function_graph: Support recording and printing the function return address

When using function_graph tracer to analyze the flow of kernel function
execution, it is often necessary to quickly locate the exact line of code
where the call occurs. While this may be easy at times, it can be more
time-consuming when some functions are inlined or the flow is too long.

This feature aims to simplify the process by recording the return address
of traced funcions and printing it when outputing trace logs.

To enhance human readability, the prefix 'ret=' is used for the kernel return
value, while '<-' serves as the prefix for the return address in trace logs to
make it look more like the function tracer.

A new trace option named 'funcgraph-retaddr' has been introduced, and the
existing option 'sym-addr' can be used to control the format of the return
address.

See below logs with both funcgraph-retval and funcgraph-retaddr enabled.

0)             | load_elf_binary() { /* <-bprm_execve+0x249/0x600 */
0)             |   load_elf_phdrs() { /* <-load_elf_binary+0x84/0x1730 */
0)             |     __kmalloc_noprof() { /* <-load_elf_phdrs+0x4a/0xb0 */
0)   3.657 us  |       __cond_resched(); /* <-__kmalloc_noprof+0x28c/0x390 ret=0x0 */
0) + 24.335 us |     } /* __kmalloc_noprof ret=0xffff8882007f3000 */
0)             |     kernel_read() { /* <-load_elf_phdrs+0x6c/0xb0 */
0)             |       rw_verify_area() { /* <-kernel_read+0x2b/0x50 */
0)             |         security_file_permission() { /* <-kernel_read+0x2b/0x50 */
0)             |           selinux_file_permission() { /* <-security_file_permission+0x26/0x40 */
0)             |             __inode_security_revalidate() { /* <-selinux_file_permission+0x6d/0x140 */
0)   2.034 us  |               __cond_resched(); /* <-__inode_security_revalidate+0x5f/0x80 ret=0x0 */
0)   6.602 us  |             } /* __inode_security_revalidate ret=0x0 */
0)   2.214 us  |             avc_policy_seqno(); /* <-selinux_file_permission+0x107/0x140 ret=0x0 */
0) + 16.670 us |           } /* selinux_file_permission ret=0x0 */
0) + 20.809 us |         } /* security_file_permission ret=0x0 */
0) + 25.217 us |       } /* rw_verify_area ret=0x0 */
0)             |       __kernel_read() { /* <-load_elf_phdrs+0x6c/0xb0 */
0)             |         ext4_file_read_iter() { /* <-__kernel_read+0x160/0x2e0 */

Then, we can use the faddr2line to locate the source code, for example:

$ ./scripts/faddr2line ./vmlinux load_elf_phdrs+0x6c/0xb0
load_elf_phdrs+0x6c/0xb0:
elf_read at fs/binfmt_elf.c:471
(inlined by) load_elf_phdrs at fs/binfmt_elf.c:531

Link: https://lore.kernel.org/20240915032912.1118397-1-dolinux.peng@gmail.com
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/oe-kbuild-all/202409150605.HgUmU8ea-lkp@intel.com/
Signed-off-by: Donglin Peng <dolinux.peng@gmail.com>
[ Rebased to handle text_delta offsets ]
Signed-off-by: Steven Rostedt (Google) <rostedt@goodmis.org>

1 files changed, 172 insertions, 44 deletions

diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c
index 5c1b150fbba3..3dd63ae2afe8 100644
--- a/kernel/trace/trace_functions_graph.c
+++ b/kernel/trace/trace_functions_graph.c
@@ -31,7 +31,10 @@ struct fgraph_data {
 	struct fgraph_cpu_data __percpu *cpu_data;
 
 	/* Place to preserve last processed entry. */
-	struct ftrace_graph_ent_entry	ent;
+	union {
+		struct ftrace_graph_ent_entry	ent;
+		struct fgraph_retaddr_ent_entry	rent;
+	} ent;
 	struct ftrace_graph_ret_entry	ret;
 	int				failed;
 	int				cpu;
@@ -64,6 +67,10 @@ static struct tracer_opt trace_opts[] = {
 	/* Display function return value in hexadecimal format ? */
 	{ TRACER_OPT(funcgraph-retval-hex, TRACE_GRAPH_PRINT_RETVAL_HEX) },
 #endif
+#ifdef CONFIG_FUNCTION_GRAPH_RETADDR
+	/* Display function return address ? */
+	{ TRACER_OPT(funcgraph-retaddr, TRACE_GRAPH_PRINT_RETADDR) },
+#endif
 	/* Include sleep time (scheduled out) between entry and return */
 	{ TRACER_OPT(sleep-time, TRACE_GRAPH_SLEEP_TIME) },
 
