RE: use-after-free in sctp_do_sm

[David Laight]

From: Vlad Yasevich
> Sent: 11 December 2015 18:38
...
> > Found a similar place in abort primitive handling like in this last
> > patch update, it's probably the issue you're still triggering.
> >
> > Also found another place that may lead to this use after free, in case
> > we receive a packet with a chunk that has no data.
> >
> > Oh my.. :)
> 
> Yes.  This is what I was worried about...  Anything that triggers
> a DELTE_TCB command has to return a code that we can trap.
> 
> The other way is to do what Dmitri suggested, but even there, we
> need to be very careful.

I'm always wary of anything that queues actions up for later processing.
It is far too easy (as found here) to end up processing actions
in invalid states, or to process actions in 'unusual' orders when
specific events happen close together.

I wonder how much fallout there'd be from getting the sctp code
to immediately action things, instead of queuing the actions for later.
It would certainly remove a lot of the unusual combinations of events.

David


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Vlad Yasevich]

On 12/14/2015 04:50 AM, David Laight wrote:
> From: Vlad Yasevich
>> Sent: 11 December 2015 18:38
> ...
>>> Found a similar place in abort primitive handling like in this last
>>> patch update, it's probably the issue you're still triggering.
>>>
>>> Also found another place that may lead to this use after free, in case
>>> we receive a packet with a chunk that has no data.
>>>
>>> Oh my.. :)
>>
>> Yes.  This is what I was worried about...  Anything that triggers
>> a DELTE_TCB command has to return a code that we can trap.
>>
>> The other way is to do what Dmitri suggested, but even there, we
>> need to be very careful.
> 
> I'm always wary of anything that queues actions up for later processing.
> It is far too easy (as found here) to end up processing actions
> in invalid states, or to process actions in 'unusual' orders when
> specific events happen close together.
> 
> I wonder how much fallout there'd be from getting the sctp code
> to immediately action things, instead of queuing the actions for later.
> It would certainly remove a lot of the unusual combinations of events.
> 

We've bandied this idea around for a while, but no one has had the time
to tackle this.  This would be rather time-consuming task, but in the end
might be a good idea.

-vlad

>   David
> 
> 

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

position SCTP_DISPOSITION_CONSUME.
>  */
> if (0 != (error = sctp_cmd_interpreter(event_type, subtype, state,
> -  ep, asoc,
> +  ep, *asoc,
>event_arg, status,
>commands, gfp)))
> goto bail;
> @@ -1174,11 +1174,12 @@ static int sctp_side_effects(sctp_event_t event_type, 
> sctp_subtype_t subtype,
> break;
>
> case SCTP_DISPOSITION_DELETE_TCB:
> +   case SCTP_DISPOSITION_ABORT:
> /* This should now be a command. */
> +   *asoc = NULL;
> break;
>
> case SCTP_DISPOSITION_CONSUME:
> -   case SCTP_DISPOSITION_ABORT:
> /*
>  * We should no longer have much work to do here as the
>  * real work has been done as explicit commands above.
> diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c
> index 6f46aa16cb76..d801e151498a 100644
> --- a/net/sctp/sm_statefuns.c
> +++ b/net/sctp/sm_statefuns.c
> @@ -4959,12 +4959,10 @@ sctp_disposition_t sctp_sf_cookie_wait_prm_abort(
> sctp_cmd_seq_t *commands)
>  {
> struct sctp_chunk *abort = arg;
> -   sctp_disposition_t retval;
>
> /* Stop T1-init timer */
> sctp_add_cmd_sf(commands, SCTP_CMD_TIMER_STOP,
> SCTP_TO(SCTP_EVENT_TIMEOUT_T1_INIT));
> -   retval = SCTP_DISPOSITION_CONSUME;
>
> sctp_add_cmd_sf(commands, SCTP_CMD_REPLY, SCTP_CHUNK(abort));
>
> @@ -4983,7 +4981,7 @@ sctp_disposition_t sctp_sf_cookie_wait_prm_abort(
> sctp_add_cmd_sf(commands, SCTP_CMD_INIT_FAILED,
> SCTP_PERR(SCTP_ERROR_USER_ABORT));
>
> -   return retval;
> +   return SCTP_DISPOSITION_ABORT;
>  }
>
>  /*


Still happens...
I am on commit aa53685549a2cfb5f175b0c4a20bc9aa1e5a1b85 with your
latest patch applied.
Can you figure out what happens now from the report below? If not I
can create a repro, it's just somewhat time consuming.


BUG: KASAN: use-after-free in sctp_do_sm+0x4bca/0x4db0 at addr 880067c600a8
Read of size 4 by task syzkaller_execu/10266
=
BUG kmalloc-4096 (Tainted: GW  ): kasan: bad access detected
-
Disabling lock debugging due to kernel taint
INFO: Allocated in sctp_association_new+0x6f/0x1da0 age=53 cpu=2 pid=10265
[<  none  >] ___slab_alloc+0x489/0x4e0 mm/slub.c:2468
[<  none  >] __slab_alloc+0x4c/0x90 mm/slub.c:2497
[< inline >] slab_alloc_node mm/slub.c:2560
[< inline >] slab_alloc mm/slub.c:2602
[<  none  >] kmem_cache_alloc_trace+0x264/0x2f0 mm/slub.c:2619
[< inline >] kmalloc include/linux/slab.h:458
[< inline >] kzalloc include/linux/slab.h:602
[<  none  >] sctp_association_new+0x6f/0x1da0 net/sctp/associola.c:302
[<  none  >] sctp_unpack_cookie+0x8b0/0x11c0
net/sctp/sm_make_chunk.c:1812
[<  none  >] sctp_sf_do_5_1D_ce+0x3ca/0x1410 net/sctp/sm_statefuns.c:702
[<  none  >] sctp_do_sm+0x20d/0x4db0 net/sctp/sm_sideeffect.c:1122
[<  none  >] sctp_endpoint_bh_rcv+0x38d/0x830 net/sctp/endpointola.c:486
[<  none  >] sctp_inq_push+0x12c/0x190 net/sctp/inqueue.c:95
[<  none  >] sctp_rcv+0x1d3b/0x2840 net/sctp/input.c:270
[<  none  >] ip_local_deliver_finish+0x2b0/0xa50 net/ipv4/ip_input.c:216
[< inline >] NF_HOOK_THRESH include/linux/netfilter.h:226
[< inline >] NF_HOOK include/linux/netfilter.h:249
[<  none  >] ip_local_deliver+0x1c4/0x2f0 net/ipv4/ip_input.c:257
[< inline >] dst_input include/net/dst.h:465
[<  none  >] ip_rcv_finish+0x5ea/0x1730 net/ipv4/ip_input.c:365
[< inline >] NF_HOOK_THRESH include/linux/netfilter.h:226
[< inline >] NF_HOOK include/linux/netfilter.h:249
[<  none  >] ip_rcv+0x963/0x1080 net/ipv4/ip_input.c:455
[<  none  >] __netif_receive_skb_core+0x1636/0x2f90 net/core/dev.c:3943
[<  none  >] __netif_receive_skb+0x2a/0x160 net/core/dev.c:3978

INFO: Freed in sctp_association_put+0x150/0x250 age=0 cpu=2 pid=10266
[<  none  >] __slab_free+0x1fc/0x320 mm/slub.c:2678
[< inline >] slab_free mm/slub.c:2833
[<  none  >] kfree+0x26a/0x290 mm/slub.c:3662
[< inline >] sctp_association_destroy net/sctp/associola.c:424
[<  none  >] sctp_association_put+0x150/0x250 net/sctp/associola.c:

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

Em 11-12-2015 11:35, Dmitry Vyukov escreveu:

On Wed, Dec 9, 2015 at 5:41 PM, Marcelo Ricardo Leitner
 wrote:

On Wed, Dec 09, 2015 at 01:03:56PM -0200, Marcelo Ricardo Leitner wrote:

On Wed, Dec 09, 2015 at 03:41:29PM +0100, Dmitry Vyukov wrote:

On Tue, Dec 8, 2015 at 8:22 PM, Dmitry Vyukov  wrote:

On Tue, Dec 8, 2015 at 6:40 PM, Marcelo Ricardo Leitner
 wrote:

...

The patches were combined already, but this last pick by Vlad is just
not yet patched. It's not necessary for your testing and I didn't want
to interrupt it in case you were already testing it.

You can use my last patch here, from 2 emails ago, the one which
contains this line:
-   case SCTP_DISPOSITION_ABORT:



You are right. I missed that they are combined. Testing with it now.





Use-after-free still happens.
I am on commit aa53685549a2cfb5f175b0c4a20bc9aa1e5a1b85 (Dec 8) plus
the following sctp-related changes:


Changes are fine.  Ugh. Ok, I'll try your new reproducer here.


Heh I wasn't going to reproduce this by myself anytime soon, I think.
It's using the same socket to connect to itself, and only happens if the
connect() gets there before the listen() call. Figured this out because
I could only reproduce it under strace at first.

Please give this other patch a try. A state command
(sctp_sf_cookie_wait_prm_abort) was issuing SCTP_CMD_INIT_FAILED, which
leads to SCTP_CMD_DELETE_TCB, but returning SCTP_DISPOSITION_CONSUME,
which fooled the patch.

---8<---
commit 9f84d50e36cee0ce66e4ce9b3b1665e0a1dbcdd3
Author: Marcelo Ricardo Leitner 
Date:   Fri Dec 4 15:30:23 2015 -0200

 sctp: fix use-after-free in pr_debug statement

 Dmitry Vyukov reported a use-after-free in the code expanded by the
 macro debug_post_sfx, which is caused by the use of the asoc pointer
 after it was freed within sctp_side_effect() scope.

 This patch fixes it by allowing sctp_side_effect to clear that asoc
 pointer when the TCB is freed.

 As Vlad explained, we also have to cover the SCTP_DISPOSITION_ABORT case
 because it will trigger DELETE_TCB too on that same loop.

 Also, there was a place issuing SCTP_CMD_INIT_FAILED but returning
 SCTP_DISPOSITION_CONSUME, which would fool the scheme above. Fix it by
 returning SCTP_DISPOSITION_ABORT instead.

 The macro is already prepared to handle such NULL pointer.

 Reported-by: Dmitry Vyukov 

diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
index 6098d4c42fa9..be23d5c2074f 100644
--- a/net/sctp/sm_sideeffect.c
+++ b/net/sctp/sm_sideeffect.c
@@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
  sctp_state_t state,
  struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
  void *event_arg,
  sctp_disposition_t status,
  sctp_cmd_seq_t *commands,
@@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
 debug_post_sfn();

 error = sctp_side_effects(event_type, subtype, state,
- ep, asoc, event_arg, status,
+ ep, , event_arg, status,
   , gfp);
 debug_post_sfx();

@@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
  sctp_state_t state,
  struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
  void *event_arg,
  sctp_disposition_t status,
  sctp_cmd_seq_t *commands,
@@ -1151,7 +1151,7 @@ static int sctp_side_effects(sctp_event_t event_type, 
sctp_subtype_t subtype,
  * disposition SCTP_DISPOSITION_CONSUME.
  */
 if (0 != (error = sctp_cmd_interpreter(event_type, subtype, state,
-  ep, asoc,
+  ep, *asoc,
event_arg, status,
commands, gfp)))
 goto bail;
@@ -1174,11 +1174,12 @@ static int sctp_side_effects(sctp_event_t event_type, 
sctp_subtype_t subtype,
 break;

 case SCTP_DISPOSITION_DELETE_TCB:
+   case SCTP_DISPOSITION_ABORT:
 /* This should now be a command. */
+   *asoc = NULL;
   

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Fri, Dec 11, 2015 at 11:51:21AM -0200, Marcelo Ricardo Leitner wrote:
> Em 11-12-2015 11:35, Dmitry Vyukov escreveu:
> >On Wed, Dec 9, 2015 at 5:41 PM, Marcelo Ricardo Leitner
> > wrote:
> >>On Wed, Dec 09, 2015 at 01:03:56PM -0200, Marcelo Ricardo Leitner wrote:
> >>>On Wed, Dec 09, 2015 at 03:41:29PM +0100, Dmitry Vyukov wrote:
> On Tue, Dec 8, 2015 at 8:22 PM, Dmitry Vyukov  wrote:
> >On Tue, Dec 8, 2015 at 6:40 PM, Marcelo Ricardo Leitner
> > wrote:
> >>>...
> >>The patches were combined already, but this last pick by Vlad is just
> >>not yet patched. It's not necessary for your testing and I didn't want
> >>to interrupt it in case you were already testing it.
> >>
> >>You can use my last patch here, from 2 emails ago, the one which
> >>contains this line:
> >>-   case SCTP_DISPOSITION_ABORT:
> >
> >
> >You are right. I missed that they are combined. Testing with it now.
> 
> 
> 
> 
> Use-after-free still happens.
> I am on commit aa53685549a2cfb5f175b0c4a20bc9aa1e5a1b85 (Dec 8) plus
> the following sctp-related changes:
> >>>
> >>>Changes are fine.  Ugh. Ok, I'll try your new reproducer here.
> >>
> >>Heh I wasn't going to reproduce this by myself anytime soon, I think.
> >>It's using the same socket to connect to itself, and only happens if the
> >>connect() gets there before the listen() call. Figured this out because
> >>I could only reproduce it under strace at first.
> >>
> >>Please give this other patch a try. A state command
> >>(sctp_sf_cookie_wait_prm_abort) was issuing SCTP_CMD_INIT_FAILED, which
> >>leads to SCTP_CMD_DELETE_TCB, but returning SCTP_DISPOSITION_CONSUME,
> >>which fooled the patch.
> >>
> >>---8<---
> >>commit 9f84d50e36cee0ce66e4ce9b3b1665e0a1dbcdd3
> >>Author: Marcelo Ricardo Leitner 
> >>Date:   Fri Dec 4 15:30:23 2015 -0200
> >>
> >> sctp: fix use-after-free in pr_debug statement
> >>
> >> Dmitry Vyukov reported a use-after-free in the code expanded by the
> >> macro debug_post_sfx, which is caused by the use of the asoc pointer
> >> after it was freed within sctp_side_effect() scope.
> >>
> >> This patch fixes it by allowing sctp_side_effect to clear that asoc
> >> pointer when the TCB is freed.
> >>
> >> As Vlad explained, we also have to cover the SCTP_DISPOSITION_ABORT 
> >> case
> >> because it will trigger DELETE_TCB too on that same loop.
> >>
> >> Also, there was a place issuing SCTP_CMD_INIT_FAILED but returning
> >> SCTP_DISPOSITION_CONSUME, which would fool the scheme above. Fix it by
> >> returning SCTP_DISPOSITION_ABORT instead.
> >>
> >> The macro is already prepared to handle such NULL pointer.
> >>
> >> Reported-by: Dmitry Vyukov 
> >>
> >>diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
> >>index 6098d4c42fa9..be23d5c2074f 100644
> >>--- a/net/sctp/sm_sideeffect.c
> >>+++ b/net/sctp/sm_sideeffect.c
> >>@@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
> >>  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t 
> >> subtype,
> >>  sctp_state_t state,
> >>  struct sctp_endpoint *ep,
> >>-struct sctp_association *asoc,
> >>+struct sctp_association **asoc,
> >>  void *event_arg,
> >>  sctp_disposition_t status,
> >>  sctp_cmd_seq_t *commands,
> >>@@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t 
> >>event_type, sctp_subtype_t subtype,
> >> debug_post_sfn();
> >>
> >> error = sctp_side_effects(event_type, subtype, state,
> >>- ep, asoc, event_arg, status,
> >>+ ep, , event_arg, status,
> >>   , gfp);
> >> debug_post_sfx();
> >>
> >>@@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t 
> >>event_type, sctp_subtype_t subtype,
> >>  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t 
> >> subtype,
> >>  sctp_state_t state,
> >>  struct sctp_endpoint *ep,
> >>-struct sctp_association *asoc,
> >>+struct sctp_association **asoc,
> >>  void *event_arg,
> >>  sctp_disposition_t status,
> >>  sctp_cmd_seq_t *commands,
> >>@@ -1151,7 +1151,7 @@ static int sctp_side_effects(sctp_event_t event_type, 
> >>sctp_subtype_t subtype,
> >>  * disposition SCTP_DISPOSITION_CONSUME.
> >>  */
> >> if (0 != (error = sctp_cmd_interpreter(event_type, subtype, state,
> >>-  

