Re: [PATCH] ptrace: kill BKL in ptrace syscall

[Frederic Weisbecker <fweisbec@gmail.com>]

On Sun, Mar 28, 2010 at 11:38:51PM +0200, Frederic Weisbecker wrote:
> From: Arnd Bergmann <arnd@arndb.de>
> 
> The comment suggests that this usage is stale. There is no bkl in the
> exec path so if there is a race lurking there, the bkl in ptrace is
> not going to help in this regard.
> 
> Overview of the possibility of "accidental" races this bkl might
> protect:
> 
> - ptrace_traceme() is protected against task removal and concurrent
> read/write on current->ptrace as it locks write tasklist_lock.
> 
> - arch_ptrace_attach() is serialized by ptrace_traceme() against
> concurrent PTRACE_TRACEME or PTRACE_ATTACH
> 
> - ptrace_attach() is protected the same way ptrace_traceme() and
> in turn serializes arch_ptrace_attach()
> 
> - ptrace_check_attach() does its own well described serializing too.
> 
> There is no obvious race here.


And the call to arch_ptrace() doesn't need any pushdown I think,
since this is serialized by nature as there is only one "ptracer"
for a task.



> 
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> ---
>  kernel/ptrace.c |   10 ----------
>  1 files changed, 0 insertions(+), 10 deletions(-)
> 
> diff --git a/kernel/ptrace.c b/kernel/ptrace.c
> index 42ad8ae..5357502 100644
> --- a/kernel/ptrace.c
> +++ b/kernel/ptrace.c
> @@ -666,10 +666,6 @@ SYSCALL_DEFINE4(ptrace, long, request, long, pid, long, addr, long, data)
>  	struct task_struct *child;
>  	long ret;
>  
> -	/*
> -	 * This lock_kernel fixes a subtle race with suid exec
> -	 */
> -	lock_kernel();
>  	if (request == PTRACE_TRACEME) {
>  		ret = ptrace_traceme();
>  		if (!ret)
> @@ -703,7 +699,6 @@ SYSCALL_DEFINE4(ptrace, long, request, long, pid, long, addr, long, data)
>   out_put_task_struct:
>  	put_task_struct(child);
>   out:
> -	unlock_kernel();
>  	return ret;
>  }
>  
> @@ -813,10 +808,6 @@ asmlinkage long compat_sys_ptrace(compat_long_t request, compat_long_t pid,
>  	struct task_struct *child;
>  	long ret;
>  
> -	/*
> -	 * This lock_kernel fixes a subtle race with suid exec
> -	 */
> -	lock_kernel();
>  	if (request == PTRACE_TRACEME) {
>  		ret = ptrace_traceme();
>  		goto out;
> @@ -846,7 +837,6 @@ asmlinkage long compat_sys_ptrace(compat_long_t request, compat_long_t pid,
>   out_put_task_struct:
>  	put_task_struct(child);
>   out:
> -	unlock_kernel();
>  	return ret;
>  }
>  #endif	/* CONFIG_COMPAT */
> -- 
> 1.6.2.3
>