Skip to contentSkip to conversation

bpf: add bpf_send_signal_thread() helper

PATCH Superseded Yonghong Song 2 participants 7 messages Jan 10, 2020 – Jan 14, 2020

Subsystems (6)
Kselftest BPF Tracing BPF selftests +2 more subsystems

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

cover bpf: add bpf_send_signal_thread() helper

Commit 8b401f9ed244 ("bpf: implement bpf_send_signal() helper") added helper bpf_send_signal() which permits bpf program to send a signal to the current process.

This patch implemented a new helper bpf_send_signal_thread() to send a signal to the current thread. This helper can simplify user space code if the thread context of bpf sending signal is needed in user space. Please see Patch #1 for details of use case and kernel implementation.

Patch #2 added some bpf self tests for the new helper.

Yonghong Song (2):
bpf: add bpf_send_signal_thread() helper tools/bpf: add a selftest for bpf_send_signal_thread()

 include/uapi/linux/bpf.h                      | 18 +++++++++--
 kernel/trace/bpf_trace.c                      | 27 +++++++++++++++--
 tools/include/uapi/linux/bpf.h                | 18 +++++++++--
 .../selftests/bpf/prog_tests/send_signal.c    | 30 ++++++++++++-------
 .../bpf/progs/test_send_signal_kern.c         |  9 ++++--
 5 files changed, 82 insertions(+), 20 deletions(-)
Signature · 1 line
2.17.1
#1 lore raw

PATCH 1/2 bpf: add bpf_send_signal_thread() helper

Commit 8b401f9ed244 ("bpf: implement bpf_send_signal() helper") added helper bpf_send_signal() which permits bpf program to send a signal to the current process.

We found a use case where sending the signal to the current thread is more preferable.
- A bpf program will collect the stack trace and then send signal to the user application.
- The user application will add some thread specific information to the just collected stack trace for later analysis.

If bpf_send_signal() is used, user application will need to check whether the thread receiving the signal matches the thread collecting the stack by checking thread id. If not, it will need to send signal to another thread through pthread_kill().

This patch proposed a new helper bpf_send_signal_thread(), which sends the signal to the current thread. This way, user space is guaranteed that bpf_program execution context and user space signal handling context are the same thread.

---
 include/uapi/linux/bpf.h | 18 ++++++++++++++++--
 kernel/trace/bpf_trace.c | 27 ++++++++++++++++++++++++---
 2 files changed, 40 insertions(+), 5 deletions(-)
include/uapi/linux/bpf.h +16 -2
index 52966e758fe5..3320f8bdfe7e 100644
@@ -2714,7 +2714,7 @@ union bpf_attr {
  *
  * int bpf_send_signal(u32 sig)
  *	Description
- *		Send signal *sig* to the current task.
+ *		Send signal *sig* to the process of the current task.
  *	Return
  *		0 on success or successfully queued.
  *
@@ -2850,6 +2850,19 @@ union bpf_attr {
  *	Return
  *		0 on success, or a negative error in case of failure.
  *
+ * int bpf_send_signal_thread(u32 sig)
+ *	Description
+ *		Send signal *sig* to the current task.
+ *	Return
+ *		0 on success or successfully queued.
+ *
+ *		**-EBUSY** if work queue under nmi is full.
+ *
+ *		**-EINVAL** if *sig* is invalid.
+ *
+ *		**-EPERM** if no permission to send the *sig*.
+ *
+ *		**-EAGAIN** if bpf program can try again.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
@@ -2968,7 +2981,8 @@ union bpf_attr {
 	FN(probe_read_kernel),		\
 	FN(probe_read_user_str),	\
 	FN(probe_read_kernel_str),	\
-	FN(tcp_send_ack),
+	FN(tcp_send_ack),		\
+	FN(send_signal_thread),
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
  * function eBPF program intends to call
kernel/trace/bpf_trace.c +24 -3
index e5ef4ae9edb5..19e793aa441a 100644
@@ -703,6 +703,7 @@ struct send_signal_irq_work {
 	struct irq_work irq_work;
 	struct task_struct *task;
 	u32 sig;
+	enum pid_type type;
 };
 
