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

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

[Alexei Starovoitov]

On Wed, Jul 13, 2016 at 03:36:11AM -0700, Sargun Dhillon wrote:
> Provides BPF programs, attached to kprobes a safe way to write to
> memory referenced by probes. This is done by making probe_kernel_write
> accessible to bpf functions via the bpf_probe_write helper.

not quite :)

> Signed-off-by: Sargun Dhillon <sargun@sargun.me>
> ---
>  include/uapi/linux/bpf.h  |  3 +++
>  kernel/trace/bpf_trace.c  | 20 ++++++++++++++++++++
>  samples/bpf/bpf_helpers.h |  2 ++
>  3 files changed, 25 insertions(+)
> 
> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> index 406459b..355b565 100644
> --- a/include/uapi/linux/bpf.h
> +++ b/include/uapi/linux/bpf.h
> @@ -313,6 +313,9 @@ enum bpf_func_id {
>   */
>   BPF_FUNC_skb_get_tunnel_opt,
>   BPF_FUNC_skb_set_tunnel_opt,
> +
> + BPF_FUNC_probe_write, /* int bpf_probe_write(void *dst, void *src,
> int size) */
> +

the patch is against some old kernel.
Please always make the patch against net-next tree and cc netdev list.

> +static u64 bpf_probe_write(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
> +{
> + void *dst = (void *) (long) r1;
> + void *unsafe_ptr = (void *) (long) r2;
> + int  size = (int) r3;
> +
> + return probe_kernel_write(dst, unsafe_ptr, size);
> +}

the patch is whitepsace mangled. Please see Documentation/networking/netdev-FAQ.txt

the main issue though that we cannot simply allow bpf to do probe_write,
since it may crash the kernel.
What might be ok is to allow writing into memory of current
user space process only. This way bpf prog will keep kernel safety guarantees,
yet it will be able to modify user process memory when necessary.
Since bpf+tracing is root only, it doesn't pose security risk.

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

[Sargun Dhillon]

On Wed, 13 Jul 2016, Alexei Starovoitov wrote:

> On Wed, Jul 13, 2016 at 03:36:11AM -0700, Sargun Dhillon wrote:
>> Provides BPF programs, attached to kprobes a safe way to write to
>> memory referenced by probes. This is done by making probe_kernel_write
>> accessible to bpf functions via the bpf_probe_write helper.
>
> not quite :)
>
>> Signed-off-by: Sargun Dhillon <sargun@sargun.me>
>> ---
>>  include/uapi/linux/bpf.h  |  3 +++
>>  kernel/trace/bpf_trace.c  | 20 ++++++++++++++++++++
>>  samples/bpf/bpf_helpers.h |  2 ++
>>  3 files changed, 25 insertions(+)
>>
>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>> index 406459b..355b565 100644
>> --- a/include/uapi/linux/bpf.h
>> +++ b/include/uapi/linux/bpf.h
>> @@ -313,6 +313,9 @@ enum bpf_func_id {
>>   */
>>   BPF_FUNC_skb_get_tunnel_opt,
>>   BPF_FUNC_skb_set_tunnel_opt,
>> +
>> + BPF_FUNC_probe_write, /* int bpf_probe_write(void *dst, void *src,
>> int size) */
>> +
>
> the patch is against some old kernel.
> Please always make the patch against net-next tree and cc netdev list.
>
Sorry, I did this against Linus's tree, not net-next. Will fix.

>> +static u64 bpf_probe_write(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
>> +{
>> + void *dst = (void *) (long) r1;
>> + void *unsafe_ptr = (void *) (long) r2;
>> + int  size = (int) r3;
>> +
>> + return probe_kernel_write(dst, unsafe_ptr, size);
>> +}
>
> the patch is whitepsace mangled. Please see Documentation/networking/netdev-FAQ.txt
Also will fix.

>
> the main issue though that we cannot simply allow bpf to do probe_write,
> since it may crash the kernel.
> What might be ok is to allow writing into memory of current
> user space process only. This way bpf prog will keep kernel safety guarantees,
> yet it will be able to modify user process memory when necessary.
> Since bpf+tracing is root only, it doesn't pose security risk.
>
>

Doesn't probe_write prevent you from writing to protected memory and 
generate an EFAULT? Or are you worried about the situation where a bpf 
program writes to some other chunk of kernel memory, or writes bad data 
to said kernel memory?

I guess when I meant "safe" -- it's safer than allowing arbitrary memcpy. 
I don't see a good way to ensure safety otherwise as we don't know 
which registers point to memory that it's reasonable for probes to 
manipulate. It's not like skb_store_bytes where we can check the pointer 
going in is the same pointer that's referenced, and with a super 
restricted datatype.

Perhaps, it would be a good idea to describe an example where I used this:
#include <uapi/linux/ptrace.h>
#include <net/sock.h>
#include <bcc/proto.h>


int trace_inet_stream_connect(struct pt_regs *ctx)
{
	if (!PT_REGS_PARM2(ctx)) {
		return 0;
	}
	struct sockaddr uaddr = {};
	struct sockaddr_in *addr_in;
	bpf_probe_read(&uaddr, sizeof(struct sockaddr), (void *)PT_REGS_PARM2(ctx));
	if (uaddr.sa_family == AF_INET) {
		// Simple cast causes LLVM weirdness
		addr_in = &uaddr;
		char fmt[] = "Connecting on port: %d\n";
		bpf_trace_printk(fmt, sizeof(fmt), ntohs(addr_in->sin_port));
		if (ntohs(addr_in->sin_port) == 80) {
			addr_in->sin_port = htons(443);
			bpf_probe_write((void *)PT_REGS_PARM2(ctx), &uaddr, sizeof(uaddr));
		}
	}
        return 0;
};

There are two reasons I want to do this:
1) Debugging - sometimes, it makes sense to divert a program's syscalls in 
order to allow for better debugging
2) Network Functions - I wrote a load balancer which intercepts 
inet_stream_connect & tcp_set_state. We can manipulate the destination 
address as neccessary at connect time. This also has the nice side effect 
that getpeername() returns the real IP that a server is connected to, and 
the performance is far better than doing "network load balancing"

(I realize this is a total hack, better approaches would be appreciated)

If we allowed manipulation of the current task's user memory by exposing 
copy_to_user, that could also work if I attach the probe to sys_connect, 
I could overwrite the address there before it gets copied into 
kernel space, but that could lead to its own weirdness.

