Skip to contentSkip to conversation

tracing: three fixes

GIT PULL In review Steven Rostedt 2 participants 7 messages Apr 8, 2013 – Apr 9, 2013

Subsystems (3)
lore
Share & tools
lore

Steven Rostedt

author maintainer
#0 lore raw

Linus,

This includes three fixes. Two fix features added in 3.9 and one fixes a long time minor bug.

The first patch fixes a race that can happen if the user switches from the irqsoff tracer to another tracer. If a irqs off latency is detected, it will try to use the snapshot buffer, but the new tracer wont have it allocated. There's a nasty warning that gets printed and the trace is ignored. Nothing crashes, just a nasty WARN_ON is shown.

The second patch fixes an issue where if the sysctl is used to disable and enable function tracing, it can put the function tracing into an unstable state.

The third patch fixes an issue with perf using the function tracer. An update was done, where the stub function could be called during the perf function tracing, and that stub function wont have the "control" flag set and cause a nasty warning when running perf.

Please pull the latest trace-fixes-3.9-rc6 tree, which can be found at:

git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace.git
trace-fixes-3.9-rc6

Head SHA1: 620bd150ec911fae0ad02e66fa003548505ec7f4

Jan Kiszka (1):
ftrace: Consistently restore trace function on sysctl enabling

Steven Rostedt (Red Hat) (2):
tracing: Fix race with update_max_tr_single and changing tracers
ftrace: Do not call stub functions in control loop

----

 include/linux/ftrace.h |    2 ++
 kernel/trace/ftrace.c  |   13 +++++--------
 kernel/trace/trace.c   |    5 ++++-
 3 files changed, 11 insertions(+), 9 deletions(-)
author maintainer
#1 lore raw

PATCH 1/3 tracing: Fix race with update_max_tr_single and changing tracers

From: "Steven Rostedt (Red Hat)" <[email protected]>

The commit 34600f0e9 "tracing: Fix race with max_tr and changing tracers" fixed the updating of the main buffers with the race of changing tracers, but left out the fix to the updating of just a per cpu buffer.

---
 kernel/trace/trace.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
kernel/trace/trace.c +4 -1
index 4f1dade..7ba7fc7 100644
@@ -744,8 +744,11 @@ update_max_tr_single(struct trace_array *tr, struct task_struct *tsk, int cpu)
 		return;
 
 	WARN_ON_ONCE(!irqs_disabled());
-	if (WARN_ON_ONCE(!current_trace->allocated_snapshot))
+	if (!current_trace->allocated_snapshot) {
+		/* Only the nop tracer should hit this when disabling */
+		WARN_ON_ONCE(current_trace != &nop_trace);
 		return;
+	}
 
 	arch_spin_lock(&ftrace_max_lock);
 
Signature · 1 line
1.7.10.4
author maintainer
#2 lore raw

PATCH 2/3 ftrace: Consistently restore trace function on sysctl enabling

In reply to Steven Rostedt

From: Jan Kiszka <[email protected]>

If we reenable ftrace via syctl, we currently set ftrace_trace_function based on the previous simplistic algorithm. This is inconsistent with what update_ftrace_function does. So better call that helper instead.

---
 kernel/trace/ftrace.c |    8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)
kernel/trace/ftrace.c +2 -6
index 6893d5a..cc4943c 100644
@@ -4555,12 +4555,8 @@ ftrace_enable_sysctl(struct ctl_table *table, int write,
 		ftrace_startup_sysctl();
 
 		/* we are starting ftrace again */
-		if (ftrace_ops_list != &ftrace_list_end) {
-			if (ftrace_ops_list->next == &ftrace_list_end)
-				ftrace_trace_function = ftrace_ops_list->func;
-			else
-				ftrace_trace_function = ftrace_ops_list_func;
-		}
+		if (ftrace_ops_list != &ftrace_list_end)
+			update_ftrace_function();
 
 	} else {
 		/* stopping ftrace calls (just send to ftrace_stub) */
Signature · 1 line
1.7.10.4
author maintainer
#3 lore raw

PATCH 3/3 ftrace: Do not call stub functions in control loop

In reply to Steven Rostedt

From: "Steven Rostedt (Red Hat)" <[email protected]>

The function tracing control loop used by perf spits out a warning if the called function is not a control function. This is because the control function references a per cpu allocated data structure on struct ftrace_ops that is not allocated for other types of functions.