 static DEFINE_PER_CPU(struct send_signal_irq_work, send_signal_work);
@@ -712,10 +713,10 @@ static void do_bpf_send_signal(struct irq_work *entry)
 	struct send_signal_irq_work *work;
 
 	work = container_of(entry, struct send_signal_irq_work, irq_work);
-	group_send_sig_info(work->sig, SEND_SIG_PRIV, work->task, PIDTYPE_TGID);
+	group_send_sig_info(work->sig, SEND_SIG_PRIV, work->task, work->type);
 }
 
-BPF_CALL_1(bpf_send_signal, u32, sig)
+static int bpf_send_signal_common(u32 sig, enum pid_type type)
 {
 	struct send_signal_irq_work *work = NULL;
 
@@ -748,11 +749,17 @@ BPF_CALL_1(bpf_send_signal, u32, sig)
 		 */
 		work->task = current;
 		work->sig = sig;
+		work->type = type;
 		irq_work_queue(&work->irq_work);
 		return 0;
 	}
 
-	return group_send_sig_info(sig, SEND_SIG_PRIV, current, PIDTYPE_TGID);
+	return group_send_sig_info(sig, SEND_SIG_PRIV, current, type);
+}
+
+BPF_CALL_1(bpf_send_signal, u32, sig)
+{
+	return bpf_send_signal_common(sig, PIDTYPE_TGID);
 }
 
 static const struct bpf_func_proto bpf_send_signal_proto = {
@@ -762,6 +769,18 @@ static const struct bpf_func_proto bpf_send_signal_proto = {
 	.arg1_type	= ARG_ANYTHING,
 };
 
+BPF_CALL_1(bpf_send_signal_thread, u32, sig)
+{
+	return bpf_send_signal_common(sig, PIDTYPE_PID);
+}
+
+static const struct bpf_func_proto bpf_send_signal_thread_proto = {
+	.func		= bpf_send_signal_thread,
+	.gpl_only	= false,
+	.ret_type	= RET_INTEGER,
+	.arg1_type	= ARG_ANYTHING,
+};
+
 static const struct bpf_func_proto *
 tracing_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
 {
@@ -822,6 +841,8 @@ tracing_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog)
 #endif
 	case BPF_FUNC_send_signal:
 		return &bpf_send_signal_proto;
+	case BPF_FUNC_send_signal_thread:
+		return &bpf_send_signal_thread_proto;
 	default:
 		return NULL;
 	}
Signature · 1 line
2.17.1
#2 lore raw
Quoted 44 lines · Yonghong Song

Commit 8b401f9ed244 ("bpf: implement bpf_send_signal() helper") added helper bpf_send_signal() which permits bpf program to send a signal to the current process.

We found a use case where sending the signal to the current thread is more preferable.
- A bpf program will collect the stack trace and then send signal to the user application.
- The user application will add some thread specific information to the just collected stack trace for later analysis.

If bpf_send_signal() is used, user application will need to check whether the thread receiving the signal matches the thread collecting the stack by checking thread id. If not, it will need to send signal to another thread through pthread_kill().

This patch proposed a new helper bpf_send_signal_thread(), which sends the signal to the current thread. This way, user space is guaranteed that bpf_program execution context and user space signal handling context are the same thread.

---
 include/uapi/linux/bpf.h | 18 ++++++++++++++++--
 kernel/trace/bpf_trace.c | 27 ++++++++++++++++++++++++---
 2 files changed, 40 insertions(+), 5 deletions(-)
include/uapi/linux/bpf.h +1 -1
index 52966e758fe5..3320f8bdfe7e 100644
@@ -2714,7 +2714,7 @@ union bpf_attr {
  *
  * int bpf_send_signal(u32 sig)
  *     Description
- *             Send signal *sig* to the current task.
+ *             Send signal *sig* to the process of the current task.
  *     Return
  *             0 on success or successfully queued.
  *

include/uapi/linux/bpf.h union bpf_attr { · line 2855

Quoted 6 lines
include/uapi/linux/bpf.h +3 -0
@@ -2850,6 +2850,19 @@ union bpf_attr {
  *     Return
  *             0 on success, or a negative error in case of failure.
  *
+ * int bpf_send_signal_thread(u32 sig)
+ *     Description
+ *             Send signal *sig* to the current task.