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Fri, Dec 11, 2015 at 3:03 PM, Marcelo Ricardo Leitner
 wrote:
> On Fri, Dec 11, 2015 at 11:51:21AM -0200, Marcelo Ricardo Leitner wrote:
>> Em 11-12-2015 11:35, Dmitry Vyukov escreveu:
>> >On Wed, Dec 9, 2015 at 5:41 PM, Marcelo Ricardo Leitner
>> > wrote:
>> >>On Wed, Dec 09, 2015 at 01:03:56PM -0200, Marcelo Ricardo Leitner wrote:
>> >>>On Wed, Dec 09, 2015 at 03:41:29PM +0100, Dmitry Vyukov wrote:
>> On Tue, Dec 8, 2015 at 8:22 PM, Dmitry Vyukov  wrote:
>> >On Tue, Dec 8, 2015 at 6:40 PM, Marcelo Ricardo Leitner
>> > wrote:
>> >>>...
>> >>The patches were combined already, but this last pick by Vlad is just
>> >>not yet patched. It's not necessary for your testing and I didn't want
>> >>to interrupt it in case you were already testing it.
>> >>
>> >>You can use my last patch here, from 2 emails ago, the one which
>> >>contains this line:
>> >>-   case SCTP_DISPOSITION_ABORT:
>> >
>> >
>> >You are right. I missed that they are combined. Testing with it now.
>> 
>> 
>> 
>> 
>> Use-after-free still happens.
>> I am on commit aa53685549a2cfb5f175b0c4a20bc9aa1e5a1b85 (Dec 8) plus
>> the following sctp-related changes:
>> >>>
>> >>>Changes are fine.  Ugh. Ok, I'll try your new reproducer here.
>> >>
>> >>Heh I wasn't going to reproduce this by myself anytime soon, I think.
>> >>It's using the same socket to connect to itself, and only happens if the
>> >>connect() gets there before the listen() call. Figured this out because
>> >>I could only reproduce it under strace at first.
>> >>
>> >>Please give this other patch a try. A state command
>> >>(sctp_sf_cookie_wait_prm_abort) was issuing SCTP_CMD_INIT_FAILED, which
>> >>leads to SCTP_CMD_DELETE_TCB, but returning SCTP_DISPOSITION_CONSUME,
>> >>which fooled the patch.
>> >>
>> >>---8<---
>> >>commit 9f84d50e36cee0ce66e4ce9b3b1665e0a1dbcdd3
>> >>Author: Marcelo Ricardo Leitner 
>> >>Date:   Fri Dec 4 15:30:23 2015 -0200
>> >>
>> >> sctp: fix use-after-free in pr_debug statement
>> >>
>> >> Dmitry Vyukov reported a use-after-free in the code expanded by the
>> >> macro debug_post_sfx, which is caused by the use of the asoc pointer
>> >> after it was freed within sctp_side_effect() scope.
>> >>
>> >> This patch fixes it by allowing sctp_side_effect to clear that asoc
>> >> pointer when the TCB is freed.
>> >>
>> >> As Vlad explained, we also have to cover the SCTP_DISPOSITION_ABORT 
>> >> case
>> >> because it will trigger DELETE_TCB too on that same loop.
>> >>
>> >> Also, there was a place issuing SCTP_CMD_INIT_FAILED but returning
>> >> SCTP_DISPOSITION_CONSUME, which would fool the scheme above. Fix it by
>> >> returning SCTP_DISPOSITION_ABORT instead.
>> >>
>> >> The macro is already prepared to handle such NULL pointer.
>> >>
>> >> Reported-by: Dmitry Vyukov 
>> >>
>> >>diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
>> >>index 6098d4c42fa9..be23d5c2074f 100644
>> >>--- a/net/sctp/sm_sideeffect.c
>> >>+++ b/net/sctp/sm_sideeffect.c
>> >>@@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
>> >>  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t 
>> >> subtype,
>> >>  sctp_state_t state,
>> >>  struct sctp_endpoint *ep,
>> >>-struct sctp_association *asoc,
>> >>+struct sctp_association **asoc,
>> >>  void *event_arg,
>> >>  sctp_disposition_t status,
>> >>  sctp_cmd_seq_t *commands,
>> >>@@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t 
>> >>event_type, sctp_subtype_t subtype,
>> >> debug_post_sfn();
>> >>
>> >> error = sctp_side_effects(event_type, subtype, state,
>> >>- ep, asoc, event_arg, status,
>> >>+ ep, , event_arg, status,
>> >>   , gfp);
>> >> debug_post_sfx();
>> >>
>> >>@@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t 
>> >>event_type, sctp_subtype_t subtype,
>> >>  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t 
>> >> subtype,
>> >>  sctp_state_t state,
>> >>  struct sctp_endpoint *ep,
>> >>-struct sctp_association *asoc,
>> >>+struct sctp_association **asoc,
>> >>  void *event_arg,
>> >>  sctp_disposition_t status,
>> >>  sctp_cmd_seq_t *commands,
>> >>@@ -1151,7 +1151,7 @@ static int sctp_side_effects(sctp_event_t 
>> >>event_type, 

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

Em 11-12-2015 12:30, Dmitry Vyukov escreveu:

On Fri, Dec 11, 2015 at 3:03 PM, Marcelo Ricardo Leitner
 wrote:

On Fri, Dec 11, 2015 at 11:51:21AM -0200, Marcelo Ricardo Leitner wrote:

Em 11-12-2015 11:35, Dmitry Vyukov escreveu:

On Wed, Dec 9, 2015 at 5:41 PM, Marcelo Ricardo Leitner
 wrote:

On Wed, Dec 09, 2015 at 01:03:56PM -0200, Marcelo Ricardo Leitner wrote:

On Wed, Dec 09, 2015 at 03:41:29PM +0100, Dmitry Vyukov wrote:

On Tue, Dec 8, 2015 at 8:22 PM, Dmitry Vyukov  wrote:

On Tue, Dec 8, 2015 at 6:40 PM, Marcelo Ricardo Leitner
 wrote:

...

The patches were combined already, but this last pick by Vlad is just
not yet patched. It's not necessary for your testing and I didn't want
to interrupt it in case you were already testing it.

You can use my last patch here, from 2 emails ago, the one which
contains this line:
-   case SCTP_DISPOSITION_ABORT:



You are right. I missed that they are combined. Testing with it now.





Use-after-free still happens.
I am on commit aa53685549a2cfb5f175b0c4a20bc9aa1e5a1b85 (Dec 8) plus
the following sctp-related changes:


Changes are fine.  Ugh. Ok, I'll try your new reproducer here.


Heh I wasn't going to reproduce this by myself anytime soon, I think.
It's using the same socket to connect to itself, and only happens if the
connect() gets there before the listen() call. Figured this out because
I could only reproduce it under strace at first.

Please give this other patch a try. A state command
(sctp_sf_cookie_wait_prm_abort) was issuing SCTP_CMD_INIT_FAILED, which
leads to SCTP_CMD_DELETE_TCB, but returning SCTP_DISPOSITION_CONSUME,
which fooled the patch.

---8<---
commit 9f84d50e36cee0ce66e4ce9b3b1665e0a1dbcdd3
Author: Marcelo Ricardo Leitner 
Date:   Fri Dec 4 15:30:23 2015 -0200

 sctp: fix use-after-free in pr_debug statement

 Dmitry Vyukov reported a use-after-free in the code expanded by the
 macro debug_post_sfx, which is caused by the use of the asoc pointer
 after it was freed within sctp_side_effect() scope.

 This patch fixes it by allowing sctp_side_effect to clear that asoc
 pointer when the TCB is freed.

 As Vlad explained, we also have to cover the SCTP_DISPOSITION_ABORT case
 because it will trigger DELETE_TCB too on that same loop.

 Also, there was a place issuing SCTP_CMD_INIT_FAILED but returning
 SCTP_DISPOSITION_CONSUME, which would fool the scheme above. Fix it by
 returning SCTP_DISPOSITION_ABORT instead.

 The macro is already prepared to handle such NULL pointer.

 Reported-by: Dmitry Vyukov 

diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
index 6098d4c42fa9..be23d5c2074f 100644
--- a/net/sctp/sm_sideeffect.c
+++ b/net/sctp/sm_sideeffect.c
@@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
  sctp_state_t state,
  struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
  void *event_arg,
  sctp_disposition_t status,
  sctp_cmd_seq_t *commands,
@@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
 debug_post_sfn();

 error = sctp_side_effects(event_type, subtype, state,
- ep, asoc, event_arg, status,
+ ep, , event_arg, status,
   , gfp);
 debug_post_sfx();

@@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
  sctp_state_t state,
  struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
  void *event_arg,
  sctp_disposition_t status,
  sctp_cmd_seq_t *commands,
@@ -1151,7 +1151,7 @@ static int sctp_side_effects(sctp_event_t event_type, 
sctp_subtype_t subtype,
  * disposition SCTP_DISPOSITION_CONSUME.
  */
 if (0 != (error = sctp_cmd_interpreter(event_type, subtype, state,
-  ep, asoc,
+  ep, *asoc,
event_arg, status,
commands, gfp)))
 goto bail;
@@ -1174,11 +1174,12 @@ static int sctp_side_effects(sctp_event_t event_type, 

Re: use-after-free in sctp_do_sm

[Vlad Yasevich]

On 12/11/2015 09:03 AM, Marcelo Ricardo Leitner wrote:
> On Fri, Dec 11, 2015 at 11:51:21AM -0200, Marcelo Ricardo Leitner wrote:
>> Em 11-12-2015 11:35, Dmitry Vyukov escreveu:
>>> On Wed, Dec 9, 2015 at 5:41 PM, Marcelo Ricardo Leitner
>>>  wrote:
 On Wed, Dec 09, 2015 at 01:03:56PM -0200, Marcelo Ricardo Leitner wrote:
> On Wed, Dec 09, 2015 at 03:41:29PM +0100, Dmitry Vyukov wrote:
>> On Tue, Dec 8, 2015 at 8:22 PM, Dmitry Vyukov  wrote:
>>> On Tue, Dec 8, 2015 at 6:40 PM, Marcelo Ricardo Leitner
>>>  wrote:
> ...
 The patches were combined already, but this last pick by Vlad is just
 not yet patched. It's not necessary for your testing and I didn't want
 to interrupt it in case you were already testing it.

 You can use my last patch here, from 2 emails ago, the one which
 contains this line:
 -   case SCTP_DISPOSITION_ABORT:
>>>
>>>
>>> You are right. I missed that they are combined. Testing with it now.
>>
>>
>>
>>
>> Use-after-free still happens.
>> I am on commit aa53685549a2cfb5f175b0c4a20bc9aa1e5a1b85 (Dec 8) plus
>> the following sctp-related changes:
>
> Changes are fine.  Ugh. Ok, I'll try your new reproducer here.

 Heh I wasn't going to reproduce this by myself anytime soon, I think.
 It's using the same socket to connect to itself, and only happens if the
 connect() gets there before the listen() call. Figured this out because
 I could only reproduce it under strace at first.

 Please give this other patch a try. A state command
 (sctp_sf_cookie_wait_prm_abort) was issuing SCTP_CMD_INIT_FAILED, which
 leads to SCTP_CMD_DELETE_TCB, but returning SCTP_DISPOSITION_CONSUME,
 which fooled the patch.

 ---8<---
 commit 9f84d50e36cee0ce66e4ce9b3b1665e0a1dbcdd3
 Author: Marcelo Ricardo Leitner 
 Date:   Fri Dec 4 15:30:23 2015 -0200

 sctp: fix use-after-free in pr_debug statement

 Dmitry Vyukov reported a use-after-free in the code expanded by the
 macro debug_post_sfx, which is caused by the use of the asoc pointer
 after it was freed within sctp_side_effect() scope.

 This patch fixes it by allowing sctp_side_effect to clear that asoc
 pointer when the TCB is freed.

 As Vlad explained, we also have to cover the SCTP_DISPOSITION_ABORT 
 case
 because it will trigger DELETE_TCB too on that same loop.

 Also, there was a place issuing SCTP_CMD_INIT_FAILED but returning
 SCTP_DISPOSITION_CONSUME, which would fool the scheme above. Fix it by
 returning SCTP_DISPOSITION_ABORT instead.

 The macro is already prepared to handle such NULL pointer.

 Reported-by: Dmitry Vyukov 

 diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
 index 6098d4c42fa9..be23d5c2074f 100644
 --- a/net/sctp/sm_sideeffect.c
 +++ b/net/sctp/sm_sideeffect.c
 @@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t 
 subtype,
  sctp_state_t state,
  struct sctp_endpoint *ep,
 -struct sctp_association *asoc,
 +struct sctp_association **asoc,
  void *event_arg,
  sctp_disposition_t status,
  sctp_cmd_seq_t *commands,
 @@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t 
 event_type, sctp_subtype_t subtype,
 debug_post_sfn();

 error = sctp_side_effects(event_type, subtype, state,
 - ep, asoc, event_arg, status,
 + ep, , event_arg, status,
   , gfp);
 debug_post_sfx();

 @@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t 
 event_type, sctp_subtype_t subtype,
  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t 
 subtype,
  sctp_state_t state,
  struct sctp_endpoint *ep,
 -struct sctp_association *asoc,
 +struct sctp_association **asoc,
  void *event_arg,
  sctp_disposition_t status,
  sctp_cmd_seq_t *commands,
 @@ -1151,7 +1151,7 @@ static int sctp_side_effects(sctp_event_t 
 event_type, sctp_subtype_t subtype,
  * disposition SCTP_DISPOSITION_CONSUME.
  */
 

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

5\x4f\xd3\xc1\xef\xc7\xbf\x1d\x0c\xe1\xf2\xc6\x64\x9d\xb5\x98\x5e\xc0\x1b\x7e\x83\xee\x06\x79\x10\x3b\xeb\x3c\x89\x9e\x30\xb6\xb5\xbd\xf9\xaa\xc1\xe0\x47\xdf\xed\x94\xda\xc5\xcb\x21\x32\x66\xbd\xc9\xa5\x84\xbc\x32\x8f\xce\x8e\xff\x1f\x76\x63\x67\x2f\x40\xc7\x42\xa3\x60\x17\xd6\x05\x45\xc2\x10\xd1\x53\x5f\x0d\x02\xcd\xf1\x44\x30",
112);
memcpy((void*)0x20004f80,
"\x0a\x00\x33\xdc\x14\x4d\x5b\xd1\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xdd\x01\xf8\xfd\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00",
128);
long r13 = syscall(SYS_sendto, r0, 0x20004f90ul, 0x70ul,
0x0ul, 0x20004f80ul, 0x80ul);
long r14 = syscall(SYS_listen, r0, 0x3ul, 0, 0, 0, 0);
long r15 = syscall(SYS_accept4, r0, 0x20003f80ul,
0x20003ab4ul, 0x80800ul, 0, 0);
*(uint64_t*)0x20003000 = 0x2;
*(uint64_t*)0x20003008 = 0x2;
*(uint64_t*)0x20003010 = 0x1;
*(uint64_t*)0x20003018 = 0x7;
*(uint64_t*)0x20003020 = 0x7;
*(uint64_t*)0x20003028 = 0x5f;
*(uint64_t*)0x20003030 = 0x9;
*(uint64_t*)0x20003038 = 0x88;
long r24 = syscall(SYS_setsockopt, r0, 0xfff7ul,
0x8ul, 0x20003000ul, 0x40ul, 0);
long r25 = syscall(SYS_dup3, r15, r0, 0x8ul, 0, 0, 0);
memcpy((void*)0x20006000,
"\xd9\x4f\xbe\x3f\x43\x89\x02\x0d\x1e\x84\x8d\x16\xe8\xdf\xdd\x27\x1f\xfe\xc6\x4a\xfa\x93\x00\xb9\xaf\xd7\x5e\xf1\x1f\x88\xc4\x57\x12\x70\xb4\xc5\xa6\xfc\xb9\x99\xd2\x80\x30\x2a\x53\xda\xd2\x57\x6d\xdc",
50);
long r27 = syscall(SYS_setsockopt, r25, 0x117ul, 0x1ul,
0x20006000ul, 0x32ul, 0);
long r28 = syscall(SYS_close, r15, 0, 0, 0, 0, 0);
return 0;
}


The use-after-free reports:

==
BUG: KASAN: use-after-free in sctp_do_sm+0x530e/0x5d90 at addr 880069d4c808
Read of size 4 by task a.out/8211
=
BUG kmalloc-4096 (Tainted: GB  ): kasan: bad access detected
-

INFO: Allocated in sctp_association_new+0xbd/0x21d0 age=9 cpu=3 pid=8211
[<  none  >] ___slab_alloc+0x648/0x8c0 mm/slub.c:2468
[<  none  >] __slab_alloc+0x4c/0x90 mm/slub.c:2497
[< inline >] slab_alloc_node mm/slub.c:2560
[< inline >] slab_alloc mm/slub.c:2602
[<  none  >] kmem_cache_alloc_trace+0x23c/0x3f0 mm/slub.c:2619
[< inline >] kmalloc include/linux/slab.h:458
[< inline >] kzalloc include/linux/slab.h:602
[<  none  >] sctp_association_new+0xbd/0x21d0 net/sctp/associola.c:302
[<  none  >] __sctp_connect+0x5e8/0xd80 net/sctp/socket.c:1161
[<  none  >] sctp_connect+0xdc/0x130 net/sctp/socket.c:3874
[<  none  >] inet_dgram_connect+0x136/0x2a0 net/ipv4/af_inet.c:528
[<  none  >] SYSC_connect+0x263/0x380 net/socket.c:1542
[<  none  >] SyS_connect+0x24/0x30 net/socket.c:1523
[<  none  >] entry_SYSCALL_64_fastpath+0x16/0x7a
arch/x86/entry/entry_64.S:185

INFO: Freed in sctp_association_put+0x179/0x2c0 age=11 cpu=3 pid=8211
[<  none  >] __slab_free+0x21e/0x3e0 mm/slub.c:2678
[< inline >] slab_free mm/slub.c:2833
[<  none  >] kfree+0x26f/0x3e0 mm/slub.c:3662
[< inline >] sctp_association_destroy net/sctp/associola.c:424
[<  none  >] sctp_association_put+0x179/0x2c0 net/sctp/associola.c:860
[<  none  >] sctp_association_free+0x416/0x5d0 net/sctp/associola.c:402
[< inline >] sctp_cmd_delete_tcb net/sctp/sm_sideeffect.c:867
[< inline >] sctp_cmd_interpreter net/sctp/sm_sideeffect.c:1287
[< inline >] sctp_side_effects net/sctp/sm_sideeffect.c:1153
[<  none  >] sctp_do_sm+0x1364/0x5d90 net/sctp/sm_sideeffect.c:1125
[<  none  >] sctp_primitive_ABORT+0xa9/0xd0 net/sctp/primitive.c:119
[<  none  >] sctp_close+0x2ad/0x9b0 net/sctp/socket.c:1517
[<  none  >] inet_release+0x111/0x270 net/ipv4/af_inet.c:413
[<  none  >] inet6_release+0x55/0x90 net/ipv6/af_inet6.c:406
[<  none  >] sock_release+0x96/0x260 net/socket.c:571
[<  none  >] sock_close+0x16/0x20 net/socket.c:1022
[<  none  >] __fput+0x244/0x860 fs/file_table.c:208
[<  none  >] fput+0x15/0x20 fs/file_table.c:244
[<  none  >] task_work_run+0x130/0x240 kernel/tas

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Wed, Dec 09, 2015 at 03:41:29PM +0100, Dmitry Vyukov wrote:
> On Tue, Dec 8, 2015 at 8:22 PM, Dmitry Vyukov  wrote:
> > On Tue, Dec 8, 2015 at 6:40 PM, Marcelo Ricardo Leitner
> >  wrote:
...
> >> The patches were combined already, but this last pick by Vlad is just
> >> not yet patched. It's not necessary for your testing and I didn't want
> >> to interrupt it in case you were already testing it.
> >>
> >> You can use my last patch here, from 2 emails ago, the one which
> >> contains this line:
> >> -   case SCTP_DISPOSITION_ABORT:
> >
> >
> > You are right. I missed that they are combined. Testing with it now.
> 
> 
> 
> 
> Use-after-free still happens.
> I am on commit aa53685549a2cfb5f175b0c4a20bc9aa1e5a1b85 (Dec 8) plus
> the following sctp-related changes:

Changes are fine.  Ugh. Ok, I'll try your new reproducer here.

  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Wed, Dec 09, 2015 at 01:03:56PM -0200, Marcelo Ricardo Leitner wrote:
> On Wed, Dec 09, 2015 at 03:41:29PM +0100, Dmitry Vyukov wrote:
> > On Tue, Dec 8, 2015 at 8:22 PM, Dmitry Vyukov  wrote:
> > > On Tue, Dec 8, 2015 at 6:40 PM, Marcelo Ricardo Leitner
> > >  wrote:
> ...
> > >> The patches were combined already, but this last pick by Vlad is just
> > >> not yet patched. It's not necessary for your testing and I didn't want
> > >> to interrupt it in case you were already testing it.
> > >>
> > >> You can use my last patch here, from 2 emails ago, the one which
> > >> contains this line:
> > >> -   case SCTP_DISPOSITION_ABORT:
> > >
> > >
> > > You are right. I missed that they are combined. Testing with it now.
> > 
> > 
> > 
> > 
> > Use-after-free still happens.
> > I am on commit aa53685549a2cfb5f175b0c4a20bc9aa1e5a1b85 (Dec 8) plus
> > the following sctp-related changes:
> 
> Changes are fine.  Ugh. Ok, I'll try your new reproducer here.