commit 0a016409e42 "ftrace: Optimize the function tracer list loop"

Had an optimization done to all function tracing loops to optimize for a single registered ops. Unfortunately, this allows for a slight race when tracing starts or ends, where the stub function might be called after the current registered ops is removed. In this case we get the following dump:

root# perf stat -e ftrace:function sleep 1

[   74.339105] WARNING: at include/linux/ftrace.h:209 ftrace_ops_control_func+0xde/0xf0()
[   74.349522] Hardware name: PRIMERGY RX200 S6
[   74.357149] Modules linked in: sg igb iTCO_wdt ptp pps_core iTCO_vendor_support i7core_edac dca lpc_ich i2c_i801 coretemp edac_core crc32c_intel mfd_core ghash_clmulni_intel dm_multipath acpi_power_meter pcspk
r microcode vhost_net tun macvtap macvlan nfsd kvm_intel kvm auth_rpcgss nfs_acl lockd sunrpc uinput xfs libcrc32c sd_mod crc_t10dif sr_mod cdrom mgag200 i2c_algo_bit drm_kms_helper ttm qla2xxx mptsas ahci drm li
bahci scsi_transport_sas mptscsih libata scsi_transport_fc i2c_core mptbase scsi_tgt dm_mirror dm_region_hash dm_log dm_mod
[   74.446233] Pid: 1377, comm: perf Tainted: G        W    3.9.0-rc1 #1
[   74.453458] Call Trace:
[   74.456233]  [<ffffffff81062e3f>] warn_slowpath_common+0x7f/0xc0
[   74.462997]  [<ffffffff810fbc60>] ? rcu_note_context_switch+0xa0/0xa0
[   74.470272]  [<ffffffff811041a2>] ? __unregister_ftrace_function+0xa2/0x1a0
[   74.478117]  [<ffffffff81062e9a>] warn_slowpath_null+0x1a/0x20
[   74.484681]  [<ffffffff81102ede>] ftrace_ops_control_func+0xde/0xf0
[   74.491760]  [<ffffffff8162f400>] ftrace_call+0x5/0x2f
[   74.497511]  [<ffffffff8162f400>] ? ftrace_call+0x5/0x2f
[   74.503486]  [<ffffffff8162f400>] ? ftrace_call+0x5/0x2f
[   74.509500]  [<ffffffff810fbc65>] ? synchronize_sched+0x5/0x50
[   74.516088]  [<ffffffff816254d5>] ? _cond_resched+0x5/0x40
[   74.522268]  [<ffffffff810fbc65>] ? synchronize_sched+0x5/0x50
[   74.528837]  [<ffffffff811041a2>] ? __unregister_ftrace_function+0xa2/0x1a0
[   74.536696]  [<ffffffff816254d5>] ? _cond_resched+0x5/0x40
[   74.542878]  [<ffffffff8162402d>] ? mutex_lock+0x1d/0x50
[   74.548869]  [<ffffffff81105c67>] unregister_ftrace_function+0x27/0x50
[   74.556243]  [<ffffffff8111eadf>] perf_ftrace_event_register+0x9f/0x140
[   74.563709]  [<ffffffff816254d5>] ? _cond_resched+0x5/0x40
[   74.569887]  [<ffffffff8162402d>] ? mutex_lock+0x1d/0x50
[   74.575898]  [<ffffffff8111e94e>] perf_trace_destroy+0x2e/0x50
[   74.582505]  [<ffffffff81127ba9>] tp_perf_event_destroy+0x9/0x10
[   74.589298]  [<ffffffff811295d0>] free_event+0x70/0x1a0
[   74.595208]  [<ffffffff8112a579>] perf_event_release_kernel+0x69/0xa0
[   74.602460]  [<ffffffff816254d5>] ? _cond_resched+0x5/0x40
[   74.608667]  [<ffffffff8112a640>] put_event+0x90/0xc0
[   74.614373]  [<ffffffff8112a740>] perf_release+0x10/0x20
[   74.620367]  [<ffffffff811a3044>] __fput+0xf4/0x280
[   74.625894]  [<ffffffff811a31de>] ____fput+0xe/0x10
[   74.631387]  [<ffffffff81083697>] task_work_run+0xa7/0xe0
[   74.637452]  [<ffffffff81014981>] do_notify_resume+0x71/0xb0
[   74.643843]  [<ffffffff8162fa92>] int_signal+0x12/0x17

To fix this a new ftrace_ops flag is added that denotes the ftrace_list_end ftrace_ops stub as just that, a stub. This flag is now checked in the control loop and the function is not called if the flag is set.

