Depends on:
http://lists.infradead.org/pipermail/linux-riscv/2018-October/001931.html
Why we don't have HAVE_SYSCALL_TRACEPOINTS in arch/riscv/Kconfig?
Signed-off-by: David Abdurachmanov <[email protected]>
---
arch/riscv/kernel/ptrace.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/riscv/kernel/ptrace.c b/arch/riscv/kernel/ptrace.c
index 9f82a7e34c64..9db7d0076375 100644
--- a/arch/riscv/kernel/ptrace.c
+++ b/arch/riscv/kernel/ptrace.c
@@ -120,6 +120,6 @@ void do_syscall_trace_exit(struct pt_regs *regs)
#ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS
if (test_thread_flag(TIF_SYSCALL_TRACEPOINT))
- trace_sys_exit(regs, regs->regs[0]);
+ trace_sys_exit(regs, regs_return_value(regs));
#endif
}
--
2.17.2
On Mon, Oct 29, 2018 at 1:15 PM David Abdurachmanov
<[email protected]> wrote:
>
> Depends on:
> http://lists.infradead.org/pipermail/linux-riscv/2018-October/001931.html
>
> Why we don't have HAVE_SYSCALL_TRACEPOINTS in arch/riscv/Kconfig?
>
I looked this morning into Documentation/trace/ftrace-design.rst again.
Of all the things the document describes we are only missing
HAVE_SYSCALL_TRACEPOINTS in Kconfig, yet we already have it
implemented in ptrace.c (+ this minor fix). Reading document, I think,
we check all the boxes needed for HAVE_SYSCALL_TRACEPOINTS.
I assume it was just missed (?) and will do v2 later on.
david
On Mon, 29 Oct 2018 05:15:16 PDT (-0700), [email protected] wrote:
> Depends on:
> http://lists.infradead.org/pipermail/linux-riscv/2018-October/001931.html
>
> Why we don't have HAVE_SYSCALL_TRACEPOINTS in arch/riscv/Kconfig?
>
> Signed-off-by: David Abdurachmanov <[email protected]>
> ---
> arch/riscv/kernel/ptrace.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/riscv/kernel/ptrace.c b/arch/riscv/kernel/ptrace.c
> index 9f82a7e34c64..9db7d0076375 100644
> --- a/arch/riscv/kernel/ptrace.c
> +++ b/arch/riscv/kernel/ptrace.c
> @@ -120,6 +120,6 @@ void do_syscall_trace_exit(struct pt_regs *regs)
>
> #ifdef CONFIG_HAVE_SYSCALL_TRACEPOINTS
> if (test_thread_flag(TIF_SYSCALL_TRACEPOINT))
> - trace_sys_exit(regs, regs->regs[0]);
> + trace_sys_exit(regs, regs_return_value(regs));
> #endif
> }
I'm dropping this onto next-audit on top of the other two.
Thanks!