This all makes sense and looks good, but I think it's very unclear why the distinction between sending signal to process vs thread. Could you extend bpf_send_signal and bpf_send_signal_thread descriptions explaining the difference (e.g., that, according to POSIX, when sending signal to a process, any thread within that process can get signal delivered, while sending to a specific thread will ensure that that specific thread will receive desired signal).

[...]

#3 lore raw
Quoted 53 lines · Andrii Nakryiko
Quoted 44 lines · Yonghong Song

Commit 8b401f9ed244 ("bpf: implement bpf_send_signal() helper") added helper bpf_send_signal() which permits bpf program to send a signal to the current process.

We found a use case where sending the signal to the current thread is more preferable.
- A bpf program will collect the stack trace and then send signal to the user application.
- The user application will add some thread specific information to the just collected stack trace for later analysis.

If bpf_send_signal() is used, user application will need to check whether the thread receiving the signal matches the thread collecting the stack by checking thread id. If not, it will need to send signal to another thread through pthread_kill().

This patch proposed a new helper bpf_send_signal_thread(), which sends the signal to the current thread. This way, user space is guaranteed that bpf_program execution context and user space signal handling context are the same thread.

---
  include/uapi/linux/bpf.h | 18 ++++++++++++++++--
  kernel/trace/bpf_trace.c | 27 ++++++++++++++++++++++++---
  2 files changed, 40 insertions(+), 5 deletions(-)
include/uapi/linux/bpf.h +1 -1
index 52966e758fe5..3320f8bdfe7e 100644
@@ -2714,7 +2714,7 @@ union bpf_attr {
   *
   * int bpf_send_signal(u32 sig)
   *     Description
- *             Send signal *sig* to the current task.
+ *             Send signal *sig* to the process of the current task.
   *     Return
   *             0 on success or successfully queued.
   *

include/uapi/linux/bpf.h union bpf_attr { · line 2855

Quoted 6 lines
include/uapi/linux/bpf.h +3 -0
@@ -2850,6 +2850,19 @@ union bpf_attr {
   *     Return
   *             0 on success, or a negative error in case of failure.
   *
+ * int bpf_send_signal_thread(u32 sig)
+ *     Description
+ *             Send signal *sig* to the current task.
Quoted 7 lines

This all makes sense and looks good, but I think it's very unclear why the distinction between sending signal to process vs thread. Could you extend bpf_send_signal and bpf_send_signal_thread descriptions explaining the difference (e.g., that, according to POSIX, when sending signal to a process, any thread within that process can get signal delivered, while sending to a specific thread will ensure that that specific thread will receive desired signal).

Sounds good. Will send v2 with better descriptions later.

Quoted 3 lines

[...]

#4 lore raw

PATCH 2/2 tools/bpf: add a selftest for bpf_send_signal_thread()

In reply to Yonghong Song

The test_progs send_signal() is amended to test
bpf_send_signal_thread() as well.