Heh I wasn't going to reproduce this by myself anytime soon, I think.
It's using the same socket to connect to itself, and only happens if the
connect() gets there before the listen() call. Figured this out because
I could only reproduce it under strace at first.

Please give this other patch a try. A state command
(sctp_sf_cookie_wait_prm_abort) was issuing SCTP_CMD_INIT_FAILED, which
leads to SCTP_CMD_DELETE_TCB, but returning SCTP_DISPOSITION_CONSUME,
which fooled the patch.

---8<---
commit 9f84d50e36cee0ce66e4ce9b3b1665e0a1dbcdd3
Author: Marcelo Ricardo Leitner 
Date:   Fri Dec 4 15:30:23 2015 -0200

sctp: fix use-after-free in pr_debug statement

Dmitry Vyukov reported a use-after-free in the code expanded by the
macro debug_post_sfx, which is caused by the use of the asoc pointer
after it was freed within sctp_side_effect() scope.

This patch fixes it by allowing sctp_side_effect to clear that asoc
pointer when the TCB is freed.

As Vlad explained, we also have to cover the SCTP_DISPOSITION_ABORT case
because it will trigger DELETE_TCB too on that same loop.

Also, there was a place issuing SCTP_CMD_INIT_FAILED but returning
SCTP_DISPOSITION_CONSUME, which would fool the scheme above. Fix it by
returning SCTP_DISPOSITION_ABORT instead.

The macro is already prepared to handle such NULL pointer.

Reported-by: Dmitry Vyukov 

diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
index 6098d4c42fa9..be23d5c2074f 100644
--- a/net/sctp/sm_sideeffect.c
+++ b/net/sctp/sm_sideeffect.c
@@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
 static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
 sctp_state_t state,
 struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
 void *event_arg,
 sctp_disposition_t status,
 sctp_cmd_seq_t *commands,
@@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
debug_post_sfn();
 
error = sctp_side_effects(event_type, subtype, state,
- ep, asoc, event_arg, status,
+ ep, , event_arg, status,
  , gfp);
debug_post_sfx();
 
@@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
 static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
 sctp_state_t state,
 struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
 void *event_arg,
 sctp_disposition_t status,
 sctp_cmd_seq_t *commands,
@@ -1151,7 +1151,7 @@ static int sctp_side_effects(sctp_event_t event_type, 
sctp_subtype_t subtype,
 * disposition SCTP_DISPOSITION_CONSUME.
 */
if (0 != (error = sctp_cmd_interpreter(event_type, subtype, state,
-  ep, asoc,
+  ep, *asoc,
   event_arg, status,
   commands, gfp)))
goto bail;
@@ -1174,11 +1174,12 @@ static int sctp_side_effects(sctp_event_t event_type, 
sctp_subtype_t subtype,
break;
 
case SCTP_DISPOSITION_DELETE_TCB:
+   case SCTP_DISPOSITION_ABORT:
/* This should now be a command. */
+   *asoc = NULL;
break;
 
case 

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Mon, Dec 7, 2015 at 9:52 PM, Marcelo Ricardo Leitner
 wrote:
> Em 07-12-2015 18:37, Vlad Yasevich escreveu:
>>
>> On 12/07/2015 02:50 PM, Marcelo Ricardo Leitner wrote:
>>>
>>> On Mon, Dec 07, 2015 at 02:33:52PM -0500, Vlad Yasevich wrote:

 On 12/07/2015 01:52 PM, Marcelo Ricardo Leitner wrote:
>
> Vlad, I reviewed the places on which it returns SCTP_DISPOSITION_ABORT,
> and if I didn't miss something in there all of them either issue
> SCTP_CMD_ASSOC_FAILED or SCTP_CMD_INIT_FAILED before returning it, thus
> delaying DELETE_TCB and with that the asoc free.


 They delay it from the perspective of the command interpreter since the
 command
 to delete the TCB happens a little later, but status code  is checked
 after all
 commands are processed and command processing doesn't change it.  So the
 'status'
 code would still be SCTP_DISPOSITION_ABORT after DELETE_TCB command was
 processed.
 So, I think we may still have an use-after-free issue here.
>>>
>>>
>>> Gotcha! That's pretty much it then. From that point of view now, there
>>> shouldn't be a case that it returns _ABORT without freeing the asoc in
>>> the same loop. (more below)
>>>
> There is one place,
> though, that may not do it that way, it's sctp_sf_abort_violation(),
> but
> then that code only runs if asoc is already NULL by then.


 I don't believe so.  The violation state function can run with a
 non-NULL association
 if we are encountering protocol violations after the association is
 established.
>>>
>>>
>>> Yup, that's correct. I just tried to reference one case on which it
>>> would return _ABORT without issuing any of those _FAILEDs before doing
>>> so (meaning the association could still be valid) but that in that case,
>>> the asoc was already NULL.
>>
>>
>> I think it is possible to hit the 'discard:' tag in that function while
>> still
>> having a valid association.  That happens when ABORT chunk is required to
>> be
>> authenticated.  This that case, instead of generating an ABORT and
>> terminating the
>> current association, we just drop the packet, but still report an _ABORT
>> disposition code.
>>
>> This probably need to change if we are going to catch the _ABORT
>> disposition and
>> clear the asoc pointer.
>
>
> Oups. Nice one. I'll switch it to SCTP_DISPOSITION_DISCARD if it hits that
> if() then. Thanks Vlad.


So I am waiting for a new patch, right?
Can you please combine all changes into a single patch (as far as I
understand the previous one must be applied on top of the first one)?
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Tue, Dec 08, 2015 at 06:30:51PM +0100, Dmitry Vyukov wrote:
> On Mon, Dec 7, 2015 at 9:52 PM, Marcelo Ricardo Leitner
>  wrote:
> > Em 07-12-2015 18:37, Vlad Yasevich escreveu:
> >>
> >> On 12/07/2015 02:50 PM, Marcelo Ricardo Leitner wrote:
> >>>
> >>> On Mon, Dec 07, 2015 at 02:33:52PM -0500, Vlad Yasevich wrote:
> 
>  On 12/07/2015 01:52 PM, Marcelo Ricardo Leitner wrote:
> >
> > Vlad, I reviewed the places on which it returns SCTP_DISPOSITION_ABORT,
> > and if I didn't miss something in there all of them either issue
> > SCTP_CMD_ASSOC_FAILED or SCTP_CMD_INIT_FAILED before returning it, thus
> > delaying DELETE_TCB and with that the asoc free.
> 
> 
>  They delay it from the perspective of the command interpreter since the
>  command
>  to delete the TCB happens a little later, but status code  is checked
>  after all
>  commands are processed and command processing doesn't change it.  So the
>  'status'
>  code would still be SCTP_DISPOSITION_ABORT after DELETE_TCB command was
>  processed.
>  So, I think we may still have an use-after-free issue here.
> >>>
> >>>
> >>> Gotcha! That's pretty much it then. From that point of view now, there
> >>> shouldn't be a case that it returns _ABORT without freeing the asoc in
> >>> the same loop. (more below)
> >>>
> > There is one place,
> > though, that may not do it that way, it's sctp_sf_abort_violation(),
> > but
> > then that code only runs if asoc is already NULL by then.
> 
> 
>  I don't believe so.  The violation state function can run with a
>  non-NULL association
>  if we are encountering protocol violations after the association is
>  established.
> >>>
> >>>
> >>> Yup, that's correct. I just tried to reference one case on which it
> >>> would return _ABORT without issuing any of those _FAILEDs before doing
> >>> so (meaning the association could still be valid) but that in that case,
> >>> the asoc was already NULL.
> >>
> >>
> >> I think it is possible to hit the 'discard:' tag in that function while
> >> still
> >> having a valid association.  That happens when ABORT chunk is required to
> >> be
> >> authenticated.  This that case, instead of generating an ABORT and
> >> terminating the
> >> current association, we just drop the packet, but still report an _ABORT
> >> disposition code.
> >>
> >> This probably need to change if we are going to catch the _ABORT
> >> disposition and
> >> clear the asoc pointer.
> >
> >
> > Oups. Nice one. I'll switch it to SCTP_DISPOSITION_DISCARD if it hits that
> > if() then. Thanks Vlad.
> 
> 
> So I am waiting for a new patch, right?
> Can you please combine all changes into a single patch (as far as I
> understand the previous one must be applied on top of the first one)?

The patches were combined already, but this last pick by Vlad is just
not yet patched. It's not necessary for your testing and I didn't want
to interrupt it in case you were already testing it.

You can use my last patch here, from 2 emails ago, the one which
contains this line:
-   case SCTP_DISPOSITION_ABORT:

  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Tue, Dec 8, 2015 at 6:40 PM, Marcelo Ricardo Leitner
 wrote:
> On Tue, Dec 08, 2015 at 06:30:51PM +0100, Dmitry Vyukov wrote:
>> On Mon, Dec 7, 2015 at 9:52 PM, Marcelo Ricardo Leitner
>>  wrote:
>> > Em 07-12-2015 18:37, Vlad Yasevich escreveu:
>> >>
>> >> On 12/07/2015 02:50 PM, Marcelo Ricardo Leitner wrote:
>> >>>
>> >>> On Mon, Dec 07, 2015 at 02:33:52PM -0500, Vlad Yasevich wrote:
>> 
>>  On 12/07/2015 01:52 PM, Marcelo Ricardo Leitner wrote:
>> >
>> > Vlad, I reviewed the places on which it returns SCTP_DISPOSITION_ABORT,
>> > and if I didn't miss something in there all of them either issue
>> > SCTP_CMD_ASSOC_FAILED or SCTP_CMD_INIT_FAILED before returning it, thus
>> > delaying DELETE_TCB and with that the asoc free.
>> 
>> 
>>  They delay it from the perspective of the command interpreter since the
>>  command
>>  to delete the TCB happens a little later, but status code  is checked
>>  after all
>>  commands are processed and command processing doesn't change it.  So the
>>  'status'
>>  code would still be SCTP_DISPOSITION_ABORT after DELETE_TCB command was
>>  processed.
>>  So, I think we may still have an use-after-free issue here.
>> >>>
>> >>>
>> >>> Gotcha! That's pretty much it then. From that point of view now, there
>> >>> shouldn't be a case that it returns _ABORT without freeing the asoc in
>> >>> the same loop. (more below)
>> >>>
>> > There is one place,
>> > though, that may not do it that way, it's sctp_sf_abort_violation(),
>> > but
>> > then that code only runs if asoc is already NULL by then.
>> 
>> 
>>  I don't believe so.  The violation state function can run with a
>>  non-NULL association
>>  if we are encountering protocol violations after the association is
>>  established.
>> >>>
>> >>>
>> >>> Yup, that's correct. I just tried to reference one case on which it
>> >>> would return _ABORT without issuing any of those _FAILEDs before doing
>> >>> so (meaning the association could still be valid) but that in that case,
>> >>> the asoc was already NULL.
>> >>
>> >>
>> >> I think it is possible to hit the 'discard:' tag in that function while
>> >> still
>> >> having a valid association.  That happens when ABORT chunk is required to
>> >> be
>> >> authenticated.  This that case, instead of generating an ABORT and
>> >> terminating the
>> >> current association, we just drop the packet, but still report an _ABORT
>> >> disposition code.
>> >>
>> >> This probably need to change if we are going to catch the _ABORT
>> >> disposition and
>> >> clear the asoc pointer.
>> >
>> >
>> > Oups. Nice one. I'll switch it to SCTP_DISPOSITION_DISCARD if it hits that
>> > if() then. Thanks Vlad.
>>
>>
>> So I am waiting for a new patch, right?
>> Can you please combine all changes into a single patch (as far as I
>> understand the previous one must be applied on top of the first one)?
>
> The patches were combined already, but this last pick by Vlad is just
> not yet patched. It's not necessary for your testing and I didn't want
> to interrupt it in case you were already testing it.
>
> You can use my last patch here, from 2 emails ago, the one which
> contains this line:
> -   case SCTP_DISPOSITION_ABORT:


You are right. I missed that they are combined. Testing with it now.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Mon, Dec 7, 2015 at 2:15 PM, Marcelo Ricardo Leitner
 wrote:
> On Mon, Dec 07, 2015 at 12:26:09PM +0100, Dmitry Vyukov wrote:
>> On Sat, Dec 5, 2015 at 5:39 PM, Vlad Yasevich  wrote:
>> > On 12/04/2015 04:34 PM, Marcelo Ricardo Leitner wrote:
>> >> On Fri, Dec 04, 2015 at 09:25:35PM +0100, Dmitry Vyukov wrote:
>> >>> On Fri, Dec 4, 2015 at 6:48 PM, Marcelo Ricardo Leitner
>> >>>  wrote:
>>  Hi Dmitry,
>> 
>>  Can you please test this patch?
>>  I'll re-post with proper subject if it works.
>> >>>
>> >>> Still happening with the same stacks.
>> >>
>> >> Then there may be another one, I'm afraid.
>> >>
>> >> I'm using the testapp you shared in the first email, with that debug line
>> >> enabled and added a new one:
>> >> +   pr_debug("%p %d\n", asoc, asoc ? asoc->state : 0);
>> >> debug_post_sfx();
>> >> (should have used %x, but ok)
>> >>
>> >> Also enabled slub_debug=PUZ, and I get:
>> >>
>> >> without the patch:
>> >> [   87.873640] sctp: 8800b71533d8 1
>> >> [   87.873647] sctp: sctp_do_sm[post-sfx]: error:0,
>> >> asoc:8800b71533d8[STATE_CLOSED]
>> >> [   87.873739] sctp: 8800b71533d8 1
>> >> [   87.873742] sctp: sctp_do_sm[post-sfx]: error:0,
>> >> asoc:8800b71533d8[STATE_CLOSED]
>> >> [   87.875149] sctp: 8800b71533d8 1802201963
>> >> [   87.875238] sctp: sctp_do_sm[post-sfx]: error:0,
>> >> asoc:8800b71533d8[STATE_CLOSED]
>> >>
>> >> 1802201963 = 0x6b6b6b6b, poison
>> >>
>> >> with the patch:
>> >> [   81.071265] sctp: 880137571148 1
>> >> [   81.071273] sctp: sctp_do_sm[post-sfx]: error:0,
>> >> asoc:880137571148[STATE_CLOSED]
>> >> [   81.071372] sctp: 880137571148 1
>> >> [   81.071375] sctp: sctp_do_sm[post-sfx]: error:0,
>> >> asoc:880137571148[STATE_CLOSED]
>> >> [   81.072423] sctp:   (null) 0
>> >> [   81.072427] sctp: sctp_do_sm[post-sfx]: error:0, asoc:
>> >> (null)[STATE_CLOSED]
>> >>
>> >> This one, at least, is gone with this patch.
>> >>
>> >>   Marcelo
>> >>
>> >
>> > Hi Marcelo
>> >
>> > I think you also need to catch the SCTP_DISPOSITION_ABORT and update
>> > the pointer.  There are some issues there though as some functions report
>> > that code without actually destroying the association.  This happens when
>> > the ABORT chunk may be dropped.
>> >
>> > I think this might be why we still see the issue.
>>
>>
>> Marcelo,
>>
>> Is this info enough for you to cook another fix?
>
> Hi, I think so. I was really wondering how you could trigger that issue
> without the timestamp fix and Vlad's comment does shed some light on it.
>
> I'll do more tests later today, but what did you have connecting to the
> listening socket? Somehow you made that accept() call to return..

Local connect in another thread I guess.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Sat, Dec 5, 2015 at 5:39 PM, Vlad Yasevich  wrote:
> On 12/04/2015 04:34 PM, Marcelo Ricardo Leitner wrote:
>> On Fri, Dec 04, 2015 at 09:25:35PM +0100, Dmitry Vyukov wrote:
>>> On Fri, Dec 4, 2015 at 6:48 PM, Marcelo Ricardo Leitner
>>>  wrote:
 Hi Dmitry,

 Can you please test this patch?
 I'll re-post with proper subject if it works.
>>>
>>> Still happening with the same stacks.
>>
>> Then there may be another one, I'm afraid.
>>
>> I'm using the testapp you shared in the first email, with that debug line
>> enabled and added a new one:
>> +   pr_debug("%p %d\n", asoc, asoc ? asoc->state : 0);
>> debug_post_sfx();
>> (should have used %x, but ok)
>>
>> Also enabled slub_debug=PUZ, and I get:
>>
>> without the patch:
>> [   87.873640] sctp: 8800b71533d8 1
>> [   87.873647] sctp: sctp_do_sm[post-sfx]: error:0,
>> asoc:8800b71533d8[STATE_CLOSED]
>> [   87.873739] sctp: 8800b71533d8 1
>> [   87.873742] sctp: sctp_do_sm[post-sfx]: error:0,
>> asoc:8800b71533d8[STATE_CLOSED]
>> [   87.875149] sctp: 8800b71533d8 1802201963
>> [   87.875238] sctp: sctp_do_sm[post-sfx]: error:0,
>> asoc:8800b71533d8[STATE_CLOSED]
>>
>> 1802201963 = 0x6b6b6b6b, poison
>>
>> with the patch:
>> [   81.071265] sctp: 880137571148 1
>> [   81.071273] sctp: sctp_do_sm[post-sfx]: error:0,
>> asoc:880137571148[STATE_CLOSED]
>> [   81.071372] sctp: 880137571148 1
>> [   81.071375] sctp: sctp_do_sm[post-sfx]: error:0,
>> asoc:880137571148[STATE_CLOSED]
>> [   81.072423] sctp:   (null) 0
>> [   81.072427] sctp: sctp_do_sm[post-sfx]: error:0, asoc:
>> (null)[STATE_CLOSED]
>>
>> This one, at least, is gone with this patch.
>>
>>   Marcelo
>>
>
> Hi Marcelo
>
> I think you also need to catch the SCTP_DISPOSITION_ABORT and update
> the pointer.  There are some issues there though as some functions report
> that code without actually destroying the association.  This happens when
> the ABORT chunk may be dropped.
>
> I think this might be why we still see the issue.