Any ideas?

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

[Alexei Starovoitov]

On Wed, Jul 13, 2016 at 01:31:57PM -0700, Sargun Dhillon wrote:
> 
> 
> On Wed, 13 Jul 2016, Alexei Starovoitov wrote:
> 
> > On Wed, Jul 13, 2016 at 03:36:11AM -0700, Sargun Dhillon wrote:
> >> Provides BPF programs, attached to kprobes a safe way to write to
> >> memory referenced by probes. This is done by making probe_kernel_write
> >> accessible to bpf functions via the bpf_probe_write helper.
> >
> > not quite :)
> >
> >> Signed-off-by: Sargun Dhillon <sargun@sargun.me>
> >> ---
> >>  include/uapi/linux/bpf.h  |  3 +++
> >>  kernel/trace/bpf_trace.c  | 20 ++++++++++++++++++++
> >>  samples/bpf/bpf_helpers.h |  2 ++
> >>  3 files changed, 25 insertions(+)
> >>
> >> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> >> index 406459b..355b565 100644
> >> --- a/include/uapi/linux/bpf.h
> >> +++ b/include/uapi/linux/bpf.h
> >> @@ -313,6 +313,9 @@ enum bpf_func_id {
> >>   */
> >>   BPF_FUNC_skb_get_tunnel_opt,
> >>   BPF_FUNC_skb_set_tunnel_opt,
> >> +
> >> + BPF_FUNC_probe_write, /* int bpf_probe_write(void *dst, void *src,
> >> int size) */
> >> +
> >
> > the patch is against some old kernel.
> > Please always make the patch against net-next tree and cc netdev list.
> >
> Sorry, I did this against Linus's tree, not net-next. Will fix.
> 
> >> +static u64 bpf_probe_write(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
> >> +{
> >> + void *dst = (void *) (long) r1;
> >> + void *unsafe_ptr = (void *) (long) r2;
> >> + int  size = (int) r3;
> >> +
> >> + return probe_kernel_write(dst, unsafe_ptr, size);
> >> +}
> >
> > the patch is whitepsace mangled. Please see Documentation/networking/netdev-FAQ.txt
> Also will fix.
> 
> >
> > the main issue though that we cannot simply allow bpf to do probe_write,
> > since it may crash the kernel.
> > What might be ok is to allow writing into memory of current
> > user space process only. This way bpf prog will keep kernel safety guarantees,
> > yet it will be able to modify user process memory when necessary.
> > Since bpf+tracing is root only, it doesn't pose security risk.
> >
> >
> 
> Doesn't probe_write prevent you from writing to protected memory and 
> generate an EFAULT? Or are you worried about the situation where a bpf 
> program writes to some other chunk of kernel memory, or writes bad data 
> to said kernel memory?
> 
> I guess when I meant "safe" -- it's safer than allowing arbitrary memcpy. 
> I don't see a good way to ensure safety otherwise as we don't know 
> which registers point to memory that it's reasonable for probes to 
> manipulate. It's not like skb_store_bytes where we can check the pointer 
> going in is the same pointer that's referenced, and with a super 
> restricted datatype.

exactly. probe_write can write anywhere in the kernel and that
will cause crashes. If we allow that bpf becomes no different than
kernel module.

> Perhaps, it would be a good idea to describe an example where I used this:
> #include <uapi/linux/ptrace.h>
> #include <net/sock.h>
> #include <bcc/proto.h>
> 
> 
> int trace_inet_stream_connect(struct pt_regs *ctx)
> {
> 	if (!PT_REGS_PARM2(ctx)) {
> 		return 0;
> 	}
> 	struct sockaddr uaddr = {};
> 	struct sockaddr_in *addr_in;
> 	bpf_probe_read(&uaddr, sizeof(struct sockaddr), (void *)PT_REGS_PARM2(ctx));
> 	if (uaddr.sa_family == AF_INET) {
> 		// Simple cast causes LLVM weirdness
> 		addr_in = &uaddr;
> 		char fmt[] = "Connecting on port: %d\n";
> 		bpf_trace_printk(fmt, sizeof(fmt), ntohs(addr_in->sin_port));
> 		if (ntohs(addr_in->sin_port) == 80) {
> 			addr_in->sin_port = htons(443);
> 			bpf_probe_write((void *)PT_REGS_PARM2(ctx), &uaddr, sizeof(uaddr));
> 		}
> 	}
>         return 0;
> };
> 
> There are two reasons I want to do this:
> 1) Debugging - sometimes, it makes sense to divert a program's syscalls in 
> order to allow for better debugging
> 2) Network Functions - I wrote a load balancer which intercepts 
> inet_stream_connect & tcp_set_state. We can manipulate the destination 
> address as neccessary at connect time. This also has the nice side effect 
> that getpeername() returns the real IP that a server is connected to, and 
> the performance is far better than doing "network load balancing"
> 
> (I realize this is a total hack, better approaches would be appreciated)

nice. interesting idea.
Have you considered ld_preload hack to do port rewrite?

> If we allowed manipulation of the current task's user memory by exposing 
> copy_to_user, that could also work if I attach the probe to sys_connect, 
> I could overwrite the address there before it gets copied into 
> kernel space, but that could lead to its own weirdness.

we cannot simply call copy_to_user from the bpf either,
but yeah, something semantically equivalent to copy_to_user should
solve your port rewriting case, right?
Could you explain little bit more on 'syscall divert' ideas?

