[PATCH 19/24] ptrace: Allow compat to use the native siginfo
Catalin Marinas
catalin.marinas at arm.com
Thu Oct 2 07:13:29 PDT 2014
On Wed, Sep 03, 2014 at 10:19:13PM +0100, Andrew Pinski wrote:
> diff --git a/kernel/ptrace.c b/kernel/ptrace.c
> index 54e7522..40ae023 100644
> --- a/kernel/ptrace.c
> +++ b/kernel/ptrace.c
> @@ -657,7 +657,7 @@ static int ptrace_peek_siginfo(struct task_struct *child,
> break;
>
> #ifdef CONFIG_COMPAT
> - if (unlikely(is_compat_task())) {
> + if (unlikely(is_compat_task() && !COMPAT_USE_NATIVE_SIGINFO)) {
> compat_siginfo_t __user *uinfo = compat_ptr(data);
>
> if (copy_siginfo_to_user32(uinfo, &info) ||
> @@ -1126,16 +1126,27 @@ int compat_ptrace_request(struct task_struct *child, compat_long_t request,
>
> case PTRACE_GETSIGINFO:
> ret = ptrace_getsiginfo(child, &siginfo);
> - if (!ret)
> - ret = copy_siginfo_to_user32(
> - (struct compat_siginfo __user *) datap,
> - &siginfo);
> + if (!ret) {
> + if (COMPAT_USE_NATIVE_SIGINFO)
> + ret = copy_siginfo_to_user(
> + (struct siginfo __user *) datap,
> + &siginfo);
> + else
> + ret = copy_siginfo_to_user32(
> + (struct compat_siginfo __user *) datap,
> + &siginfo);
> + }
I haven't read the subsequent patch yet but couldn't you call
ptrace_request directly from the arm64 code for ILP32 tasks instead of
compat_ptrace_request()? Is the ILP32 ptrace support the same with the
LP64 one?
--
Catalin
More information about the linux-arm-kernel
mailing list