra_msr)) - return -1; + t->tsc = rdtsc(); /* we are running on local CPU of interest */ + + if (has_aperf) { + if (get_msr(cpu, MSR_APERF, &t->aperf)) + return -3; + if (get_msr(cpu, MSR_MPERF, &t->mperf)) + return -4; + } + + if (extra_msr_offset) + if (get_msr(cpu, extra_msr_offset, &t->extra_msr)) + return -5; + + /* collect core counters only for 1st thread in core */ + if (!(t->flags & CPU_IS_FIRST_THREAD_IN_CORE)) + return 0; + + if (do_nhm_cstates) { + if (get_msr(cpu, MSR_CORE_C3_RESIDENCY, &c->c3)) + return -6; + if (get_msr(cpu, MSR_CORE_C6_RESIDENCY, &c->c6)) + return -7; + } + + if (do_snb_cstates) + if (get_msr(cpu, MSR_CORE_C7_RESIDENCY, &c->c7)) + return -8; + + /* collect package counters only for 1st core in package */ + if (!(t->flags & CPU_IS_FIRST_CORE_IN_PACKAGE)) + return 0; + + if (do_nhm_cstates) { + if (get_msr(cpu, MSR_PKG_C3_RESIDENCY, &p->pc3)) + return -9; + if (get_msr(cpu, MSR_PKG_C6_RESIDENCY, &p->pc6)) + return -10; + } + if (do_snb_cstates) { + if (get_msr(cpu, MSR_PKG_C2_RESIDENCY, &p->pc2)) + return -11; + if (get_msr(cpu, MSR_PKG_C7_RESIDENCY, &p->pc7)) + return -12; } return 0; } -void print_nehalem_info(void) +void print_verbose_header(void) { unsigned long long msr; unsigned int ratio; @@ -615,143 +715,82 @@ void print_nehalem_info(void) } -void free_counter_list(struct counters *list) +void free_all_buffers(void) { - struct counters *p; + CPU_FREE(cpu_present_set); + cpu_present_set = NULL; + cpu_present_set = 0; - for (p = list; p; ) { - struct counters *free_me; + CPU_FREE(cpu_affinity_set); + cpu_affinity_set = NULL; + cpu_affinity_setsize = 0; - free_me = p; - p = p->next; - free(free_me); - } -} + free(thread_even); + free(core_even); + free(package_even); -void free_all_counters(void) -{ - free_counter_list(cnt_even); - cnt_even = NULL; + thread_even = NULL; + core_even = NULL; + package_even = NULL; - free_counter_list(cnt_odd); - cnt_odd = NULL; + free(thread_odd); + free(core_odd); + free(package_odd); - free_counter_list(cnt_delta); - cnt_delta = NULL; + thread_odd = NULL; + core_odd = NULL; + package_odd = NULL; - free_counter_list(cnt_average); - cnt_average = NULL; + free(output_buffer); + output_buffer = NULL; + outp = NULL; } -void insert_counters(struct counters **list, - struct counters *new) +/* + * cpu_is_first_sibling_in_core(cpu) + * return 1 if given CPU is 1st HT sibling in the core + */ +int cpu_is_first_sibling_in_core(int cpu) { - struct counters *prev; - - /* - * list was empty - */ - if (*list == NULL) { - new->next = *list; - *list = new; - return; - } - - if (!summary_only) - show_cpu = 1; /* there is more than one CPU */ - - /* - * insert on front of list. - * It is sorted by ascending package#, core#, cpu# - */ - if (((*list)->pkg > new->pkg) || - (((*list)->pkg == new->pkg) && ((*list)->core > new->core)) || - (((*list)->pkg == new->pkg) && ((*list)->core == new->core) && ((*list)->cpu > new->cpu))) { - new->next = *list; - *list = new; - return; - } - - prev = *list; - - while (prev->next && (prev->next->pkg < new->pkg)) { - prev = prev->next; - if (!summary_only) - show_pkg = 1; /* there is more than 1 package */ - } - - while (prev->next && (prev->next->pkg == new->pkg) - && (prev->next->core < new->core)) { - prev = prev->next; - if (!summary_only) - show_core = 1; /* there is more than 1 core */ - } + char path[64]; + FILE *filep; + int first_cpu; - while (prev->next && (prev->next->pkg == new->pkg) - && (prev->next->core == new->core) - && (prev->next->cpu < new->cpu)) { - prev = prev->next; + sprintf(path, "/sys/devices/system/cpu/cpu%d/topology/thread_siblings_list", cpu); + filep = fopen(path, "r"); + if (filep == NULL) { + perror(path); + exit(1); } - - /* - * insert after "prev" - */ - new->next = prev->next; - prev->next = new; + fscanf(filep, "%d", &first_cpu); + fclose(filep); + return (cpu == first_cpu); } -void alloc_new_counters(int pkg, int core, int cpu) +/* + * cpu_is_first_core_in_package(cpu) + * return 1 if given CPU is 1st core in package + */ +int cpu_is_first_core_in_package(int cpu) { - struct counters *new; - - if (verbose > 1) - printf("pkg%d core%d, cpu%d\n", pkg, core, cpu); - - new = (struct counters *)calloc(1, sizeof(struct counters)); - if (new == NULL) { - perror("calloc"); - exit(1); - } - new->pkg = pkg; - new->core = core; - new->cpu = cpu; - insert_counters(&cnt_odd, new); - - new = (struct counters *)calloc(1, - sizeof(struct counters)); - if (new == NULL) { - perror("calloc"); - exit(1); - } - new->pkg = pkg; - new->core = core; - new->cpu = cpu; - insert_counters(&cnt_even, new); - - new = (struct counters *)calloc(1, sizeof(struct counters)); - if (new == NULL) { - perror("calloc"); - exit(1); - } - new->pkg = pkg; - new->core = core; - new->cpu = cpu; - insert_counters(&cnt_delta, new); + char path[64]; + FILE *filep; + int first_cpu; - new = (struct counters *)calloc(1, sizeof(struct counters)); - if (new == NULL) { - perror("calloc"); + sprintf(path, "/sys/devices/system/cpu/cpu%d/topology/core_siblings_list", cpu); + filep = fopen(path, "r"); + if (filep == NULL) { + perror(path); exit(1); } - new->pkg = pkg; - new->core = core; - new->cpu = cpu; - cnt_average = new; + fscanf(filep, "%d", &first_cpu); + fclose(filep); + return (cpu == first_cpu); } int get_physical_package_id(int cpu) { - char path[64]; + char path[80]; FILE *filep; int pkg; @@ -768,7 +807,7 @@ int get_physical_package_id(int cpu) int get_core_id(int cpu) { - char path[64]; + char path[80]; FILE *filep; int core; @@ -783,14 +822,87 @@ int get_core_id(int cpu) return core; } +int get_num_ht_siblings(int cpu) +{ + char path[80]; + FILE *filep; + int sib1, sib2; + int matches; + char character; + + sprintf(path, "/sys/devices/system/cpu/cpu%d/topology/thread_siblings_list", cpu); + filep = fopen(path, "r"); + if (filep == NULL) { + perror(path); + exit(1); + } + /* + * file format: + * if a pair of number with a character between: 2 siblings (eg. 1-2, or 1,4) + * otherwinse 1 sibling (self). + */ + matches = fscanf(filep, "%d%c%d\n", &sib1, &character, &sib2); + + fclose(filep); + + if (matches == 3) + return 2; + else + return 1; +} + /* - * run func(pkg, core, cpu) on every cpu in /proc/stat + * run func(thread, core, package) in topology order + * skip non-present cpus */ -int for_all_cpus(void (func)(int, int, int)) +int for_all_cpus_2(int (func)(struct thread_data *, struct core_data *, + struct pkg_data *, struct thread_data *, struct core_data *, + struct pkg_data *), struct thread_data *thread_base, + struct core_data *core_base, struct pkg_data *pkg_base, + struct thread_data *thread_base2, struct core_data *core_base2, + struct pkg_data *pkg_base2) +{ + int retval, pkg_no, core_no, thread_no; + + for (pkg_no = 0; pkg_no < topo.num_packages; ++pkg_no) { + for (core_no = 0; core_no < topo.num_cores_per_pkg; ++core_no) { + for (thread_no = 0; thread_no < + topo.num_threads_per_core; ++thread_no) { + struct thread_data *t, *t2; + struct core_data *c, *c2; + struct pkg_data *p, *p2; + + t = GET_THREAD(thread_base, thread_no, core_no, pkg_no); + + if (cpu_is_not_present(t->cpu_id)) + continue; + + t2 = GET_THREAD(thread_base2, thread_no, core_no, pkg_no); + + c = GET_CORE(core_base, core_no, pkg_no); + c2 = GET_CORE(core_base2, core_no, pkg_no); + + p = GET_PKG(pkg_base, pkg_no); + p2 = GET_PKG(pkg_base2, pkg_no); + + retval = func(t, c, p, t2, c2, p2); + if (retval) + return retval; + } + } + } + return 0; +} + +/* + * run func(cpu) on every cpu in /proc/stat + * return max_cpu number + */ +int for_all_proc_cpus(int (func)(int)) { FILE *fp; - int cpu_count; + int cpu_num; int retval; fp = fopen(proc_stat, "r"); @@ -805,78 +917,88 @@ int for_all_cpus(void (func)(int, int, int)) exit(1); } - for (cpu_count = 0; ; cpu_count++) { - int cpu; - - retval = fscanf(fp, "cpu%u %*d %*d %*d %*d %*d %*d %*d %*d %*d %*d\n", &cpu); + while (1) { + retval = fscanf(fp, "cpu%u %*d %*d %*d %*d %*d %*d %*d %*d %*d %*d\n", &cpu_num); if (retval != 1) break; - func(get_physical_package_id(cpu), get_core_id(cpu), cpu); + retval = func(cpu_num); + if (retval) { + fclose(fp); + return(retval); + } } fclose(fp); - return cpu_count; + return 0; } void re_initialize(void) { - free_all_counters(); - num_cpus = for_all_cpus(alloc_new_counters); - cpu_mask_uninit(); - cpu_mask_init(num_cpus); - printf("turbostat: re-initialized with num_cpus %d\n", num_cpus); + free_all_buffers(); + setup_all_buffers(); + printf("turbostat: re-initialized with num_cpus %d\n", topo.num_cpus); } -void dummy(int pkg, int core, int cpu) { return; } + /* - * check to see if a cpu came on-line + * count_cpus() + * remember the last one seen, it will be the max */ -int verify_num_cpus(void) +int count_cpus(int cpu) { - int new_num_cpus; - - new_num_cpus = for_all_cpus(dummy); + if (topo.max_cpu_num < cpu) + topo.max_cpu_num = cpu; - if (new_num_cpus != num_cpus) { - if (verbose) - printf("num_cpus was %d, is now %d\n", - num_cpus, new_num_cpus); - return -1; - } + topo.num_cpus += 1; + return 0; +} +int mark_cpu_present(int cpu) +{ + CPU_SET_S(cpu, cpu_present_setsize, cpu_present_set); return 0; } void turbostat_loop() { + int retval; + restart: - get_counters(cnt_even); + retval = for_all_cpus(get_counters, EVEN_COUNTERS); + if (retval) { + re_initialize(); + goto restart; + } gettimeofday(&tv_even, (struct timezone *)NULL); while (1) { - if (verify_num_cpus()) { + if (for_all_proc_cpus(cpu_is_not_present)) { re_initialize(); goto restart; } sleep(interval_sec); - if (get_counters(cnt_odd)) { + retval = for_all_cpus(get_counters, ODD_COUNTERS); + if (retval) { re_initialize(); goto restart; } gettimeofday(&tv_odd, (struct timezone *)NULL); - compute_delta(cnt_odd, cnt_even, cnt_delta); timersub(&tv_odd, &tv_even, &tv_delta); - compute_average(cnt_delta, cnt_average); - print_counters(cnt_delta); + for_all_cpus_2(delta_cpu, ODD_COUNTERS, EVEN_COUNTERS); + compute_average(EVEN_COUNTERS); + format_all_counters(EVEN_COUNTERS); + flush_stdout(); sleep(interval_sec); - if (get_counters(cnt_even)) { + retval = for_all_cpus(get_counters, EVEN_COUNTERS); + if (retval) { re_initialize(); goto restart; } gettimeofday(&tv_even, (struct timezone *)NULL); - compute_delta(cnt_even, cnt_odd, cnt_delta); timersub(&tv_even, &tv_odd, &tv_delta); - compute_average(cnt_delta, cnt_average); - print_counters(cnt_delta); + for_all_cpus_2(delta_cpu, EVEN_COUNTERS, ODD_COUNTERS); + compute_average(ODD_COUNTERS); + format_all_counters(ODD_COUNTERS); + flush_stdout(); } } @@ -1051,6 +1173,208 @@ int open_dev_cpu_msr(int dummy1) return 0; } +void topology_probe() +{ + int i; + int max_core_id = 0; + int max_package_id = 0; + int max_siblings = 0; + struct cpu_topology { + int core_id; + int physical_package_id; + } *cpus; + + /* Initialize num_cpus, max_cpu_num */ + topo.num_cpus = 0; + topo.max_cpu_num = 0; + for_all_proc_cpus(count_cpus); + if (!summary_only && topo.num_cpus > 1) + show_cpu = 1; + + if (verbose > 1) + fprintf(stderr, "num_cpus %d max_cpu_num %d\n", topo.num_cpus, topo.max_cpu_num); + + cpus = calloc(1, (topo.max_cpu_num + 1) * sizeof(struct cpu_topology)); + if (cpus == NULL) { + perror("calloc cpus"); + exit(1); + } + + /* + * Allocate and initialize cpu_present_set + */ + cpu_present_set = CPU_ALLOC((topo.max_cpu_num + 1)); + if (cpu_present_set == NULL) { + perror("CPU_ALLOC"); + exit(3); + } + cpu_present_setsize = CPU_ALLOC_SIZE((topo.max_cpu_num + 1)); + CPU_ZERO_S(cpu_present_setsize, cpu_present_set); + for_all_proc_cpus(mark_cpu_present); + + /* + * Allocate and initialize cpu_affinity_set + */ + cpu_affinity_set = CPU_ALLOC((topo.max_cpu_num + 1)); + if (cpu_affinity_set == NULL) { + perror("CPU_ALLOC"); + exit(3); + } + cpu_affinity_setsize = CPU_ALLOC_SIZE((topo.max_cpu_num + 1)); + CPU_ZERO_S(cpu_affinity_setsize, cpu_affinity_set); + + + /* + * For online cpus + * find max_core_id, max_package_id + */ + for (i = 0; i <= topo.max_cpu_num; ++i) { + int siblings; + + if (cpu_is_not_present(i)) { + if (verbose > 1) + fprintf(stderr, "cpu%d NOT PRESENT\n", i); + continue; + } + cpus[i].core_id = get_core_id(i); + if (cpus[i].core_id > max_core_id) + max_core_id = cpus[i].core_id; + + cpus[i].physical_package_id = get_physical_package_id(i); + if (cpus[i].physical_package_id > max_package_id) + max_package_id = cpus[i].physical_package_id; + + siblings = get_num_ht_siblings(i); + if (siblings > max_siblings) + max_siblings = siblings; + if (verbose > 1) + fprintf(stderr, "cpu %d pkg %d core %d\n", + i, cpus[i].physical_package_id, cpus[i].core_id); + } + topo.num_cores_per_pkg = max_core_id + 1; + if (verbose > 1) + fprintf(stderr, "max_core_id %d, sizing for %d cores per package\n", + max_core_id, topo.num_cores_per_pkg); + if (!summary_only && topo.num_cores_per_pkg > 1) + show_core = 1; + + topo.num_packages = max_package_id + 1; + if (verbose > 1) + fprintf(stderr, "max_package_id %d, sizing for %d packages\n", + max_package_id, topo.num_packages); + if (!summary_only && topo.num_packages > 1) + show_pkg = 1; + + topo.num_threads_per_core = max_siblings; + if (verbose > 1) + fprintf(stderr, "max_siblings %d\n", max_siblings); + + free(cpus); +} + +void +allocate_counters(struct thread_data **t, struct core_data **c, struct pkg_data **p) +{ + int i; + + *t = calloc(topo.num_threads_per_core * topo.num_cores_per_pkg * + topo.num_packages, sizeof(struct thread_data)); + if (*t == NULL) + goto error; + + for (i = 0; i < topo.num_threads_per_core * + topo.num_cores_per_pkg * topo.num_packages; i++) + (*t)[i].cpu_id = -1; + + *c = calloc(topo.num_cores_per_pkg * topo.num_packages, + sizeof(struct core_data)); + if (*c == NULL) + goto error; + + for (i = 0; i < topo.num_cores_per_pkg * topo.num_packages; i++) + (*c)[i].core_id = -1; + + *p = calloc(topo.num_packages, sizeof(struct pkg_data)); + if (*p == NULL) + goto error; + + for (i = 0; i < topo.num_packages; i++) + (*p)[i].package_id = i; + + return; +error: + perror("calloc counters"); + exit(1); +} +/* + * init_counter() + * + * set cpu_id, core_num, pkg_num + * set FIRST_THREAD_IN_CORE and FIRST_CORE_IN_PACKAGE + * + * increment topo.num_cores when 1st core in pkg seen + */ +void init_counter(struct thread_data *thread_base, struct core_data *core_base, + struct pkg_data *pkg_base, int thread_num, int core_num, + int pkg_num, int cpu_id) +{ + struct thread_data *t; + struct core_data *c; + struct pkg_data *p; + + t = GET_THREAD(thread_base, thread_num, core_num, pkg_num); + c = GET_CORE(core_base, core_num, pkg_num); + p = GET_PKG(pkg_base, pkg_num); + + t->cpu_id = cpu_id; + if (thread_num == 0) { + t->flags |= CPU_IS_FIRST_THREAD_IN_CORE; + if (cpu_is_first_core_in_package(cpu_id)) + t->flags |= CPU_IS_FIRST_CORE_IN_PACKAGE; + } + + c->core_id = core_num; + p->package_id = pkg_num; +} + + +int initialize_counters(int cpu_id) +{ + int my_thread_id, my_core_id, my_package_id; + + my_package_id = get_physical_package_id(cpu_id); + my_core_id = get_core_id(cpu_id); + + if (cpu_is_first_sibling_in_core(cpu_id)) { + my_thread_id = 0; + topo.num_cores++; + } else { + my_thread_id = 1; + } + + init_counter(EVEN_COUNTERS, my_thread_id, my_core_id, my_package_id, cpu_id); + init_counter(ODD_COUNTERS, my_thread_id, my_core_id, my_package_id, cpu_id); + return 0; +} + +void allocate_output_buffer() +{ + output_buffer = calloc(1, (1 + topo.num_cpus) * 128); + outp = output_buffer; + if (outp == NULL) { + perror("calloc"); + exit(-1); + } +} + +void setup_all_buffers(void) +{ + topology_probe(); + allocate_counters(&thread_even, &core_even, &package_even); + allocate_counters(&thread_odd, &core_odd, &package_odd); + allocate_output_buffer(); + for_all_proc_cpus(initialize_counters); +} void turbostat_init() { check_cpuid(); @@ -1058,21 +1382,19 @@ void turbostat_init() check_dev_msr(); check_super_user(); - num_cpus = for_all_cpus(alloc_new_counters); - cpu_mask_init(num_cpus); + setup_all_buffers(); if (verbose) - print_nehalem_info(); + print_verbose_header(); } int fork_it(char **argv) { - int retval; pid_t child_pid; - get_counters(cnt_even); - /* clear affinity side-effect of get_counters() */ - sched_setaffinity(0, cpu_present_setsize, cpu_present_set); + for_all_cpus(get_counters, EVEN_COUNTERS); + /* clear affinity side-effect of get_counters() */ + sched_setaffinity(0, cpu_present_setsize, cpu_present_set); gettimeofday(&tv_even, (struct timezone *)NULL); child_pid = fork(); @@ -1095,14 +1417,17 @@ int fork_it(char **argv) exit(1); } } - get_counters(cnt_odd); + /* + * n.b. fork_it() does not check for errors from for_all_cpus() + * because re-starting is problematic when forking + */ + for_all_cpus(get_counters, ODD_COUNTERS); gettimeofday(&tv_odd, (struct timezone *)NULL); - retval = compute_delta(cnt_odd, cnt_even, cnt_delta); - timersub(&tv_odd, &tv_even, &tv_delta); - compute_average(cnt_delta, cnt_average); - if (!retval) - print_counters(cnt_delta); + for_all_cpus_2(delta_cpu, ODD_COUNTERS, EVEN_COUNTERS); + compute_average(EVEN_COUNTERS); + format_all_counters(EVEN_COUNTERS); + flush_stderr(); fprintf(stderr, "%.6f sec\n", tv_delta.tv_sec + tv_delta.tv_usec/1000000.0); @@ -1115,8 +1440,14 @@ void cmdline(int argc, char **argv) progname = argv[0]; - while ((opt = getopt(argc, argv, "+svi:M:")) != -1) { + while ((opt = getopt(argc, argv, "+cpsvi:M:")) != -1) { switch (opt) { + case 'c': + show_core_only++; + break; + case 'p': + show_pkg_only++; + break; case 's': summary_only++; break; @@ -1142,10 +1473,8 @@ int main(int argc, char **argv) cmdline(argc, argv); if (verbose > 1) - fprintf(stderr, "turbostat Dec 6, 2010" + fprintf(stderr, "turbostat v2.0 May 16, 2012" " - Len Brown <lenb@kernel.org>\n"); - if (verbose > 1) - fprintf(stderr, "http://userweb.kernel.org/~lenb/acpi/utils/pmtools/turbostat/\n"); turbostat_init(); diff --git a/tools/testing/fault-injection/failcmd.sh b/tools/testing/fault-injection/failcmd.sh new file mode 100644 index 00000000000..78a9ed7fecd --- /dev/null +++ b/tools/testing/fault-injection/failcmd.sh @@ -0,0 +1,219 @@ +#!