@@ -83,6 +90,11 @@ static struct tracer_flags tracer_flags = {
 	.opts = trace_opts
 };
 
+u32 graph_tracer_flags_get(u32 flags)
+{
+	return tracer_flags.val & flags;
+}
+
 /*
  * DURATION column is being also used to display IRQ signs,
  * following values are used by print_graph_irq and others
@@ -119,6 +131,40 @@ int __trace_graph_entry(struct trace_array *tr,
 	return 1;
 }
 
+#ifdef CONFIG_FUNCTION_GRAPH_RETADDR
+int __trace_graph_retaddr_entry(struct trace_array *tr,
+				struct ftrace_graph_ent *trace,
+				unsigned int trace_ctx,
+				unsigned long retaddr)
+{
+	struct trace_event_call *call = &event_fgraph_retaddr_entry;
+	struct ring_buffer_event *event;
+	struct trace_buffer *buffer = tr->array_buffer.buffer;
+	struct fgraph_retaddr_ent_entry *entry;
+
+	event = trace_buffer_lock_reserve(buffer, TRACE_GRAPH_RETADDR_ENT,
+					  sizeof(*entry), trace_ctx);
+	if (!event)
+		return 0;
+	entry	= ring_buffer_event_data(event);
+	entry->graph_ent.func = trace->func;
+	entry->graph_ent.depth = trace->depth;
+	entry->graph_ent.retaddr = retaddr;
+	if (!call_filter_check_discard(call, entry, buffer, event))
+		trace_buffer_unlock_commit_nostack(buffer, event);
+
+	return 1;
+}
+#else
+int __trace_graph_retaddr_entry(struct trace_array *tr,
+				struct ftrace_graph_ent *trace,
+				unsigned int trace_ctx,
+				unsigned long retaddr)
+{
+	return 1;
+}
+#endif
+
 static inline int ftrace_graph_ignore_irqs(void)
 {
 	if (!ftrace_graph_skip_irqs || trace_recursion_test(TRACE_IRQ_BIT))
@@ -133,7 +179,8 @@ struct fgraph_times {
 };
 
 int trace_graph_entry(struct ftrace_graph_ent *trace,
-		      struct fgraph_ops *gops)
+		      struct fgraph_ops *gops,
+		      struct fgraph_extras *extras)
 {
 	unsigned long *task_var = fgraph_get_task_var(gops);
 	struct trace_array *tr = gops->private;
@@ -199,7 +246,12 @@ int trace_graph_entry(struct ftrace_graph_ent *trace,
 	disabled = atomic_inc_return(&data->disabled);
 	if (likely(disabled == 1)) {
 		trace_ctx = tracing_gen_ctx_flags(flags);
-		ret = __trace_graph_entry(tr, trace, trace_ctx);
+		if (unlikely(IS_ENABLED(CONFIG_FUNCTION_GRAPH_RETADDR) && extras
+				&& (extras->flags & TRACE_GRAPH_PRINT_RETADDR)))
+			ret = __trace_graph_retaddr_entry(tr, trace, trace_ctx,
+							  extras->retaddr);
+		else
+			ret = __trace_graph_entry(tr, trace, trace_ctx);
 	} else {
 		ret = 0;
 	}
@@ -507,7 +559,7 @@ get_return_for_leaf(struct trace_iterator *iter,
 	 * then we just reuse the data from before.
 	 */
 	if (data && data->failed) {
-		curr = &data->ent;
+		curr = &data->ent.ent;
 		next = &data->ret;
 	} else {
 
@@ -537,7 +589,10 @@ get_return_for_leaf(struct trace_iterator *iter,
 			 * Save current and next entries for later reference
 			 * if the output fails.
 			 */
-			data->ent = *curr;
+			if (unlikely(curr->ent.type == TRACE_GRAPH_RETADDR_ENT))
+				data->ent.rent = *(struct fgraph_retaddr_ent_entry *)curr;
+			else
+				data->ent.ent = *curr;
 			/*
 			 * If the next event is not a return type, then
 			 * we only care about what type it is. Otherwise we can
@@ -701,52 +756,96 @@ print_graph_duration(struct trace_array *tr, unsigned long long duration,
 }
 
 #ifdef CONFIG_FUNCTION_GRAPH_RETVAL
-
 #define __TRACE_GRAPH_PRINT_RETVAL TRACE_GRAPH_PRINT_RETVAL
+#else
+#define __TRACE_GRAPH_PRINT_RETVAL 0
+#endif
 
