[PATCH v2] riscv: Dump user opcode bytes on fatal faults

Yunhui Cui cuiyunhui at bytedance.com
Wed Mar 29 01:29:50 PDT 2023


We encountered such a problem that when the system starts to execute
init, init exits unexpectedly with error message: "unhandled signal 4
code 0x1 ...".

We are more curious about which instruction execution caused the
exception. After dumping it through show_opcodes(), we found that it
was caused by a floating-point instruction.

In this way, we found the problem: in the system bringup , it is
precisely that we have not enabled the floating point function(CONFIG_FPU
is set, but not enalbe COMPAT_HWCAP_ISA_F/D in the dts or acpi).

Like commit ba54d856a9d8 ("x86/fault: Dump user opcode bytes on fatal
faults"), when an exception occurs, it is necessary to dump the
instruction that caused the exception.

Signed-off-by: Yunhui Cui <cuiyunhui at bytedance.com>
---
 arch/riscv/include/asm/bug.h |  1 +
 arch/riscv/kernel/process.c  | 30 ++++++++++++++++++++++++++++++
 arch/riscv/kernel/traps.c    |  1 +
 3 files changed, 32 insertions(+)

diff --git a/arch/riscv/include/asm/bug.h b/arch/riscv/include/asm/bug.h
index 1aaea81fb141..56dab998d05d 100644
--- a/arch/riscv/include/asm/bug.h
+++ b/arch/riscv/include/asm/bug.h
@@ -86,6 +86,7 @@ struct pt_regs;
 struct task_struct;
 
 void __show_regs(struct pt_regs *regs);
+void show_opcodes(struct pt_regs *regs);
 void die(struct pt_regs *regs, const char *str);
 void do_trap(struct pt_regs *regs, int signo, int code, unsigned long addr);
 
diff --git a/arch/riscv/kernel/process.c b/arch/riscv/kernel/process.c
index 774ffde386ab..9ba9f8719605 100644
--- a/arch/riscv/kernel/process.c
+++ b/arch/riscv/kernel/process.c
@@ -83,6 +83,36 @@ void show_regs(struct pt_regs *regs)
 		dump_backtrace(regs, NULL, KERN_DEFAULT);
 }
 
+static int copy_code(struct pt_regs *regs, u8 *buf, unsigned long src,
+		     unsigned int nbytes)
+{
+	if (!user_mode(regs))
+		return copy_from_kernel_nofault(buf, (u8 *)src, nbytes);
+
+	/* The user space code from other tasks cannot be accessed. */
+	if (regs != task_pt_regs(current))
+		return -EPERM;
+
+	return copy_from_user_nofault(buf, (void __user *)src, nbytes);
+}
+
+void show_opcodes(struct pt_regs *regs)
+{
+	u8 opcodes[4];
+
+	switch (copy_code(regs, opcodes, regs->epc, sizeof(opcodes))) {
+	case 0:
+		pr_info("Opcode: %4ph", opcodes);
+		break;
+	case -EPERM:
+		pr_err("Unable to access userspace of other tasks");
+		break;
+	default:
+		pr_err("Failed to access opcode");
+		break;
+	}
+}
+
 #ifdef CONFIG_COMPAT
 static bool compat_mode_supported __read_mostly;
 
diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
index f6fda94e8e59..892826234ee9 100644
--- a/arch/riscv/kernel/traps.c
+++ b/arch/riscv/kernel/traps.c
@@ -100,6 +100,7 @@ void do_trap(struct pt_regs *regs, int signo, int code, unsigned long addr)
 		print_vma_addr(KERN_CONT " in ", instruction_pointer(regs));
 		pr_cont("\n");
 		__show_regs(regs);
+		show_opcodes(regs);
 	}
 
 	force_sig_fault(signo, code, (void __user *)addr);
-- 
2.20.1




More information about the linux-riscv mailing list