perf events: Clean up pid passing
The kernel perf event creation path shouldn't use find_task_by_vpid()
because a vpid exists in a specific namespace. find_task_by_vpid() uses
current's pid namespace which isn't always the correct namespace to use
for the vpid in all the places perf_event_create_kernel_counter() (and
thus find_get_context()) is called.
The goal is to clean up pid namespace handling and prevent bugs like:
https://bugzilla.kernel.org/show_bug.cgi?id=17281
Instead of using pids switch find_get_context() to use task struct
pointers directly. The syscall is responsible for resolving the pid to
a task struct. This moves the pid namespace resolution into the syscall
much like every other syscall that takes pid parameters.
Signed-off-by: Matt Helsley <matthltc@us.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Robin Green <greenrd@greenrd.org>
Cc: Prasad <prasad@linux.vnet.ibm.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
LKML-Reference: <a134e5e392ab0204961fd1a62c84a222bf5874a9.1284407763.git.matthltc@us.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
diff --git a/arch/arm/oprofile/common.c b/arch/arm/oprofile/common.c
index 0691176..aad63e6 100644
--- a/arch/arm/oprofile/common.c
+++ b/arch/arm/oprofile/common.c
@@ -96,7 +96,7 @@
return ret;
pevent = perf_event_create_kernel_counter(&counter_config[event].attr,
- cpu, -1,
+ cpu, NULL,
op_overflow_handler);
if (IS_ERR(pevent)) {
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 93bf53a..39d8860 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -902,7 +902,7 @@
extern struct perf_event *
perf_event_create_kernel_counter(struct perf_event_attr *attr,
int cpu,
- pid_t pid,
+ struct task_struct *task,
perf_overflow_handler_t callback);
extern u64 perf_event_read_value(struct perf_event *event,
u64 *enabled, u64 *running);
diff --git a/kernel/hw_breakpoint.c b/kernel/hw_breakpoint.c
index 6122f02..3b714e8 100644
--- a/kernel/hw_breakpoint.c
+++ b/kernel/hw_breakpoint.c
@@ -433,8 +433,7 @@
perf_overflow_handler_t triggered,
struct task_struct *tsk)
{
- return perf_event_create_kernel_counter(attr, -1, task_pid_vnr(tsk),
- triggered);
+ return perf_event_create_kernel_counter(attr, -1, tsk, triggered);
}
EXPORT_SYMBOL_GPL(register_user_hw_breakpoint);
@@ -516,7 +515,7 @@
get_online_cpus();
for_each_online_cpu(cpu) {
pevent = per_cpu_ptr(cpu_events, cpu);
- bp = perf_event_create_kernel_counter(attr, cpu, -1, triggered);
+ bp = perf_event_create_kernel_counter(attr, cpu, NULL, triggered);
*pevent = bp;
diff --git a/kernel/perf_event.c b/kernel/perf_event.c
index 3f5309d..86f394e 100644
--- a/kernel/perf_event.c
+++ b/kernel/perf_event.c
@@ -2053,15 +2053,14 @@
}
static struct perf_event_context *
-find_get_context(struct pmu *pmu, pid_t pid, int cpu)
+find_get_context(struct pmu *pmu, struct task_struct *task, int cpu)
{
struct perf_event_context *ctx;
struct perf_cpu_context *cpuctx;
- struct task_struct *task;
unsigned long flags;
int ctxn, err;
- if (pid == -1 && cpu != -1) {
+ if (!task && cpu != -1) {
/* Must be root to operate on a CPU event: */
if (perf_paranoid_cpu() && !capable(CAP_SYS_ADMIN))
return ERR_PTR(-EACCES);
@@ -2084,10 +2083,6 @@
return ctx;
}
- task = find_lively_task_by_vpid(pid);
- if (IS_ERR(task))
- return (void*)task;
-
err = -EINVAL;
ctxn = pmu->task_ctx_nr;
if (ctxn < 0)
@@ -5527,6 +5522,7 @@
struct perf_event_context *ctx;
struct file *event_file = NULL;
struct file *group_file = NULL;
+ struct task_struct *task = NULL;
struct pmu *pmu;
int event_fd;
int fput_needed = 0;
@@ -5581,10 +5577,13 @@
if ((pmu->task_ctx_nr == perf_sw_context) && group_leader)
pmu = group_leader->pmu;
+ if (pid != -1)
+ task = find_lively_task_by_vpid(pid);
+
/*
* Get the target context (task or percpu):
*/
- ctx = find_get_context(pmu, pid, cpu);
+ ctx = find_get_context(pmu, task, cpu);
if (IS_ERR(ctx)) {
err = PTR_ERR(ctx);
goto err_group_fd;
@@ -5666,11 +5665,11 @@
*
* @attr: attributes of the counter to create
* @cpu: cpu in which the counter is bound
- * @pid: task to profile
+ * @task: task to profile (NULL for percpu)
*/
struct perf_event *
perf_event_create_kernel_counter(struct perf_event_attr *attr, int cpu,
- pid_t pid,
+ struct task_struct *task,
perf_overflow_handler_t overflow_handler)
{
struct perf_event_context *ctx;
@@ -5687,7 +5686,7 @@
goto err;
}
- ctx = find_get_context(event->pmu, pid, cpu);
+ ctx = find_get_context(event->pmu, task, cpu);
if (IS_ERR(ctx)) {
err = PTR_ERR(ctx);
goto err_free;
diff --git a/kernel/watchdog.c b/kernel/watchdog.c
index 89eadbb..dc8e168 100644
--- a/kernel/watchdog.c
+++ b/kernel/watchdog.c
@@ -358,7 +358,7 @@
/* Try to register using hardware perf events */
wd_attr = &wd_hw_attr;
wd_attr->sample_period = hw_nmi_get_sample_period();
- event = perf_event_create_kernel_counter(wd_attr, cpu, -1, watchdog_overflow_callback);
+ event = perf_event_create_kernel_counter(wd_attr, cpu, NULL, watchdog_overflow_callback);
if (!IS_ERR(event)) {
printk(KERN_INFO "NMI watchdog enabled, takes one hw-pmu counter.\n");
goto out_save;