-static void print_graph_retval(struct trace_seq *s, unsigned long retval,
-				bool leaf, void *func, bool hex_format)
+#ifdef CONFIG_FUNCTION_GRAPH_RETADDR
+#define __TRACE_GRAPH_PRINT_RETADDR TRACE_GRAPH_PRINT_RETADDR
+static void print_graph_retaddr(struct trace_seq *s, struct fgraph_retaddr_ent_entry *entry,
+				u32 trace_flags, bool comment)
+{
+	if (comment)
+		trace_seq_puts(s, " /*");
+
+	trace_seq_puts(s, " <-");
+	seq_print_ip_sym(s, entry->graph_ent.retaddr, trace_flags | TRACE_ITER_SYM_OFFSET);
+
+	if (comment)
+		trace_seq_puts(s, " */");
+}
+#else
+#define __TRACE_GRAPH_PRINT_RETADDR 0
+#define print_graph_retaddr(_seq, _entry, _tflags, _comment)		do { } while (0)
+#endif
+
+#if defined(CONFIG_FUNCTION_GRAPH_RETVAL) || defined(CONFIG_FUNCTION_GRAPH_RETADDR)
+
+static void print_graph_retval(struct trace_seq *s, struct ftrace_graph_ent_entry *entry,
+				struct ftrace_graph_ret *graph_ret, void *func,
+				u32 opt_flags, u32 trace_flags)
 {
 	unsigned long err_code = 0;
+	unsigned long retval = 0;
+	bool print_retaddr = false;
+	bool print_retval = false;
+	bool hex_format = !!(opt_flags & TRACE_GRAPH_PRINT_RETVAL_HEX);
 
-	if (retval == 0 || hex_format)
-		goto done;
+#ifdef CONFIG_FUNCTION_GRAPH_RETVAL
+	retval = graph_ret->retval;
+	print_retval = !!(opt_flags & TRACE_GRAPH_PRINT_RETVAL);
+#endif
 
-	/* Check if the return value matches the negative format */
-	if (IS_ENABLED(CONFIG_64BIT) && (retval & BIT(31)) &&
-		(((u64)retval) >> 32) == 0) {
-		/* sign extension */
-		err_code = (unsigned long)(s32)retval;
-	} else {
-		err_code = retval;
+#ifdef CONFIG_FUNCTION_GRAPH_RETADDR
+	print_retaddr = !!(opt_flags & TRACE_GRAPH_PRINT_RETADDR);
+#endif
+
+	if (print_retval && retval && !hex_format) {
+		/* Check if the return value matches the negative format */
+		if (IS_ENABLED(CONFIG_64BIT) && (retval & BIT(31)) &&
+			(((u64)retval) >> 32) == 0) {
+			err_code = sign_extend64(retval, 31);
+		} else {
+			err_code = retval;
+		}
+
+		if (!IS_ERR_VALUE(err_code))
+			err_code = 0;
 	}
 
-	if (!IS_ERR_VALUE(err_code))
-		err_code = 0;
+	if (entry) {
+		if (entry->ent.type != TRACE_GRAPH_RETADDR_ENT)
+			print_retaddr = false;
 
-done:
-	if (leaf) {
-		if (hex_format || (err_code == 0))
-			trace_seq_printf(s, "%ps(); /* = 0x%lx */\n",
-					func, retval);
+		trace_seq_printf(s, "%ps();", func);
+		if (print_retval || print_retaddr)
+			trace_seq_puts(s, " /*");
 		else
-			trace_seq_printf(s, "%ps(); /* = %ld */\n",
-					func, err_code);
+			trace_seq_putc(s, '\n');
 	} else {
+		print_retaddr = false;
+		trace_seq_printf(s, "} /* %ps", func);
+	}
+
+	if (print_retaddr)
+		print_graph_retaddr(s, (struct fgraph_retaddr_ent_entry *)entry,
+				    trace_flags, false);
+
+	if (print_retval) {
 		if (hex_format || (err_code == 0))
-			trace_seq_printf(s, "} /* %ps = 0x%lx */\n",
-					func, retval);
+			trace_seq_printf(s, " ret=0x%lx", retval);
 		else
-			trace_seq_printf(s, "} /* %ps = %ld */\n",
-					func, err_code);
+			trace_seq_printf(s, " ret=%ld", err_code);
 	}
+
+	if (!entry || print_retval || print_retaddr)
+		trace_seq_puts(s, " */\n");
 }
 
 #else
 
-#define __TRACE_GRAPH_PRINT_RETVAL 0
-
-#define print_graph_retval(_seq, _retval, _leaf, _func, _format) do {} while (0)
+#define print_graph_retval(_seq, _ent, _ret, _func, _opt_flags, _trace_flags) do {} while (0)
 
 #endif
 
@@ -798,14 +897,15 @@ print_graph_entry_leaf(struct trace_iterator *iter,
 		trace_seq_putc(s, ' ');
 
