Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1445377

Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write

Show all headers | View raw

On Sun, 17 Jul 2016, Alexei Starovoitov wrote:

> On Sun, Jul 17, 2016 at 03:19:13AM -0700, Sargun Dhillon wrote:
>>
>> +static u64 bpf_copy_to_user(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
>> +{
>> +	void *to = (void *) (long) r1;
>> +	void *from = (void *) (long) r2;
>> +	int  size = (int) r3;
>> +
>> +	/* check if we're in a user context */
>> +	if (unlikely(in_interrupt()))
>> +		return -EINVAL;
>> +	if (unlikely(!current->pid))
>> +		return -EINVAL;
>> +
>> +	return copy_to_user(to, from, size);
>> +}
>
> thanks for the patch, unfortunately it's not that straightforward.
> copy_to_user might fault. Try enabling CONFIG_DEBUG_ATOMIC_SLEEP and
> you'll see the splat since bpf programs are protected by rcu.
> Also 'current' can be null and I'm not sure what current->pid does.
> So the writing to user memory either has to be verified to avoid
> sleeping and faults or we need to use something like task_work_add
> mechanism. Ideas are certainly welcome.
>
>
From casual inspection, I can't find where current can be null when 
in_interrupt() is false. Although, we can check before dereferencing it. 
When not in a user context, the pid of the task struct returns 0.

As far as preventing sleep, would the following alteration do? Or do we 
actually need something more sophisticated?
diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
index be89c148..45878f3 100644
--- a/kernel/trace/bpf_trace.c
+++ b/kernel/trace/bpf_trace.c
@@ -86,14 +86,19 @@ static u64 bpf_copy_to_user(u64 r1, u64 r2, u64 r3, 
u64 r4, u64 r5)
         void *to = (void *) (long) r1;
         void *from = (void *) (long) r2;
         int  size = (int) r3;
+       struct task_struct *task = current;

         /* check if we're in a user context */
         if (unlikely(in_interrupt()))
                 return -EINVAL;
-       if (unlikely(!current->pid))
+       if (unlikely(!task || !task->pid))
                 return -EINVAL;

-       return copy_to_user(to, from, size);
+       /* Is this a user address, or a kernel address? */
+       if (!access_ok(VERIFY_WRITE, to, size))
+               return -EINVAL;
+
+       return probe_kernel_write(to, from, size);
  }

  static const struct bpf_func_proto bpf_copy_to_user_proto = {


probe_kernel_write doesn't block, and this will disallow BPF programs to 
write to kernel memory. This turns off the pagefault handler under the 
hood, unblocking us.

Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread

Thread

[PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Sargun Dhillon <sargun@sargun.me> - 2016-07-13 12:40 +0200
  Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Alexei Starovoitov <alexei.starovoitov@gmail.com> - 2016-07-13 19:20 +0200
    Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Sargun Dhillon <sargun@sargun.me> - 2016-07-13 22:40 +0200
      Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Alexei Starovoitov <alexei.starovoitov@gmail.com> - 2016-07-15 07:50 +0200
        Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Sargun Dhillon <sargun@sargun.me> - 2016-07-16 04:20 +0200
          Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Alexei Starovoitov <alexei.starovoitov@gmail.com> - 2016-07-16 04:40 +0200
            Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Sargun Dhillon <sargun@sargun.me> - 2016-07-17 12:20 +0200
              Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Alexei Starovoitov <alexei.starovoitov@gmail.com> - 2016-07-18 06:20 +0200
                Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Sargun Dhillon <sargun@sargun.me> - 2016-07-18 13:00 +0200
                Re: [PATCH 1/1] tracing, bpf: Implement function bpf_probe_write Alexei Starovoitov <alexei.starovoitov@gmail.com> - 2016-07-19 08:20 +0200

csiph-web