Marcelo,

Is this info enough for you to cook another fix?
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Mon, Dec 07, 2015 at 12:26:09PM +0100, Dmitry Vyukov wrote:
> On Sat, Dec 5, 2015 at 5:39 PM, Vlad Yasevich  wrote:
> > On 12/04/2015 04:34 PM, Marcelo Ricardo Leitner wrote:
> >> On Fri, Dec 04, 2015 at 09:25:35PM +0100, Dmitry Vyukov wrote:
> >>> On Fri, Dec 4, 2015 at 6:48 PM, Marcelo Ricardo Leitner
> >>>  wrote:
>  Hi Dmitry,
> 
>  Can you please test this patch?
>  I'll re-post with proper subject if it works.
> >>>
> >>> Still happening with the same stacks.
> >>
> >> Then there may be another one, I'm afraid.
> >>
> >> I'm using the testapp you shared in the first email, with that debug line
> >> enabled and added a new one:
> >> +   pr_debug("%p %d\n", asoc, asoc ? asoc->state : 0);
> >> debug_post_sfx();
> >> (should have used %x, but ok)
> >>
> >> Also enabled slub_debug=PUZ, and I get:
> >>
> >> without the patch:
> >> [   87.873640] sctp: 8800b71533d8 1
> >> [   87.873647] sctp: sctp_do_sm[post-sfx]: error:0,
> >> asoc:8800b71533d8[STATE_CLOSED]
> >> [   87.873739] sctp: 8800b71533d8 1
> >> [   87.873742] sctp: sctp_do_sm[post-sfx]: error:0,
> >> asoc:8800b71533d8[STATE_CLOSED]
> >> [   87.875149] sctp: 8800b71533d8 1802201963
> >> [   87.875238] sctp: sctp_do_sm[post-sfx]: error:0,
> >> asoc:8800b71533d8[STATE_CLOSED]
> >>
> >> 1802201963 = 0x6b6b6b6b, poison
> >>
> >> with the patch:
> >> [   81.071265] sctp: 880137571148 1
> >> [   81.071273] sctp: sctp_do_sm[post-sfx]: error:0,
> >> asoc:880137571148[STATE_CLOSED]
> >> [   81.071372] sctp: 880137571148 1
> >> [   81.071375] sctp: sctp_do_sm[post-sfx]: error:0,
> >> asoc:880137571148[STATE_CLOSED]
> >> [   81.072423] sctp:   (null) 0
> >> [   81.072427] sctp: sctp_do_sm[post-sfx]: error:0, asoc:
> >> (null)[STATE_CLOSED]
> >>
> >> This one, at least, is gone with this patch.
> >>
> >>   Marcelo
> >>
> >
> > Hi Marcelo
> >
> > I think you also need to catch the SCTP_DISPOSITION_ABORT and update
> > the pointer.  There are some issues there though as some functions report
> > that code without actually destroying the association.  This happens when
> > the ABORT chunk may be dropped.
> >
> > I think this might be why we still see the issue.
> 
> 
> Marcelo,
> 
> Is this info enough for you to cook another fix?

Hi, I think so. I was really wondering how you could trigger that issue
without the timestamp fix and Vlad's comment does shed some light on it.

I'll do more tests later today, but what did you have connecting to the
listening socket? Somehow you made that accept() call to return..

  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Vlad Yasevich]

On 12/07/2015 01:52 PM, Marcelo Ricardo Leitner wrote:
> On Mon, Dec 07, 2015 at 02:20:47PM +0100, Dmitry Vyukov wrote:
>> On Mon, Dec 7, 2015 at 2:15 PM, Marcelo Ricardo Leitner
>>  wrote:
>>> On Mon, Dec 07, 2015 at 12:26:09PM +0100, Dmitry Vyukov wrote:
 On Sat, Dec 5, 2015 at 5:39 PM, Vlad Yasevich  wrote:
> ...
> Hi Marcelo
>
> I think you also need to catch the SCTP_DISPOSITION_ABORT and update
> the pointer.  There are some issues there though as some functions report
> that code without actually destroying the association.  This happens when
> the ABORT chunk may be dropped.
>
> I think this might be why we still see the issue.


 Marcelo,

 Is this info enough for you to cook another fix?
>>>
>>> Hi, I think so. I was really wondering how you could trigger that issue
>>> without the timestamp fix and Vlad's comment does shed some light on it.
>>>
>>> I'll do more tests later today, but what did you have connecting to the
>>> listening socket? Somehow you made that accept() call to return..
>>
>> Local connect in another thread I guess.
> 
> Vlad, I reviewed the places on which it returns SCTP_DISPOSITION_ABORT,
> and if I didn't miss something in there all of them either issue
> SCTP_CMD_ASSOC_FAILED or SCTP_CMD_INIT_FAILED before returning it, thus
> delaying DELETE_TCB and with that the asoc free.

They delay it from the perspective of the command interpreter since the command
to delete the TCB happens a little later, but status code  is checked after all
commands are processed and command processing doesn't change it.  So the 
'status'
code would still be SCTP_DISPOSITION_ABORT after DELETE_TCB command was 
processed.
So, I think we may still have an use-after-free issue here.

> There is one place,
> though, that may not do it that way, it's sctp_sf_abort_violation(), but
> then that code only runs if asoc is already NULL by then.

I don't believe so.  The violation state function can run with a non-NULL 
association
if we are encountering protocol violations after the association is established.

-vlad

> 
> Dmitry, still no luck here, cannot reproduce another hit.
> I'm using sctp_test and a custom test of mine, both on localhost so I
> would catch it in server or client side, nothing..
> 
> I need more info. Please enable the pr_debug() on debug_post_sfn() macro
> and see which status is being reported when you trigger the issue.
> And/or share a traffic capture so we can see what's going on with the
> association.
> 
>   Marcelo
> 

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Mon, Dec 07, 2015 at 02:33:52PM -0500, Vlad Yasevich wrote:
> On 12/07/2015 01:52 PM, Marcelo Ricardo Leitner wrote:
> > On Mon, Dec 07, 2015 at 02:20:47PM +0100, Dmitry Vyukov wrote:
> >> On Mon, Dec 7, 2015 at 2:15 PM, Marcelo Ricardo Leitner
> >>  wrote:
> >>> On Mon, Dec 07, 2015 at 12:26:09PM +0100, Dmitry Vyukov wrote:
>  On Sat, Dec 5, 2015 at 5:39 PM, Vlad Yasevich  
>  wrote:
> > ...
> > Hi Marcelo
> >
> > I think you also need to catch the SCTP_DISPOSITION_ABORT and update
> > the pointer.  There are some issues there though as some functions 
> > report
> > that code without actually destroying the association.  This happens 
> > when
> > the ABORT chunk may be dropped.
> >
> > I think this might be why we still see the issue.
> 
> 
>  Marcelo,
> 
>  Is this info enough for you to cook another fix?
> >>>
> >>> Hi, I think so. I was really wondering how you could trigger that issue
> >>> without the timestamp fix and Vlad's comment does shed some light on it.
> >>>
> >>> I'll do more tests later today, but what did you have connecting to the
> >>> listening socket? Somehow you made that accept() call to return..
> >>
> >> Local connect in another thread I guess.
> > 
> > Vlad, I reviewed the places on which it returns SCTP_DISPOSITION_ABORT,
> > and if I didn't miss something in there all of them either issue
> > SCTP_CMD_ASSOC_FAILED or SCTP_CMD_INIT_FAILED before returning it, thus
> > delaying DELETE_TCB and with that the asoc free.
> 
> They delay it from the perspective of the command interpreter since the 
> command
> to delete the TCB happens a little later, but status code  is checked after 
> all
> commands are processed and command processing doesn't change it.  So the 
> 'status'
> code would still be SCTP_DISPOSITION_ABORT after DELETE_TCB command was 
> processed.
> So, I think we may still have an use-after-free issue here.

Gotcha! That's pretty much it then. From that point of view now, there
shouldn't be a case that it returns _ABORT without freeing the asoc in
the same loop. (more below)

> > There is one place,
> > though, that may not do it that way, it's sctp_sf_abort_violation(), but
> > then that code only runs if asoc is already NULL by then.
> 
> I don't believe so.  The violation state function can run with a non-NULL 
> association
> if we are encountering protocol violations after the association is 
> established.

Yup, that's correct. I just tried to reference one case on which it
would return _ABORT without issuing any of those _FAILEDs before doing
so (meaning the association could still be valid) but that in that case,
the asoc was already NULL.

Dmitry, please give this one a run, as I still cannot reproduce your use
case..

---8<---

commit b63ad8dc45257dd6c536ac0227fcc623efd9328b
Author: Marcelo Ricardo Leitner 
Date:   Fri Dec 4 15:30:23 2015 -0200

sctp: fix use-after-free in pr_debug statement

Dmitry Vyukov reported a use-after-free in the code expanded by the
macro debug_post_sfx, which is caused by the use of the asoc pointer
after it was freed within sctp_side_effect() scope.

This patch fixes it by allowing sctp_side_effect to clear that asoc
pointer when the TCB is freed.

As Vlad explained, we also have to cover the SCTP_DISPOSITION_ABORT case
because it will trigger DELETE_TCB too on that same loop.

The macro is already prepared to handle such NULL pointer.

Reported-by: Dmitry Vyukov 

diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
index 6098d4c42fa9..be23d5c2074f 100644
--- a/net/sctp/sm_sideeffect.c
+++ b/net/sctp/sm_sideeffect.c
@@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
 static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
 sctp_state_t state,
 struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
 void *event_arg,
 sctp_disposition_t status,
 sctp_cmd_seq_t *commands,
@@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
debug_post_sfn();
 
error = sctp_side_effects(event_type, subtype, state,
- ep, asoc, event_arg, status,
+ ep, , event_arg, status,
  , gfp);
debug_post_sfx();
 
@@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
 static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
 sctp_state_t state,
 struct 

Re: use-after-free in sctp_do_sm

[Vlad Yasevich]

On 12/07/2015 02:50 PM, Marcelo Ricardo Leitner wrote:
> On Mon, Dec 07, 2015 at 02:33:52PM -0500, Vlad Yasevich wrote:
>> On 12/07/2015 01:52 PM, Marcelo Ricardo Leitner wrote:
>>> On Mon, Dec 07, 2015 at 02:20:47PM +0100, Dmitry Vyukov wrote:
 On Mon, Dec 7, 2015 at 2:15 PM, Marcelo Ricardo Leitner
  wrote:
> On Mon, Dec 07, 2015 at 12:26:09PM +0100, Dmitry Vyukov wrote:
>> On Sat, Dec 5, 2015 at 5:39 PM, Vlad Yasevich  
>> wrote:
>>> ...
>>> Hi Marcelo
>>>
>>> I think you also need to catch the SCTP_DISPOSITION_ABORT and update
>>> the pointer.  There are some issues there though as some functions 
>>> report
>>> that code without actually destroying the association.  This happens 
>>> when
>>> the ABORT chunk may be dropped.
>>>
>>> I think this might be why we still see the issue.
>>
>>
>> Marcelo,
>>
>> Is this info enough for you to cook another fix?
>
> Hi, I think so. I was really wondering how you could trigger that issue
> without the timestamp fix and Vlad's comment does shed some light on it.
>
> I'll do more tests later today, but what did you have connecting to the
> listening socket? Somehow you made that accept() call to return..

 Local connect in another thread I guess.
>>>
>>> Vlad, I reviewed the places on which it returns SCTP_DISPOSITION_ABORT,
>>> and if I didn't miss something in there all of them either issue
>>> SCTP_CMD_ASSOC_FAILED or SCTP_CMD_INIT_FAILED before returning it, thus
>>> delaying DELETE_TCB and with that the asoc free.
>>
>> They delay it from the perspective of the command interpreter since the 
>> command
>> to delete the TCB happens a little later, but status code  is checked after 
>> all
>> commands are processed and command processing doesn't change it.  So the 
>> 'status'
>> code would still be SCTP_DISPOSITION_ABORT after DELETE_TCB command was 
>> processed.
>> So, I think we may still have an use-after-free issue here.
> 
> Gotcha! That's pretty much it then. From that point of view now, there
> shouldn't be a case that it returns _ABORT without freeing the asoc in
> the same loop. (more below)
> 
>>> There is one place,
>>> though, that may not do it that way, it's sctp_sf_abort_violation(), but
>>> then that code only runs if asoc is already NULL by then.
>>
>> I don't believe so.  The violation state function can run with a non-NULL 
>> association
>> if we are encountering protocol violations after the association is 
>> established.
> 
> Yup, that's correct. I just tried to reference one case on which it
> would return _ABORT without issuing any of those _FAILEDs before doing
> so (meaning the association could still be valid) but that in that case,
> the asoc was already NULL.

I think it is possible to hit the 'discard:' tag in that function while still
having a valid association.  That happens when ABORT chunk is required to be
authenticated.  This that case, instead of generating an ABORT and terminating 
the
current association, we just drop the packet, but still report an _ABORT 
disposition code.

This probably need to change if we are going to catch the _ABORT disposition and
clear the asoc pointer.

-vlad

> 
> Dmitry, please give this one a run, as I still cannot reproduce your use
> case..
> 
> ---8<---
> 
> commit b63ad8dc45257dd6c536ac0227fcc623efd9328b
> Author: Marcelo Ricardo Leitner 
> Date:   Fri Dec 4 15:30:23 2015 -0200
> 
> sctp: fix use-after-free in pr_debug statement
> 
> Dmitry Vyukov reported a use-after-free in the code expanded by the
> macro debug_post_sfx, which is caused by the use of the asoc pointer
> after it was freed within sctp_side_effect() scope.
> 
> This patch fixes it by allowing sctp_side_effect to clear that asoc
> pointer when the TCB is freed.
> 
> As Vlad explained, we also have to cover the SCTP_DISPOSITION_ABORT case
> because it will trigger DELETE_TCB too on that same loop.
> 
> The macro is already prepared to handle such NULL pointer.
> 
> Reported-by: Dmitry Vyukov 
> 
> diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
> index 6098d4c42fa9..be23d5c2074f 100644
> --- a/net/sctp/sm_sideeffect.c
> +++ b/net/sctp/sm_sideeffect.c
> @@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
>  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
>sctp_state_t state,
>struct sctp_endpoint *ep,
> -  struct sctp_association *asoc,
> +  struct sctp_association **asoc,
>void *event_arg,
>sctp_disposition_t status,
>sctp_cmd_seq_t *commands,
> @@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, 

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

Em 07-12-2015 18:37, Vlad Yasevich escreveu:

On 12/07/2015 02:50 PM, Marcelo Ricardo Leitner wrote:

On Mon, Dec 07, 2015 at 02:33:52PM -0500, Vlad Yasevich wrote:

On 12/07/2015 01:52 PM, Marcelo Ricardo Leitner wrote:

Vlad, I reviewed the places on which it returns SCTP_DISPOSITION_ABORT,
and if I didn't miss something in there all of them either issue
SCTP_CMD_ASSOC_FAILED or SCTP_CMD_INIT_FAILED before returning it, thus
delaying DELETE_TCB and with that the asoc free.


They delay it from the perspective of the command interpreter since the command
to delete the TCB happens a little later, but status code  is checked after all
commands are processed and command processing doesn't change it.  So the 
'status'
code would still be SCTP_DISPOSITION_ABORT after DELETE_TCB command was 
processed.
So, I think we may still have an use-after-free issue here.


Gotcha! That's pretty much it then. From that point of view now, there
shouldn't be a case that it returns _ABORT without freeing the asoc in
the same loop. (more below)


There is one place,
though, that may not do it that way, it's sctp_sf_abort_violation(), but
then that code only runs if asoc is already NULL by then.


I don't believe so.  The violation state function can run with a non-NULL 
association
if we are encountering protocol violations after the association is established.


Yup, that's correct. I just tried to reference one case on which it
would return _ABORT without issuing any of those _FAILEDs before doing
so (meaning the association could still be valid) but that in that case,
the asoc was already NULL.


I think it is possible to hit the 'discard:' tag in that function while still
having a valid association.  That happens when ABORT chunk is required to be
authenticated.  This that case, instead of generating an ABORT and terminating 
the
current association, we just drop the packet, but still report an _ABORT 
disposition code.

This probably need to change if we are going to catch the _ABORT disposition and
clear the asoc pointer.


Oups. Nice one. I'll switch it to SCTP_DISPOSITION_DISCARD if it hits 
that if() then. Thanks Vlad.


  Marcelo
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Mon, Dec 07, 2015 at 02:20:47PM +0100, Dmitry Vyukov wrote:
> On Mon, Dec 7, 2015 at 2:15 PM, Marcelo Ricardo Leitner
>  wrote:
> > On Mon, Dec 07, 2015 at 12:26:09PM +0100, Dmitry Vyukov wrote:
> >> On Sat, Dec 5, 2015 at 5:39 PM, Vlad Yasevich  wrote:
...
> >> > Hi Marcelo
> >> >
> >> > I think you also need to catch the SCTP_DISPOSITION_ABORT and update
> >> > the pointer.  There are some issues there though as some functions report
> >> > that code without actually destroying the association.  This happens when
> >> > the ABORT chunk may be dropped.
> >> >
> >> > I think this might be why we still see the issue.
> >>
> >>
> >> Marcelo,
> >>
> >> Is this info enough for you to cook another fix?
> >
> > Hi, I think so. I was really wondering how you could trigger that issue
> > without the timestamp fix and Vlad's comment does shed some light on it.
> >
> > I'll do more tests later today, but what did you have connecting to the
> > listening socket? Somehow you made that accept() call to return..
> 
> Local connect in another thread I guess.

Vlad, I reviewed the places on which it returns SCTP_DISPOSITION_ABORT,
and if I didn't miss something in there all of them either issue
SCTP_CMD_ASSOC_FAILED or SCTP_CMD_INIT_FAILED before returning it, thus
delaying DELETE_TCB and with that the asoc free. There is one place,
though, that may not do it that way, it's sctp_sf_abort_violation(), but
then that code only runs if asoc is already NULL by then.

Dmitry, still no luck here, cannot reproduce another hit.
I'm using sctp_test and a custom test of mine, both on localhost so I
would catch it in server or client side, nothing..

I need more info. Please enable the pr_debug() on debug_post_sfn() macro
and see which status is being reported when you trigger the issue.
And/or share a traffic capture so we can see what's going on with the
association.