 	/*
-	 * Write out the function return value if the option function-retval is
-	 * enabled.
+	 * Write out the function return value or return address
 	 */
-	if (flags & __TRACE_GRAPH_PRINT_RETVAL)
-		print_graph_retval(s, graph_ret->retval, true, (void *)func,
-				!!(flags & TRACE_GRAPH_PRINT_RETVAL_HEX));
-	else
+	if (flags & (__TRACE_GRAPH_PRINT_RETVAL | __TRACE_GRAPH_PRINT_RETADDR)) {
+		print_graph_retval(s, entry, graph_ret,
+				   (void *)graph_ret->func + iter->tr->text_delta,
+				   flags, tr->trace_flags);
+	} else {
 		trace_seq_printf(s, "%ps();\n", (void *)func);
+	}
 
 	print_graph_irq(iter, graph_ret->func, TRACE_GRAPH_RET,
 			cpu, iter->ent->pid, flags);
@@ -846,7 +946,12 @@ print_graph_entry_nested(struct trace_iterator *iter,
 
 	func = call->func + iter->tr->text_delta;
 
-	trace_seq_printf(s, "%ps() {\n", (void *)func);
+	trace_seq_printf(s, "%ps() {", (void *)func);
+	if (flags & __TRACE_GRAPH_PRINT_RETADDR  &&
+		entry->ent.type == TRACE_GRAPH_RETADDR_ENT)
+		print_graph_retaddr(s, (struct fgraph_retaddr_ent_entry *)entry,
+			tr->trace_flags, true);
+	trace_seq_putc(s, '\n');
 
 	if (trace_seq_has_overflowed(s))
 		return TRACE_TYPE_PARTIAL_LINE;
@@ -1093,11 +1198,10 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s,
 
 	/*
 	 * Always write out the function name and its return value if the
-	 * function-retval option is enabled.
+	 * funcgraph-retval option is enabled.
 	 */
 	if (flags & __TRACE_GRAPH_PRINT_RETVAL) {
-		print_graph_retval(s, trace->retval, false, (void *)func,
-			!!(flags & TRACE_GRAPH_PRINT_RETVAL_HEX));
+		print_graph_retval(s, NULL, trace, (void *)func, flags, tr->trace_flags);
 	} else {
 		/*
 		 * If the return function does not have a matching entry,
@@ -1212,7 +1316,7 @@ print_graph_function_flags(struct trace_iterator *iter, u32 flags)
 	 * to print out the missing entry which would never go out.
 	 */
 	if (data && data->failed) {
-		field = &data->ent;
+		field = &data->ent.ent;
 		iter->cpu = data->cpu;
 		ret = print_graph_entry(field, s, iter, flags);
 		if (ret == TRACE_TYPE_HANDLED && iter->cpu != cpu) {
@@ -1236,6 +1340,16 @@ print_graph_function_flags(struct trace_iterator *iter, u32 flags)
 		saved = *field;
 		return print_graph_entry(&saved, s, iter, flags);
 	}
+#ifdef CONFIG_FUNCTION_GRAPH_RETADDR
+	case TRACE_GRAPH_RETADDR_ENT: {
+		struct fgraph_retaddr_ent_entry saved;
+		struct fgraph_retaddr_ent_entry *rfield;
+
+		trace_assign_type(rfield, entry);
+		saved = *rfield;
+		return print_graph_entry((struct ftrace_graph_ent_entry *)&saved, s, iter, flags);
+	}
+#endif
 	case TRACE_GRAPH_RET: {
 		struct ftrace_graph_ret_entry *field;
 		trace_assign_type(field, entry);
@@ -1430,6 +1544,13 @@ static struct trace_event graph_trace_entry_event = {
 	.funcs		= &graph_functions,
 };
 
+#ifdef CONFIG_FUNCTION_GRAPH_RETADDR
+static struct trace_event graph_trace_retaddr_entry_event = {
+	.type		= TRACE_GRAPH_RETADDR_ENT,
+	.funcs		= &graph_functions,
+};
+#endif
+
 static struct trace_event graph_trace_ret_event = {
 	.type		= TRACE_GRAPH_RET,
 	.funcs		= &graph_functions
@@ -1516,6 +1637,13 @@ static __init int init_graph_trace(void)
 		return 1;
 	}
 
+#ifdef CONFIG_FUNCTION_GRAPH_RETADDR
+	if (!register_trace_event(&graph_trace_retaddr_entry_event)) {
+		pr_warn("Warning: could not register graph trace retaddr events\n");
+		return 1;
+	}
+#endif
+
 	if (!register_trace_event(&graph_trace_ret_event)) {
 		pr_warn("Warning: could not register graph trace events\n");
 		return 1;