diff mbox series

[07/10] arm64: stacktrace: report source of unwind data

Message ID 20241010101510.1487477-8-mark.rutland@arm.com (mailing list archive)
State New, archived
Headers show
Series arm64: stacktrace: improve unwind reporting | expand

Commit Message

Mark Rutland Oct. 10, 2024, 10:15 a.m. UTC
When analysing a stacktrace it can be useful to know where an unwound PC
came from, as in some situations certain sources may be suspect or known
to be unreliable. In future it would also be useful to track this so
that certain unwind steps can be performed in a stateful manner. For
example when unwinding across an exception boundary, we'd ideally unwind
pt_regs::pc, then pt_regs::lr, then the next frame record.

This patch adds an enumerated set of unwind sources, tracks this during
the unwind, and updates dump_backtrace() to log these for interesting
unwind steps.

The interesting sources recorded are:

 "C" - the PC came from the caller of an unwind function.
 "T" - the PC came from thread_saved_pc() for a blocked task.
 "P" - the PC came from a pt_regs::pc.
 "U" - the PC came from an unknown source (indicates an unwinder error).

... with nothing recorded when the PC came from a frame_record::pc as
this is the vastly common case and logging this would make it difficult
to spot the more interesting cases.

For example, when triggering a backtrace via magic-sysrq + L, the CPU
handling the sysrq will have a backtrace whose first element is the
caller (C) of dump_backtrace():

| Call trace:
|  show_stack+0x18/0x30 (C)
|  dump_stack_lvl+0x60/0x80
|  dump_stack+0x18/0x24
|  nmi_cpu_backtrace+0xfc/0x140
| ...

... and other CPUs will have a backtrace whose first element is their
pt_regs::pc (P) at the instant the backtrace IPI was taken:

| Call trace:
|  _raw_spin_unlock_irqrestore+0x8/0x50 (P)
|  wake_up_process+0x18/0x24
|  process_timeout+0x14/0x20
|  call_timer_fn.isra.0+0x24/0x80
| ...

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Josh Poimboeuf <jpoimboe@kernel.org>
Cc: Kalesh Singh <kaleshsingh@google.com>
Cc: Madhavan T. Venkataraman <madvenka@linux.microsoft.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Miroslav Benes <mbenes@suse.cz>
Cc: Puranjay Mohan <puranjay12@gmail.com>
Cc: Will Deacon <will@kernel.org>
---
 arch/arm64/kernel/stacktrace.c | 47 +++++++++++++++++++++++++++++++---
 1 file changed, 43 insertions(+), 4 deletions(-)

Comments

Mark Brown Oct. 10, 2024, 3:13 p.m. UTC | #1
On Thu, Oct 10, 2024 at 11:15:07AM +0100, Mark Rutland wrote:

> When analysing a stacktrace it can be useful to know where an unwound PC
> came from, as in some situations certain sources may be suspect or known
> to be unreliable. In future it would also be useful to track this so
> that certain unwind steps can be performed in a stateful manner. For
> example when unwinding across an exception boundary, we'd ideally unwind
> pt_regs::pc, then pt_regs::lr, then the next frame record.

Reviewed-by: Mark Brown <broonie@kernel.org>
diff mbox series

Patch

diff --git a/arch/arm64/kernel/stacktrace.c b/arch/arm64/kernel/stacktrace.c
index 22f8709f0e763..83ab37e13159e 100644
--- a/arch/arm64/kernel/stacktrace.c
+++ b/arch/arm64/kernel/stacktrace.c
@@ -20,6 +20,14 @@ 
 #include <asm/stack_pointer.h>
 #include <asm/stacktrace.h>
 
+enum kunwind_source {
+	KUNWIND_SOURCE_UNKNOWN,
+	KUNWIND_SOURCE_FRAME,
+	KUNWIND_SOURCE_CALLER,
+	KUNWIND_SOURCE_TASK,
+	KUNWIND_SOURCE_REGS_PC,
+};
+
 /*
  * Kernel unwind state
  *
@@ -37,6 +45,7 @@  struct kunwind_state {
 #ifdef CONFIG_KRETPROBES
 	struct llist_node *kr_cur;
 #endif
+	enum kunwind_source source;
 };
 
 static __always_inline void
@@ -45,6 +54,7 @@  kunwind_init(struct kunwind_state *state,
 {
 	unwind_init_common(&state->common);
 	state->task = task;
+	state->source = KUNWIND_SOURCE_UNKNOWN;
 }
 
 /*
@@ -62,6 +72,7 @@  kunwind_init_from_regs(struct kunwind_state *state,
 
 	state->common.fp = regs->regs[29];
 	state->common.pc = regs->pc;
+	state->source = KUNWIND_SOURCE_REGS_PC;
 }
 
 /*
@@ -79,6 +90,7 @@  kunwind_init_from_caller(struct kunwind_state *state)
 
 	state->common.fp = (unsigned long)__builtin_frame_address(1);
 	state->common.pc = (unsigned long)__builtin_return_address(0);
+	state->source = KUNWIND_SOURCE_CALLER;
 }
 
 /*
@@ -99,6 +111,7 @@  kunwind_init_from_task(struct kunwind_state *state,
 
 	state->common.fp = thread_saved_fp(task);
 	state->common.pc = thread_saved_pc(task);
+	state->source = KUNWIND_SOURCE_TASK;
 }
 
 static __always_inline int
@@ -148,9 +161,19 @@  kunwind_next(struct kunwind_state *state)
 	if (fp == (unsigned long)&task_pt_regs(tsk)->stackframe)
 		return -ENOENT;
 
-	err = unwind_next_frame_record(&state->common);
-	if (err)
-		return err;
+	switch (state->source) {
+	case KUNWIND_SOURCE_FRAME:
+	case KUNWIND_SOURCE_CALLER:
+	case KUNWIND_SOURCE_TASK:
+	case KUNWIND_SOURCE_REGS_PC:
+		err = unwind_next_frame_record(&state->common);
+		if (err)
+			return err;
+		state->source = KUNWIND_SOURCE_FRAME;
+		break;
+	default:
+		return -EINVAL;
+	}
 
 	state->common.pc = ptrauth_strip_kernel_insn_pac(state->common.pc);
 
@@ -294,10 +317,26 @@  noinline noinstr void arch_bpf_stack_walk(bool (*consume_entry)(void *cookie, u6
 	kunwind_stack_walk(arch_bpf_unwind_consume_entry, &data, current, NULL);
 }
 
+static const char *state_source_string(const struct kunwind_state *state)
+{
+	switch (state->source) {
+	case KUNWIND_SOURCE_FRAME:	return NULL;
+	case KUNWIND_SOURCE_CALLER:	return "C";
+	case KUNWIND_SOURCE_TASK:	return "T";
+	case KUNWIND_SOURCE_REGS_PC:	return "P";
+	default:			return "U";
+	}
+}
+
 static bool dump_backtrace_entry(const struct kunwind_state *state, void *arg)
 {
+	const char *source = state_source_string(state);
 	char *loglvl = arg;
-	printk("%s %pSb\n", loglvl, (void *)state->common.pc);
+	printk("%s %pSb%s%s%s\n", loglvl,
+		(void *)state->common.pc,
+		source ? " (" : "",
+		source ? source : "",
+		source ? ")" : "");
 	return true;
 }