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

[Sargun Dhillon]

On Thu, 14 Jul 2016, Alexei Starovoitov wrote:

> On Wed, Jul 13, 2016 at 01:31:57PM -0700, Sargun Dhillon wrote:
>>
>>
>> On Wed, 13 Jul 2016, Alexei Starovoitov wrote:
>>
>>> On Wed, Jul 13, 2016 at 03:36:11AM -0700, Sargun Dhillon wrote:
>>>> Provides BPF programs, attached to kprobes a safe way to write to
>>>> memory referenced by probes. This is done by making probe_kernel_write
>>>> accessible to bpf functions via the bpf_probe_write helper.
>>>
>>> not quite :)
>>>
>>>> Signed-off-by: Sargun Dhillon <sargun@sargun.me>
>>>> ---
>>>>  include/uapi/linux/bpf.h  |  3 +++
>>>>  kernel/trace/bpf_trace.c  | 20 ++++++++++++++++++++
>>>>  samples/bpf/bpf_helpers.h |  2 ++
>>>>  3 files changed, 25 insertions(+)
>>>>
>>>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>>>> index 406459b..355b565 100644
>>>> --- a/include/uapi/linux/bpf.h
>>>> +++ b/include/uapi/linux/bpf.h
>>>> @@ -313,6 +313,9 @@ enum bpf_func_id {
>>>>   */
>>>>   BPF_FUNC_skb_get_tunnel_opt,
>>>>   BPF_FUNC_skb_set_tunnel_opt,
>>>> +
>>>> + BPF_FUNC_probe_write, /* int bpf_probe_write(void *dst, void *src,
>>>> int size) */
>>>> +
>>>
>>> the patch is against some old kernel.
>>> Please always make the patch against net-next tree and cc netdev list.
>>>
>> Sorry, I did this against Linus's tree, not net-next. Will fix.
>>
>>>> +static u64 bpf_probe_write(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
>>>> +{
>>>> + void *dst = (void *) (long) r1;
>>>> + void *unsafe_ptr = (void *) (long) r2;
>>>> + int  size = (int) r3;
>>>> +
>>>> + return probe_kernel_write(dst, unsafe_ptr, size);
>>>> +}
>>>
>>> the patch is whitepsace mangled. Please see Documentation/networking/netdev-FAQ.txt
>> Also will fix.
>>
>>>
>>> the main issue though that we cannot simply allow bpf to do probe_write,
>>> since it may crash the kernel.
>>> What might be ok is to allow writing into memory of current
>>> user space process only. This way bpf prog will keep kernel safety guarantees,
>>> yet it will be able to modify user process memory when necessary.
>>> Since bpf+tracing is root only, it doesn't pose security risk.
>>>
>>>
>>
>> Doesn't probe_write prevent you from writing to protected memory and
>> generate an EFAULT? Or are you worried about the situation where a bpf
>> program writes to some other chunk of kernel memory, or writes bad data
>> to said kernel memory?
>>
>> I guess when I meant "safe" -- it's safer than allowing arbitrary memcpy.
>> I don't see a good way to ensure safety otherwise as we don't know
>> which registers point to memory that it's reasonable for probes to
>> manipulate. It's not like skb_store_bytes where we can check the pointer
>> going in is the same pointer that's referenced, and with a super
>> restricted datatype.
>
> exactly. probe_write can write anywhere in the kernel and that
> will cause crashes. If we allow that bpf becomes no different than
> kernel module.
>
>> Perhaps, it would be a good idea to describe an example where I used this:
>> #include <uapi/linux/ptrace.h>
>> #include <net/sock.h>
>> #include <bcc/proto.h>
>>
>>
>> int trace_inet_stream_connect(struct pt_regs *ctx)
>> {
>> 	if (!PT_REGS_PARM2(ctx)) {
>> 		return 0;
>> 	}
>> 	struct sockaddr uaddr = {};
>> 	struct sockaddr_in *addr_in;
>> 	bpf_probe_read(&uaddr, sizeof(struct sockaddr), (void *)PT_REGS_PARM2(ctx));
>> 	if (uaddr.sa_family == AF_INET) {
>> 		// Simple cast causes LLVM weirdness
>> 		addr_in = &uaddr;
>> 		char fmt[] = "Connecting on port: %d\n";
>> 		bpf_trace_printk(fmt, sizeof(fmt), ntohs(addr_in->sin_port));
>> 		if (ntohs(addr_in->sin_port) == 80) {
>> 			addr_in->sin_port = htons(443);
>> 			bpf_probe_write((void *)PT_REGS_PARM2(ctx), &uaddr, sizeof(uaddr));
>> 		}
>> 	}
>>         return 0;
>> };
>>
>> There are two reasons I want to do this:
>> 1) Debugging - sometimes, it makes sense to divert a program's syscalls in
>> order to allow for better debugging
>> 2) Network Functions - I wrote a load balancer which intercepts
>> inet_stream_connect & tcp_set_state. We can manipulate the destination
>> address as neccessary at connect time. This also has the nice side effect
>> that getpeername() returns the real IP that a server is connected to, and
>> the performance is far better than doing "network load balancing"
>>
>> (I realize this is a total hack, better approaches would be appreciated)
>
> nice. interesting idea.
> Have you considered ld_preload hack to do port rewrite?
>
We've thought about it. It wont really work for us, because we're doing 
this to manipulate 3rd party runtimes, many of which are written in 
languages that don't play nice with LD_PRELOAD. Go is the primary problem 
child in this case. We also looked at using SECCOMP + ptrace, but again, 
not all runtimes play nice with ptrace.