  Marcelo
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Vlad Yasevich]

On 12/04/2015 04:34 PM, Marcelo Ricardo Leitner wrote:
> On Fri, Dec 04, 2015 at 09:25:35PM +0100, Dmitry Vyukov wrote:
>> On Fri, Dec 4, 2015 at 6:48 PM, Marcelo Ricardo Leitner
>>  wrote:
>>> Hi Dmitry,
>>>
>>> Can you please test this patch?
>>> I'll re-post with proper subject if it works.
>>
>> Still happening with the same stacks.
> 
> Then there may be another one, I'm afraid.
> 
> I'm using the testapp you shared in the first email, with that debug line
> enabled and added a new one:
> +   pr_debug("%p %d\n", asoc, asoc ? asoc->state : 0);
> debug_post_sfx();
> (should have used %x, but ok)
> 
> Also enabled slub_debug=PUZ, and I get:
> 
> without the patch:
> [   87.873640] sctp: 8800b71533d8 1
> [   87.873647] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:8800b71533d8[STATE_CLOSED]
> [   87.873739] sctp: 8800b71533d8 1
> [   87.873742] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:8800b71533d8[STATE_CLOSED]
> [   87.875149] sctp: 8800b71533d8 1802201963
> [   87.875238] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:8800b71533d8[STATE_CLOSED]
> 
> 1802201963 = 0x6b6b6b6b, poison
> 
> with the patch:
> [   81.071265] sctp: 880137571148 1
> [   81.071273] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:880137571148[STATE_CLOSED]
> [   81.071372] sctp: 880137571148 1
> [   81.071375] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:880137571148[STATE_CLOSED]
> [   81.072423] sctp:   (null) 0
> [   81.072427] sctp: sctp_do_sm[post-sfx]: error:0, asoc:
> (null)[STATE_CLOSED]
> 
> This one, at least, is gone with this patch.
> 
>   Marcelo
> 

Hi Marcelo

I think you also need to catch the SCTP_DISPOSITION_ABORT and update
the pointer.  There are some issues there though as some functions report
that code without actually destroying the association.  This happens when
the ABORT chunk may be dropped.

I think this might be why we still see the issue.

-vlad
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Thu, Dec 3, 2015 at 6:02 PM, Eric Dumazet  wrote:
> On Thu, Dec 3, 2015 at 7:55 AM, Dmitry Vyukov  wrote:
>> On Thu, Dec 3, 2015 at 3:48 PM, Eric Dumazet  wrote:

 No, I don't. But pr_debug always computes its arguments. See no_printk
 in printk.h. So this use-after-free happens for all users.
>>>
>>> Hmm.
>>>
>>> pr_debug() should be a nop unless either DEBUG or CONFIG_DYNAMIC_DEBUG are 
>>> set
>>>
>>> On our production kernels, pr_debug() is a nop.
>>>
>>> Can you double check ? Thanks !
>>
>>
>> Why should it be nop? no_printk thing in printk.h pretty much
>> explicitly makes it not a nop...
>>
>> Double-checked: debug_post_sfx leads to some generated code:
>>
>> debug_post_sfx();
>> 8229f256:   48 8b 85 58 fe ff ffmov-0x1a8(%rbp),%rax
>> 8229f25d:   48 85 c0test   %rax,%rax
>> 8229f260:   74 24   je
>> 8229f286 
>> 8229f262:   8b b0 a8 00 00 00   mov0xa8(%rax),%esi
>> 8229f268:   48 8b 85 60 fe ff ffmov-0x1a0(%rbp),%rax
>> 8229f26f:   44 89 85 74 fe ff ffmov%r8d,-0x18c(%rbp)
>> 8229f276:   48 8b 78 20 mov0x20(%rax),%rdi
>> 8229f27a:   e8 71 28 01 00  callq
>> 822b1af0 
>> 8229f27f:   44 8b 85 74 fe ff ffmov-0x18c(%rbp),%r8d
>>
>> return error;
>> }
>> 8229f286:   48 81 c4 a0 01 00 00add$0x1a0,%rsp
>> 8229f28d:   44 89 c0mov%r8d,%eax
>> 8229f290:   5b  pop%rbx
>> 8229f291:   41 5c   pop%r12
>> 8229f293:   41 5d   pop%r13
>> 8229f295:   41 5e   pop%r14
>> 8229f297:   41 5f   pop%r15
>> 8229f299:   5d  pop%rbp
>> 8229f29a:   c3  retq
>
> This is a serious concern, because we let in the past lot of patches
> converting traditional
>
> #ifdef DEBUG
> # define some_hand_coded_ugly_debug()  printk( _
> #else
> # define some_hand_coded_ugly_debug()
> #endif
>
> On the premise pr_debug() would be a nop.
>
> It seems it is not always the case. This is a very serious problem.
>
> We probably have hundred of potential bugs, because few people
> actually make sure all debugging stuff is correct,
> like comments can be wrong because they are not updated properly as time 
> flies.


FWIW I enabled CONFIG_DYNAMIC_DEBUG on my fuzzer. Not that it gives
any particular guarantees, but still can catch some of these.



> It is definitely a nop for many cases.
>
> +void eric_test_pr_debug(struct sock *sk)
> +{
> +   if (atomic_read(>sk_omem_alloc))
> +   pr_debug("%s: optmem leakage for sock %p\n",
> +__func__, sk);
> +}
>
> ->
>
> 4740 :
> 4740: e8 00 00 00 00   callq  4745 
> 4741: R_X86_64_PC32 __fentry__-0x4
> 4745: 55   push   %rbp
> 4746: 8b 87 24 01 00 00 mov0x124(%rdi),%eax //
> atomic_read()  but nothing follows
> 474c: 48 89 e5 mov%rsp,%rbp
> 474f: 5d   pop%rbp
> 4750: c3   retq
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Thu, Dec 3, 2015 at 9:51 PM, Joe Perches  wrote:
> (adding lkml as this is likely better discussed there)
>
> On Thu, 2015-12-03 at 15:42 -0500, Jason Baron wrote:
>> On 12/03/2015 03:24 PM, Joe Perches wrote:
>> > On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
>> > > On 12/03/2015 03:03 PM, Joe Perches wrote:
>> > > > On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
>> > > > > On 12/03/2015 01:52 PM, Aaron Conole wrote:
>> > > > > > I think that as a minimum, the following patch should be evaluted,
>> > > > > > but am unsure to whom I should submit it (after I test):
>> > > > []
>> > > > > Agreed - the intention here is certainly to have no side effects. It
>> > > > > looks like 'no_printk()' is used in quite a few other places that 
>> > > > > would
>> > > > > benefit from this change. So we probably want a generic
>> > > > > 'really_no_printk()' macro.
>> > > >
>> > > > https://lkml.org/lkml/2012/6/17/231
>> > >
>> > > I don't see this in the tree.
>> >
>> > It never got applied.
>> >
>> > > Also maybe we should just convert
>> > > no_printk() to do what your 'eliminated_printk()'.
>> >
>> > Some of them at least.
>> >
>> > > So we can convert all users with this change?
>> >
>> > I don't think so, I think there are some
>> > function evaluation/side effects that are
>> > required.  I believe some do hardware I/O.
>> >
>> > It'd be good to at least isolate them.
>> >
>> > I'm not sure how to find them via some
>> > automated tool/mechanism though.
>> >
>> > I asked Julia Lawall about it once in this
>> > thread:  https://lkml.org/lkml/2014/12/3/696
>> >
>>
>> Seems rather fragile to have side effects that we rely
>> upon hidden in a printk().
>
> Yup.
>
>> Just convert them and see what breaks :)
>
> I appreciate your optimism.  It's very 1995.
> Try it and see what happens.


Whatever is the resolution for pr_debug, we still need to fix this
particular use-after-free. It affects stability of debug builds, gives
invalid debug output, prevents us from finding more bugs in SCTP. And
maybe somebody uses CONFIG_DYNAMIC_DEBUG in production.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Vlad Yasevich]

On 12/04/2015 07:55 AM, Marcelo Ricardo Leitner wrote:
> On Fri, Dec 04, 2015 at 11:40:02AM +0100, Dmitry Vyukov wrote:
>> On Thu, Dec 3, 2015 at 9:51 PM, Joe Perches  wrote:
>>> (adding lkml as this is likely better discussed there)
>>>
>>> On Thu, 2015-12-03 at 15:42 -0500, Jason Baron wrote:
 On 12/03/2015 03:24 PM, Joe Perches wrote:
> On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
>> On 12/03/2015 03:03 PM, Joe Perches wrote:
>>> On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
 On 12/03/2015 01:52 PM, Aaron Conole wrote:
> I think that as a minimum, the following patch should be evaluted,
> but am unsure to whom I should submit it (after I test):
>>> []
 Agreed - the intention here is certainly to have no side effects. It
 looks like 'no_printk()' is used in quite a few other places that would
 benefit from this change. So we probably want a generic
 'really_no_printk()' macro.
>>>
>>> https://lkml.org/lkml/2012/6/17/231
>>
>> I don't see this in the tree.
>
> It never got applied.
>
>> Also maybe we should just convert
>> no_printk() to do what your 'eliminated_printk()'.
>
> Some of them at least.
>
>> So we can convert all users with this change?
>
> I don't think so, I think there are some
> function evaluation/side effects that are
> required.  I believe some do hardware I/O.
>
> It'd be good to at least isolate them.
>
> I'm not sure how to find them via some
> automated tool/mechanism though.
>
> I asked Julia Lawall about it once in this
> thread:  https://lkml.org/lkml/2014/12/3/696
>

 Seems rather fragile to have side effects that we rely
 upon hidden in a printk().
>>>
>>> Yup.
>>>
 Just convert them and see what breaks :)
>>>
>>> I appreciate your optimism.  It's very 1995.
>>> Try it and see what happens.
>>
>>
>> Whatever is the resolution for pr_debug, we still need to fix this
>> particular use-after-free. It affects stability of debug builds, gives
>> invalid debug output, prevents us from finding more bugs in SCTP. And
>> maybe somebody uses CONFIG_DYNAMIC_DEBUG in production.
> 
> Agreed. I'm already working on a fix for this particular use-after-free.
> 
> Another interesting thing about this is that sctp_do_sm() is called for
> nearly every movement that happens on a sctp socket. Said that, that
> always-running IDR search hidden on that debug statement do have some
> nasty performance impact, specially because it's serialized on a
> spinlock.

YUCK!  I didn't really pay much attention to those debug macros before, but
debug_post_sfx() is truly awful.

This wasn't such a bad thing where these macros depended on CONFIG_SCTP_DEBUG,
but now that they are always built, we need fix them.

-vlad



> This wouldn't be happening if it was fully ellided and would
> be ok if that pr_debug() was really being printed, but not as it is.
> Kudos to this report that I could notice this. I'm trying to fix this on
> SCTP-side as well.
> 
>   Marcelo
> 

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Aaron Conole]

Vlad Yasevich  writes:
> On 12/04/2015 07:55 AM, Marcelo Ricardo Leitner wrote:
>> On Fri, Dec 04, 2015 at 11:40:02AM +0100, Dmitry Vyukov wrote:
>>> On Thu, Dec 3, 2015 at 9:51 PM, Joe Perches  wrote:
 (adding lkml as this is likely better discussed there)

 On Thu, 2015-12-03 at 15:42 -0500, Jason Baron wrote:
> On 12/03/2015 03:24 PM, Joe Perches wrote:
>> On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
>>> On 12/03/2015 03:03 PM, Joe Perches wrote:
 On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
> On 12/03/2015 01:52 PM, Aaron Conole wrote:
>> I think that as a minimum, the following patch should be evaluted,
>> but am unsure to whom I should submit it (after I test):
 []
> Agreed - the intention here is certainly to have no side effects. It
> looks like 'no_printk()' is used in quite a few other places that 
> would
> benefit from this change. So we probably want a generic
> 'really_no_printk()' macro.

 https://lkml.org/lkml/2012/6/17/231
>>>
>>> I don't see this in the tree.
>>
>> It never got applied.
>>
>>> Also maybe we should just convert
>>> no_printk() to do what your 'eliminated_printk()'.
>>
>> Some of them at least.
>>
>>> So we can convert all users with this change?
>>
>> I don't think so, I think there are some
>> function evaluation/side effects that are
>> required.  I believe some do hardware I/O.
>>
>> It'd be good to at least isolate them.
>>
>> I'm not sure how to find them via some
>> automated tool/mechanism though.
>>
>> I asked Julia Lawall about it once in this
>> thread:  https://lkml.org/lkml/2014/12/3/696
>>
>
> Seems rather fragile to have side effects that we rely
> upon hidden in a printk().

 Yup.

> Just convert them and see what breaks :)

 I appreciate your optimism.  It's very 1995.
 Try it and see what happens.
>>>
>>>
>>> Whatever is the resolution for pr_debug, we still need to fix this
>>> particular use-after-free. It affects stability of debug builds, gives
>>> invalid debug output, prevents us from finding more bugs in SCTP. And
>>> maybe somebody uses CONFIG_DYNAMIC_DEBUG in production.
>> 
>> Agreed. I'm already working on a fix for this particular use-after-free.
>> 
>> Another interesting thing about this is that sctp_do_sm() is called for
>> nearly every movement that happens on a sctp socket. Said that, that
>> always-running IDR search hidden on that debug statement do have some
>> nasty performance impact, specially because it's serialized on a
>> spinlock.
>
> YUCK!  I didn't really pay much attention to those debug macros before, but
> debug_post_sfx() is truly awful.
>
> This wasn't such a bad thing where these macros depended on CONFIG_SCTP_DEBUG,
> but now that they are always built, we need fix them.

I've proposed a patch to linux-kernel to fix them, but I don't think
it's really as bad as folks imagine. Ubuntu, RHEL, and Fedora all use
DYNAMIC_DEBUG configuration option, which means that the code is getting
emitted anyway (correctly, I'll add) and is shunted out by a dynamic
debug flag. So for the average user, it's not even really a blip.

That does mean there's a cool side-effect of the entire print-macro setup
which implies we execute less code when running with DYNAMIC_DEBUG=y in
the "normal" case. "Turn on the dynamic debugging config and watch
everything get better" isn't the worst mantra, is it? :)

> -vlad
>
>
>
>> This wouldn't be happening if it was fully ellided and would
>> be ok if that pr_debug() was really being printed, but not as it is.
>> Kudos to this report that I could notice this. I'm trying to fix this on
>> SCTP-side as well.
>> 
>>   Marcelo
>> 
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Fri, Dec 04, 2015 at 11:40:02AM +0100, Dmitry Vyukov wrote:
> On Thu, Dec 3, 2015 at 9:51 PM, Joe Perches  wrote:
> > (adding lkml as this is likely better discussed there)
> >
> > On Thu, 2015-12-03 at 15:42 -0500, Jason Baron wrote:
> >> On 12/03/2015 03:24 PM, Joe Perches wrote:
> >> > On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
> >> > > On 12/03/2015 03:03 PM, Joe Perches wrote:
> >> > > > On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
> >> > > > > On 12/03/2015 01:52 PM, Aaron Conole wrote:
> >> > > > > > I think that as a minimum, the following patch should be 
> >> > > > > > evaluted,
> >> > > > > > but am unsure to whom I should submit it (after I test):
> >> > > > []
> >> > > > > Agreed - the intention here is certainly to have no side effects. 
> >> > > > > It
> >> > > > > looks like 'no_printk()' is used in quite a few other places that 
> >> > > > > would
> >> > > > > benefit from this change. So we probably want a generic
> >> > > > > 'really_no_printk()' macro.
> >> > > >
> >> > > > https://lkml.org/lkml/2012/6/17/231
> >> > >
> >> > > I don't see this in the tree.
> >> >
> >> > It never got applied.
> >> >
> >> > > Also maybe we should just convert
> >> > > no_printk() to do what your 'eliminated_printk()'.
> >> >
> >> > Some of them at least.
> >> >
> >> > > So we can convert all users with this change?
> >> >
> >> > I don't think so, I think there are some
> >> > function evaluation/side effects that are
> >> > required.  I believe some do hardware I/O.
> >> >
> >> > It'd be good to at least isolate them.
> >> >
> >> > I'm not sure how to find them via some
> >> > automated tool/mechanism though.
> >> >
> >> > I asked Julia Lawall about it once in this
> >> > thread:  https://lkml.org/lkml/2014/12/3/696
> >> >
> >>
> >> Seems rather fragile to have side effects that we rely
> >> upon hidden in a printk().
> >
> > Yup.
> >
> >> Just convert them and see what breaks :)
> >
> > I appreciate your optimism.  It's very 1995.
> > Try it and see what happens.
> 
> 
> Whatever is the resolution for pr_debug, we still need to fix this
> particular use-after-free. It affects stability of debug builds, gives
> invalid debug output, prevents us from finding more bugs in SCTP. And
> maybe somebody uses CONFIG_DYNAMIC_DEBUG in production.

Agreed. I'm already working on a fix for this particular use-after-free.

Another interesting thing about this is that sctp_do_sm() is called for
nearly every movement that happens on a sctp socket. Said that, that
always-running IDR search hidden on that debug statement do have some
nasty performance impact, specially because it's serialized on a
spinlock. This wouldn't be happening if it was fully ellided and would
be ok if that pr_debug() was really being printed, but not as it is.
Kudos to this report that I could notice this. I'm trying to fix this on
SCTP-side as well.

  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

Hi Dmitry,

Can you please test this patch?
I'll re-post with proper subject if it works.

Thanks.

---8<---

Dmitry Vyukov reported a use-after-free in the code expanded by the
macro debug_post_sfx, which is caused by the use of the asoc pointer
after it was freed within sctp_side_effect() scope.

This patch fixes it by allowing sctp_side_effect to clear that asoc
pointer when the TCB is freed.

The macro is already prepared to handle such NULL pointer.

Reported-by: Dmitry Vyukov 
Signed-off-by: Marcelo Ricardo Leitner 
---
 net/sctp/sm_sideeffect.c | 9 +
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
index 
6098d4c42fa91287d3cde36ac05d860f76d4fe32..05594dcd93e0d649cace5215d225bef2713f9310
 100644
--- a/net/sctp/sm_sideeffect.c
+++ b/net/sctp/sm_sideeffect.c
@@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
 static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
 sctp_state_t state,
 struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
 void *event_arg,
 sctp_disposition_t status,
 sctp_cmd_seq_t *commands,
@@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
debug_post_sfn();
 
error = sctp_side_effects(event_type, subtype, state,
- ep, asoc, event_arg, status,
+ ep, , event_arg, status,
  , gfp);
debug_post_sfx();
 
@@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t event_type, 
sctp_subtype_t subtype,
 static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
 sctp_state_t state,
 struct sctp_endpoint *ep,
-struct sctp_association *asoc,
+struct sctp_association **asoc,
 void *event_arg,
 sctp_disposition_t status,
 sctp_cmd_seq_t *commands,
@@ -1151,7 +1151,7 @@ static int sctp_side_effects(sctp_event_t event_type, 
sctp_subtype_t subtype,
 * disposition SCTP_DISPOSITION_CONSUME.
 */
if (0 != (error = sctp_cmd_interpreter(event_type, subtype, state,
-  ep, asoc,
+  ep, *asoc,
   event_arg, status,
   commands, gfp)))
goto bail;
@@ -1175,6 +1175,7 @@ static int sctp_side_effects(sctp_event_t event_type, 
sctp_subtype_t subtype,
 
case SCTP_DISPOSITION_DELETE_TCB:
/* This should now be a command. */
+   *asoc = NULL;
break;
 
case SCTP_DISPOSITION_CONSUME:
-- 
2.5.0

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Jason Baron]

On 12/04/2015 12:03 PM, Joe Perches wrote:
> On Fri, 2015-12-04 at 11:47 -0500, Jason Baron wrote:
>> When DYNAMIC_DEBUG is enabled we have this wrapper from
>> include/linux/dynamic_debug.h:
>>
>> if (unlikely(descriptor.flags & _DPRINTK_FLAGS_PRINT))
>>  
>>
>> So the compiler is not emitting the side-effects in this
>> case.
> 
> Huh?  Do I misunderstand what you are writing?