  $ ./test_progs -n 40
  #40/1 send_signal_tracepoint:OK
  #40/2 send_signal_perf:OK
  #40/3 send_signal_nmi:OK
  #40/4 send_signal_tracepoint_thread:OK
  #40/5 send_signal_perf_thread:OK
  #40/6 send_signal_nmi_thread:OK
  #40 send_signal:OK
  Summary: 1/6 PASSED, 0 SKIPPED, 0 FAILED
---
 tools/include/uapi/linux/bpf.h                | 18 +++++++++--
 .../selftests/bpf/prog_tests/send_signal.c    | 30 ++++++++++++-------
 .../bpf/progs/test_send_signal_kern.c         |  9 ++++--
 3 files changed, 42 insertions(+), 15 deletions(-)
tools/include/uapi/linux/bpf.h +16 -2
index 52966e758fe5..3320f8bdfe7e 100644
@@ -2714,7 +2714,7 @@ union bpf_attr {
  *
  * int bpf_send_signal(u32 sig)
  *	Description
- *		Send signal *sig* to the current task.
+ *		Send signal *sig* to the process of the current task.
  *	Return
  *		0 on success or successfully queued.
  *
@@ -2850,6 +2850,19 @@ union bpf_attr {
  *	Return
  *		0 on success, or a negative error in case of failure.
  *
+ * int bpf_send_signal_thread(u32 sig)
+ *	Description
+ *		Send signal *sig* to the current task.
+ *	Return
+ *		0 on success or successfully queued.
+ *
+ *		**-EBUSY** if work queue under nmi is full.
+ *
+ *		**-EINVAL** if *sig* is invalid.
+ *
+ *		**-EPERM** if no permission to send the *sig*.
+ *
+ *		**-EAGAIN** if bpf program can try again.
  */
 #define __BPF_FUNC_MAPPER(FN)		\
 	FN(unspec),			\
@@ -2968,7 +2981,8 @@ union bpf_attr {
 	FN(probe_read_kernel),		\
 	FN(probe_read_user_str),	\
 	FN(probe_read_kernel_str),	\
-	FN(tcp_send_ack),
+	FN(tcp_send_ack),		\
+	FN(send_signal_thread),
 
 /* integer value in 'imm' field of BPF_CALL instruction selects which helper
  * function eBPF program intends to call
tools/testing/selftests/bpf/prog_tests/send_signal.c +19 -11
index b607112c64e7..b12350832be3 100644
@@ -10,7 +10,8 @@ static void sigusr1_handler(int signum)
 
 static void test_send_signal_common(struct perf_event_attr *attr,
 				    int prog_type,
-				    const char *test_name)
+				    const char *test_name,
+				    bool send_thread)
 {
 	int err = -1, pmu_fd, prog_fd, info_map_fd, status_map_fd;
 	const char *file = "./test_send_signal_kern.o";
@@ -110,7 +111,7 @@ static void test_send_signal_common(struct perf_event_attr *attr,
 
 	/* trigger the bpf send_signal */
 	key = 0;
-	val = (((__u64)(SIGUSR1)) << 32) | pid;
+	val = (((__u64)send_thread) << 48) | (((__u64)(SIGUSR1)) << 32) | pid;
 	bpf_map_update_elem(info_map_fd, &key, &val, 0);
 
 	/* notify child that bpf program can send_signal now */
@@ -140,7 +141,7 @@ static void test_send_signal_common(struct perf_event_attr *attr,
 	wait(NULL);
 }
 
-static void test_send_signal_tracepoint(void)
+static void test_send_signal_tracepoint(bool send_thread)
 {
 	const char *id_path = "/sys/kernel/debug/tracing/events/syscalls/sys_enter_nanosleep/id";
 	struct perf_event_attr attr = {
@@ -168,10 +169,11 @@ static void test_send_signal_tracepoint(void)
 
 	attr.config = strtol(buf, NULL, 0);
 
-	test_send_signal_common(&attr, BPF_PROG_TYPE_TRACEPOINT, "tracepoint");
+	test_send_signal_common(&attr, BPF_PROG_TYPE_TRACEPOINT, "tracepoint",
+				send_thread);
 }
 
-static void test_send_signal_perf(void)
+static void test_send_signal_perf(bool send_thread)
 {
 	struct perf_event_attr attr = {
 		.sample_period = 1,
@@ -180,10 +182,10 @@ static void test_send_signal_perf(void)
 	};
 
 	test_send_signal_common(&attr, BPF_PROG_TYPE_PERF_EVENT,
-				"perf_sw_event");
+				"perf_sw_event", send_thread);
 }
 