>> If we allowed manipulation of the current task's user memory by exposing
>> copy_to_user, that could also work if I attach the probe to sys_connect,
>> I could overwrite the address there before it gets copied into
>> kernel space, but that could lead to its own weirdness.
>
> we cannot simply call copy_to_user from the bpf either,
> but yeah, something semantically equivalent to copy_to_user should
> solve your port rewriting case, right?
> Could you explain little bit more on 'syscall divert' ideas?
>
>
If we had a "safe" copy_to_user which checked if BPF programs were running 
in the user context, that would work right? I mean, you could still make 
user programs crash, but that's better than making the kernel fall over. 
We would need both copy_from_user, and copy_to_user. If you look at the 
example program, it first checks what the user is connecting to -- so it'd 
have to check the address the user is passing to the syscall.

Syscall Divert:
The idea here is to try to "lie" to the program about the environment. A 
specific example is where I want to bypass certain calls like prctl during 
debugging, to allow certain instructions to execute. I don't always have 
access to the source of the containerizer I'm running, and it's nice to 
turn them off during debugging.

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

[Alexei Starovoitov]

On Fri, Jul 15, 2016 at 07:16:01PM -0700, Sargun Dhillon wrote:
> 
> 
> On Thu, 14 Jul 2016, Alexei Starovoitov wrote:
> 
> >On Wed, Jul 13, 2016 at 01:31:57PM -0700, Sargun Dhillon wrote:
> >>
> >>
> >>On Wed, 13 Jul 2016, Alexei Starovoitov wrote:
> >>
> >>>On Wed, Jul 13, 2016 at 03:36:11AM -0700, Sargun Dhillon wrote:
> >>>>Provides BPF programs, attached to kprobes a safe way to write to
> >>>>memory referenced by probes. This is done by making probe_kernel_write
> >>>>accessible to bpf functions via the bpf_probe_write helper.
> >>>
> >>>not quite :)
> >>>
> >>>>Signed-off-by: Sargun Dhillon <sargun@sargun.me>
> >>>>---
> >>>> include/uapi/linux/bpf.h  |  3 +++
> >>>> kernel/trace/bpf_trace.c  | 20 ++++++++++++++++++++
> >>>> samples/bpf/bpf_helpers.h |  2 ++
> >>>> 3 files changed, 25 insertions(+)
> >>>>
> >>>>diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
> >>>>index 406459b..355b565 100644
> >>>>--- a/include/uapi/linux/bpf.h
> >>>>+++ b/include/uapi/linux/bpf.h
> >>>>@@ -313,6 +313,9 @@ enum bpf_func_id {
> >>>>  */
> >>>>  BPF_FUNC_skb_get_tunnel_opt,
> >>>>  BPF_FUNC_skb_set_tunnel_opt,
> >>>>+
> >>>>+ BPF_FUNC_probe_write, /* int bpf_probe_write(void *dst, void *src,
> >>>>int size) */
> >>>>+
> >>>
> >>>the patch is against some old kernel.
> >>>Please always make the patch against net-next tree and cc netdev list.
> >>>
> >>Sorry, I did this against Linus's tree, not net-next. Will fix.
> >>
> >>>>+static u64 bpf_probe_write(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
> >>>>+{
> >>>>+ void *dst = (void *) (long) r1;
> >>>>+ void *unsafe_ptr = (void *) (long) r2;
> >>>>+ int  size = (int) r3;
> >>>>+
> >>>>+ return probe_kernel_write(dst, unsafe_ptr, size);
> >>>>+}
> >>>
> >>>the patch is whitepsace mangled. Please see Documentation/networking/netdev-FAQ.txt
> >>Also will fix.
> >>
> >>>
> >>>the main issue though that we cannot simply allow bpf to do probe_write,
> >>>since it may crash the kernel.
> >>>What might be ok is to allow writing into memory of current
> >>>user space process only. This way bpf prog will keep kernel safety guarantees,
> >>>yet it will be able to modify user process memory when necessary.
> >>>Since bpf+tracing is root only, it doesn't pose security risk.
> >>>
> >>>
> >>
> >>Doesn't probe_write prevent you from writing to protected memory and
> >>generate an EFAULT? Or are you worried about the situation where a bpf
> >>program writes to some other chunk of kernel memory, or writes bad data
> >>to said kernel memory?
> >>
> >>I guess when I meant "safe" -- it's safer than allowing arbitrary memcpy.
> >>I don't see a good way to ensure safety otherwise as we don't know
> >>which registers point to memory that it's reasonable for probes to
> >>manipulate. It's not like skb_store_bytes where we can check the pointer
> >>going in is the same pointer that's referenced, and with a super
> >>restricted datatype.
> >
> >exactly. probe_write can write anywhere in the kernel and that
> >will cause crashes. If we allow that bpf becomes no different than
> >kernel module.
> >
> >>Perhaps, it would be a good idea to describe an example where I used this:
> >>#include <uapi/linux/ptrace.h>
> >>#include <net/sock.h>
> >>#include <bcc/proto.h>
> >>
> >>
> >>int trace_inet_stream_connect(struct pt_regs *ctx)
> >>{
> >>	if (!PT_REGS_PARM2(ctx)) {
> >>		return 0;
> >>	}
> >>	struct sockaddr uaddr = {};
> >>	struct sockaddr_in *addr_in;
> >>	bpf_probe_read(&uaddr, sizeof(struct sockaddr), (void *)PT_REGS_PARM2(ctx));
> >>	if (uaddr.sa_family == AF_INET) {
> >>		// Simple cast causes LLVM weirdness
> >>		addr_in = &uaddr;
> >>		char fmt[] = "Connecting on port: %d\n";
> >>		bpf_trace_printk(fmt, sizeof(fmt), ntohs(addr_in->sin_port));
> >>		if (ntohs(addr_in->sin_port) == 80) {
> >>			addr_in->sin_port = htons(443);
> >>			bpf_probe_write((void *)PT_REGS_PARM2(ctx), &uaddr, sizeof(uaddr));
> >>		}
> >>	}
> >>        return 0;
> >>};
> >>
> >>There are two reasons I want to do this:
> >>1) Debugging - sometimes, it makes sense to divert a program's syscalls in
> >>order to allow for better debugging
> >>2) Network Functions - I wrote a load balancer which intercepts
> >>inet_stream_connect & tcp_set_state. We can manipulate the destination
> >>address as neccessary at connect time. This also has the nice side effect
> >>that getpeername() returns the real IP that a server is connected to, and
> >>the performance is far better than doing "network load balancing"
> >>
> >>(I realize this is a total hack, better approaches would be appreciated)
> >
> >nice. interesting idea.
> >Have you considered ld_preload hack to do port rewrite?
> >
> We've thought about it. It wont really work for us, because we're doing this
> to manipulate 3rd party runtimes, many of which are written in languages
> that don't play nice with LD_PRELOAD. Go is the primary problem child in
> this case. We also looked at using SECCOMP + ptrace, but again, not all
> runtimes play nice with ptrace.