/bin/bash +# +# NAME +# failcmd.sh - run a command with injecting slab/page allocation failures +# +# SYNOPSIS +# failcmd.sh --help +# failcmd.sh [<options>] command [arguments] +# +# DESCRIPTION +# Run command with injecting slab/page allocation failures by fault +# injection. +# +# NOTE: you need to run this script as root. +# + +usage() +{ + cat >&2 <<EOF +Usage: $0 [options] command [arguments] + +OPTIONS + -p percent + --probability=percent + likelihood of failure injection, in percent. + Default value is 1 + + -t value + --times=value + specifies how many times failures may happen at most. + Default value is 1 + + --oom-kill-allocating-task=value + set /proc/sys/vm/oom_kill_allocating_task to specified value + before running the command. + Default value is 1 + + -h, --help + Display a usage message and exit + + --interval=value, --space=value, --verbose=value, --task-filter=value, + --stacktrace-depth=value, --require-start=value, --require-end=value, + --reject-start=value, --reject-end=value, --ignore-gfp-wait=value + See Documentation/fault-injection/fault-injection.txt for more + information + + failslab options: + --cache-filter=value + + fail_page_alloc options: + --ignore-gfp-highmem=value, --min-order=value + +ENVIRONMENT + FAILCMD_TYPE + The following values for FAILCMD_TYPE are recognized: + + failslab + inject slab allocation failures + fail_page_alloc + inject page allocation failures + + If FAILCMD_TYPE is not defined, then failslab is used. +EOF +} + +if [ $UID != 0 ]; then + echo must be run as root >&2 + exit 1 +fi + +DEBUGFS=`mount -t debugfs | head -1 | awk '{ print $3}'` + +if [ ! -d "$DEBUGFS" ]; then + echo debugfs is not mounted >&2 + exit 1 +fi + +FAILCMD_TYPE=${FAILCMD_TYPE:-failslab} +FAULTATTR=$DEBUGFS/$FAILCMD_TYPE + +if [ ! -d $FAULTATTR ]; then + echo $FAILCMD_TYPE is not available >&2 + exit 1 +fi + +LONGOPTS=probability:,interval:,times:,space:,verbose:,task-filter: +LONGOPTS=$LONGOPTS,stacktrace-depth:,require-start:,require-end: +LONGOPTS=$LONGOPTS,reject-start:,reject-end:,oom-kill-allocating-task:,help + +if [ $FAILCMD_TYPE = failslab ]; then + LONGOPTS=$LONGOPTS,ignore-gfp-wait:,cache-filter: +elif [ $FAILCMD_TYPE = fail_page_alloc ]; then + LONGOPTS=$LONGOPTS,ignore-gfp-wait:,ignore-gfp-highmem:,min-order: +fi + +TEMP=`getopt -o p:i:t:s:v:h --long $LONGOPTS -n 'failcmd.sh' -- "$@"` + +if [ $? != 0 ]; then + usage + exit 1 +fi + +eval set -- "$TEMP" + +fault_attr_default() +{ + echo N > $FAULTATTR/task-filter + echo 0 > $FAULTATTR/probability + echo 1 > $FAULTATTR/times +} + +fault_attr_default + +oom_kill_allocating_task_saved=`cat /proc/sys/vm/oom_kill_allocating_task` + +restore_values() +{ + fault_attr_default + echo $oom_kill_allocating_task_saved \ + > /proc/sys/vm/oom_kill_allocating_task +} + +# +# Default options +# +declare -i oom_kill_allocating_task=1 +declare task_filter=Y +declare -i probability=1 +declare -i times=1 + +while true; do + case "$1" in + -p|--probability) + probability=$2 + shift 2 + ;; + -i|--interval) + echo $2 > $FAULTATTR/interval + shift 2 + ;; + -t|--times) + times=$2 + shift 2 + ;; + -s|--space) + echo $2 > $FAULTATTR/space + shift 2 + ;; + -v|--verbose) + echo $2 > $FAULTATTR/verbose + shift 2 + ;; + --task-filter) + task_filter=$2 + shift 2 + ;; + --stacktrace-depth) + echo $2 > $FAULTATTR/stacktrace-depth + shift 2 + ;; + --require-start) + echo $2 > $FAULTATTR/require-start + shift 2 + ;; + --require-end) + echo $2 > $FAULTATTR/require-end + shift 2 + ;; + --reject-start) + echo $2 > $FAULTATTR/reject-start + shift 2 + ;; + --reject-end) + echo $2 > $FAULTATTR/reject-end + shift 2 + ;; + --oom-kill-allocating-task) + oom_kill_allocating_task=$2 + shift 2 + ;; + --ignore-gfp-wait) + echo $2 > $FAULTATTR/ignore-gfp-wait + shift 2 + ;; + --cache-filter) + echo $2 > $FAULTATTR/cache_filter + shift 2 + ;; + --ignore-gfp-highmem) + echo $2 > $FAULTATTR/ignore-gfp-highmem + shift 2 + ;; + --min-order) + echo $2 > $FAULTATTR/min-order + shift 2 + ;; + -h|--help) + usage + exit 0 + shift + ;; + --) + shift + break + ;; + esac +done + +[ -z "$1" ] && exit 0 + +echo $oom_kill_allocating_task > /proc/sys/vm/oom_kill_allocating_task +echo $task_filter > $FAULTATTR/task-filter +echo $probability > $FAULTATTR/probability +echo $times > $FAULTATTR/times + +trap "restore_values" SIGINT SIGTERM EXIT + +cmd="echo 1 > /proc/self/make-it-fail && exec $@" +bash -c "$cmd" diff --git a/tools/testing/ktest/ktest.pl b/tools/testing/ktest/ktest.pl index 292b13ad03f..52b7959cd51 100755 --- a/tools/testing/ktest/ktest.pl +++ b/tools/testing/ktest/ktest.pl @@ -52,6 +52,7 @@ my %default = ( "STOP_AFTER_SUCCESS" => 10, "STOP_AFTER_FAILURE" => 60, "STOP_TEST_AFTER" => 600, + "MAX_MONITOR_WAIT" => 1800, # required, and we will ask users if they don't have them but we keep the default # value something that is common. @@ -77,6 +78,11 @@ my $output_config; my $test_type; my $build_type; my $build_options; +my $final_post_ktest; +my $pre_ktest; +my $post_ktest; +my $pre_test; +my $post_test; my $pre_build; my $post_build; my $pre_build_die; @@ -93,6 +99,7 @@ my $reboot_on_success; my $die_on_failure; my $powercycle_after_reboot; my $poweroff_after_halt; +my $max_monitor_wait; my $ssh_exec; my $scp_to_target; my $scp_to_target_install; @@ -101,6 +108,7 @@ my $grub_menu; my $grub_number; my $target; my $make; +my $pre_install; my $post_install; my $no_install; my $noclean; @@ -167,6 +175,7 @@ my $bisect_check; my $config_bisect; my $config_bisect_type; +my $config_bisect_check; my $patchcheck_type; my $patchcheck_start; @@ -182,6 +191,9 @@ my $newconfig = 0; my %entered_configs; my %config_help; my %variable; + +# force_config is the list of configs that we force enabled (or disabled) +# in a .config file. The MIN_CONFIG and ADD_CONFIG configs. my %force_config; # do not force reboots on config problems @@ -197,6 +209,10 @@ my %option_map = ( "OUTPUT_DIR" => \$outputdir, "BUILD_DIR" => \$builddir, "TEST_TYPE" => \$test_type, + "PRE_KTEST" => \$pre_ktest, + "POST_KTEST" => \$post_ktest, + "PRE_TEST" => \$pre_test, + "POST_TEST" => \$post_test, "BUILD_TYPE" => \$build_type, "BUILD_OPTIONS" => \$build_options, "PRE_BUILD" => \$pre_build, @@ -216,6 +232,7 @@ my %option_map = ( "ADD_CONFIG" => \$addconfig, "REBOOT_TYPE" => \$reboot_type, "GRUB_MENU" => \$grub_menu, + "PRE_INSTALL" => \$pre_install, "POST_INSTALL" => \$post_install, "NO_INSTALL" => \$no_install, "REBOOT_SCRIPT" => \$reboot_script, @@ -228,6 +245,7 @@ my %option_map = ( "POWER_OFF" => \$power_off, "POWERCYCLE_AFTER_REBOOT" => \$powercycle_after_reboot, "POWEROFF_AFTER_HALT" => \$poweroff_after_halt, + "MAX_MONITOR_WAIT" => \$max_monitor_wait, "SLEEP_TIME" => \$sleep_time, "BISECT_SLEEP_TIME" => \$bisect_sleep_time, "PATCHCHECK_SLEEP_TIME" => \$patchcheck_sleep_time, @@ -272,6 +290,7 @@ my %option_map = ( "CONFIG_BISECT" => \$config_bisect, "CONFIG_BISECT_TYPE" => \$config_bisect_type, + "CONFIG_BISECT_CHECK" => \$config_bisect_check, "PATCHCHECK_TYPE" => \$patchcheck_type, "PATCHCHECK_START" => \$patchcheck_start, @@ -604,6 +623,10 @@ sub process_compare { return $lval eq $rval; } elsif ($cmp eq "!