Yes, I wasn't terribly clear - I was trying to say that the
'side-effects', in this case the debug code and use-after-free, are
hidden behind the branch. They aren't invoked unless we enable the debug
statement.

Thanks,

-Jason

> 
> You are testing a variable that is not generally set
> so the call is not being performed in the general case,
> but the compiler can not elide the code.
> 
> If the variable was enabled via the control file, the
> __dynamic_pr_debug would be performed with the
> use-after-free.
> 
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Jason Baron]

On 12/04/2015 11:12 AM, Dmitry Vyukov wrote:
> On Thu, Dec 3, 2015 at 9:51 PM, Joe Perches  wrote:
>> (adding lkml as this is likely better discussed there)
>>
>> On Thu, 2015-12-03 at 15:42 -0500, Jason Baron wrote:
>>> On 12/03/2015 03:24 PM, Joe Perches wrote:
 On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
> On 12/03/2015 03:03 PM, Joe Perches wrote:
>> On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
>>> On 12/03/2015 01:52 PM, Aaron Conole wrote:
 I think that as a minimum, the following patch should be evaluted,
 but am unsure to whom I should submit it (after I test):
>> []
>>> Agreed - the intention here is certainly to have no side effects. It
>>> looks like 'no_printk()' is used in quite a few other places that would
>>> benefit from this change. So we probably want a generic
>>> 'really_no_printk()' macro.
>>
>> https://lkml.org/lkml/2012/6/17/231
>
> I don't see this in the tree.

 It never got applied.

> Also maybe we should just convert
> no_printk() to do what your 'eliminated_printk()'.

 Some of them at least.

> So we can convert all users with this change?

 I don't think so, I think there are some
 function evaluation/side effects that are
 required.  I believe some do hardware I/O.

 It'd be good to at least isolate them.

 I'm not sure how to find them via some
 automated tool/mechanism though.

 I asked Julia Lawall about it once in this
 thread:  https://lkml.org/lkml/2014/12/3/696

>>>
>>> Seems rather fragile to have side effects that we rely
>>> upon hidden in a printk().
>>
>> Yup.
>>
>>> Just convert them and see what breaks :)
>>
>> I appreciate your optimism.  It's very 1995.
>> Try it and see what happens.
> 
> 
> But Aaron says that DYNAMIC_DEBUG is enabled in most major
> distributions, and all these side-effects don't happen with
> DYNAMIC_DEBUG.

When DYNAMIC_DEBUG is enabled we have this wrapper from
include/linux/dynamic_debug.h:

if (unlikely(descriptor.flags & _DPRINTK_FLAGS_PRINT))


So the compiler is not emitting the side-effects in this
case.

>This suggests that we can make these side-effects not
> happen without DYNAMIC_DEBUG as well.
> Or I am missing something here?
> 

When DYNAMIC_DEBUG is disabled we are instead replacing
pr_debug() with the 'no_printk()' function as you've pointed
out. We are changing this to emit no code at all:

http://marc.info/?l=linux-kernel=144918276518878=2

Thanks,

-Jason
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Joe Perches]

On Fri, 2015-12-04 at 11:47 -0500, Jason Baron wrote:
> When DYNAMIC_DEBUG is enabled we have this wrapper from
> include/linux/dynamic_debug.h:
> 
> if (unlikely(descriptor.flags & _DPRINTK_FLAGS_PRINT))
>   
> 
> So the compiler is not emitting the side-effects in this
> case.

Huh?  Do I misunderstand what you are writing?

You are testing a variable that is not generally set
so the call is not being performed in the general case,
but the compiler can not elide the code.

If the variable was enabled via the control file, the
__dynamic_pr_debug would be performed with the
use-after-free.

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Thu, Dec 3, 2015 at 9:51 PM, Joe Perches  wrote:
> (adding lkml as this is likely better discussed there)
>
> On Thu, 2015-12-03 at 15:42 -0500, Jason Baron wrote:
>> On 12/03/2015 03:24 PM, Joe Perches wrote:
>> > On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
>> > > On 12/03/2015 03:03 PM, Joe Perches wrote:
>> > > > On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
>> > > > > On 12/03/2015 01:52 PM, Aaron Conole wrote:
>> > > > > > I think that as a minimum, the following patch should be evaluted,
>> > > > > > but am unsure to whom I should submit it (after I test):
>> > > > []
>> > > > > Agreed - the intention here is certainly to have no side effects. It
>> > > > > looks like 'no_printk()' is used in quite a few other places that 
>> > > > > would
>> > > > > benefit from this change. So we probably want a generic
>> > > > > 'really_no_printk()' macro.
>> > > >
>> > > > https://lkml.org/lkml/2012/6/17/231
>> > >
>> > > I don't see this in the tree.
>> >
>> > It never got applied.
>> >
>> > > Also maybe we should just convert
>> > > no_printk() to do what your 'eliminated_printk()'.
>> >
>> > Some of them at least.
>> >
>> > > So we can convert all users with this change?
>> >
>> > I don't think so, I think there are some
>> > function evaluation/side effects that are
>> > required.  I believe some do hardware I/O.
>> >
>> > It'd be good to at least isolate them.
>> >
>> > I'm not sure how to find them via some
>> > automated tool/mechanism though.
>> >
>> > I asked Julia Lawall about it once in this
>> > thread:  https://lkml.org/lkml/2014/12/3/696
>> >
>>
>> Seems rather fragile to have side effects that we rely
>> upon hidden in a printk().
>
> Yup.
>
>> Just convert them and see what breaks :)
>
> I appreciate your optimism.  It's very 1995.
> Try it and see what happens.


But Aaron says that DYNAMIC_DEBUG is enabled in most major
distributions, and all these side-effects don't happen with
DYNAMIC_DEBUG. This suggests that we can make these side-effects not
happen without DYNAMIC_DEBUG as well.
Or I am missing something here?
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Fri, Dec 4, 2015 at 6:48 PM, Marcelo Ricardo Leitner
 wrote:
> Hi Dmitry,
>
> Can you please test this patch?
> I'll re-post with proper subject if it works.

Still happening with the same stacks.


> ---8<---
>
> Dmitry Vyukov reported a use-after-free in the code expanded by the
> macro debug_post_sfx, which is caused by the use of the asoc pointer
> after it was freed within sctp_side_effect() scope.
>
> This patch fixes it by allowing sctp_side_effect to clear that asoc
> pointer when the TCB is freed.
>
> The macro is already prepared to handle such NULL pointer.
>
> Reported-by: Dmitry Vyukov 
> Signed-off-by: Marcelo Ricardo Leitner 
> ---
>  net/sctp/sm_sideeffect.c | 9 +
>  1 file changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c
> index 
> 6098d4c42fa91287d3cde36ac05d860f76d4fe32..05594dcd93e0d649cace5215d225bef2713f9310
>  100644
> --- a/net/sctp/sm_sideeffect.c
> +++ b/net/sctp/sm_sideeffect.c
> @@ -63,7 +63,7 @@ static int sctp_cmd_interpreter(sctp_event_t event_type,
>  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
>  sctp_state_t state,
>  struct sctp_endpoint *ep,
> -struct sctp_association *asoc,
> +struct sctp_association **asoc,
>  void *event_arg,
>  sctp_disposition_t status,
>  sctp_cmd_seq_t *commands,
> @@ -1123,7 +1123,7 @@ int sctp_do_sm(struct net *net, sctp_event_t 
> event_type, sctp_subtype_t subtype,
> debug_post_sfn();
>
> error = sctp_side_effects(event_type, subtype, state,
> - ep, asoc, event_arg, status,
> + ep, , event_arg, status,
>   , gfp);
> debug_post_sfx();
>
> @@ -1136,7 +1136,7 @@ int sctp_do_sm(struct net *net, sctp_event_t 
> event_type, sctp_subtype_t subtype,
>  static int sctp_side_effects(sctp_event_t event_type, sctp_subtype_t subtype,
>  sctp_state_t state,
>  struct sctp_endpoint *ep,
> -struct sctp_association *asoc,
> +struct sctp_association **asoc,
>  void *event_arg,
>  sctp_disposition_t status,
>  sctp_cmd_seq_t *commands,
> @@ -1151,7 +1151,7 @@ static int sctp_side_effects(sctp_event_t event_type, 
> sctp_subtype_t subtype,
>  * disposition SCTP_DISPOSITION_CONSUME.
>  */
> if (0 != (error = sctp_cmd_interpreter(event_type, subtype, state,
> -  ep, asoc,
> +  ep, *asoc,
>event_arg, status,
>commands, gfp)))
> goto bail;
> @@ -1175,6 +1175,7 @@ static int sctp_side_effects(sctp_event_t event_type, 
> sctp_subtype_t subtype,
>
> case SCTP_DISPOSITION_DELETE_TCB:
> /* This should now be a command. */
> +   *asoc = NULL;
> break;
>
> case SCTP_DISPOSITION_CONSUME:
> --
> 2.5.0
>
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Fri, Dec 4, 2015 at 10:34 PM, Marcelo Ricardo Leitner
 wrote:
> On Fri, Dec 04, 2015 at 09:25:35PM +0100, Dmitry Vyukov wrote:
>> On Fri, Dec 4, 2015 at 6:48 PM, Marcelo Ricardo Leitner
>>  wrote:
>> > Hi Dmitry,
>> >
>> > Can you please test this patch?
>> > I'll re-post with proper subject if it works.
>>
>> Still happening with the same stacks.
>
> Then there may be another one, I'm afraid.
>
> I'm using the testapp you shared in the first email, with that debug line
> enabled and added a new one:
> +   pr_debug("%p %d\n", asoc, asoc ? asoc->state : 0);
> debug_post_sfx();
> (should have used %x, but ok)
>
> Also enabled slub_debug=PUZ, and I get:
>
> without the patch:
> [   87.873640] sctp: 8800b71533d8 1
> [   87.873647] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:8800b71533d8[STATE_CLOSED]
> [   87.873739] sctp: 8800b71533d8 1
> [   87.873742] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:8800b71533d8[STATE_CLOSED]
> [   87.875149] sctp: 8800b71533d8 1802201963
> [   87.875238] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:8800b71533d8[STATE_CLOSED]
>
> 1802201963 = 0x6b6b6b6b, poison
>
> with the patch:
> [   81.071265] sctp: 880137571148 1
> [   81.071273] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:880137571148[STATE_CLOSED]
> [   81.071372] sctp: 880137571148 1
> [   81.071375] sctp: sctp_do_sm[post-sfx]: error:0,
> asoc:880137571148[STATE_CLOSED]
> [   81.072423] sctp:   (null) 0
> [   81.072427] sctp: sctp_do_sm[post-sfx]: error:0, asoc:
> (null)[STATE_CLOSED]
>
> This one, at least, is gone with this patch.


I will try to extract reproducer next week.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Fri, Dec 04, 2015 at 09:25:35PM +0100, Dmitry Vyukov wrote:
> On Fri, Dec 4, 2015 at 6:48 PM, Marcelo Ricardo Leitner
>  wrote:
> > Hi Dmitry,
> >
> > Can you please test this patch?
> > I'll re-post with proper subject if it works.
> 
> Still happening with the same stacks.

Then there may be another one, I'm afraid.

I'm using the testapp you shared in the first email, with that debug line
enabled and added a new one:
+   pr_debug("%p %d\n", asoc, asoc ? asoc->state : 0);
debug_post_sfx();
(should have used %x, but ok)

Also enabled slub_debug=PUZ, and I get:

without the patch:
[   87.873640] sctp: 8800b71533d8 1
[   87.873647] sctp: sctp_do_sm[post-sfx]: error:0,
asoc:8800b71533d8[STATE_CLOSED]
[   87.873739] sctp: 8800b71533d8 1
[   87.873742] sctp: sctp_do_sm[post-sfx]: error:0,
asoc:8800b71533d8[STATE_CLOSED]
[   87.875149] sctp: 8800b71533d8 1802201963
[   87.875238] sctp: sctp_do_sm[post-sfx]: error:0,
asoc:8800b71533d8[STATE_CLOSED]

1802201963 = 0x6b6b6b6b, poison

with the patch:
[   81.071265] sctp: 880137571148 1
[   81.071273] sctp: sctp_do_sm[post-sfx]: error:0,
asoc:880137571148[STATE_CLOSED]
[   81.071372] sctp: 880137571148 1
[   81.071375] sctp: sctp_do_sm[post-sfx]: error:0,
asoc:880137571148[STATE_CLOSED]
[   81.072423] sctp:   (null) 0
[   81.072427] sctp: sctp_do_sm[post-sfx]: error:0, asoc:
(null)[STATE_CLOSED]

This one, at least, is gone with this patch.

  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Thu, Dec 3, 2015 at 2:05 PM, Marcelo Ricardo Leitner
 wrote:
> Hi,
>
> On Tue, Nov 24, 2015 at 10:15:57AM +0100, Dmitry Vyukov wrote:
>>
>> Call Trace:
>>  [] __asan_report_load4_noabort+0x3e/0x40
>>  [] sctp_do_sm+0x42f6/0x4f60
>>  [] sctp_primitive_SHUTDOWN+0xa9/0xd0
>>  [] sctp_close+0x616/0x790
>>  [] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
>>  [] inet6_release+0x50/0x70 ./net/ipv6/af_inet6.c:416
>>  [< inline >] constant_test_bit ././arch/x86/include/asm/bitops.h:321
>>  [] sock_release+0x8d/0x200 ./net/socket.c:601
>>  [] sock_close+0x16/0x20 ./net/socket.c:1188
>>  [] __fput+0x21d/0x6e0 ./fs/file_table.c:265
>>  [] fput+0x15/0x20 ./fs/file_table.c:84
>>  [] task_work_run+0x163/0x1f0 
>> ./include/trace/events/rcu.h:20
>>  [< inline >] __list_add ./include/linux/list.h:42
>
> By any chance, did you have the pr_debug()s enabled?
> Because that would trigger a use-after-free on debug_post_sfx()
> macro expansion when the asoc is freed:
>
> #define debug_post_sfx() \
> pr_debug("%s[post-sfx]: error:%d, asoc:%p[%s]\n", __func__, error, \
>  asoc, sctp_state_tbl[(asoc && sctp_id2assoc(ep->base.sk, \
>  sctp_assoc2id(asoc))) ? asoc->state : SCTP_STATE_CLOSED])


No, I don't. But pr_debug always computes its arguments. See no_printk
in printk.h. So this use-after-free happens for all users.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

Hi,

On Tue, Nov 24, 2015 at 10:15:57AM +0100, Dmitry Vyukov wrote:
> 
> Call Trace:
>  [] __asan_report_load4_noabort+0x3e/0x40
>  [] sctp_do_sm+0x42f6/0x4f60
>  [] sctp_primitive_SHUTDOWN+0xa9/0xd0
>  [] sctp_close+0x616/0x790
>  [] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
>  [] inet6_release+0x50/0x70 ./net/ipv6/af_inet6.c:416
>  [< inline >] constant_test_bit ././arch/x86/include/asm/bitops.h:321
>  [] sock_release+0x8d/0x200 ./net/socket.c:601
>  [] sock_close+0x16/0x20 ./net/socket.c:1188
>  [] __fput+0x21d/0x6e0 ./fs/file_table.c:265
>  [] fput+0x15/0x20 ./fs/file_table.c:84
>  [] task_work_run+0x163/0x1f0 
> ./include/trace/events/rcu.h:20
>  [< inline >] __list_add ./include/linux/list.h:42

By any chance, did you have the pr_debug()s enabled?
Because that would trigger a use-after-free on debug_post_sfx()
macro expansion when the asoc is freed:

#define debug_post_sfx() \
pr_debug("%s[post-sfx]: error:%d, asoc:%p[%s]\n", __func__, error, \
 asoc, sctp_state_tbl[(asoc && sctp_id2assoc(ep->base.sk, \
 sctp_assoc2id(asoc))) ? asoc->state : SCTP_STATE_CLOSED])

  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Sat, Nov 28, 2015 at 04:50:56PM +0100, Dmitry Vyukov wrote:
> This also seems to lead the the following WARNINGS:
> 
> [ cut here ]
> WARNING: CPU: 3 PID: 21734 at kernel/jump_label.c:77
> __static_key_slow_dec+0xfb/0x120()
> jump label: negative count!
> Modules linked in:
> CPU: 3 PID: 21734 Comm: executor Tainted: GB   W   4.4.0-rc2+ #3
> Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
>   88006083f660 82719fc6 88006083f6d0
>  88003bbf8000 85a612e0 88006083f6a0 81244ec9
>  8152c54b ed000c107ed6 85a612e0 004d
> Call Trace:
>  [< inline >] __dump_stack lib/dump_stack.c:15
>  [] dump_stack+0x68/0x92 lib/dump_stack.c:50
>  [] warn_slowpath_common+0xd9/0x140 kernel/panic.c:460
>  [] warn_slowpath_fmt+0xa9/0xd0 kernel/panic.c:472
>  [] __static_key_slow_dec+0xfb/0x120 kernel/jump_label.c:76
>  [] static_key_slow_dec+0x51/0x90 kernel/jump_label.c:100
>  [] net_disable_timestamp+0x3b/0x50 net/core/dev.c:1709
>  [] sock_disable_timestamp+0x93/0xb0 net/core/sock.c:444
>  [] sk_destruct+0xec/0x440 net/core/sock.c:1457
>  [] __sk_free+0x57/0x200 net/core/sock.c:1476
>  [] sk_free+0x30/0x40 net/core/sock.c:1487
>  [< inline >] sock_put include/net/sock.h:1623
>  [] sctp_close+0x628/0x790 net/sctp/socket.c:1546
>  [] inet_release+0xed/0x1c0 net/ipv4/af_inet.c:413
>  [] inet6_release+0x50/0x70 net/ipv6/af_inet6.c:406
>  [] sock_release+0x8d/0x1d0 net/socket.c:571
>  [] sock_close+0x16/0x20 net/socket.c:1022
>  [] __fput+0x220/0x770 fs/file_table.c:208
>  [] fput+0x15/0x20 fs/file_table.c:244
>  [] task_work_run+0x163/0x1f0 kernel/task_work.c:115
>  [< inline >] exit_task_work include/linux/task_work.h:21
>  [] do_exit+0x809/0x2ae0 kernel/exit.c:750
>  [] do_group_exit+0x108/0x320 kernel/exit.c:880
>  [] get_signal+0x597/0x1630 kernel/signal.c:2307
>  [] do_signal+0x7f/0x18e0 arch/x86/kernel/signal.c:709
>  [] exit_to_usermode_loop+0xf1/0x1a0
> arch/x86/entry/common.c:247
>  [< inline >] prepare_exit_to_usermode arch/x86/entry/common.c:282
>  [] syscall_return_slowpath+0x19f/0x210
> arch/x86/entry/common.c:344
>  [] int_ret_from_sys_call+0x25/0x9f
> arch/x86/entry/entry_64.S:281
> ---[ end trace 3e42717665ff2020 ]---
> 
> 
> These WARNINGS always go with the original use-after-free reports. And
> I was not able to reproduce this WARNING with commented out
> sctp_association_destroy.
> 
> For the reference here is syzkaller program that triggers the WARNING.
> 
> r0 = socket(0xa, 0x1, 0x84)
> mmap(&(0x7f00)=nil, (0x1), 0x3, 0x32, 0x, 0x0)
> bind(r0, 
> &(0x7f00)="0a0033e049d02e714c37ffc4",
> 0x1c)
> connect(r0, 
> &(0x7f001000)="020033d97f01",
> 0x80)
> setsockopt$sock_int(r0, 0x1, 0x1d, &(0x7f001000+0x336)=0x1, 0x4)
> listen(r0, 0xbb3)
> r1 = accept(r0, &(0x7f003000+0xfd6)=nil, &(0x7f004000-0x2)=nil)

These two are unrelated, actually.

