Merge branch 'perf/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux-2.6 into perf/urgent
diff --git a/arch/x86/kernel/cpu/perf_event_intel.c b/arch/x86/kernel/cpu/perf_event_intel.c
index 214ac86..d8d86d0 100644
--- a/arch/x86/kernel/cpu/perf_event_intel.c
+++ b/arch/x86/kernel/cpu/perf_event_intel.c
@@ -491,33 +491,78 @@
* Intel Errata AAP53 (model 30)
* Intel Errata BD53 (model 44)
*
- * These chips need to be 'reset' when adding counters by programming
- * the magic three (non counting) events 0x4300D2, 0x4300B1 and 0x4300B5
- * either in sequence on the same PMC or on different PMCs.
+ * The official story:
+ * These chips need to be 'reset' when adding counters by programming the
+ * magic three (non-counting) events 0x4300B5, 0x4300D2, and 0x4300B1 either
+ * in sequence on the same PMC or on different PMCs.
+ *
+ * In practise it appears some of these events do in fact count, and
+ * we need to programm all 4 events.
*/
+static void intel_pmu_nhm_workaround(void)
+{
+ struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
+ static const unsigned long nhm_magic[4] = {
+ 0x4300B5,
+ 0x4300D2,
+ 0x4300B1,
+ 0x4300B1
+ };
+ struct perf_event *event;
+ int i;
+
+ /*
+ * The Errata requires below steps:
+ * 1) Clear MSR_IA32_PEBS_ENABLE and MSR_CORE_PERF_GLOBAL_CTRL;
+ * 2) Configure 4 PERFEVTSELx with the magic events and clear
+ * the corresponding PMCx;
+ * 3) set bit0~bit3 of MSR_CORE_PERF_GLOBAL_CTRL;
+ * 4) Clear MSR_CORE_PERF_GLOBAL_CTRL;
+ * 5) Clear 4 pairs of ERFEVTSELx and PMCx;
+ */
+
+ /*
+ * The real steps we choose are a little different from above.
+ * A) To reduce MSR operations, we don't run step 1) as they
+ * are already cleared before this function is called;
+ * B) Call x86_perf_event_update to save PMCx before configuring
+ * PERFEVTSELx with magic number;
+ * C) With step 5), we do clear only when the PERFEVTSELx is
+ * not used currently.
+ * D) Call x86_perf_event_set_period to restore PMCx;
+ */
+
+ /* We always operate 4 pairs of PERF Counters */
+ for (i = 0; i < 4; i++) {
+ event = cpuc->events[i];
+ if (event)
+ x86_perf_event_update(event);
+ }
+
+ for (i = 0; i < 4; i++) {
+ wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + i, nhm_magic[i]);
+ wrmsrl(MSR_ARCH_PERFMON_PERFCTR0 + i, 0x0);
+ }
+
+ wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0xf);
+ wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0x0);
+
+ for (i = 0; i < 4; i++) {
+ event = cpuc->events[i];
+
+ if (event) {
+ x86_perf_event_set_period(event);
+ __x86_pmu_enable_event(&event->hw,
+ ARCH_PERFMON_EVENTSEL_ENABLE);
+ } else
+ wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + i, 0x0);
+ }
+}
+
static void intel_pmu_nhm_enable_all(int added)
{
- if (added) {
- struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
- int i;
-
- wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + 0, 0x4300D2);
- wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + 1, 0x4300B1);
- wrmsrl(MSR_ARCH_PERFMON_EVENTSEL0 + 2, 0x4300B5);
-
- wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0x3);
- wrmsrl(MSR_CORE_PERF_GLOBAL_CTRL, 0x0);