-static void test_send_signal_nmi(void)
+static void test_send_signal_nmi(bool send_thread)
 {
 	struct perf_event_attr attr = {
 		.sample_freq = 50,
@@ -211,15 +213,21 @@ static void test_send_signal_nmi(void)
 	}
 
 	test_send_signal_common(&attr, BPF_PROG_TYPE_PERF_EVENT,
-				"perf_hw_event");
+				"perf_hw_event", send_thread);
 }
 
 void test_send_signal(void)
 {
 	if (test__start_subtest("send_signal_tracepoint"))
-		test_send_signal_tracepoint();
+		test_send_signal_tracepoint(false);
 	if (test__start_subtest("send_signal_perf"))
-		test_send_signal_perf();
+		test_send_signal_perf(false);
 	if (test__start_subtest("send_signal_nmi"))
-		test_send_signal_nmi();
+		test_send_signal_nmi(false);
+	if (test__start_subtest("send_signal_tracepoint_thread"))
+		test_send_signal_tracepoint(true);
+	if (test__start_subtest("send_signal_perf_thread"))
+		test_send_signal_perf(true);
+	if (test__start_subtest("send_signal_nmi_thread"))
+		test_send_signal_nmi(true);
 }
tools/testing/selftests/bpf/progs/test_send_signal_kern.c +7 -2
index 0e6be01157e6..4a722024c32b 100644
@@ -23,6 +23,7 @@ int bpf_send_signal_test(void *ctx)
 {
 	__u64 *info_val, *status_val;
 	__u32 key = 0, pid, sig;
+	int use_signal_thread;
 	int ret;
 
 	status_val = bpf_map_lookup_elem(&status_map, &key);
@@ -33,11 +34,15 @@ int bpf_send_signal_test(void *ctx)
 	if (!info_val || *info_val == 0)
 		return 0;
 
-	sig = *info_val >> 32;
+	use_signal_thread = *info_val >> 48;
+	sig = *info_val >> 32 & 0xFFFF;
 	pid = *info_val & 0xffffFFFF;
 
 	if ((bpf_get_current_pid_tgid() >> 32) == pid) {
-		ret = bpf_send_signal(sig);
+		if (use_signal_thread)
+			ret = bpf_send_signal_thread(sig);
+		else
+			ret = bpf_send_signal(sig);
 		if (ret == 0)
 			*status_val = 1;
 	}
Signature · 1 line
2.17.1
#5 lore raw

Re: [PATCH bpf-next 2/2] tools/bpf: add a selftest for bpf_send_signal_thread()

Quoted 14 lines · Yonghong Song

The test_progs send_signal() is amended to test
bpf_send_signal_thread() as well.

  $ ./test_progs -n 40
  #40/1 send_signal_tracepoint:OK
  #40/2 send_signal_perf:OK
  #40/3 send_signal_nmi:OK
  #40/4 send_signal_tracepoint_thread:OK
  #40/5 send_signal_perf_thread:OK
  #40/6 send_signal_nmi_thread:OK
  #40 send_signal:OK
  Summary: 1/6 PASSED, 0 SKIPPED, 0 FAILED
Quoted 3 lines
---
 tools/include/uapi/linux/bpf.h                | 18 +++++++++--

maybe do tools/uapi header sync in a first patch, along the original change?

Quoted 4 lines
 .../selftests/bpf/prog_tests/send_signal.c    | 30 ++++++++++++-------
 .../bpf/progs/test_send_signal_kern.c         |  9 ++++--
 3 files changed, 42 insertions(+), 15 deletions(-)

[...]

Quoted 13 lines
tools/testing/selftests/bpf/progs/test_send_signal_kern.c +1 -0
index 0e6be01157e6..4a722024c32b 100644
@@ -23,6 +23,7 @@ int bpf_send_signal_test(void *ctx)
 {
        __u64 *info_val, *status_val;
        __u32 key = 0, pid, sig;
+       int use_signal_thread;
        int ret;
 
        status_val = bpf_map_lookup_elem(&status_map, &key);

tools/testing/selftests/bpf/progs/test_send_signal_kern.c int bpf_send_signal_test(void *ctx) · line 39

Quoted 7 lines
tools/testing/selftests/bpf/progs/test_send_signal_kern.c +2 -1
@@ -33,11 +34,15 @@ int bpf_send_signal_test(void *ctx)
        if (!info_val || *info_val == 0)
                return 0;
 
-       sig = *info_val >> 32;
+       use_signal_thread = *info_val >> 48;
+       sig = *info_val >> 32 & 0xFFFF;
        pid = *info_val & 0xffffFFFF;

Would you mind rewriting this test w/ BPF skeleton and global data? It would make it cleaner without all this masking stuff?

