Skip to contentSkip to conversation

perf fixes

GIT PULL Ingo Molnar 1 participant 1 message Dec 18, 2009

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

Linus,

Please pull the latest perf-fixes-for-linus git tree from:

git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git perf-fixes-for-linus

Thanks,

Ingo

------------------>
Arnaldo Carvalho de Melo (1):
perf session: Make events_stats u64 to avoid overflow on 32-bit arches

Frederic Weisbecker (3):
perf events, x86/stacktrace: Make stack walking optional
perf events, x86/stacktrace: Fix performance/softlockup by providing a special frame pointer-only stack walker
hw-breakpoints: Fix hardware breakpoints -> perf events dependency

Masami Hiramatsu (4):
perf probe: Fix libdwarf include path for Debian perf probe: Check whether debugfs path is correct
kprobe-tracer: Check new event/group name perf probe: Check new event name

Peter Zijlstra (1):
perf events: Dont report side-band events on each cpu for per-task-per-cpu events

Robert P. J. Day (1):
perf events: Remove unused perf_counter.h header file

 arch/Kconfig                      |    4 +-
 arch/x86/Kconfig                  |    2 +
 arch/x86/include/asm/stacktrace.h |   24 ++
 arch/x86/kernel/cpu/perf_event.c  |    1 +
 arch/x86/kernel/dumpstack.c       |   33 +++-
 arch/x86/kernel/dumpstack.h       |    6 -
 arch/x86/kernel/dumpstack_32.c    |    2 +-
 arch/x86/kernel/dumpstack_64.c    |    4 +-
 arch/x86/kernel/stacktrace.c      |   18 +-
 arch/x86/oprofile/backtrace.c     |    9 +-
 include/linux/perf_counter.h      |  444 -------------------------------------
 kernel/perf_event.c               |   32 ++--
 kernel/trace/trace_kprobe.c       |   31 ++-
 kernel/trace/trace_sysprof.c      |    1 +
 tools/perf/Makefile               |    3 +-
 tools/perf/builtin-probe.c        |    4 +
 tools/perf/builtin-report.c       |    2 +-
 tools/perf/util/event.h           |    4 +-
 tools/perf/util/probe-event.c     |   15 ++
 tools/perf/util/probe-finder.h    |   59 +++---
 20 files changed, 175 insertions(+), 523 deletions(-)
 delete mode 100644 include/linux/perf_counter.h
arch/Kconfig +1 -3
index d828758..9d055b4 100644
@@ -135,9 +135,7 @@ config HAVE_DEFAULT_NO_SPIN_MUTEXES
 
 config HAVE_HW_BREAKPOINT
 	bool
-	depends on HAVE_PERF_EVENTS
-	select ANON_INODES
-	select PERF_EVENTS
+	depends on PERF_EVENTS
 
 config HAVE_USER_RETURN_NOTIFIER
 	bool
arch/x86/Kconfig +2 -0
index 3b2a5ac..55298e8 100644
@@ -50,6 +50,8 @@ config X86
 	select HAVE_KERNEL_BZIP2
 	select HAVE_KERNEL_LZMA
 	select HAVE_HW_BREAKPOINT
+	select PERF_EVENTS
+	select ANON_INODES
 	select HAVE_ARCH_KMEMCHECK
 	select HAVE_USER_RETURN_NOTIFIER
 
arch/x86/include/asm/stacktrace.h +24 -0
index cf86a5e..35e8912 100644
@@ -5,6 +5,29 @@ extern int kstack_depth_to_print;
 
 int x86_is_stack_id(int id, char *name);
 
+struct thread_info;
+struct stacktrace_ops;
+
+typedef unsigned long (*walk_stack_t)(struct thread_info *tinfo,
+				      unsigned long *stack,
+				      unsigned long bp,
+				      const struct stacktrace_ops *ops,
+				      void *data,
+				      unsigned long *end,
+				      int *graph);
+
+extern unsigned long
+print_context_stack(struct thread_info *tinfo,
+		    unsigned long *stack, unsigned long bp,
+		    const struct stacktrace_ops *ops, void *data,
+		    unsigned long *end, int *graph);
+
+extern unsigned long
+print_context_stack_bp(struct thread_info *tinfo,
+		       unsigned long *stack, unsigned long bp,
+		       const struct stacktrace_ops *ops, void *data,
+		       unsigned long *end, int *graph);
+
 /* Generic stack tracer with callbacks */
 
 struct stacktrace_ops {
@@ -14,6 +37,7 @@ struct stacktrace_ops {
 	void (*address)(void *data, unsigned long address, int reliable);
 	/* On negative return stop dumping */
 	int (*stack)(void *data, char *name);
+	walk_stack_t	walk_stack;
 };
 
 void dump_trace(struct task_struct *tsk, struct pt_regs *regs,
arch/x86/kernel/cpu/perf_event.c +1 -0
index 45506d5..c223b7e 100644
@@ -2336,6 +2336,7 @@ static const struct stacktrace_ops backtrace_ops = {
 	.warning_symbol		= backtrace_warning_symbol,
 	.stack			= backtrace_stack,
 	.address		= backtrace_address,
+	.walk_stack		= print_context_stack_bp,
 };
 