=") { return $lval ne $rval; + } elsif ($cmp eq "=~") { + return $lval =~ m/$rval/; + } elsif ($cmp eq "!~") { + return $lval !~ m/$rval/; } my $statement = "$lval $cmp $rval"; @@ -659,7 +682,7 @@ sub process_expression { } } - if ($val =~ /(.*)(==|\!=|>=|<=|>|<)(.*)/) { + if ($val =~ /(.*)(==|\!=|>=|<=|>|<|=~|\!~)(.*)/) { my $ret = process_compare($1, $2, $3); if ($ret < 0) { die "$name: $.: Unable to process comparison\n"; @@ -1117,7 +1140,11 @@ sub reboot { } if (defined($time)) { - wait_for_monitor($time, $reboot_success_line); + if (wait_for_monitor($time, $reboot_success_line)) { + # reboot got stuck? + doprint "Reboot did not finish. Forcing power cycle\n"; + run_command "$power_cycle"; + } end_monitor; } } @@ -1212,6 +1239,11 @@ sub wait_for_monitor { my $full_line = ""; my $line; my $booted = 0; + my $start_time = time; + my $skip_call_trace = 0; + my $bug = 0; + my $bug_ignored = 0; + my $now; doprint "** Wait for monitor to settle down **\n"; @@ -1227,11 +1259,39 @@ sub wait_for_monitor { $booted = 1; } + if ($full_line =~ /\[ backtrace testing \]/) { + $skip_call_trace = 1; + } + + if ($full_line =~ /call trace:/i) { + if (!$bug && !$skip_call_trace) { + if ($ignore_errors) { + $bug_ignored = 1; + } else { + $bug = 1; + } + } + } + + if ($full_line =~ /\[ end of backtrace testing \]/) { + $skip_call_trace = 0; + } + + if ($full_line =~ /Kernel panic -/) { + $bug = 1; + } + if ($line =~ /\n/) { $full_line = ""; } + $now = time; + if ($now - $start_time >= $max_monitor_wait) { + doprint "Exiting monitor flush due to hitting MAX_MONITOR_WAIT\n"; + return 1; + } } print "** Monitor flushed **\n"; + return $bug; } sub save_logs { @@ -1273,6 +1333,10 @@ sub save_logs { sub fail { + if (defined($post_test)) { + run_command $post_test; + } + if ($die_on_failure) { dodie @_; } @@ -1656,6 +1720,12 @@ sub install { return if ($no_install); + if (defined($pre_install)) { + my $cp_pre_install = eval_kernel_version $pre_install; + run_command "$cp_pre_install" or + dodie "Failed to run pre install"; + } + my $cp_target = eval_kernel_version $target_image; run_scp_install "$outputdir/$build_target", "$cp_target" or @@ -1814,6 +1884,7 @@ sub make_oldconfig { sub load_force_config { my ($config) = @_; + doprint "Loading force configs from $config\n"; open(IN, $config) or dodie "failed to read $config"; while (<IN>) { @@ -1937,6 +2008,10 @@ sub halt { sub success { my ($i) = @_; + if (defined($post_test)) { + run_command $post_test; + } + $successes++; my $name = ""; @@ -2003,6 +2078,7 @@ sub do_run_test { my $line; my $full_line; my $bug = 0; + my $bug_ignored = 0; wait_for_monitor 1; @@ -2027,7 +2103,11 @@ sub do_run_test { doprint $line; if ($full_line =~ /call trace:/i) { - $bug = 1; + if ($ignore_errors) { + $bug_ignored = 1; + } else { + $bug = 1; + } } if ($full_line =~ /Kernel panic -/) { @@ -2040,6 +2120,10 @@ sub do_run_test { } } while (!$child_done && !$bug); + if (!$bug && $bug_ignored) { + doprint "WARNING: Call Trace detected but ignored due to IGNORE_ERRORS=1\n"; + } + if ($bug) { my $failure_start = time; my $now; @@ -2362,9 +2446,24 @@ sub bisect { success $i; } +# config_ignore holds the configs that were set (or unset) for +# a good config and we will ignore these configs for the rest +# of a config bisect. These configs stay as they were. my %config_ignore; + +# config_set holds what all configs were set as. my %config_set; +# config_off holds the set of configs that the bad config had disabled. +# We need to record them and set them in the .config when running +# oldnoconfig, because oldnoconfig does not turn off new symbols, but +# instead just keeps the defaults. +my %config_off; + +# config_off_tmp holds a set of configs to turn off for now +my @config_off_tmp; + +# config_list is the set of configs that are being tested my %config_list; my %null_config; @@ -2443,12 +2542,21 @@ sub create_config { } } + # turn off configs to keep off + foreach my $config (keys %config_off) { + print OUT "# $config is not set\n"; + } + + # turn off configs that should be off for now + foreach my $config (@config_off_tmp) { + print OUT "# $config is not set\n"; + } + foreach my $config (keys %config_ignore) { print OUT "$config_ignore{$config}\n"; } close(OUT); -# exit; make_oldconfig; } @@ -2525,6 +2633,13 @@ sub run_config_bisect { do { my @tophalf = @start_list[0 .. $half]; + # keep the bottom half off + if ($half < $#start_list) { + @config_off_tmp = @start_list[$half + 1 .. $#start_list]; + } else { + @config_off_tmp = (); + } + create_config @tophalf; read_current_config \%current_config; @@ -2541,7 +2656,11 @@ sub run_config_bisect { if (!$found) { # try the other half doprint "Top half produced no set configs, trying bottom half\n"; + + # keep the top half off + @config_off_tmp = @tophalf; @tophalf = @start_list[$half + 1 .. $#start_list]; + create_config @tophalf; read_current_config \%current_config; foreach my $config (@tophalf) { @@ -2679,6 +2798,10 @@ sub config_bisect { $added_configs{$2} = $1; $config_list{$2} = $1; } + } elsif (/^# ((CONFIG\S*).