Re: Buggy __free(kfree) usage pattern already in tree

[Peter Zijlstra]

On Tue, Sep 19, 2023 at 02:59:54PM +0200, Peter Zijlstra wrote:


> +	scoped_guard (mutex_intr, &task->signal->cred_guard_mutex) {
>  
> +		scoped_guard (task_lock, task) {
> +			retval = __ptrace_may_access(task, PTRACE_MODE_ATTACH_REALCREDS);
> +			if (retval)
> +				return retval;
> +		}
>  
> +		scoped_guard (write_lock, &tasklist_lock) {
> +			if (unlikely(task->exit_state))
> +				return -EPERM;
> +			if (task->ptrace)
> +				return -EPERM;
>  
> +			task->ptrace = flags;
>  
> +			ptrace_link(task, current);
> +
> +			/* SEIZE doesn't trap tracee on attach */
> +			if (!seize)
> +				send_sig_info(SIGSTOP, SEND_SIG_PRIV, task);
> +
> +			ptrace_set_stopped(task);
> +
> +		}
> +
> +		goto success;
>  	}
> +	return -ERESTARTNOINTR;
>  
> +success:
> +	/*
> +	 * We do not bother to change retval or clear JOBCTL_TRAPPING
> +	 * if wait_on_bit() was interrupted by SIGKILL. The tracer will
> +	 * not return to user-mode, it will exit and clear this bit in
> +	 * __ptrace_unlink() if it wasn't already cleared by the tracee;
> +	 * and until then nobody can ptrace this task.
> +	 */
> +	wait_on_bit(&task->jobctl, JOBCTL_TRAPPING_BIT, TASK_KILLABLE);
> +	proc_ptrace_connector(task, PTRACE_ATTACH);
> +
> +	return 0;

This isn't exactly nice..

I tried something like:

	scoped_cond_guard (mutex_intr, return -EINTR, &task->signal->cred_guard_mutex) {
		...
	}

Which I can make work, but then I also tried to capture my other case:

	scoped_cond_guard (rwsem_down_intr, if (task) return -EINTR,
			   task ? &task->signal->exec_guard_mutex : NULL)  {

		...
	}

But I can't get that to work because of that extra if, the not case
doesn't fall through and do the body.

Anyway, I'll poke more..