 #include "../dumpstack.h"
arch/x86/kernel/dumpstack.c +29 -4
index 0a0aa1c..c56bc28 100644
@@ -109,6 +109,30 @@ print_context_stack(struct thread_info *tinfo,
 	}
 	return bp;
 }
+EXPORT_SYMBOL_GPL(print_context_stack);
+
+unsigned long
+print_context_stack_bp(struct thread_info *tinfo,
+		       unsigned long *stack, unsigned long bp,
+		       const struct stacktrace_ops *ops, void *data,
+		       unsigned long *end, int *graph)
+{
+	struct stack_frame *frame = (struct stack_frame *)bp;
+	unsigned long *ret_addr = &frame->return_address;
+
+	while (valid_stack_ptr(tinfo, ret_addr, sizeof(*ret_addr), end)) {
+		unsigned long addr = *ret_addr;
+
+		if (__kernel_text_address(addr)) {
+			ops->address(data, addr, 1);
+			frame = frame->next_frame;
+			ret_addr = &frame->return_address;
+			print_ftrace_graph_addr(addr, data, ops, tinfo, graph);
+		}
+	}
+	return (unsigned long)frame;
+}
+EXPORT_SYMBOL_GPL(print_context_stack_bp);
 
 
 static void
@@ -141,10 +165,11 @@ static void print_trace_address(void *data, unsigned long addr, int reliable)
 }
 
 static const struct stacktrace_ops print_trace_ops = {
-	.warning = print_trace_warning,
-	.warning_symbol = print_trace_warning_symbol,
-	.stack = print_trace_stack,
-	.address = print_trace_address,
+	.warning		= print_trace_warning,
+	.warning_symbol		= print_trace_warning_symbol,
+	.stack			= print_trace_stack,
+	.address		= print_trace_address,
+	.walk_stack		= print_context_stack,
 };
 
 void
arch/x86/kernel/dumpstack.h +0 -6
index 81086c2..4fd1420 100644
@@ -14,12 +14,6 @@
 #define get_bp(bp) asm("movq %%rbp, %0" : "=r" (bp) :)
 #endif
 
-extern unsigned long
-print_context_stack(struct thread_info *tinfo,
-		unsigned long *stack, unsigned long bp,
-		const struct stacktrace_ops *ops, void *data,
-		unsigned long *end, int *graph);
-
 extern void
 show_trace_log_lvl(struct task_struct *task, struct pt_regs *regs,
 		unsigned long *stack, unsigned long bp, char *log_lvl);
arch/x86/kernel/dumpstack_32.c +1 -1
index e0ed4c7..ae775ca 100644
@@ -58,7 +58,7 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs,
 
 		context = (struct thread_info *)
 			((unsigned long)stack & (~(THREAD_SIZE - 1)));
-		bp = print_context_stack(context, stack, bp, ops, data, NULL, &graph);
+		bp = ops->walk_stack(context, stack, bp, ops, data, NULL, &graph);
 
 		stack = (unsigned long *)context->previous_esp;
 		if (!stack)
arch/x86/kernel/dumpstack_64.c +2 -2
index b13af53..0ad9597 100644
@@ -188,8 +188,8 @@ void dump_trace(struct task_struct *task, struct pt_regs *regs,
 			if (ops->stack(data, id) < 0)
 				break;
 