Do you know if this accept() returned something? Seems so.
Seems to be originated on
sctp_v6_create_accept_sk() -> sctp_copy_sock():

void sctp_copy_sock(struct sock *newsk, struct sock *sk,
struct sctp_association *asoc)
{
struct inet_sock *inet = inet_sk(sk);
struct inet_sock *newinet;

newsk->sk_type = sk->sk_type;
newsk->sk_bound_dev_if = sk->sk_bound_dev_if;
newsk->sk_flags = sk->sk_flags;   <---

As it enabled SO_TIMESTAMP on listening socket, this flag will be copied and
will trigger the second net_disable_timestamp() by the time the second
socket is destroyed, because it never had its enable counterpart called.

This also happens via sctp peeloff operation.

  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Thu, Dec 3, 2015 at 6:02 PM, Eric Dumazet  wrote:
> On Thu, Dec 3, 2015 at 7:55 AM, Dmitry Vyukov  wrote:
>> On Thu, Dec 3, 2015 at 3:48 PM, Eric Dumazet  wrote:

 No, I don't. But pr_debug always computes its arguments. See no_printk
 in printk.h. So this use-after-free happens for all users.
>>>
>>> Hmm.
>>>
>>> pr_debug() should be a nop unless either DEBUG or CONFIG_DYNAMIC_DEBUG are 
>>> set
>>>
>>> On our production kernels, pr_debug() is a nop.
>>>
>>> Can you double check ? Thanks !
>>
>>
>> Why should it be nop? no_printk thing in printk.h pretty much
>> explicitly makes it not a nop...
>>
>> Double-checked: debug_post_sfx leads to some generated code:
>>
>> debug_post_sfx();
>> 8229f256:   48 8b 85 58 fe ff ffmov-0x1a8(%rbp),%rax
>> 8229f25d:   48 85 c0test   %rax,%rax
>> 8229f260:   74 24   je
>> 8229f286 
>> 8229f262:   8b b0 a8 00 00 00   mov0xa8(%rax),%esi
>> 8229f268:   48 8b 85 60 fe ff ffmov-0x1a0(%rbp),%rax
>> 8229f26f:   44 89 85 74 fe ff ffmov%r8d,-0x18c(%rbp)
>> 8229f276:   48 8b 78 20 mov0x20(%rax),%rdi
>> 8229f27a:   e8 71 28 01 00  callq
>> 822b1af0 
>> 8229f27f:   44 8b 85 74 fe ff ffmov-0x18c(%rbp),%r8d
>>
>> return error;
>> }
>> 8229f286:   48 81 c4 a0 01 00 00add$0x1a0,%rsp
>> 8229f28d:   44 89 c0mov%r8d,%eax
>> 8229f290:   5b  pop%rbx
>> 8229f291:   41 5c   pop%r12
>> 8229f293:   41 5d   pop%r13
>> 8229f295:   41 5e   pop%r14
>> 8229f297:   41 5f   pop%r15
>> 8229f299:   5d  pop%rbp
>> 8229f29a:   c3  retq
>
> This is a serious concern, because we let in the past lot of patches
> converting traditional
>
> #ifdef DEBUG
> # define some_hand_coded_ugly_debug()  printk( _
> #else
> # define some_hand_coded_ugly_debug()
> #endif
>
> On the premise pr_debug() would be a nop.
>
> It seems it is not always the case. This is a very serious problem.
>
> We probably have hundred of potential bugs, because few people
> actually make sure all debugging stuff is correct,
> like comments can be wrong because they are not updated properly as time 
> flies.
>
> It is definitely a nop for many cases.
>
> +void eric_test_pr_debug(struct sock *sk)
> +{
> +   if (atomic_read(>sk_omem_alloc))
> +   pr_debug("%s: optmem leakage for sock %p\n",
> +__func__, sk);
> +}
>
> ->
>
> 4740 :
> 4740: e8 00 00 00 00   callq  4745 
> 4741: R_X86_64_PC32 __fentry__-0x4
> 4745: 55   push   %rbp
> 4746: 8b 87 24 01 00 00 mov0x124(%rdi),%eax //
> atomic_read()  but nothing follows
> 474c: 48 89 e5 mov%rsp,%rbp
> 474f: 5d   pop%rbp
> 4750: c3   retq



I would expect that it is nop when argument evaluation does not have
side-effects. For example, for a load of a variable compiler will most
likely elide it (though, it does not have to elide it, because the
load is spelled in the code, so it can also legally emit the load and
doesn't use the result).
But if argument computation has side-effect (or compiler can't prove
otherwise), it must emit code. It must emit code for function calls
when the function is defined in a different translation unit, and for
volatile accesses (most likely including atomic accesses), etc
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Eric Dumazet]

On Thu, Dec 3, 2015 at 7:55 AM, Dmitry Vyukov  wrote:
> On Thu, Dec 3, 2015 at 3:48 PM, Eric Dumazet  wrote:
>>>
>>> No, I don't. But pr_debug always computes its arguments. See no_printk
>>> in printk.h. So this use-after-free happens for all users.
>>
>> Hmm.
>>
>> pr_debug() should be a nop unless either DEBUG or CONFIG_DYNAMIC_DEBUG are 
>> set
>>
>> On our production kernels, pr_debug() is a nop.
>>
>> Can you double check ? Thanks !
>
>
> Why should it be nop? no_printk thing in printk.h pretty much
> explicitly makes it not a nop...
>
> Double-checked: debug_post_sfx leads to some generated code:
>
> debug_post_sfx();
> 8229f256:   48 8b 85 58 fe ff ffmov-0x1a8(%rbp),%rax
> 8229f25d:   48 85 c0test   %rax,%rax
> 8229f260:   74 24   je
> 8229f286 
> 8229f262:   8b b0 a8 00 00 00   mov0xa8(%rax),%esi
> 8229f268:   48 8b 85 60 fe ff ffmov-0x1a0(%rbp),%rax
> 8229f26f:   44 89 85 74 fe ff ffmov%r8d,-0x18c(%rbp)
> 8229f276:   48 8b 78 20 mov0x20(%rax),%rdi
> 8229f27a:   e8 71 28 01 00  callq
> 822b1af0 
> 8229f27f:   44 8b 85 74 fe ff ffmov-0x18c(%rbp),%r8d
>
> return error;
> }
> 8229f286:   48 81 c4 a0 01 00 00add$0x1a0,%rsp
> 8229f28d:   44 89 c0mov%r8d,%eax
> 8229f290:   5b  pop%rbx
> 8229f291:   41 5c   pop%r12
> 8229f293:   41 5d   pop%r13
> 8229f295:   41 5e   pop%r14
> 8229f297:   41 5f   pop%r15
> 8229f299:   5d  pop%rbp
> 8229f29a:   c3  retq

This is a serious concern, because we let in the past lot of patches
converting traditional

#ifdef DEBUG
# define some_hand_coded_ugly_debug()  printk( _
#else
# define some_hand_coded_ugly_debug()
#endif

On the premise pr_debug() would be a nop.

It seems it is not always the case. This is a very serious problem.

We probably have hundred of potential bugs, because few people
actually make sure all debugging stuff is correct,
like comments can be wrong because they are not updated properly as time flies.

It is definitely a nop for many cases.

+void eric_test_pr_debug(struct sock *sk)
+{
+   if (atomic_read(>sk_omem_alloc))
+   pr_debug("%s: optmem leakage for sock %p\n",
+__func__, sk);
+}

->

4740 :
4740: e8 00 00 00 00   callq  4745 
4741: R_X86_64_PC32 __fentry__-0x4
4745: 55   push   %rbp
4746: 8b 87 24 01 00 00 mov0x124(%rdi),%eax //
atomic_read()  but nothing follows
474c: 48 89 e5 mov%rsp,%rbp
474f: 5d   pop%rbp
4750: c3   retq
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Thu, Dec 03, 2015 at 02:51:33PM -0200, Marcelo Ricardo Leitner wrote:
> On Sat, Nov 28, 2015 at 04:50:56PM +0100, Dmitry Vyukov wrote:
> > This also seems to lead the the following WARNINGS:
> > 
> > [ cut here ]
> > WARNING: CPU: 3 PID: 21734 at kernel/jump_label.c:77
> > __static_key_slow_dec+0xfb/0x120()
> > jump label: negative count!
> > Modules linked in:
> > CPU: 3 PID: 21734 Comm: executor Tainted: GB   W   4.4.0-rc2+ #3
> > Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
> >   88006083f660 82719fc6 88006083f6d0
> >  88003bbf8000 85a612e0 88006083f6a0 81244ec9
> >  8152c54b ed000c107ed6 85a612e0 004d
> > Call Trace:
> >  [< inline >] __dump_stack lib/dump_stack.c:15
> >  [] dump_stack+0x68/0x92 lib/dump_stack.c:50
> >  [] warn_slowpath_common+0xd9/0x140 kernel/panic.c:460
> >  [] warn_slowpath_fmt+0xa9/0xd0 kernel/panic.c:472
> >  [] __static_key_slow_dec+0xfb/0x120 
> > kernel/jump_label.c:76
> >  [] static_key_slow_dec+0x51/0x90 kernel/jump_label.c:100
> >  [] net_disable_timestamp+0x3b/0x50 net/core/dev.c:1709
> >  [] sock_disable_timestamp+0x93/0xb0 net/core/sock.c:444
> >  [] sk_destruct+0xec/0x440 net/core/sock.c:1457
> >  [] __sk_free+0x57/0x200 net/core/sock.c:1476
> >  [] sk_free+0x30/0x40 net/core/sock.c:1487
> >  [< inline >] sock_put include/net/sock.h:1623
> >  [] sctp_close+0x628/0x790 net/sctp/socket.c:1546
> >  [] inet_release+0xed/0x1c0 net/ipv4/af_inet.c:413
> >  [] inet6_release+0x50/0x70 net/ipv6/af_inet6.c:406
> >  [] sock_release+0x8d/0x1d0 net/socket.c:571
> >  [] sock_close+0x16/0x20 net/socket.c:1022
> >  [] __fput+0x220/0x770 fs/file_table.c:208
> >  [] fput+0x15/0x20 fs/file_table.c:244
> >  [] task_work_run+0x163/0x1f0 kernel/task_work.c:115
> >  [< inline >] exit_task_work include/linux/task_work.h:21
> >  [] do_exit+0x809/0x2ae0 kernel/exit.c:750
> >  [] do_group_exit+0x108/0x320 kernel/exit.c:880
> >  [] get_signal+0x597/0x1630 kernel/signal.c:2307
> >  [] do_signal+0x7f/0x18e0 arch/x86/kernel/signal.c:709
> >  [] exit_to_usermode_loop+0xf1/0x1a0
> > arch/x86/entry/common.c:247
> >  [< inline >] prepare_exit_to_usermode arch/x86/entry/common.c:282
> >  [] syscall_return_slowpath+0x19f/0x210
> > arch/x86/entry/common.c:344
> >  [] int_ret_from_sys_call+0x25/0x9f
> > arch/x86/entry/entry_64.S:281
> > ---[ end trace 3e42717665ff2020 ]---
> > 
> > 
> > These WARNINGS always go with the original use-after-free reports. And
> > I was not able to reproduce this WARNING with commented out
> > sctp_association_destroy.
> > 
> > For the reference here is syzkaller program that triggers the WARNING.
> > 
> > r0 = socket(0xa, 0x1, 0x84)
> > mmap(&(0x7f00)=nil, (0x1), 0x3, 0x32, 0x, 0x0)
> > bind(r0, 
> > &(0x7f00)="0a0033e049d02e714c37ffc4",
> > 0x1c)
> > connect(r0, 
> > &(0x7f001000)="020033d97f01",
> > 0x80)
> > setsockopt$sock_int(r0, 0x1, 0x1d, &(0x7f001000+0x336)=0x1, 0x4)
> > listen(r0, 0xbb3)
> > r1 = accept(r0, &(0x7f003000+0xfd6)=nil, &(0x7f004000-0x2)=nil)
> 
> These two are unrelated, actually.
> 
> Do you know if this accept() returned something? Seems so.
> Seems to be originated on
> sctp_v6_create_accept_sk() -> sctp_copy_sock():
> 
> void sctp_copy_sock(struct sock *newsk, struct sock *sk,
> struct sctp_association *asoc)
> {
> struct inet_sock *inet = inet_sk(sk);
> struct inet_sock *newinet;
> 
> newsk->sk_type = sk->sk_type;
> newsk->sk_bound_dev_if = sk->sk_bound_dev_if;
> newsk->sk_flags = sk->sk_flags;   <---
> 
> As it enabled SO_TIMESTAMP on listening socket, this flag will be copied and
> will trigger the second net_disable_timestamp() by the time the second
> socket is destroyed, because it never had its enable counterpart called.
> 
> This also happens via sctp peeloff operation.

Vlad, others,

It's been a long time but this was introduced by commit 914e1c8b6980
("sctp: Inherit all socket options from parent correctly."). This is not
very consistent with how other protocols work and it will be hard to
keep tracking a negative mask of flags that we can't copy.

I reviewed the list of options and I'm thinking that only
SO_BINDTODEVICE is worth copying, leaving the others for the application
to re-set, as it is for other protocols. So I'm thinking on simply:

-   newsk->sk_flags = sk->sk_flags;
+   newsk->sk_flags = sk->sk_flags & SO_BINDTODEVICE;

in the above.

What do you think?

  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the 

Re: use-after-free in sctp_do_sm

[Marcelo Ricardo Leitner]

On Thu, Dec 03, 2015 at 04:55:44PM +0100, Dmitry Vyukov wrote:
> On Thu, Dec 3, 2015 at 3:48 PM, Eric Dumazet  wrote:
> >>
> >> No, I don't. But pr_debug always computes its arguments. See no_printk
> >> in printk.h. So this use-after-free happens for all users.
> >
> > Hmm.
> >
> > pr_debug() should be a nop unless either DEBUG or CONFIG_DYNAMIC_DEBUG are 
> > set
> >
> > On our production kernels, pr_debug() is a nop.
> >
> > Can you double check ? Thanks !
> 
> 
> Why should it be nop? no_printk thing in printk.h pretty much
> explicitly makes it not a nop...
> 
> Double-checked: debug_post_sfx leads to some generated code:

Oups. I was under that impression too, that it would do sanity-check
while being optimized out.

I'll think on a fix for this.

Thanks,
  Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Eric Dumazet]

>
> No, I don't. But pr_debug always computes its arguments. See no_printk
> in printk.h. So this use-after-free happens for all users.

Hmm.

pr_debug() should be a nop unless either DEBUG or CONFIG_DYNAMIC_DEBUG are set

On our production kernels, pr_debug() is a nop.

Can you double check ? Thanks !
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Thu, Dec 3, 2015 at 3:48 PM, Eric Dumazet  wrote:
>>
>> No, I don't. But pr_debug always computes its arguments. See no_printk
>> in printk.h. So this use-after-free happens for all users.
>
> Hmm.
>
> pr_debug() should be a nop unless either DEBUG or CONFIG_DYNAMIC_DEBUG are set
>
> On our production kernels, pr_debug() is a nop.
>
> Can you double check ? Thanks !


Why should it be nop? no_printk thing in printk.h pretty much
explicitly makes it not a nop...

Double-checked: debug_post_sfx leads to some generated code:

debug_post_sfx();
8229f256:   48 8b 85 58 fe ff ffmov-0x1a8(%rbp),%rax
8229f25d:   48 85 c0test   %rax,%rax
8229f260:   74 24   je
8229f286 
8229f262:   8b b0 a8 00 00 00   mov0xa8(%rax),%esi
8229f268:   48 8b 85 60 fe ff ffmov-0x1a0(%rbp),%rax
8229f26f:   44 89 85 74 fe ff ffmov%r8d,-0x18c(%rbp)
8229f276:   48 8b 78 20 mov0x20(%rax),%rdi
8229f27a:   e8 71 28 01 00  callq
822b1af0 
8229f27f:   44 8b 85 74 fe ff ffmov-0x18c(%rbp),%r8d

return error;
}
8229f286:   48 81 c4 a0 01 00 00add$0x1a0,%rsp
8229f28d:   44 89 c0mov%r8d,%eax
8229f290:   5b  pop%rbx
8229f291:   41 5c   pop%r12
8229f293:   41 5d   pop%r13
8229f295:   41 5e   pop%r14
8229f297:   41 5f   pop%r15
8229f299:   5d  pop%rbp
8229f29a:   c3  retq
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Aaron Conole]

Dmitry Vyukov  writes:
> On Thu, Dec 3, 2015 at 6:02 PM, Eric Dumazet  wrote:
>> On Thu, Dec 3, 2015 at 7:55 AM, Dmitry Vyukov  wrote:
>>> On Thu, Dec 3, 2015 at 3:48 PM, Eric Dumazet  wrote:
>
> No, I don't. But pr_debug always computes its arguments. See no_printk
> in printk.h. So this use-after-free happens for all users.

 Hmm.

 pr_debug() should be a nop unless either DEBUG or
 CONFIG_DYNAMIC_DEBUG are set

 On our production kernels, pr_debug() is a nop.