interesting!
I was about to suggest to hack write support into seccomp, since few
folks were interested in it as well. Why seccomp won't work?
You cannot have a centralized daemon that launches all the processes?

> >>If we allowed manipulation of the current task's user memory by exposing
> >>copy_to_user, that could also work if I attach the probe to sys_connect,
> >>I could overwrite the address there before it gets copied into
> >>kernel space, but that could lead to its own weirdness.
> >
> >we cannot simply call copy_to_user from the bpf either,
> >but yeah, something semantically equivalent to copy_to_user should
> >solve your port rewriting case, right?
> >Could you explain little bit more on 'syscall divert' ideas?
> >
> >
> If we had a "safe" copy_to_user which checked if BPF programs were running
> in the user context, that would work right? I mean, you could still make
> user programs crash, but that's better than making the kernel fall over. We
> would need both copy_from_user, and copy_to_user. If you look at the example
> program, it first checks what the user is connecting to -- so it'd have to
> check the address the user is passing to the syscall.

yep. imo 'safe' copy_to_user is no different in user impact from seccomp.
Both can make user process inoperable if there is a bug in bpf program,
but both are always safe from kernel point of view.
there is no need in copy_from_user. bpf_probe_read can read user memory just
fine or you actually want to make even safer version of probe_read that
can only read current task memory instead of any memory?
Makes sense to me.

> Syscall Divert:
> The idea here is to try to "lie" to the program about the environment. A
> specific example is where I want to bypass certain calls like prctl during
> debugging, to allow certain instructions to execute. I don't always have
> access to the source of the containerizer I'm running, and it's nice to turn
> them off during debugging.

makes sense too.
so seccomp with write support is also not option here, because you actually
want to divert seccomp itself?

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

[Sargun Dhillon]

On Fri, 15 Jul 2016, Alexei Starovoitov wrote:

> On Fri, Jul 15, 2016 at 07:16:01PM -0700, Sargun Dhillon wrote:
>>
>>
>> On Thu, 14 Jul 2016, Alexei Starovoitov wrote:
>>
>>> On Wed, Jul 13, 2016 at 01:31:57PM -0700, Sargun Dhillon wrote:
>>>>
>>>>
>>>> On Wed, 13 Jul 2016, Alexei Starovoitov wrote:
>>>>
>>>>> On Wed, Jul 13, 2016 at 03:36:11AM -0700, Sargun Dhillon wrote:
>>>>>> Provides BPF programs, attached to kprobes a safe way to write to
>>>>>> memory referenced by probes. This is done by making probe_kernel_write
>>>>>> accessible to bpf functions via the bpf_probe_write helper.
>>>>>
>>>>> not quite :)
>>>>>
>>>>>> Signed-off-by: Sargun Dhillon <sargun@sargun.me>
>>>>>> ---
>>>>>> include/uapi/linux/bpf.h  |  3 +++
>>>>>> kernel/trace/bpf_trace.c  | 20 ++++++++++++++++++++
>>>>>> samples/bpf/bpf_helpers.h |  2 ++
>>>>>> 3 files changed, 25 insertions(+)
>>>>>>
>>>>>> diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
>>>>>> index 406459b..355b565 100644
>>>>>> --- a/include/uapi/linux/bpf.h
>>>>>> +++ b/include/uapi/linux/bpf.h
>>>>>> @@ -313,6 +313,9 @@ enum bpf_func_id {
>>>>>>  */
>>>>>>  BPF_FUNC_skb_get_tunnel_opt,
>>>>>>  BPF_FUNC_skb_set_tunnel_opt,
>>>>>> +
>>>>>> + BPF_FUNC_probe_write, /* int bpf_probe_write(void *dst, void *src,
>>>>>> int size) */
>>>>>> +
>>>>>
>>>>> the patch is against some old kernel.
>>>>> Please always make the patch against net-next tree and cc netdev list.
>>>>>
>>>> Sorry, I did this against Linus's tree, not net-next. Will fix.
>>>>
>>>>>> +static u64 bpf_probe_write(u64 r1, u64 r2, u64 r3, u64 r4, u64 r5)
>>>>>> +{
>>>>>> + void *dst = (void *) (long) r1;
>>>>>> + void *unsafe_ptr = (void *) (long) r2;
>>>>>> + int  size = (int) r3;
>>>>>> +
>>>>>> + return probe_kernel_write(dst, unsafe_ptr, size);
>>>>>> +}
>>>>>
>>>>> the patch is whitepsace mangled. Please see Documentation/networking/netdev-FAQ.txt
>>>> Also will fix.
>>>>
>>>>>
>>>>> the main issue though that we cannot simply allow bpf to do probe_write,
>>>>> since it may crash the kernel.
>>>>> What might be ok is to allow writing into memory of current
>>>>> user space process only. This way bpf prog will keep kernel safety guarantees,
>>>>> yet it will be able to modify user process memory when necessary.
>>>>> Since bpf+tracing is root only, it doesn't pose security risk.
>>>>>
>>>>>
>>>>
>>>> Doesn't probe_write prevent you from writing to protected memory and
>>>> generate an EFAULT? Or are you worried about the situation where a bpf
>>>> program writes to some other chunk of kernel memory, or writes bad data
>>>> to said kernel memory?
>>>>
>>>> I guess when I meant "safe" -- it's safer than allowing arbitrary memcpy.
>>>> I don't see a good way to ensure safety otherwise as we don't know
>>>> which registers point to memory that it's reasonable for probes to
>>>> manipulate. It's not like skb_store_bytes where we can check the pointer
>>>> going in is the same pointer that's referenced, and with a super
>>>> restricted datatype.
>>>
>>> exactly. probe_write can write anywhere in the kernel and that
>>> will cause crashes. If we allow that bpf becomes no different than
>>> kernel module.
>>>
>>>> Perhaps, it would be a good idea to describe an example where I used this:
>>>> #include <uapi/linux/ptrace.h>
>>>> #include <net/sock.h>
>>>> #include <bcc/proto.h>
>>>>
>>>>
>>>> int trace_inet_stream_connect(struct pt_regs *ctx)
>>>> {
>>>> 	if (!PT_REGS_PARM2(ctx)) {
>>>> 		return 0;
>>>> 	}
>>>> 	struct sockaddr uaddr = {};
>>>> 	struct sockaddr_in *addr_in;
>>>> 	bpf_probe_read(&uaddr, sizeof(struct sockaddr), (void *)PT_REGS_PARM2(ctx));
>>>> 	if (uaddr.sa_family == AF_INET) {
>>>> 		// Simple cast causes LLVM weirdness
>>>> 		addr_in = &uaddr;
>>>> 		char fmt[] = "Connecting on port: %d\n";
>>>> 		bpf_trace_printk(fmt, sizeof(fmt), ntohs(addr_in->sin_port));
>>>> 		if (ntohs(addr_in->sin_port) == 80) {
>>>> 			addr_in->sin_port = htons(443);
>>>> 			bpf_probe_write((void *)PT_REGS_PARM2(ctx), &uaddr, sizeof(uaddr));
>>>> 		}
>>>> 	}
>>>>        return 0;
>>>> };
>>>>
>>>> There are two reasons I want to do this:
>>>> 1) Debugging - sometimes, it makes sense to divert a program's syscalls in
>>>> order to allow for better debugging
>>>> 2) Network Functions - I wrote a load balancer which intercepts
>>>> inet_stream_connect & tcp_set_state. We can manipulate the destination
>>>> address as neccessary at connect time. This also has the nice side effect
>>>> that getpeername() returns the real IP that a server is connected to, and
>>>> the performance is far better than doing "network load balancing"
>>>>
>>>> (I realize this is a total hack, better approaches would be appreciated)
>>>
>>> nice. interesting idea.
>>> Have you considered ld_preload hack to do port rewrite?
>>>
>> We've thought about it. It wont really work for us, because we're doing this
>> to manipulate 3rd party runtimes, many of which are written in languages
>> that don't play nice with LD_PRELOAD. Go is the primary problem child in
>> this case. We also looked at using SECCOMP + ptrace, but again, not all
>> runtimes play nice with ptrace.
>
> interesting!
> I was about to suggest to hack write support into seccomp, since few
> folks were interested in it as well. Why seccomp won't work?
> You cannot have a centralized daemon that launches all the processes?
>
>>>> If we allowed manipulation of the current task's user memory by exposing
>>>> copy_to_user, that could also work if I attach the probe to sys_connect,
>>>> I could overwrite the address there before it gets copied into
>>>> kernel space, but that could lead to its own weirdness.
>>>
>>> we cannot simply call copy_to_user from the bpf either,
>>> but yeah, something semantically equivalent to copy_to_user should
>>> solve your port rewriting case, right?
>>> Could you explain little bit more on 'syscall divert' ideas?
>>>
>>>
>> If we had a "safe" copy_to_user which checked if BPF programs were running
>> in the user context, that would work right? I mean, you could still make
>> user programs crash, but that's better than making the kernel fall over. We
>> would need both copy_from_user, and copy_to_user. If you look at the example
>> program, it first checks what the user is connecting to -- so it'd have to
>> check the address the user is passing to the syscall.
>
> yep. imo 'safe' copy_to_user is no different in user impact from seccomp.
> Both can make user process inoperable if there is a bug in bpf program,
> but both are always safe from kernel point of view.
> there is no need in copy_from_user. bpf_probe_read can read user memory just
> fine or you actually want to make even safer version of probe_read that
> can only read current task memory instead of any memory?
> Makes sense to me.
>
>> Syscall Divert:
>> The idea here is to try to "lie" to the program about the environment. A
>> specific example is where I want to bypass certain calls like prctl during
>> debugging, to allow certain instructions to execute. I don't always have
>> access to the source of the containerizer I'm running, and it's nice to turn
>> them off during debugging.
>
> makes sense too.
> so seccomp with write support is also not option here, because you actually
> want to divert seccomp itself?
>
>
Yeah, the only difference with seccomp is that the impact is isolated to a 
group of processes versus the entire system. Though, kprobes require root, 
whereas seccomp is accessible to others. It would be really nice to be 
able to call bpf_probe_read from seccomp filters, and be able to filter on 
things like the endpoint that people are connecting to. As it stands right 
now, it would open up seccomp to a TOCTOU attack. I'm not sure


If seccomp had full eBPF, and write support that'd solve most of my use 
cases. As far as diverting diverting future syscall behaviour, if I could 
just intercept those, and rewrite seccomp filters that'd solve the problem 
too.