-
- for (i = 0; i < 3; i++) {
- struct perf_event *event = cpuc->events[i];
-
- if (!event)
- continue;
-
- __x86_pmu_enable_event(&event->hw,
- ARCH_PERFMON_EVENTSEL_ENABLE);
- }
- }
+ if (added)
+ intel_pmu_nhm_workaround();
intel_pmu_enable_all(added);
}
diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c
index 3632ce8..19cccc3 100644
--- a/kernel/trace/ring_buffer.c
+++ b/kernel/trace/ring_buffer.c
@@ -3846,6 +3846,9 @@
rpos = reader->read;
pos += size;
+ if (rpos >= commit)
+ break;
+
event = rb_reader_event(cpu_buffer);
size = rb_event_length(event);
} while (len > size);
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index ba14a22..9ec59f5 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -3463,6 +3463,7 @@
size_t cnt, loff_t *fpos)
{
char *buf;
+ size_t written;
if (tracing_disabled)
return -EINVAL;
@@ -3484,11 +3485,15 @@
} else
buf[cnt] = '\0';
- cnt = mark_printk("%s", buf);
+ written = mark_printk("%s", buf);
kfree(buf);
- *fpos += cnt;
+ *fpos += written;
- return cnt;
+ /* don't tell userspace we wrote more - it might confuse them */
+ if (written > cnt)
+ written = cnt;
+
+ return written;
}
static int tracing_clock_show(struct seq_file *m, void *v)
diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
index 09b4fa6..4c758f1 100644
--- a/kernel/trace/trace_events.c
+++ b/kernel/trace/trace_events.c
@@ -598,88 +598,165 @@
return ret;
}
-static void print_event_fields(struct trace_seq *s, struct list_head *head)
+enum {
+ FORMAT_HEADER = 1,
+ FORMAT_PRINTFMT = 2,
+};
+
+static void *f_next(struct seq_file *m, void *v, loff_t *pos)
{
+ struct ftrace_event_call *call = m->private;
struct ftrace_event_field *field;
+ struct list_head *head;
- list_for_each_entry_reverse(field, head, link) {
- /*
- * Smartly shows the array type(except dynamic array).
- * Normal:
- * field:TYPE VAR
- * If TYPE := TYPE[LEN], it is shown:
- * field:TYPE VAR[LEN]
- */
- const char *array_descriptor = strchr(field->type, '[');
+ (*pos)++;
- if (!strncmp(field->type, "__data_loc", 10))
- array_descriptor = NULL;
+ switch ((unsigned long)v) {
+ case FORMAT_HEADER:
+ head = &ftrace_common_fields;
- if (!array_descriptor) {
- trace_seq_printf(s, "\tfield:%s %s;\toffset:%u;"
- "\tsize:%u;\tsigned:%d;\n",
- field->type, field->name, field->offset,
- field->size, !!field->is_signed);
- } else {
- trace_seq_printf(s, "\tfield:%.*s %s%s;\toffset:%u;"
- "\tsize:%u;\tsigned:%d;\n",
- (int)(array_descriptor - field->type),
- field->type, field->name,
- array_descriptor, field->offset,
- field->size, !!field->is_signed);
- }
+ if (unlikely(list_empty(head)))
+ return NULL;
+
+ field = list_entry(head->prev, struct ftrace_event_field, link);
+ return field;
+
+ case FORMAT_PRINTFMT:
+ /* all done */
+ return NULL;
}
+
+ head = trace_get_fields(call);
+
+ /*
+ * To separate common fields from event fields, the
+ * LSB is set on the first event field. Clear it in case.
+ */
+ v = (void *)((unsigned long)v & ~1L);
+
+ field = v;
+ /*
+ * If this is a common field, and at the end of the list, then
+ * continue with main list.