-			bp = print_context_stack(tinfo, stack, bp, ops,
-						 data, estack_end, &graph);
+			bp = ops->walk_stack(tinfo, stack, bp, ops,
+					     data, estack_end, &graph);
 			ops->stack(data, "<EOE>");
 			/*
 			 * We link to the next stack via the
arch/x86/kernel/stacktrace.c +10 -8
index c3eb207..922eefb 100644
@@ -53,17 +53,19 @@ save_stack_address_nosched(void *data, unsigned long addr, int reliable)
 }
 
 static const struct stacktrace_ops save_stack_ops = {
-	.warning = save_stack_warning,
-	.warning_symbol = save_stack_warning_symbol,
-	.stack = save_stack_stack,
-	.address = save_stack_address,
+	.warning	= save_stack_warning,
+	.warning_symbol	= save_stack_warning_symbol,
+	.stack		= save_stack_stack,
+	.address	= save_stack_address,
+	.walk_stack	= print_context_stack,
 };
 
 static const struct stacktrace_ops save_stack_ops_nosched = {
-	.warning = save_stack_warning,
-	.warning_symbol = save_stack_warning_symbol,
-	.stack = save_stack_stack,
-	.address = save_stack_address_nosched,
+	.warning	= save_stack_warning,
+	.warning_symbol	= save_stack_warning_symbol,
+	.stack		= save_stack_stack,
+	.address	= save_stack_address_nosched,
+	.walk_stack	= print_context_stack,
 };
 
 /*
arch/x86/oprofile/backtrace.c +5 -4
index 044897b..3855096 100644
@@ -41,10 +41,11 @@ static void backtrace_address(void *data, unsigned long addr, int reliable)
 }
 
 static struct stacktrace_ops backtrace_ops = {
-	.warning = backtrace_warning,
-	.warning_symbol = backtrace_warning_symbol,
-	.stack = backtrace_stack,
-	.address = backtrace_address,
+	.warning	= backtrace_warning,
+	.warning_symbol	= backtrace_warning_symbol,
+	.stack		= backtrace_stack,
+	.address	= backtrace_address,
+	.walk_stack	= print_context_stack,
 };
 
 struct frame_head {
include/linux/perf_counter.h deleted +0 -444

444 lines not on this page · raw message

kernel/perf_event.c +18 -14
index 8ab8698..03cc061 100644
@@ -1381,6 +1381,9 @@ static void perf_ctx_adjust_freq(struct perf_event_context *ctx)
 		if (event->state != PERF_EVENT_STATE_ACTIVE)
 			continue;
 
+		if (event->cpu != -1 && event->cpu != smp_processor_id())
+			continue;
+
 		hwc = &event->hw;
 
 		interrupts = hwc->interrupts;
@@ -3265,6 +3268,9 @@ static void perf_event_task_output(struct perf_event *event,
 
 static int perf_event_task_match(struct perf_event *event)
 {
+	if (event->cpu != -1 && event->cpu != smp_processor_id())
+		return 0;
+
 	if (event->attr.comm || event->attr.mmap || event->attr.task)
 		return 1;
 
@@ -3290,12 +3296,11 @@ static void perf_event_task_event(struct perf_task_event *task_event)
 	rcu_read_lock();
 	cpuctx = &get_cpu_var(perf_cpu_context);
 	perf_event_task_ctx(&cpuctx->ctx, task_event);
-	put_cpu_var(perf_cpu_context);
-
 	if (!ctx)
 		ctx = rcu_dereference(task_event->task->perf_event_ctxp);
 	if (ctx)
 		perf_event_task_ctx(ctx, task_event);
+	put_cpu_var(perf_cpu_context);
 	rcu_read_unlock();
 }
 
@@ -3372,6 +3377,9 @@ static void perf_event_comm_output(struct perf_event *event,
 
 static int perf_event_comm_match(struct perf_event *event)
 {
+	if (event->cpu != -1 && event->cpu != smp_processor_id())
+		return 0;
+
 	if (event->attr.comm)
 		return 1;
 
@@ -3408,15 +3416,10 @@ static void perf_event_comm_event(struct perf_comm_event *comm_event)
 	rcu_read_lock();
 	cpuctx = &get_cpu_var(perf_cpu_context);
 	perf_event_comm_ctx(&cpuctx->ctx, comm_event);
-	put_cpu_var(perf_cpu_context);
-
-	/*
-	 * doesn't really matter which of the child contexts the
-	 * events ends up in.
-	 */
 	ctx = rcu_dereference(current->perf_event_ctxp);
 	if (ctx)
 		perf_event_comm_ctx(ctx, comm_event);
+	put_cpu_var(perf_cpu_context);
 	rcu_read_unlock();
 }
 
@@ -3491,6 +3494,9 @@ static void perf_event_mmap_output(struct perf_event *event,
 static int perf_event_mmap_match(struct perf_event *event,
 				   struct perf_mmap_event *mmap_event)
 {
+	if (event->cpu != -1 && event->cpu != smp_processor_id())
+		return 0;
+
 	if (event->attr.mmap)
 		return 1;
 
