In the absence of other debug facilities dumping user code around the
unhandled exception address may help debugging the issue.
Signed-off-by: Max Filippov <[email protected]>
---
arch/xtensa/Kconfig.debug | 8 ++++++++
arch/xtensa/kernel/traps.c | 18 ++++++++++++++++++
2 files changed, 26 insertions(+)
diff --git a/arch/xtensa/Kconfig.debug b/arch/xtensa/Kconfig.debug
index 83cc8d12fa0e..e84172a7763c 100644
--- a/arch/xtensa/Kconfig.debug
+++ b/arch/xtensa/Kconfig.debug
@@ -38,3 +38,11 @@ config PRINT_STACK_DEPTH
help
This option allows you to set the stack depth that the kernel
prints in stack traces.
+
+config PRINT_USER_CODE_ON_UNHANDLED_EXCEPTION
+ bool "Dump user code around unhandled exception address"
+ help
+ Enable this option to display user code around PC of the unhandled
+ exception (starting at address aligned on 16 byte boundary).
+ This may simplify finding faulting code in the absence of other
+ debug facilities.
diff --git a/arch/xtensa/kernel/traps.c b/arch/xtensa/kernel/traps.c
index a2a9a460ec9e..17eb180eff7c 100644
--- a/arch/xtensa/kernel/traps.c
+++ b/arch/xtensa/kernel/traps.c
@@ -175,6 +175,23 @@ __die_if_kernel(const char *str, struct pt_regs *regs, long err)
die(str, regs, err);
}
+#ifdef CONFIG_PRINT_USER_CODE_ON_UNHANDLED_EXCEPTION
+static inline void dump_user_code(struct pt_regs *regs)
+{
+ char buf[32];
+
+ if (copy_from_user(buf, (void __user *)(regs->pc & -16), sizeof(buf)) == 0) {
+ print_hex_dump(KERN_INFO, " ", DUMP_PREFIX_NONE,
+ 32, 1, buf, sizeof(buf), false);
+
+ }
+}
+#else
+static inline void dump_user_code(struct pt_regs *regs)
+{
+}
+#endif
+
/*
* Unhandled Exceptions. Kill user task or panic if in kernel space.
*/
@@ -190,6 +207,7 @@ void do_unhandled(struct pt_regs *regs)
"\tEXCCAUSE is %ld\n",
current->comm, task_pid_nr(current), regs->pc,
regs->exccause);
+ dump_user_code(regs);
force_sig(SIGILL);
}
--
2.30.2
Hi Marcus,
On Sun, Jun 25, 2023 at 9:36 AM Markus Elfring <[email protected]> wrote:
>
> > In the absence of other debug facilities dumping user code around the
> > unhandled exception address may help debugging the issue.
>
> Please choose an imperative change suggestion.
The change itself is so small that it's fully described in the
subject, and I believe
that that description is imperative. This part is the rationale. From
my experience
I cannot guarantee that dumping user code will help, but sometimes it
does, hence
the 'may help'.
Please let me know if you have better wording.
--
Thanks.
-- Max