I'm thinking of submitting something along the lines of the following 
patch [WIP]. What do you think?
diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h
index c4d9224..d435f7c 100644
--- a/include/uapi/linux/bpf.h
+++ b/include/uapi/linux/bpf.h
@@ -364,6 +364,17 @@ enum bpf_func_id {
  	 */
  	BPF_FUNC_get_current_task,

+	/**
+	 * copy_to_user(to, from, len) - Copy a block of data into user 
space.
+	 * @to: destination address in userspace
+	 * @from: source address on stack
+	 * @len: number of bytes to copy
+	 * Return:
+	 *   Returns number of bytes that could not be copied.
+	 *   On success, this will be zero
+	 */
+	BPF_FUNC_copy_to_user,
+
  	__BPF_FUNC_MAX_ID,
  };

diff --git a/kernel/trace/bpf_trace.c b/kernel/trace/bpf_trace.c
index ebfbb7d..be89c148 100644
--- a/kernel/trace/bpf_trace.c
+++ b/kernel/trace/bpf_trace.c
@@ -81,6 +81,30 @@ static const struct bpf_func_proto bpf_probe_read_proto 
= {
  	.arg3_type	= ARG_ANYTHING,
  };

+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);
+}
+
+static const struct bpf_func_proto bpf_copy_to_user_proto = {
+	.func = bpf_copy_to_user,
+	.gpl_only = false,
+	.ret_type = RET_INTEGER,
+	.arg1_type = ARG_ANYTHING,
+	.arg2_type = ARG_PTR_TO_RAW_STACK,
+	.arg3_type = ARG_CONST_STACK_SIZE,
+};
+
  /*
   * limited trace_printk()
   * only %d %u %x %ld %lu %lx %lld %llu %llx %p %s conversion specifiers 
allowed
@@ -360,6 +384,8 @@ static const struct bpf_func_proto 
*tracing_func_proto(enum bpf_func_id func_id)
  		return &bpf_get_smp_processor_id_proto;
  	case BPF_FUNC_perf_event_read:
  		return &bpf_perf_event_read_proto;
+	case BPF_FUNC_copy_to_user:
+		return &bpf_copy_to_user_proto;
  	default:
  		return NULL;
  	}
diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile
index a98b780..be2bab9 100644
--- a/samples/bpf/Makefile
+++ b/samples/bpf/Makefile
@@ -14,6 +14,7 @@ hostprogs-y += tracex3
  hostprogs-y += tracex4
  hostprogs-y += tracex5
  hostprogs-y += tracex6
+hostprogs-y += tracex7
  hostprogs-y += trace_output
  hostprogs-y += lathist
  hostprogs-y += offwaketime
@@ -35,6 +36,7 @@ tracex3-objs := bpf_load.o libbpf.o tracex3_user.o
  tracex4-objs := bpf_load.o libbpf.o tracex4_user.o
  tracex5-objs := bpf_load.o libbpf.o tracex5_user.o
  tracex6-objs := bpf_load.o libbpf.o tracex6_user.o
+tracex7-objs := bpf_load.o libbpf.o tracex7_user.o
  trace_output-objs := bpf_load.o libbpf.o trace_output_user.o
  lathist-objs := bpf_load.o libbpf.o lathist_user.o
  offwaketime-objs := bpf_load.o libbpf.o offwaketime_user.o
@@ -54,6 +56,7 @@ always += tracex3_kern.o
  always += tracex4_kern.o
  always += tracex5_kern.o
  always += tracex6_kern.o
+always += tracex7_kern.o
  always += trace_output_kern.o
  always += tcbpf1_kern.o
  always += lathist_kern.o
@@ -78,6 +81,7 @@ HOSTLOADLIBES_tracex3 += -lelf
  HOSTLOADLIBES_tracex4 += -lelf -lrt
  HOSTLOADLIBES_tracex5 += -lelf
  HOSTLOADLIBES_tracex6 += -lelf
+HOSTLOADLIBES_tracex7 += -lelf
  HOSTLOADLIBES_trace_output += -lelf -lrt
  HOSTLOADLIBES_lathist += -lelf
  HOSTLOADLIBES_offwaketime += -lelf
diff --git a/samples/bpf/bpf_helpers.h b/samples/bpf/bpf_helpers.h
index 84e3fd9..a54a26c 100644
--- a/samples/bpf/bpf_helpers.h
+++ b/samples/bpf/bpf_helpers.h
@@ -41,6 +41,8 @@ static int (*bpf_perf_event_output)(void *ctx, void 
*map, int index, void *data,
  	(void *) BPF_FUNC_perf_event_output;
  static int (*bpf_get_stackid)(void *ctx, void *map, int flags) =
  	(void *) BPF_FUNC_get_stackid;
+static int (*bpf_copy_to_user)(void *to, void *from, int size) =
+	(void *) BPF_FUNC_copy_to_user;