+ */
+ if (field->link.prev == &ftrace_common_fields) {
+ if (unlikely(list_empty(head)))
+ return NULL;
+ field = list_entry(head->prev, struct ftrace_event_field, link);
+ /* Set the LSB to notify f_show to print an extra newline */
+ field = (struct ftrace_event_field *)
+ ((unsigned long)field | 1);
+ return field;
+ }
+
+ /* If we are done tell f_show to print the format */
+ if (field->link.prev == head)
+ return (void *)FORMAT_PRINTFMT;
+
+ field = list_entry(field->link.prev, struct ftrace_event_field, link);
+
+ return field;
}
-static ssize_t
-event_format_read(struct file *filp, char __user *ubuf, size_t cnt,
- loff_t *ppos)
+static void *f_start(struct seq_file *m, loff_t *pos)
{
- struct ftrace_event_call *call = filp->private_data;
- struct list_head *head;
- struct trace_seq *s;
- char *buf;
- int r;
+ loff_t l = 0;
+ void *p;
- if (*ppos)
+ /* Start by showing the header */
+ if (!*pos)
+ return (void *)FORMAT_HEADER;
+
+ p = (void *)FORMAT_HEADER;
+ do {
+ p = f_next(m, p, &l);
+ } while (p && l < *pos);
+
+ return p;
+}
+
+static int f_show(struct seq_file *m, void *v)
+{
+ struct ftrace_event_call *call = m->private;
+ struct ftrace_event_field *field;
+ const char *array_descriptor;
+
+ switch ((unsigned long)v) {
+ case FORMAT_HEADER:
+ seq_printf(m, "name: %s\n", call->name);
+ seq_printf(m, "ID: %d\n", call->event.type);
+ seq_printf(m, "format:\n");
return 0;
- s = kmalloc(sizeof(*s), GFP_KERNEL);
- if (!s)
- return -ENOMEM;
-
- trace_seq_init(s);
-
- trace_seq_printf(s, "name: %s\n", call->name);
- trace_seq_printf(s, "ID: %d\n", call->event.type);
- trace_seq_printf(s, "format:\n");
-
- /* print common fields */
- print_event_fields(s, &ftrace_common_fields);
-
- trace_seq_putc(s, '\n');
-
- /* print event specific fields */
- head = trace_get_fields(call);
- print_event_fields(s, head);
-
- r = trace_seq_printf(s, "\nprint fmt: %s\n", call->print_fmt);
-
- if (!r) {
- /*
- * ug! The format output is bigger than a PAGE!!
- */
- buf = "FORMAT TOO BIG\n";
- r = simple_read_from_buffer(ubuf, cnt, ppos,
- buf, strlen(buf));
- goto out;
+ case FORMAT_PRINTFMT:
+ seq_printf(m, "\nprint fmt: %s\n",
+ call->print_fmt);
+ return 0;
}
- r = simple_read_from_buffer(ubuf, cnt, ppos,
- s->buffer, s->len);
- out:
- kfree(s);
- return r;
+ /*
+ * To separate common fields from event fields, the
+ * LSB is set on the first event field. Clear it and
+ * print a newline if it is set.
+ */
+ if ((unsigned long)v & 1) {
+ seq_putc(m, '\n');
+ v = (void *)((unsigned long)v & ~1L);
+ }
+
+ field = v;
+
+ /*
+ * Smartly shows the array type(except dynamic array).
+ * Normal:
+ * field:TYPE VAR
+ * If TYPE := TYPE[LEN], it is shown:
+ * field:TYPE VAR[LEN]
+ */
+ array_descriptor = strchr(field->type, '[');
+
+ if (!strncmp(field->type, "__data_loc", 10))
+ array_descriptor = NULL;
+
+ if (!array_descriptor)
+ seq_printf(m, "\tfield:%s %s;\toffset:%u;\tsize:%u;\tsigned:%d;\n",
+ field->type, field->name, field->offset,
+ field->size, !!field->is_signed);
+ else
+ seq_printf(m, "\tfield:%.*s %s%s;\toffset:%u;\tsize:%u;\tsigned:%d;\n",
+ (int)(array_descriptor - field->type),
+ field->type, field->name,
+ array_descriptor, field->offset,
+ field->size, !!field->is_signed);
+
+ return 0;
+}
+
+static void f_stop(struct seq_file *m, void *p)
+{
+}
+
+static const struct seq_operations trace_format_seq_ops = {
+ .start = f_start,
+ .next = f_next,
+ .stop = f_stop,
+ .show = f_show,
+};
+
+static int trace_format_open(struct inode *inode, struct file *file)
+{
+ struct ftrace_event_call *call = inode->i_private;
+ struct seq_file *m;
+ int ret;
+
+ ret = seq_open(file, &trace_format_seq_ops);
+ if (ret < 0)
+ return ret;
+
+ m = file->private_data;
+ m->private = call;
+
+ return 0;
}
static ssize_t
@@ -877,8 +954,10 @@
};
static const struct file_operations ftrace_event_format_fops = {
- .open = tracing_open_generic,
- .read = event_format_read,
+ .open = trace_format_open,
+ .read = seq_read,
+ .llseek = seq_lseek,
+ .release = seq_release,
};
static const struct file_operations ftrace_event_id_fops = {
diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c
index 6bff236..6f23369 100644
--- a/kernel/trace/trace_functions_graph.c
+++ b/kernel/trace/trace_functions_graph.c
@@ -507,7 +507,15 @@
* if the output fails.