*)/) { + # Keep these configs disabled + $config_set{$2} = $1; + $config_off{$2} = $1; } } close(IN); @@ -2701,6 +2824,8 @@ sub config_bisect { my %config_test; my $once = 0; + @config_off_tmp = (); + # Sometimes kconfig does weird things. We must make sure # that the config we autocreate has everything we need # to test, otherwise we may miss testing configs, or @@ -2719,6 +2844,18 @@ sub config_bisect { } } my $ret; + + if (defined($config_bisect_check) && $config_bisect_check) { + doprint " Checking to make sure bad config with min config fails\n"; + create_config keys %config_list; + $ret = run_config_bisect_test $config_bisect_type; + if ($ret) { + doprint " FAILED! Bad config with min config boots fine\n"; + return -1; + } + doprint " Bad config with min config fails as expected\n"; + } + do { $ret = run_config_bisect; } while (!$ret); @@ -3510,6 +3647,8 @@ for (my $i = 1; $i <= $opt{"NUM_TESTS"}; $i++) { $iteration = $i; + undef %force_config; + my $makecmd = set_test_option("MAKE_CMD", $i); # Load all the options into their mapped variable names @@ -3519,6 +3658,18 @@ for (my $i = 1; $i <= $opt{"NUM_TESTS"}; $i++) { $start_minconfig_defined = 1; + # The first test may override the PRE_KTEST option + if (defined($pre_ktest) && $i == 1) { + doprint "\n"; + run_command $pre_ktest; + } + + # Any test can override the POST_KTEST option + # The last test takes precedence. + if (defined($post_ktest)) { + $final_post_ktest = $post_ktest; + } + if (!defined($start_minconfig)) { $start_minconfig_defined = 0; $start_minconfig = $minconfig; @@ -3573,6 +3724,10 @@ for (my $i = 1; $i <= $opt{"NUM_TESTS"}; $i++) { doprint "\n\n"; doprint "RUNNING TEST $i of $opt{NUM_TESTS} with option $test_type $run_type$installme\n\n"; + if (defined($pre_test)) { + run_command $pre_test; + } + unlink $dmesg; unlink $buildlog; unlink $testlog; @@ -3638,6 +3793,10 @@ for (my $i = 1; $i <= $opt{"NUM_TESTS"}; $i++) { success $i; } +if (defined($final_post_ktest)) { + run_command $final_post_ktest; +} + if ($opt{"POWEROFF_ON_SUCCESS"}) { halt; } elsif ($opt{"REBOOT_ON_SUCCESS"} && !do_not_reboot && $reboot_success) { diff --git a/tools/testing/ktest/sample.conf b/tools/testing/ktest/sample.conf index cf362b3d1ec..de28a0a3b8f 100644 --- a/tools/testing/ktest/sample.conf +++ b/tools/testing/ktest/sample.conf @@ -376,6 +376,24 @@ # DEFAULTS # DEFAULTS SKIP +# If you want to execute some command before the first test runs +# you can set this option. Note, it can be set as a default option +# or an option in the first test case. All other test cases will +# ignore it. If both the default and first test have this option +# set, then the first test will take precedence. +# +# default (undefined) +#PRE_KTEST = ${SSH} ~/set_up_test + +# If you want to execute some command after all the tests have +# completed, you can set this option. Note, it can be set as a +# default or any test case can override it. If multiple test cases +# set this option, then the last test case that set it will take +# precedence +# +# default (undefined) +#POST_KTEST = ${SSH} ~/dismantle_test + # The default test type (default test) # The test types may be: # build - only build the kernel, do nothing else @@ -408,6 +426,14 @@ # (default "") #BUILD_OPTIONS = -j20 +# If you need to do some special handling before installing +# you can add a script with this option. +# The environment variable KERNEL_VERSION will be set to the +# kernel version that is used. +# +# default (undefined) +#PRE_INSTALL = ssh user@target rm -rf '/lib/modules/*-test*' + # If you need an initrd, you can add a script or code here to install # it. The environment variable KERNEL_VERSION will be set to the # kernel version that is used. Remember to add the initrd line @@ -426,6 +452,18 @@ # (default 0) #NO_INSTALL = 1 +# If there is a command that you want to run before the individual test +# case executes, then you can set this option +# +# default (undefined) +#PRE_TEST = ${SSH} reboot_to_special_kernel + +# If there is a command you want to run after the individual test case +# completes, then you can set this option. +# +# default (undefined) +#POST_TEST = cd ${BUILD_DIR}; git reset --hard + # If there is a script that you require to run before the build is done # you can specify it with PRE_BUILD. # @@ -657,6 +695,14 @@ # (default 60) #BISECT_SLEEP_TIME = 60 +# The max wait time (in seconds) for waiting for the console to finish. +# If for some reason, the console is outputting content without +# ever finishing, this will cause ktest to get stuck. This +# option is the max time ktest will wait for the monitor (console) +# to settle down before continuing. +# (default 1800) +#MAX_MONITOR_WAIT + # The time in between patch checks to sleep (in seconds) # (default 60) #PATCHCHECK_SLEEP_TIME = 60 @@ -1039,6 +1085,12 @@ # can specify it with CONFIG_BISECT_GOOD. Otherwise # the MIN_CONFIG is the base. # +# CONFIG_BISECT_CHECK (optional) +# Set this to 1 if you want to confirm that the config ktest +# generates (the bad config with the min config) is still bad. +# It may be that the min config fixes what broke the bad config +# and the test will not return a result. +# # Example: # TEST_START # TEST_TYPE = config_bisect diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile index a4162e15c25..85baf11e2ac 100644 --- a/tools/testing/selftests/Makefile +++ b/tools/testing/selftests/Makefile @@ -1,4 +1,4 @@ -TARGETS = breakpoints kcmp mqueue vm +TARGETS = breakpoints kcmp mqueue vm cpu-hotplug memory-hotplug all: for TARGET in $(TARGETS); do \ diff --git a/tools/testing/selftests/cpu-hotplug/Makefile b/tools/testing/selftests/cpu-hotplug/Makefile new file mode 100644 index 00000000000..7c9c20ff578 --- /dev/null +++ b/tools/testing/selftests/cpu-hotplug/Makefile @@ -0,0 +1,6 @@ +all: + +run_tests: + ./on-off-test.sh + +clean: diff --git a/tools/testing/selftests/cpu-hotplug/on-off-test.sh b/tools/testing/selftests/cpu-hotplug/on-off-test.sh new file mode 100644 index 00000000000..bdde7cf428b --- /dev/null +++ b/tools/testing/selftests/cpu-hotplug/on-off-test.sh @@ -0,0 +1,221 @@ +#!/bin/bash + +SYSFS= + +prerequisite() +{ + msg="skip all tests:" + + if [ $UID != 0 ]; then + echo $msg must be run as root >&2 + exit 0 + fi + + SYSFS=`mount -t sysfs | head -1 | awk '{ print $3 }'` + + if [ ! -d "$SYSFS" ]; then + echo $msg sysfs is not mounted >&2 + exit 0 + fi + + if ! ls $SYSFS/devices/system/cpu/cpu* > /dev/null 2>&1; then + echo $msg cpu hotplug is not supported >&2 + exit 0 + fi +} + +# +# list all hot-pluggable CPUs +# +hotpluggable_cpus() +{ + local state=${1:-.