Thanks to Jovi for not just reporting the bug, but also pointing out where the bug was in the code.

---
 include/linux/ftrace.h |    2 ++
 kernel/trace/ftrace.c  |    5 +++--
 2 files changed, 5 insertions(+), 2 deletions(-)
include/linux/ftrace.h +2 -0
index e5ca8ef..167abf9 100644
@@ -89,6 +89,7 @@ typedef void (*ftrace_func_t)(unsigned long ip, unsigned long parent_ip,
  *            that the call back has its own recursion protection. If it does
  *            not set this, then the ftrace infrastructure will add recursion
  *            protection for the caller.
+ * STUB   - The ftrace_ops is just a place holder.
  */
 enum {
 	FTRACE_OPS_FL_ENABLED			= 1 << 0,
@@ -98,6 +99,7 @@ enum {
 	FTRACE_OPS_FL_SAVE_REGS			= 1 << 4,
 	FTRACE_OPS_FL_SAVE_REGS_IF_SUPPORTED	= 1 << 5,
 	FTRACE_OPS_FL_RECURSION_SAFE		= 1 << 6,
+	FTRACE_OPS_FL_STUB			= 1 << 7,
 };
 
 struct ftrace_ops {
kernel/trace/ftrace.c +3 -2
index cc4943c..7e89710 100644
@@ -66,7 +66,7 @@
 
 static struct ftrace_ops ftrace_list_end __read_mostly = {
 	.func		= ftrace_stub,
-	.flags		= FTRACE_OPS_FL_RECURSION_SAFE,
+	.flags		= FTRACE_OPS_FL_RECURSION_SAFE | FTRACE_OPS_FL_STUB,
 };
 
 /* ftrace_enabled is a method to turn ftrace on or off */
@@ -4131,7 +4131,8 @@ ftrace_ops_control_func(unsigned long ip, unsigned long parent_ip,
 	preempt_disable_notrace();
 	trace_recursion_set(TRACE_CONTROL_BIT);
 	do_for_each_ftrace_op(op, ftrace_control_list) {
-		if (!ftrace_function_local_disabled(op) &&
+		if (!(op->flags & FTRACE_OPS_FL_STUB) &&
+		    !ftrace_function_local_disabled(op) &&
 		    ftrace_ops_test(op, ip))
 			op->func(ip, parent_ip, op, regs);
 	} while_for_each_ftrace_op(op);
Signature · 1 line
1.7.10.4
#4 lore raw

Re: [PATCH 3/3] ftrace: Do not call stub functions in control loop

Quoted 62 lines · Steven Rostedt

From: "Steven Rostedt (Red Hat)" <[email protected]>

The function tracing control loop used by perf spits out a warning if the called function is not a control function. This is because the control function references a per cpu allocated data structure on struct ftrace_ops that is not allocated for other types of functions.

commit 0a016409e42 "ftrace: Optimize the function tracer list loop"

Had an optimization done to all function tracing loops to optimize for a single registered ops. Unfortunately, this allows for a slight race when tracing starts or ends, where the stub function might be called after the current registered ops is removed. In this case we get the following dump:

root# perf stat -e ftrace:function sleep 1

[   74.339105] WARNING: at include/linux/ftrace.h:209 ftrace_ops_control_func+0xde/0xf0()
[   74.349522] Hardware name: PRIMERGY RX200 S6
[   74.357149] Modules linked in: sg igb iTCO_wdt ptp pps_core iTCO_vendor_support i7core_edac dca lpc_ich i2c_i801 coretemp edac_core crc32c_intel mfd_core ghash_clmulni_intel dm_multipath acpi_power_meter pcspk
r microcode vhost_net tun macvtap macvlan nfsd kvm_intel kvm auth_rpcgss nfs_acl lockd sunrpc uinput xfs libcrc32c sd_mod crc_t10dif sr_mod cdrom mgag200 i2c_algo_bit drm_kms_helper ttm qla2xxx mptsas ahci drm li
bahci scsi_transport_sas mptscsih libata scsi_transport_fc i2c_core mptbase scsi_tgt dm_mirror dm_region_hash dm_log dm_mod
[   74.446233] Pid: 1377, comm: perf Tainted: G        W    3.9.0-rc1 #1
[   74.453458] Call Trace:
[   74.456233]  [<ffffffff81062e3f>] warn_slowpath_common+0x7f/0xc0
[   74.462997]  [<ffffffff810fbc60>] ? rcu_note_context_switch+0xa0/0xa0
[   74.470272]  [<ffffffff811041a2>] ? __unregister_ftrace_function+0xa2/0x1a0
[   74.478117]  [<ffffffff81062e9a>] warn_slowpath_null+0x1a/0x20
[   74.484681]  [<ffffffff81102ede>] ftrace_ops_control_func+0xde/0xf0
[   74.491760]  [<ffffffff8162f400>] ftrace_call+0x5/0x2f
[   74.497511]  [<ffffffff8162f400>] ? ftrace_call+0x5/0x2f
[   74.503486]  [<ffffffff8162f400>] ? ftrace_call+0x5/0x2f
[   74.509500]  [<ffffffff810fbc65>] ? synchronize_sched+0x5/0x50
[   74.516088]  [<ffffffff816254d5>] ? _cond_resched+0x5/0x40
[   74.522268]  [<ffffffff810fbc65>] ? synchronize_sched+0x5/0x50
[   74.528837]  [<ffffffff811041a2>] ? __unregister_ftrace_function+0xa2/0x1a0
[   74.536696]  [<ffffffff816254d5>] ? _cond_resched+0x5/0x40
[   74.542878]  [<ffffffff8162402d>] ? mutex_lock+0x1d/0x50
[   74.548869]  [<ffffffff81105c67>] unregister_ftrace_function+0x27/0x50
[   74.556243]  [<ffffffff8111eadf>] perf_ftrace_event_register+0x9f/0x140
[   74.563709]  [<ffffffff816254d5>] ? _cond_resched+0x5/0x40
[   74.569887]  [<ffffffff8162402d>] ? mutex_lock+0x1d/0x50
[   74.575898]  [<ffffffff8111e94e>] perf_trace_destroy+0x2e/0x50
[   74.582505]  [<ffffffff81127ba9>] tp_perf_event_destroy+0x9/0x10
[   74.589298]  [<ffffffff811295d0>] free_event+0x70/0x1a0
[   74.595208]  [<ffffffff8112a579>] perf_event_release_kernel+0x69/0xa0
[   74.602460]  [<ffffffff816254d5>] ? _cond_resched+0x5/0x40
[   74.608667]  [<ffffffff8112a640>] put_event+0x90/0xc0
[   74.614373]  [<ffffffff8112a740>] perf_release+0x10/0x20
[   74.620367]  [<ffffffff811a3044>] __fput+0xf4/0x280
[   74.625894]  [<ffffffff811a31de>] ____fput+0xe/0x10
[   74.631387]  [<ffffffff81083697>] task_work_run+0xa7/0xe0
[   74.637452]  [<ffffffff81014981>] do_notify_resume+0x71/0xb0
[   74.643843]  [<ffffffff8162fa92>] int_signal+0x12/0x17

To fix this a new ftrace_ops flag is added that denotes the ftrace_list_end ftrace_ops stub as just that, a stub. This flag is now checked in the control loop and the function is not called if the flag is set.

Thanks to Jovi for not just reporting the bug, but also pointing out where the bug was in the code.

Quoted 7 lines

Involve stable? 3.8 kernel user would need this fix.

Quoted 49 lines
---
 include/linux/ftrace.h |    2 ++
 kernel/trace/ftrace.c  |    5 +++--
 2 files changed, 5 insertions(+), 2 deletions(-)
include/linux/ftrace.h +2 -0
index e5ca8ef..167abf9 100644
@@ -89,6 +89,7 @@ typedef void (*ftrace_func_t)(unsigned long ip, unsigned long parent_ip,
  *            that the call back has its own recursion protection. If it does
  *            not set this, then the ftrace infrastructure will add recursion
  *            protection for the caller.
+ * STUB   - The ftrace_ops is just a place holder.
  */
 enum {
 	FTRACE_OPS_FL_ENABLED			= 1 << 0,
@@ -98,6 +99,7 @@ enum {
 	FTRACE_OPS_FL_SAVE_REGS			= 1 << 4,
 	FTRACE_OPS_FL_SAVE_REGS_IF_SUPPORTED	= 1 << 5,
 	FTRACE_OPS_FL_RECURSION_SAFE		= 1 << 6,
+	FTRACE_OPS_FL_STUB			= 1 << 7,
 };
 