*/
data->ent = *curr;
- data->ret = *next;
+ /*
+ * If the next event is not a return type, then
+ * we only care about what type it is. Otherwise we can
+ * safely copy the entire event.
+ */
+ if (next->ent.type == TRACE_GRAPH_RET)
+ data->ret = *next;
+ else
+ data->ret.ent.type = next->ent.type;
}
}
diff --git a/lib/Kconfig.debug b/lib/Kconfig.debug
index 9e06b7f..1b4afd2 100644
--- a/lib/Kconfig.debug
+++ b/lib/Kconfig.debug
@@ -994,13 +994,16 @@
config LATENCYTOP
bool "Latency measuring infrastructure"
+ depends on HAVE_LATENCYTOP_SUPPORT
+ depends on DEBUG_KERNEL
+ depends on STACKTRACE_SUPPORT
+ depends on PROC_FS
select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE
select KALLSYMS
select KALLSYMS_ALL
select STACKTRACE
select SCHEDSTATS
select SCHED_DEBUG
- depends on HAVE_LATENCYTOP_SUPPORT
help
Enable this option if you want to use the LatencyTOP tool
to find out which userspace is blocking on what kernel operations.
diff --git a/scripts/recordmcount.pl b/scripts/recordmcount.pl
index 0171060..e67f054 100755
--- a/scripts/recordmcount.pl
+++ b/scripts/recordmcount.pl
@@ -159,6 +159,7 @@
my $function_regex; # Find the name of a function
# (return offset and func name)
my $mcount_regex; # Find the call site to mcount (return offset)
+my $mcount_adjust; # Address adjustment to mcount offset
my $alignment; # The .align value to use for $mcount_section
my $section_type; # Section header plus possible alignment command
my $can_use_local = 0; # If we can use local function references
@@ -213,6 +214,7 @@
$function_regex = "^([0-9a-fA-F]+)\\s+<(.*?)>:";
$mcount_regex = "^\\s*([0-9a-fA-F]+):.*\\smcount\$";
$section_type = '@progbits';
+$mcount_adjust = 0;
$type = ".long";
if ($arch eq "x86_64") {
@@ -351,6 +353,9 @@
} elsif ($arch eq "microblaze") {
# Microblaze calls '_mcount' instead of plain 'mcount'.
$mcount_regex = "^\\s*([0-9a-fA-F]+):.*\\s_mcount\$";
+} elsif ($arch eq "blackfin") {
+ $mcount_regex = "^\\s*([0-9a-fA-F]+):.*\\s__mcount\$";
+ $mcount_adjust = -4;
} else {
die "Arch $arch is not supported with CONFIG_FTRACE_MCOUNT_RECORD";
}
@@ -511,7 +516,7 @@
}
# is this a call site to mcount? If so, record it to print later
if ($text_found && /$mcount_regex/) {
- push(@offsets, hex $1);
+ push(@offsets, (hex $1) + $mcount_adjust);
}
}