@@ -3564,15 +3570,10 @@ got_name:
 	rcu_read_lock();
 	cpuctx = &get_cpu_var(perf_cpu_context);
 	perf_event_mmap_ctx(&cpuctx->ctx, mmap_event);
-	put_cpu_var(perf_cpu_context);
-
-	/*
-	 * doesn't really matter which of the child contexts the
-	 * events ends up in.
-	 */
 	ctx = rcu_dereference(current->perf_event_ctxp);
 	if (ctx)
 		perf_event_mmap_ctx(ctx, mmap_event);
+	put_cpu_var(perf_cpu_context);
 	rcu_read_unlock();
 
 	kfree(buf);
@@ -3863,6 +3864,9 @@ static int perf_swevent_match(struct perf_event *event,
 				struct perf_sample_data *data,
 				struct pt_regs *regs)
 {
+	if (event->cpu != -1 && event->cpu != smp_processor_id())
+		return 0;
+
 	if (!perf_swevent_is_counting(event))
 		return 0;
 
kernel/trace/trace_kprobe.c +25 -6
index 7ecab06..375f81a 100644
@@ -282,6 +282,18 @@ static int kprobe_dispatcher(struct kprobe *kp, struct pt_regs *regs);
 static int kretprobe_dispatcher(struct kretprobe_instance *ri,
 				struct pt_regs *regs);
 
+/* Check the name is good for event/group */
+static int check_event_name(const char *name)
+{
+	if (!isalpha(*name) && *name != '_')
+		return 0;
+	while (*++name != '\0') {
+		if (!isalpha(*name) && !isdigit(*name) && *name != '_')
+			return 0;
+	}
+	return 1;
+}
+
 /*
  * Allocate new trace_probe and initialize it (including kprobes).
  */
@@ -293,10 +305,11 @@ static struct trace_probe *alloc_trace_probe(const char *group,
 					     int nargs, int is_return)
 {
 	struct trace_probe *tp;
+	int ret = -ENOMEM;
 
 	tp = kzalloc(SIZEOF_TRACE_PROBE(nargs), GFP_KERNEL);
 	if (!tp)
-		return ERR_PTR(-ENOMEM);
+		return ERR_PTR(ret);
 
 	if (symbol) {
 		tp->symbol = kstrdup(symbol, GFP_KERNEL);
@@ -312,14 +325,20 @@ static struct trace_probe *alloc_trace_probe(const char *group,
 	else
 		tp->rp.kp.pre_handler = kprobe_dispatcher;
 
-	if (!event)
+	if (!event || !check_event_name(event)) {
+		ret = -EINVAL;
 		goto error;
+	}
+
 	tp->call.name = kstrdup(event, GFP_KERNEL);
 	if (!tp->call.name)
 		goto error;
 
-	if (!group)
+	if (!group || !check_event_name(group)) {
+		ret = -EINVAL;
 		goto error;
+	}
+
 	tp->call.system = kstrdup(group, GFP_KERNEL);
 	if (!tp->call.system)
 		goto error;
@@ -330,7 +349,7 @@ error:
 	kfree(tp->call.name);
 	kfree(tp->symbol);
 	kfree(tp);
-	return ERR_PTR(-ENOMEM);
+	return ERR_PTR(ret);
 }
 
 static void free_probe_arg(struct probe_arg *arg)
@@ -695,10 +714,10 @@ static int create_trace_probe(int argc, char **argv)
 	if (!event) {
 		/* Make a new event name */
 		if (symbol)
-			snprintf(buf, MAX_EVENT_NAME_LEN, "%c@%s%+ld",
+			snprintf(buf, MAX_EVENT_NAME_LEN, "%c_%s_%ld",
 				 is_return ? 'r' : 'p', symbol, offset);
 		else
-			snprintf(buf, MAX_EVENT_NAME_LEN, "%c@0x%p",
+			snprintf(buf, MAX_EVENT_NAME_LEN, "%c_0x%p",
 				 is_return ? 'r' : 'p', addr);
 		event = buf;
 	}
kernel/trace/trace_sysprof.c +1 -0
index f669396..a7974a5 100644
@@ -93,6 +93,7 @@ static const struct stacktrace_ops backtrace_ops = {
 	.warning_symbol		= backtrace_warning_symbol,
 	.stack			= backtrace_stack,
 	.address		= backtrace_address,
+	.walk_stack		= print_context_stack,
 };
 
 static int
tools/perf/Makefile +2 -1
index 7814dbb..4390d22 100644
@@ -487,10 +487,11 @@ else
 	msg := $(error No libelf.h/libelf found, please install libelf-dev/elfutils-libelf-devel and glibc-dev[el]);
 endif
 