 struct ftrace_ops {
kernel/trace/ftrace.c +3 -2
index cc4943c..7e89710 100644
@@ -66,7 +66,7 @@
 
 static struct ftrace_ops ftrace_list_end __read_mostly = {
 	.func		= ftrace_stub,
-	.flags		= FTRACE_OPS_FL_RECURSION_SAFE,
+	.flags		= FTRACE_OPS_FL_RECURSION_SAFE | FTRACE_OPS_FL_STUB,
 };
 
 /* ftrace_enabled is a method to turn ftrace on or off */
@@ -4131,7 +4131,8 @@ ftrace_ops_control_func(unsigned long ip, unsigned long parent_ip,
 	preempt_disable_notrace();
 	trace_recursion_set(TRACE_CONTROL_BIT);
 	do_for_each_ftrace_op(op, ftrace_control_list) {
-		if (!ftrace_function_local_disabled(op) &&
+		if (!(op->flags & FTRACE_OPS_FL_STUB) &&
+		    !ftrace_function_local_disabled(op) &&
 		    ftrace_ops_test(op, ip))
 			op->func(ip, parent_ip, op, regs);
 	} while_for_each_ftrace_op(op);
author maintainer
#5 lore raw

Re: [PATCH 3/3] ftrace: Do not call stub functions in control loop

Quoted 11 lines · zhangwei(Jovi)
Quoted 10 lines · Steven Rostedt

From: "Steven Rostedt (Red Hat)" <[email protected]>

The function tracing control loop used by perf spits out a warning if the called function is not a control function. This is because the control function references a per cpu allocated data structure on struct ftrace_ops that is not allocated for other types of functions.

commit 0a016409e42 "ftrace: Optimize the function tracer list loop"

Quoted 8 lines
Quoted 6 lines

Had an optimization done to all function tracing loops to optimize for a single registered ops. Unfortunately, this allows for a slight race when tracing starts or ends, where the stub function might be called after the current registered ops is removed. In this case we get the following dump:

Involve stable? 3.8 kernel user would need this fix.

Why? commit 0a016409e42 "ftrace: Optimize the function tracer list loop" was added for 3.9. 3.8 has:

        op = rcu_dereference_raw(ftrace_control_list);
        while (op != &ftrace_list_end) {
                if (!ftrace_function_local_disabled(op) &&
                    ftrace_ops_test(op, ip))
                        op->func(ip, parent_ip, op, regs);

                op = rcu_dereference_raw(op->next);
        };

The stub function will never get called.

-- Steve

#6 lore raw

Re: [PATCH 3/3] ftrace: Do not call stub functions in control loop

Quoted 33 lines · Steven Rostedt
Quoted 11 lines · zhangwei(Jovi)
Quoted 10 lines · Steven Rostedt

From: "Steven Rostedt (Red Hat)" <[email protected]>

The function tracing control loop used by perf spits out a warning if the called function is not a control function. This is because the control function references a per cpu allocated data structure on struct ftrace_ops that is not allocated for other types of functions.

commit 0a016409e42 "ftrace: Optimize the function tracer list loop"

Quoted 8 lines
Quoted 6 lines

Had an optimization done to all function tracing loops to optimize for a single registered ops. Unfortunately, this allows for a slight race when tracing starts or ends, where the stub function might be called after the current registered ops is removed. In this case we get the following dump:

Involve stable? 3.8 kernel user would need this fix.

Why? commit 0a016409e42 "ftrace: Optimize the function tracer list loop" was added for 3.9. 3.8 has:

        op = rcu_dereference_raw(ftrace_control_list);
        while (op != &ftrace_list_end) {
                if (!ftrace_function_local_disabled(op) &&
                    ftrace_ops_test(op, ip))
                        op->func(ip, parent_ip, op, regs);

                op = rcu_dereference_raw(op->next);
        };
Quoted 3 lines

The stub function will never get called.

-- Steve

Hmm, You are right, I misunderstand tag info got by git describe in my local box

[root@jovi linux]# git describe 0a016409e42
trace-3.8-rc4-fix-19-g0a01640

Quoted 6 lines

.