[PATCH] riscv: stacktrace: fix backtracing through exceptions

[PATCH] riscv: stacktrace: fix backtracing through exceptions

[Clément Léger]

Prior to commit 5d5fc33ce58e ("riscv: Improve exception and system call
latency"), backtrace through exception worked since ra was filled with
ret_from_exception symbol address and the stacktrace code checked 'pc' to
be equal to that symbol. Now that handle_exception uses regular 'call'
instructions, this isn't working anymore and backtrace stops at
handle_exception(). Since there are multiple call site to C code in the
exception handling path, rather than checking multiple potential return
addresses, add a new symbol at the end of exception handling and check pc
to be in that range.

Fixes: 5d5fc33ce58e ("riscv: Improve exception and system call latency")
Signed-off-by: Clément Léger <cleger@rivosinc.com>

---
 arch/riscv/kernel/entry.S      | 1 +
 arch/riscv/kernel/stacktrace.c | 4 +++-
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/arch/riscv/kernel/entry.S b/arch/riscv/kernel/entry.S
index 216581835eb0..33a5a9f2a0d4 100644
--- a/arch/riscv/kernel/entry.S
+++ b/arch/riscv/kernel/entry.S
@@ -278,6 +278,7 @@ SYM_CODE_START_NOALIGN(ret_from_exception)
 #else
 	sret
 #endif
+SYM_INNER_LABEL(ret_from_exception_end, SYM_L_GLOBAL)
 SYM_CODE_END(ret_from_exception)
 ASM_NOKPROBE(ret_from_exception)
 
diff --git a/arch/riscv/kernel/stacktrace.c b/arch/riscv/kernel/stacktrace.c
index 153a2db4c5fa..d4355c770c36 100644
--- a/arch/riscv/kernel/stacktrace.c
+++ b/arch/riscv/kernel/stacktrace.c
@@ -17,6 +17,7 @@
 #ifdef CONFIG_FRAME_POINTER
 
 extern asmlinkage void handle_exception(void);
+extern unsigned long ret_from_exception_end;
 
 static inline int fp_is_valid(unsigned long fp, unsigned long sp)
 {
@@ -71,7 +72,8 @@ void notrace walk_stackframe(struct task_struct *task, struct pt_regs *regs,
 			fp = frame->fp;
 			pc = ftrace_graph_ret_addr(current, &graph_idx, frame->ra,
 						   &frame->ra);
-			if (pc == (unsigned long)handle_exception) {
+			if (pc >= (unsigned long)handle_exception &&
+			    pc < (unsigned long)&ret_from_exception_end) {
 				if (unlikely(!__kernel_text_address(pc) || !fn(arg, pc)))
 					break;
 
-- 
2.45.2

Re: [PATCH] riscv: stacktrace: fix backtracing through exceptions

[Alexandre Ghiti]

Hi Clément,

On 09/12/2024 16:57, Clément Léger wrote:
> Prior to commit 5d5fc33ce58e ("riscv: Improve exception and system call
> latency"), backtrace through exception worked since ra was filled with
> ret_from_exception symbol address and the stacktrace code checked 'pc' to
> be equal to that symbol. Now that handle_exception uses regular 'call'
> instructions, this isn't working anymore and backtrace stops at
> handle_exception(). Since there are multiple call site to C code in the
> exception handling path, rather than checking multiple potential return
> addresses, add a new symbol at the end of exception handling and check pc
> to be in that range.
>
> Fixes: 5d5fc33ce58e ("riscv: Improve exception and system call latency")
> Signed-off-by: Clément Léger <cleger@rivosinc.com>
>
> ---
>   arch/riscv/kernel/entry.S      | 1 +
>   arch/riscv/kernel/stacktrace.c | 4 +++-
>   2 files changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/arch/riscv/kernel/entry.S b/arch/riscv/kernel/entry.S
> index 216581835eb0..33a5a9f2a0d4 100644
> --- a/arch/riscv/kernel/entry.S
> +++ b/arch/riscv/kernel/entry.S
> @@ -278,6 +278,7 @@ SYM_CODE_START_NOALIGN(ret_from_exception)
>   #else
>   	sret
>   #endif
> +SYM_INNER_LABEL(ret_from_exception_end, SYM_L_GLOBAL)
>   SYM_CODE_END(ret_from_exception)
>   ASM_NOKPROBE(ret_from_exception)
>   
> diff --git a/arch/riscv/kernel/stacktrace.c b/arch/riscv/kernel/stacktrace.c
> index 153a2db4c5fa..d4355c770c36 100644
> --- a/arch/riscv/kernel/stacktrace.c
> +++ b/arch/riscv/kernel/stacktrace.c
> @@ -17,6 +17,7 @@
>   #ifdef CONFIG_FRAME_POINTER
>   
>   extern asmlinkage void handle_exception(void);
> +extern unsigned long ret_from_exception_end;
>   
>   static inline int fp_is_valid(unsigned long fp, unsigned long sp)
>   {
> @@ -71,7 +72,8 @@ void notrace walk_stackframe(struct task_struct *task, struct pt_regs *regs,
>   			fp = frame->fp;
>   			pc = ftrace_graph_ret_addr(current, &graph_idx, frame->ra,
>   						   &frame->ra);
> -			if (pc == (unsigned long)handle_exception) {
> +			if (pc >= (unsigned long)handle_exception &&
> +			    pc < (unsigned long)&ret_from_exception_end) {
>   				if (unlikely(!__kernel_text_address(pc) || !fn(arg, pc)))
>   					break;
>   


You can add:

Tested-by: Alexandre Ghiti <alexghiti@rivosinc.com>
Reviewed-by: Alexandre Ghiti <alexghiti@rivosinc.com>

Thanks,

Alex

Re: [PATCH] riscv: stacktrace: fix backtracing through exceptions

[patchwork-bot+linux-riscv]

Hello:

This patch was applied to riscv/linux.git (fixes)
by Palmer Dabbelt <palmer@rivosinc.com>:

On Mon,  9 Dec 2024 16:57:12 +0100 you wrote:
> Prior to commit 5d5fc33ce58e ("riscv: Improve exception and system call
> latency"), backtrace through exception worked since ra was filled with
> ret_from_exception symbol address and the stacktrace code checked 'pc' to
> be equal to that symbol. Now that handle_exception uses regular 'call'
> instructions, this isn't working anymore and backtrace stops at
> handle_exception(). Since there are multiple call site to C code in the
> exception handling path, rather than checking multiple potential return
> addresses, add a new symbol at the end of exception handling and check pc
> to be in that range.
> 
> [...]

Here is the summary with links:
  - riscv: stacktrace: fix backtracing through exceptions
    https://git.kernel.org/riscv/c/51356ce60e59

You are awesome, thank you!
-- 
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html