Quoted 13 lines
tools/testing/selftests/bpf/progs/test_send_signal_kern.c +4 -2
        if ((bpf_get_current_pid_tgid() >> 32) == pid) {
-               ret = bpf_send_signal(sig);
+               if (use_signal_thread)
+                       ret = bpf_send_signal_thread(sig);
+               else
+                       ret = bpf_send_signal(sig);
                if (ret == 0)
                        *status_val = 1;
        }
--

2.17.1

#6 lore raw

Re: [PATCH bpf-next 2/2] tools/bpf: add a selftest for bpf_send_signal_thread()

Quoted 15 lines · Andrii Nakryiko
Quoted 14 lines · Yonghong Song

The test_progs send_signal() is amended to test
bpf_send_signal_thread() as well.

   $ ./test_progs -n 40
   #40/1 send_signal_tracepoint:OK
   #40/2 send_signal_perf:OK
   #40/3 send_signal_nmi:OK
   #40/4 send_signal_tracepoint_thread:OK
   #40/5 send_signal_perf_thread:OK
   #40/6 send_signal_nmi_thread:OK
   #40 send_signal:OK
   Summary: 1/6 PASSED, 0 SKIPPED, 0 FAILED
Quoted 5 lines
Quoted 3 lines
---
  tools/include/uapi/linux/bpf.h                | 18 +++++++++--

maybe do tools/uapi header sync in a first patch, along the original change?

Will do.

Quoted 24 lines
Quoted 5 lines
  .../selftests/bpf/prog_tests/send_signal.c    | 30 ++++++++++++-------
  .../bpf/progs/test_send_signal_kern.c         |  9 ++++--
  3 files changed, 42 insertions(+), 15 deletions(-)

[...]

Quoted 16 lines
tools/testing/selftests/bpf/progs/test_send_signal_kern.c +1 -0
index 0e6be01157e6..4a722024c32b 100644
@@ -23,6 +23,7 @@ int bpf_send_signal_test(void *ctx)
  {
         __u64 *info_val, *status_val;
         __u32 key = 0, pid, sig;
+       int use_signal_thread;
         int ret;
 
         status_val = bpf_map_lookup_elem(&status_map, &key);
@@ -33,11 +34,15 @@ int bpf_send_signal_test(void *ctx)
         if (!info_val || *info_val == 0)
                 return 0;
 
Quoted 7 lines

tools/testing/selftests/bpf/progs/test_send_signal_kern.c int bpf_send_signal_test(void *ctx) · line 39

Quoted 4 lines
tools/testing/selftests/bpf/progs/test_send_signal_kern.c +2 -1
-       sig = *info_val >> 32;
+       use_signal_thread = *info_val >> 48;
+       sig = *info_val >> 32 & 0xFFFF;
         pid = *info_val & 0xffffFFFF;

Would you mind rewriting this test w/ BPF skeleton and global data? It would make it cleaner without all this masking stuff?

Previously I made the change to minimize the number of changed lines. But since you are mentioning rewriting here, I will do it in v2.

Quoted 14 lines
Quoted 14 lines
tools/testing/selftests/bpf/progs/test_send_signal_kern.c +4 -2
         if ((bpf_get_current_pid_tgid() >> 32) == pid) {
-               ret = bpf_send_signal(sig);
+               if (use_signal_thread)
+                       ret = bpf_send_signal_thread(sig);
+               else
+                       ret = bpf_send_signal(sig);
                 if (ret == 0)
                         *status_val = 1;
         }
--

2.17.1