\*} + + for cpu in $SYSFS/devices/system/cpu/cpu*; do + if [ -f $cpu/online ] && grep -q $state $cpu/online; then + echo ${cpu##/*/cpu} + fi + done +} + +hotplaggable_offline_cpus() +{ + hotpluggable_cpus 0 +} + +hotpluggable_online_cpus() +{ + hotpluggable_cpus 1 +} + +cpu_is_online() +{ + grep -q 1 $SYSFS/devices/system/cpu/cpu$1/online +} + +cpu_is_offline() +{ + grep -q 0 $SYSFS/devices/system/cpu/cpu$1/online +} + +online_cpu() +{ + echo 1 > $SYSFS/devices/system/cpu/cpu$1/online +} + +offline_cpu() +{ + echo 0 > $SYSFS/devices/system/cpu/cpu$1/online +} + +online_cpu_expect_success() +{ + local cpu=$1 + + if ! online_cpu $cpu; then + echo $FUNCNAME $cpu: unexpected fail >&2 + elif ! cpu_is_online $cpu; then + echo $FUNCNAME $cpu: unexpected offline >&2 + fi +} + +online_cpu_expect_fail() +{ + local cpu=$1 + + if online_cpu $cpu 2> /dev/null; then + echo $FUNCNAME $cpu: unexpected success >&2 + elif ! cpu_is_offline $cpu; then + echo $FUNCNAME $cpu: unexpected online >&2 + fi +} + +offline_cpu_expect_success() +{ + local cpu=$1 + + if ! offline_cpu $cpu; then + echo $FUNCNAME $cpu: unexpected fail >&2 + elif ! cpu_is_offline $cpu; then + echo $FUNCNAME $cpu: unexpected offline >&2 + fi +} + +offline_cpu_expect_fail() +{ + local cpu=$1 + + if offline_cpu $cpu 2> /dev/null; then + echo $FUNCNAME $cpu: unexpected success >&2 + elif ! cpu_is_online $cpu; then + echo $FUNCNAME $cpu: unexpected offline >&2 + fi +} + +error=-12 +priority=0 + +while getopts e:hp: opt; do + case $opt in + e) + error=$OPTARG + ;; + h) + echo "Usage $0 [ -e errno ] [ -p notifier-priority ]" + exit + ;; + p) + priority=$OPTARG + ;; + esac +done + +if ! [ "$error" -ge -4095 -a "$error" -lt 0 ]; then + echo "error code must be -4095 <= errno < 0" >&2 + exit 1 +fi + +prerequisite + +# +# Online all hot-pluggable CPUs +# +for cpu in `hotplaggable_offline_cpus`; do + online_cpu_expect_success $cpu +done + +# +# Offline all hot-pluggable CPUs +# +for cpu in `hotpluggable_online_cpus`; do + offline_cpu_expect_success $cpu +done + +# +# Online all hot-pluggable CPUs again +# +for cpu in `hotplaggable_offline_cpus`; do + online_cpu_expect_success $cpu +done + +# +# Test with cpu notifier error injection +# + +DEBUGFS=`mount -t debugfs | head -1 | awk '{ print $3 }'` +NOTIFIER_ERR_INJECT_DIR=$DEBUGFS/notifier-error-inject/cpu + +prerequisite_extra() +{ + msg="skip extra tests:" + + /sbin/modprobe -q -r cpu-notifier-error-inject + /sbin/modprobe -q cpu-notifier-error-inject priority=$priority + + if [ ! -d "$DEBUGFS" ]; then + echo $msg debugfs is not mounted >&2 + exit 0 + fi + + if [ ! -d $NOTIFIER_ERR_INJECT_DIR ]; then + echo $msg cpu-notifier-error-inject module is not available >&2 + exit 0 + fi +} + +prerequisite_extra + +# +# Offline all hot-pluggable CPUs +# +echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_DOWN_PREPARE/error +for cpu in `hotpluggable_online_cpus`; do + offline_cpu_expect_success $cpu +done + +# +# Test CPU hot-add error handling (offline => online) +# +echo $error > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_UP_PREPARE/error +for cpu in `hotplaggable_offline_cpus`; do + online_cpu_expect_fail $cpu +done + +# +# Online all hot-pluggable CPUs +# +echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_UP_PREPARE/error +for cpu in `hotplaggable_offline_cpus`; do + online_cpu_expect_success $cpu +done + +# +# Test CPU hot-remove error handling (online => offline) +# +echo $error > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_DOWN_PREPARE/error +for cpu in `hotpluggable_online_cpus`; do + offline_cpu_expect_fail $cpu +done + +echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/CPU_DOWN_PREPARE/error +/sbin/modprobe -q -r cpu-notifier-error-inject diff --git a/tools/testing/selftests/memory-hotplug/Makefile b/tools/testing/selftests/memory-hotplug/Makefile new file mode 100644 index 00000000000..7c9c20ff578 --- /dev/null +++ b/tools/testing/selftests/memory-hotplug/Makefile @@ -0,0 +1,6 @@ +all: + +run_tests: + ./on-off-test.sh + +clean: diff --git a/tools/testing/selftests/memory-hotplug/on-off-test.sh b/tools/testing/selftests/memory-hotplug/on-off-test.sh new file mode 100644 index 00000000000..a2816f63154 --- /dev/null +++ b/tools/testing/selftests/memory-hotplug/on-off-test.sh @@ -0,0 +1,230 @@ +#!/bin/bash + +SYSFS= + +prerequisite() +{ + msg="skip all tests:" + + if [ $UID != 0 ]; then + echo $msg must be run as root >&2 + exit 0 + fi + + SYSFS=`mount -t sysfs | head -1 | awk '{ print $3 }'` + + if [ ! -d "$SYSFS" ]; then + echo $msg sysfs is not mounted >&2 + exit 0 + fi + + if ! ls $SYSFS/devices/system/memory/memory* > /dev/null 2>&1; then + echo $msg memory hotplug is not supported >&2 + exit 0 + fi +} + +# +# list all hot-pluggable memory +# +hotpluggable_memory() +{ + local state=${1:-.\*} + + for memory in $SYSFS/devices/system/memory/memory*; do + if grep -q 1 $memory/removable && + grep -q $state $memory/state; then + echo ${memory##/*/memory} + fi + done +} + +hotplaggable_offline_memory() +{ + hotpluggable_memory offline +} + +hotpluggable_online_memory() +{ + hotpluggable_memory online +} + +memory_is_online() +{ + grep -q online $SYSFS/devices/system/memory/memory$1/state +} + +memory_is_offline() +{ + grep -q offline $SYSFS/devices/system/memory/memory$1/state +} + +online_memory() +{ + echo online > $SYSFS/devices/system/memory/memory$1/state +} + +offline_memory() +{ + echo offline > $SYSFS/devices/system/memory/memory$1/state +} + +online_memory_expect_success() +{ + local memory=$1 + + if ! online_memory $memory; then + echo $FUNCNAME $memory: unexpected fail >&2 + elif ! memory_is_online $memory; then + echo $FUNCNAME $memory: unexpected offline >&2 + fi +} + +online_memory_expect_fail() +{ + local memory=$1 + + if online_memory $memory 2> /dev/null; then + echo $FUNCNAME $memory: unexpected success >&2 + elif ! memory_is_offline $memory; then + echo $FUNCNAME $memory: unexpected online >&2 + fi +} + +offline_memory_expect_success() +{ + local memory=$1 + + if ! offline_memory $memory; then + echo $FUNCNAME $memory: unexpected fail >&2 + elif ! memory_is_offline $memory; then + echo $FUNCNAME $memory: unexpected offline >&2 + fi +} + +offline_memory_expect_fail() +{ + local memory=$1 + + if offline_memory $memory 2> /dev/null; then + echo $FUNCNAME $memory: unexpected success >&2 + elif ! memory_is_online $memory; then + echo $FUNCNAME $memory: unexpected offline >&2 + fi +} + +error=-12 +priority=0 +ratio=10 + +while getopts e:hp:r: opt; do + case $opt in + e) + error=$OPTARG + ;; + h) + echo "Usage $0 [ -e errno ] [ -p notifier-priority ] [ -r percent-of-memory-to-offline ]" + exit + ;; + p) + priority=$OPTARG + ;; + r) + ratio=$OPTARG + ;; + esac +done + +if ! [ "$error" -ge -4095 -a "$error" -lt 0 ]; then + echo "error code must be -4095 <= errno < 0" >&2 + exit 1 +fi + +prerequisite + +# +# Online all hot-pluggable memory +# +for memory in `hotplaggable_offline_memory`; do + online_memory_expect_success $memory +done + +# +# Offline $ratio percent of hot-pluggable memory +# +for memory in `hotpluggable_online_memory`; do + if [ $((RANDOM % 100)) -lt $ratio ]; then + offline_memory_expect_success $memory + fi +done + +# +# Online all hot-pluggable memory again +# +for memory in `hotplaggable_offline_memory`; do + online_memory_expect_success $memory +done + +# +# Test with memory notifier error injection +# + +DEBUGFS=`mount -t debugfs | head -1 | awk '{ print $3 }'` +NOTIFIER_ERR_INJECT_DIR=$DEBUGFS/notifier-error-inject/memory + +prerequisite_extra() +{ + msg="skip extra tests:" + + /sbin/modprobe -q -r memory-notifier-error-inject + /sbin/modprobe -q memory-notifier-error-inject priority=$priority + + if [ ! -d "$DEBUGFS" ]; then + echo $msg debugfs is not mounted >&2 + exit 0 + fi + + if [ ! -d $NOTIFIER_ERR_INJECT_DIR ]; then + echo $msg memory-notifier-error-inject module is not available >&2 + exit 0 + fi +} + +prerequisite_extra + +# +# Offline $ratio percent of hot-pluggable memory +# +echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_OFFLINE/error +for memory in `hotpluggable_online_memory`; do + if [ $((RANDOM % 100)) -lt $ratio ]; then + offline_memory_expect_success $memory + fi +done + +# +# Test memory hot-add error handling (offline => online) +# +echo $error > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_ONLINE/error +for memory in `hotplaggable_offline_memory`; do + online_memory_expect_fail $memory +done + +# +# Online all hot-pluggable memory +# +echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_ONLINE/error +for memory in `hotplaggable_offline_memory`; do + online_memory_expect_success $memory +done + +# +# Test memory hot-remove error handling (online => offline) +# +echo $error > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_OFFLINE/error +for memory in `hotpluggable_online_memory`; do + offline_memory_expect_fail $memory +done + +echo 0 > $NOTIFIER_ERR_INJECT_DIR/actions/MEM_GOING_OFFLINE/error +/sbin/modprobe -q -r memory-notifier-error-inject diff --git a/tools/usb/testusb.c b/tools/usb/testusb.c index 82d7c590c02..b0adb2710c0 100644 --- a/tools/usb/testusb.c +++ b/tools/usb/testusb.c @@ -425,7 +425,7 @@ int main (int argc, char **argv) /* for easy use when hotplugging */ device = getenv ("DEVICE"); - while ((c = getopt (argc, argv, "D:aA:c:g:hns:t:v:")) != EOF) + while ((c = getopt (argc, argv, "D:aA:c:g:hlns:t:v:")) != EOF) switch (c) { case 'D': /* device, if only one */ device = optarg; @@ -468,10 +468,21 @@ int main (int argc, char **argv) case 'h': default: usage: - fprintf (stderr, "usage: %s [-n] [-D dev | -a | -A usbfs-dir]\n" - "\t[-c iterations] [-t testnum]\n" - "\t[-s packetsize] [-g sglen] [-v vary]\n", - argv [0]); + fprintf (stderr, + "usage: %s [options]\n" + "Options:\n" + "\t-D dev only test specific device\n" + "\t-A usbfs-dir\n" + "\t-a test all recognized devices\n" + "\t-l loop forever(for stress test)\n" + "\t-t testnum only run specified case\n" + "\t-n no test running, show devices to be tested\n" + "Case arguments:\n" + "\t-c iterations default 1000\n" + "\t-s packetsize default 512\n" + "\t-g sglen default 32\n" + "\t-v vary default 512\n", + argv[0]); return 1; } if (optind != argc) diff --git a/tools/vm/slabinfo.c b/tools/vm/slabinfo.c index 164cbcf6110..808d5a9d5dc 100644 --- a/tools/vm/slabinfo.c +++ b/tools/vm/slabinfo.c @@ -437,34 +437,34 @@ static void slab_stats(struct slabinfo *s) printf("Fastpath %8lu %8lu %3lu %3lu\n", s->alloc_fastpath, s->free_fastpath, s->alloc_fastpath * 100 / total_alloc, - s->free_fastpath * 100 / total_free); + total_free ? s->free_fastpath * 100 / total_free : 0); printf("Slowpath %8lu %8lu %3lu %3lu\n", total_alloc - s->alloc_fastpath, s->free_slowpath, (total_alloc - s->alloc_fastpath) * 100 / total_alloc, - s->free_slowpath * 100 / total_free); + total_free ? s->free_slowpath * 100 / total_free : 0); printf("Page Alloc %8lu %8lu %3lu %3lu\n", s->alloc_slab, s->free_slab, s->alloc_slab * 100 / total_alloc, - s->free_slab * 100 / total_free); + total_free ? s->free_slab * 100 / total_free : 0); printf("Add partial %8lu %8lu %3lu %3lu\n", s->deactivate_to_head + s->deactivate_to_tail, s->free_add_partial, (s->deactivate_to_head + s->deactivate_to_tail) * 100 / total_alloc, - s->free_add_partial * 100 / total_free); + total_free ? s->free_add_partial * 100 / total_free : 0); printf("Remove partial %8lu %8lu %3lu %3lu\n", s->alloc_from_partial, s->free_remove_partial, s->alloc_from_partial * 100 / total_alloc, - s->free_remove_partial * 100 / total_free); + total_free ? s->free_remove_partial * 100 / total_free : 0); printf("Cpu partial list %8lu %8lu %3lu %3lu\n", s->cpu_partial_alloc, s->cpu_partial_free, s->cpu_partial_alloc * 100 / total_alloc, - s->cpu_partial_free * 100 / total_free); + total_free ? s->cpu_partial_free * 100 / total_free : 0); printf("RemoteObj/SlabFrozen %8lu %8lu %3lu %3lu\n", s->deactivate_remote_frees, s->free_frozen, s->deactivate_remote_frees * 100 / total_alloc, - s->free_frozen * 100 / total_free); + total_free ? s->free_frozen * 100 / total_free : 0); printf("Total %8lu %8lu\n\n", total_alloc, total_free); |