-ifneq ($(shell sh -c "(echo '\#include <libdwarf/dwarf.h>'; echo '\#include <libdwarf/libdwarf.h>'; echo 'int main(void) { Dwarf_Debug dbg; Dwarf_Error err; Dwarf_Ranges *rng; dwarf_init(0, DW_DLC_READ, 0, 0, &dbg, &err); dwarf_get_ranges(dbg, 0, &rng, 0, 0, &err); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -ldwarf -lelf -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y)
+ifneq ($(shell sh -c "(echo '\#ifndef _MIPS_SZLONG'; echo '\#define _MIPS_SZLONG 0'; echo '\#endif'; echo '\#include <dwarf.h>'; echo '\#include <libdwarf.h>'; echo 'int main(void) { Dwarf_Debug dbg; Dwarf_Error err; Dwarf_Ranges *rng; dwarf_init(0, DW_DLC_READ, 0, 0, &dbg, &err); dwarf_get_ranges(dbg, 0, &rng, 0, 0, &err); return (long)dbg; }') | $(CC) -x c - $(ALL_CFLAGS) -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -I/usr/include/libdwarf -ldwarf -lelf -o /dev/null $(ALL_LDFLAGS) $(EXTLIBS) "$(QUIET_STDERR)" && echo y"), y)
 	msg := $(warning No libdwarf.h found or old libdwarf.h found, disables dwarf support. Please install libdwarf-dev/libdwarf-devel >= 20081231);
 	BASIC_CFLAGS += -DNO_LIBDWARF
 else
+	BASIC_CFLAGS += -I/usr/include/libdwarf
 	EXTLIBS += -lelf -ldwarf
 	LIB_OBJS += util/probe-finder.o
 endif
tools/perf/builtin-probe.c +4 -0
index 7e741f5..c1e6774 100644
@@ -38,6 +38,7 @@
 #include "util/strlist.h"
 #include "util/event.h"
 #include "util/debug.h"
+#include "util/debugfs.h"
 #include "util/symbol.h"
 #include "util/thread.h"
 #include "util/session.h"
@@ -205,6 +206,9 @@ int cmd_probe(int argc, const char **argv, const char *prefix __used)
 	if ((!session.nr_probe && !session.dellist && !session.list_events))
 		usage_with_options(probe_usage, options);
 
+	if (debugfs_valid_mountpoint(debugfs_path) < 0)
+		die("Failed to find debugfs path.");
+
 	if (session.list_events) {
 		if (session.nr_probe != 0 || session.dellist) {
 			pr_warning("  Error: Don't use --list with"
tools/perf/builtin-report.c +1 -1
index e50a6b1..5c2ab53 100644
@@ -224,7 +224,7 @@ static int __cmd_report(void)
 
 	perf_session__collapse_resort(session);
 	perf_session__output_resort(session, session->events_stats.total);
-	fprintf(stdout, "# Samples: %ld\n#\n", session->events_stats.total);
+	fprintf(stdout, "# Samples: %Ld\n#\n", session->events_stats.total);
 	perf_session__fprintf_hists(session, NULL, false, stdout);
 	if (sort_order == default_sort_order &&
 	    parent_pattern == default_parent_pattern)
tools/perf/util/event.h +2 -2
index 8027309..690a96d 100644
@@ -95,8 +95,8 @@ typedef union event_union {
 } event_t;
 
 struct events_stats {
-	unsigned long total;
-	unsigned long lost;
+	u64 total;
+	u64 lost;
 };
 
 void event__print_totals(void);
tools/perf/util/probe-event.c +15 -0
index 2ca6215..29465d4 100644
@@ -62,6 +62,18 @@ static int e_snprintf(char *str, size_t size, const char *format, ...)
 	return ret;
 }
 
+/* Check the name is good for event/group */
+static bool check_event_name(const char *name)
+{
+	if (!isalpha(*name) && *name != '_')
+		return false;
+	while (*++name != '\0') {
+		if (!isalpha(*name) && !isdigit(*name) && *name != '_')
+			return false;
+	}
+	return true;
+}
+
 /* Parse probepoint definition. */
 static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp)
 {
@@ -82,6 +94,9 @@ static void parse_perf_probe_probepoint(char *arg, struct probe_point *pp)
 		ptr = strchr(arg, ':');
 		if (ptr)	/* Group name is not supported yet. */
 			semantic_error("Group name is not supported yet.");
+		if (!check_event_name(arg))
+			semantic_error("%s is bad for event name -it must "
+				       "follow C symbol-naming rule.", arg);
 		pp->event = strdup(arg);
 		arg = tmp;
 	}
tools/perf/util/probe-finder.h +32 -27

89 lines not on this page · raw message