  /* llvm builtin functions that eBPF C program may use to
   * emit BPF_LD_ABS and BPF_LD_IND instructions
diff --git a/samples/bpf/tracex7_kern.c b/samples/bpf/tracex7_kern.c
new file mode 100644
index 0000000..c341a0a
--- /dev/null
+++ b/samples/bpf/tracex7_kern.c
@@ -0,0 +1,53 @@
+/* Copyright (c) 2013-2015 PLUMgrid, http://plumgrid.com
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ */
+#include <linux/skbuff.h>
+#include <linux/netdevice.h>
+#include <uapi/linux/bpf.h>
+#include <linux/version.h>
+#include "bpf_helpers.h"
+
+
+struct bpf_map_def SEC("maps") dnat_map = {
+	.type = BPF_MAP_TYPE_HASH,
+	.key_size = sizeof(struct sockaddr_in),
+	.value_size = sizeof(struct sockaddr_in),
+	.max_entries = 256,
+};
+
+
+/* kprobe is NOT a stable ABI
+ * kernel functions can be removed, renamed or completely change 
semantics.
+ * Number of arguments and their positions can change, etc.
+ * In such case this bpf+kprobe example will no longer be meaningful
+ *
+ * This example sits on a syscall, and the syscall ABI is relatively 
stable
+ * of course, across platforms, and over time, the ABI may change.
+ */
+SEC("kprobe/sys_connect")
+int bpf_prog1(struct pt_regs *ctx)
+{
+	struct sockaddr_in new_addr, orig_addr = {};
+	struct sockaddr_in *mapped_addr;
+	void *sockaddr_arg = (void *)PT_REGS_PARM2(ctx);
+	int sockaddr_len = (int)PT_REGS_PARM3(ctx);
+
+	if (sockaddr_len > sizeof(orig_addr))
+		return 0;
+
+	if (bpf_probe_read(&orig_addr, sizeof(orig_addr), sockaddr_arg) != 
0)
+		return 0;
+
+	mapped_addr = bpf_map_lookup_elem(&dnat_map, &orig_addr);
+	if (mapped_addr != NULL) {
+		memcpy(&new_addr, mapped_addr, sizeof(new_addr));
+		bpf_copy_to_user(sockaddr_arg, &new_addr, 
sizeof(new_addr));
+	}
+	return 0;
+}
+
+char _license[] SEC("license") = "GPL";
+u32 _version SEC("version") = LINUX_VERSION_CODE;
diff --git a/samples/bpf/tracex7_user.c b/samples/bpf/tracex7_user.c
new file mode 100644
index 0000000..167ee40
--- /dev/null
+++ b/samples/bpf/tracex7_user.c
@@ -0,0 +1,76 @@
+#include <stdio.h>
+#include <assert.h>
+#include <linux/bpf.h>
+#include <unistd.h>
+#include "libbpf.h"
+#include "bpf_load.h"
+#include <sys/socket.h>
+#include <string.h>
+#include <netinet/in.h>
+#include <arpa/inet.h>
+
+int main(int ac, char **argv)
+{
+	int serverfd, serverconnfd, clientfd;
+	socklen_t sockaddr_len;
+	struct sockaddr serv_addr, mapped_addr, tmp_addr;
+	struct sockaddr_in *serv_addr_in, *mapped_addr_in, *tmp_addr_in;
+	char filename[256];
+	char *ip;
+
+	serv_addr_in = (struct sockaddr_in *)&serv_addr;
+	mapped_addr_in = (struct sockaddr_in *)&mapped_addr;
+	tmp_addr_in = (struct sockaddr_in *)&tmp_addr;
+
+	snprintf(filename, sizeof(filename), "%s_kern.o", argv[0]);
+
+	if (load_bpf_file(filename)) {
+		printf("%s", bpf_log_buf);
+		return 1;
+	}
+
+	assert((serverfd = socket(AF_INET, SOCK_STREAM, 0)) > 0);
+	assert((clientfd = socket(AF_INET, SOCK_STREAM, 0)) > 0);
+
+	/* Bind server to ephemeral port on lo */
+	memset(&serv_addr, 0, sizeof(serv_addr));
+	serv_addr_in->sin_family = AF_INET;
+	serv_addr_in->sin_port = 0;
+	serv_addr_in->sin_addr.s_addr = htonl(INADDR_LOOPBACK);
+
+	assert(bind(serverfd, &serv_addr, sizeof(serv_addr)) == 0);
+
+	sockaddr_len = sizeof(serv_addr);
+	assert(getsockname(serverfd, &serv_addr, &sockaddr_len) == 0);
+	ip = inet_ntoa(serv_addr_in->sin_addr);
+	printf("Server bound to: %s:%d\n", ip, 
ntohs(serv_addr_in->sin_port));
+
+	memset(&mapped_addr, 0, sizeof(mapped_addr));
+	mapped_addr_in->sin_family = AF_INET;
+	mapped_addr_in->sin_port = htons(5555);
+	mapped_addr_in->sin_addr.s_addr = htonl(INADDR_LOOPBACK);
+
+	assert(bpf_update_elem(map_fd[0], &mapped_addr, &serv_addr,
+		BPF_ANY) == 0);
+
+	assert(listen(serverfd, 5) == 0);
+
+	ip = inet_ntoa(mapped_addr_in->sin_addr);
+	printf("Client connecting to: %s:%d\n", ip,
+		ntohs(mapped_addr_in->sin_port));
+	assert(connect(clientfd, &mapped_addr, sizeof(mapped_addr)) == 0);
+
+	sockaddr_len = sizeof(tmp_addr);
+	ip = inet_ntoa(tmp_addr_in->sin_addr);
+	assert((serverconnfd = accept(serverfd, &tmp_addr, &sockaddr_len)) 
> 0);
+	printf("Server received connection from: %s:%d\n", ip,
+		ntohs(tmp_addr_in->sin_port));
+
+	sockaddr_len = sizeof(tmp_addr);
+	assert(getpeername(clientfd, &tmp_addr, &sockaddr_len) == 0);
+	ip = inet_ntoa(tmp_addr_in->sin_addr);
+	printf("Client's peer address: %s:%d\n", ip,
+		ntohs(tmp_addr_in->sin_port));
+
+	return 0;
+}

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

[Alexei Starovoitov]

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.

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

[Sargun Dhillon]

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.

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

[Alexei Starovoitov]

On Mon, Jul 18, 2016 at 03:57:17AM -0700, Sargun Dhillon wrote:
> 
> 
> 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);
>  }

I think it can actually work. The only concern is that comment
in access_ok() says that it may sleep whereas I couldn't find
any arch where that would be the case.
Could you please send an official patch with detailed commit log?