 Can you double check ? Thanks !
>>>
>>>
>>> Why should it be nop? no_printk thing in printk.h pretty much
>>> explicitly makes it not a nop...

Because it was until commit 5264f2f75d8. It also violates my reading of
the following from printk.h:

 * All of these will print unconditionally, although note that pr_debug()
 * and other debug macros are compiled out unless either DEBUG is defined
 * or CONFIG_DYNAMIC_DEBUG is set.

>>>
>>> Double-checked: debug_post_sfx leads to some generated code:
>>>
>>> debug_post_sfx();
>>> 8229f256:   48 8b 85 58 fe ff ffmov-0x1a8(%rbp),%rax
>>> 8229f25d:   48 85 c0test   %rax,%rax
>>> 8229f260:   74 24   je
>>> 8229f286 
>>> 8229f262:   8b b0 a8 00 00 00   mov0xa8(%rax),%esi
>>> 8229f268:   48 8b 85 60 fe ff ffmov-0x1a0(%rbp),%rax
>>> 8229f26f:   44 89 85 74 fe ff ffmov%r8d,-0x18c(%rbp)
>>> 8229f276:   48 8b 78 20 mov0x20(%rax),%rdi
>>> 8229f27a:   e8 71 28 01 00  callq
>>> 822b1af0 
>>> 8229f27f:   44 8b 85 74 fe ff ffmov-0x18c(%rbp),%r8d
>>>
>>> return error;
>>> }
>>> 8229f286:   48 81 c4 a0 01 00 00add$0x1a0,%rsp
>>> 8229f28d:   44 89 c0mov%r8d,%eax
>>> 8229f290:   5b  pop%rbx
>>> 8229f291:   41 5c   pop%r12
>>> 8229f293:   41 5d   pop%r13
>>> 8229f295:   41 5e   pop%r14
>>> 8229f297:   41 5f   pop%r15
>>> 8229f299:   5d  pop%rbp
>>> 8229f29a:   c3  retq
>>
>> This is a serious concern, because we let in the past lot of patches
>> converting traditional

+1

>> #ifdef DEBUG
>> # define some_hand_coded_ugly_debug()  printk( _
>> #else
>> # define some_hand_coded_ugly_debug()
>> #endif
>>
>> On the premise pr_debug() would be a nop.
>>
>> It seems it is not always the case. This is a very serious problem.

+1

>> We probably have hundred of potential bugs, because few people
>> actually make sure all debugging stuff is correct,
>> like comments can be wrong because they are not updated properly as
>> time flies.
>>
>> It is definitely a nop for many cases.
>>
>> +void eric_test_pr_debug(struct sock *sk)
>> +{
>> +   if (atomic_read(>sk_omem_alloc))
>> +   pr_debug("%s: optmem leakage for sock %p\n",
>> +__func__, sk);
>> +}
>>
>> ->
>>
>> 4740 :
>> 4740: e8 00 00 00 00   callq  4745 
>> 4741: R_X86_64_PC32 __fentry__-0x4
>> 4745: 55   push   %rbp
>> 4746: 8b 87 24 01 00 00 mov0x124(%rdi),%eax //
>> atomic_read()  but nothing follows
>> 474c: 48 89 e5 mov%rsp,%rbp
>> 474f: 5d   pop%rbp
>> 4750: c3   retq
>
>
>
> I would expect that it is nop when argument evaluation does not have
> side-effects. For example, for a load of a variable compiler will most
> likely elide it (though, it does not have to elide it, because the
> load is spelled in the code, so it can also legally emit the load and
> doesn't use the result).
> But if argument computation has side-effect (or compiler can't prove
> otherwise), it must emit code. It must emit code for function calls
> when the function is defined in a different translation unit, and for
> volatile accesses (most likely including atomic accesses), etc

This isn't 100% true. As you state, in order to reach the return 0, all
side effects must be evaluated. Load generally does not have side
effects, so it can be safely elided, but function() must be emitted.

However, that is _not_ required to get the desired warning emission on a
printf argument function, see http://pastebin.com/UHuaydkj for an
example.

I think that as a minimum, the following patch should be evaluted, but am
unsure to whom I should submit it (after I test):

diff --git a/include/linux/printk.h b/include/linux/printk.h
index 9729565..cd24d2d 100644
--- a/include/linux/printk.h
+++ b/include/linux/printk.h
@@ -286,7 +286,7 @@ extern asmlinkage void dump_stack(void) 

Re: use-after-free in sctp_do_sm

[Jason Baron]

On 12/03/2015 01:52 PM, Aaron Conole wrote:
> Dmitry Vyukov  writes:
>> On Thu, Dec 3, 2015 at 6:02 PM, Eric Dumazet  wrote:
>>> On Thu, Dec 3, 2015 at 7:55 AM, Dmitry Vyukov  wrote:
 On Thu, Dec 3, 2015 at 3:48 PM, Eric Dumazet  wrote:
>>
>> No, I don't. But pr_debug always computes its arguments. See no_printk
>> in printk.h. So this use-after-free happens for all users.
>
> Hmm.
>
> pr_debug() should be a nop unless either DEBUG or
> CONFIG_DYNAMIC_DEBUG are set
>
> On our production kernels, pr_debug() is a nop.
>
> Can you double check ? Thanks !


 Why should it be nop? no_printk thing in printk.h pretty much
 explicitly makes it not a nop...
> 
> Because it was until commit 5264f2f75d8. It also violates my reading of
> the following from printk.h:
> 
>  * All of these will print unconditionally, although note that pr_debug()
>  * and other debug macros are compiled out unless either DEBUG is defined
>  * or CONFIG_DYNAMIC_DEBUG is set.
> 

 Double-checked: debug_post_sfx leads to some generated code:

 debug_post_sfx();
 8229f256:   48 8b 85 58 fe ff ffmov-0x1a8(%rbp),%rax
 8229f25d:   48 85 c0test   %rax,%rax
 8229f260:   74 24   je
 8229f286 
 8229f262:   8b b0 a8 00 00 00   mov0xa8(%rax),%esi
 8229f268:   48 8b 85 60 fe ff ffmov-0x1a0(%rbp),%rax
 8229f26f:   44 89 85 74 fe ff ffmov%r8d,-0x18c(%rbp)
 8229f276:   48 8b 78 20 mov0x20(%rax),%rdi
 8229f27a:   e8 71 28 01 00  callq
 822b1af0 
 8229f27f:   44 8b 85 74 fe ff ffmov-0x18c(%rbp),%r8d

 return error;
 }
 8229f286:   48 81 c4 a0 01 00 00add$0x1a0,%rsp
 8229f28d:   44 89 c0mov%r8d,%eax
 8229f290:   5b  pop%rbx
 8229f291:   41 5c   pop%r12
 8229f293:   41 5d   pop%r13
 8229f295:   41 5e   pop%r14
 8229f297:   41 5f   pop%r15
 8229f299:   5d  pop%rbp
 8229f29a:   c3  retq
>>>
>>> This is a serious concern, because we let in the past lot of patches
>>> converting traditional
> 
> +1
> 
>>> #ifdef DEBUG
>>> # define some_hand_coded_ugly_debug()  printk( _
>>> #else
>>> # define some_hand_coded_ugly_debug()
>>> #endif
>>>
>>> On the premise pr_debug() would be a nop.
>>>
>>> It seems it is not always the case. This is a very serious problem.
> 
> +1
> 
>>> We probably have hundred of potential bugs, because few people
>>> actually make sure all debugging stuff is correct,
>>> like comments can be wrong because they are not updated properly as
>>> time flies.
>>>
>>> It is definitely a nop for many cases.
>>>
>>> +void eric_test_pr_debug(struct sock *sk)
>>> +{
>>> +   if (atomic_read(>sk_omem_alloc))
>>> +   pr_debug("%s: optmem leakage for sock %p\n",
>>> +__func__, sk);
>>> +}
>>>
>>> ->
>>>
>>> 4740 :
>>> 4740: e8 00 00 00 00   callq  4745 
>>> 4741: R_X86_64_PC32 __fentry__-0x4
>>> 4745: 55   push   %rbp
>>> 4746: 8b 87 24 01 00 00 mov0x124(%rdi),%eax //
>>> atomic_read()  but nothing follows
>>> 474c: 48 89 e5 mov%rsp,%rbp
>>> 474f: 5d   pop%rbp
>>> 4750: c3   retq
>>
>>
>>
>> I would expect that it is nop when argument evaluation does not have
>> side-effects. For example, for a load of a variable compiler will most
>> likely elide it (though, it does not have to elide it, because the
>> load is spelled in the code, so it can also legally emit the load and
>> doesn't use the result).
>> But if argument computation has side-effect (or compiler can't prove
>> otherwise), it must emit code. It must emit code for function calls
>> when the function is defined in a different translation unit, and for
>> volatile accesses (most likely including atomic accesses), etc
> 
> This isn't 100% true. As you state, in order to reach the return 0, all
> side effects must be evaluated. Load generally does not have side
> effects, so it can be safely elided, but function() must be emitted.
> 
> However, that is _not_ required to get the desired warning emission on a
> printf argument function, see http://pastebin.com/UHuaydkj for an
> example.
> 
> I think that as a minimum, the following patch should be evaluted, but am
> unsure to whom I should submit it (after I test):

Agreed - the 

Re: use-after-free in sctp_do_sm

[Joe Perches]

On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
> On 12/03/2015 03:03 PM, Joe Perches wrote:
> > On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
> > > On 12/03/2015 01:52 PM, Aaron Conole wrote:
> > > > I think that as a minimum, the following patch should be evaluted,
> > > > but am unsure to whom I should submit it (after I test):
> > []
> > > Agreed - the intention here is certainly to have no side effects. It
> > > looks like 'no_printk()' is used in quite a few other places that would
> > > benefit from this change. So we probably want a generic
> > > 'really_no_printk()' macro.
> > 
> > https://lkml.org/lkml/2012/6/17/231
> 
> I don't see this in the tree.

It never got applied.

> Also maybe we should just convert
> no_printk() to do what your 'eliminated_printk()'.

Some of them at least.

> So we can convert all users with this change?

I don't think so, I think there are some
function evaluation/side effects that are
required.  I believe some do hardware I/O.

It'd be good to at least isolate them.

I'm not sure how to find them via some
automated tool/mechanism though.

I asked Julia Lawall about it once in this
thread:  https://lkml.org/lkml/2014/12/3/696

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Joe Perches]

On Thu, 2015-12-03 at 13:52 -0500, Aaron Conole wrote:
> Dmitry Vyukov  writes:
> > On Thu, Dec 3, 2015 at 6:02 PM, Eric Dumazet  wrote:
> > > On Thu, Dec 3, 2015 at 7:55 AM, Dmitry Vyukov  wrote:
> > > > On Thu, Dec 3, 2015 at 3:48 PM, Eric Dumazet  wrote:
> > > > > > 
> > > > > > No, I don't. But pr_debug always computes its arguments. See 
> > > > > > no_printk
> > > > > > in printk.h. So this use-after-free happens for all users.
> > > > > 
> > > > > Hmm.
> > > > > 
> > > > > pr_debug() should be a nop unless either DEBUG or
> > > > > CONFIG_DYNAMIC_DEBUG are set
> > > > > 
> > > > > On our production kernels, pr_debug() is a nop.
> > > > > 
> > > > > Can you double check ? Thanks !
> > > > 
> > > > 
> > > > Why should it be nop? no_printk thing in printk.h pretty much
> > > > explicitly makes it not a nop...
> 
> Because it was until commit 5264f2f75d8. It also violates my reading of
> the following from printk.h:
> 
>  * All of these will print unconditionally, although note that pr_debug()
>  * and other debug macros are compiled out unless either DEBUG is defined
>  * or CONFIG_DYNAMIC_DEBUG is set.
> 
> > > > 
> > > > Double-checked: debug_post_sfx leads to some generated code:
> > > > 
> > > > debug_post_sfx();
> > > > 8229f256:   48 8b 85 58 fe ff ffmov-0x1a8(%rbp),%rax
> > > > 8229f25d:   48 85 c0test   %rax,%rax
> > > > 8229f260:   74 24   je
> > > > 8229f286 
> > > > 8229f262:   8b b0 a8 00 00 00   mov0xa8(%rax),%esi
> > > > 8229f268:   48 8b 85 60 fe ff ffmov-0x1a0(%rbp),%rax
> > > > 8229f26f:   44 89 85 74 fe ff ffmov%r8d,-0x18c(%rbp)
> > > > 8229f276:   48 8b 78 20 mov0x20(%rax),%rdi
> > > > 8229f27a:   e8 71 28 01 00  callq
> > > > 822b1af0 
> > > > 8229f27f:   44 8b 85 74 fe ff ffmov-0x18c(%rbp),%r8d
> > > > 
> > > > return error;
> > > > }
> > > > 8229f286:   48 81 c4 a0 01 00 00add$0x1a0,%rsp
> > > > 8229f28d:   44 89 c0mov%r8d,%eax
> > > > 8229f290:   5b  pop%rbx
> > > > 8229f291:   41 5c   pop%r12
> > > > 8229f293:   41 5d   pop%r13
> > > > 8229f295:   41 5e   pop%r14
> > > > 8229f297:   41 5f   pop%r15
> > > > 8229f299:   5d  pop%rbp
> > > > 8229f29a:   c3  retq
> > > 
> > > This is a serious concern, because we let in the past lot of patches
> > > converting traditional
> 
> +1
> 
> > > #ifdef DEBUG
> > > # define some_hand_coded_ugly_debug()  printk( _
> > > #else
> > > # define some_hand_coded_ugly_debug()
> > > #endif
> > > 
> > > On the premise pr_debug() would be a nop.
> > > 
> > > It seems it is not always the case. This is a very serious problem.
> 
> +1
> 
> > > We probably have hundred of potential bugs, because few people
> > > actually make sure all debugging stuff is correct,
> > > like comments can be wrong because they are not updated properly as
> > > time flies.
> > > 
> > > It is definitely a nop for many cases.
> > > 
> > > +void eric_test_pr_debug(struct sock *sk)
> > > +{
> > > +   if (atomic_read(>sk_omem_alloc))
> > > +   pr_debug("%s: optmem leakage for sock %p\n",
> > > +__func__, sk);
> > > +}
> > > 
> > > ->
> > > 
> > > 4740 :
> > > 4740: e8 00 00 00 00   callq  4745 
> > > 4741: R_X86_64_PC32 __fentry__-0x4
> > > 4745: 55   push   %rbp
> > > 4746: 8b 87 24 01 00 00 mov0x124(%rdi),%eax //
> > > atomic_read()  but nothing follows
> > > 474c: 48 89 e5 mov%rsp,%rbp
> > > 474f: 5d   pop%rbp
> > > 4750: c3   retq
> > 
> > 
> > 
> > I would expect that it is nop when argument evaluation does not have
> > side-effects. For example, for a load of a variable compiler will most
> > likely elide it (though, it does not have to elide it, because the
> > load is spelled in the code, so it can also legally emit the load and
> > doesn't use the result).
> > But if argument computation has side-effect (or compiler can't prove
> > otherwise), it must emit code. It must emit code for function calls
> > when the function is defined in a different translation unit, and for
> > volatile accesses (most likely including atomic accesses), etc
> 
> This isn't 100% true. As you state, in order to reach the return 0, all
> side effects must be evaluated. Load generally does not have side
> effects, so it can be safely elided, but function() must be emitted.
> 
> However, that is _not_ required to get the desired warning emission on a
> printf 

Re: use-after-free in sctp_do_sm

[Joe Perches]

On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
> On 12/03/2015 01:52 PM, Aaron Conole wrote:
> > I think that as a minimum, the following patch should be evaluted,
> > but am unsure to whom I should submit it (after I test):
[]
> Agreed - the intention here is certainly to have no side effects. It
> looks like 'no_printk()' is used in quite a few other places that would
> benefit from this change. So we probably want a generic
> 'really_no_printk()' macro.

https://lkml.org/lkml/2012/6/17/231

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Joe Perches]

(adding lkml as this is likely better discussed there)

On Thu, 2015-12-03 at 15:42 -0500, Jason Baron wrote:
> On 12/03/2015 03:24 PM, Joe Perches wrote:
> > On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
> > > On 12/03/2015 03:03 PM, Joe Perches wrote:
> > > > On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
> > > > > On 12/03/2015 01:52 PM, Aaron Conole wrote:
> > > > > > I think that as a minimum, the following patch should be evaluted,
> > > > > > but am unsure to whom I should submit it (after I test):
> > > > []
> > > > > Agreed - the intention here is certainly to have no side effects. It
> > > > > looks like 'no_printk()' is used in quite a few other places that 
> > > > > would
> > > > > benefit from this change. So we probably want a generic
> > > > > 'really_no_printk()' macro.
> > > > 
> > > > https://lkml.org/lkml/2012/6/17/231
> > > 
> > > I don't see this in the tree.
> > 
> > It never got applied.
> > 
> > > Also maybe we should just convert
> > > no_printk() to do what your 'eliminated_printk()'.
> > 
> > Some of them at least.
> > 
> > > So we can convert all users with this change?
> > 
> > I don't think so, I think there are some
> > function evaluation/side effects that are
> > required.  I believe some do hardware I/O.
> > 
> > It'd be good to at least isolate them.
> > 
> > I'm not sure how to find them via some
> > automated tool/mechanism though.
> > 
> > I asked Julia Lawall about it once in this
> > thread:  https://lkml.org/lkml/2014/12/3/696
> > 
> 
> Seems rather fragile to have side effects that we rely
> upon hidden in a printk().

Yup.

> Just convert them and see what breaks :)

I appreciate your optimism.  It's very 1995.
Try it and see what happens.

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Jason Baron]

On 12/03/2015 03:03 PM, Joe Perches wrote:
> On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
>> On 12/03/2015 01:52 PM, Aaron Conole wrote:
>>> I think that as a minimum, the following patch should be evaluted,
>>> but am unsure to whom I should submit it (after I test):
> []
>> Agreed - the intention here is certainly to have no side effects. It
>> looks like 'no_printk()' is used in quite a few other places that would
>> benefit from this change. So we probably want a generic
>> 'really_no_printk()' macro.
> 
> https://lkml.org/lkml/2012/6/17/231
> 

I don't see this in the tree. Also maybe we should just convert
no_printk() to do what your 'eliminated_printk()'. So we can convert all
users with this change?

Thanks,

-Jason
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Jason Baron]

On 12/03/2015 03:24 PM, Joe Perches wrote:
> On Thu, 2015-12-03 at 15:10 -0500, Jason Baron wrote:
>> On 12/03/2015 03:03 PM, Joe Perches wrote:
>>> On Thu, 2015-12-03 at 14:32 -0500, Jason Baron wrote:
 On 12/03/2015 01:52 PM, Aaron Conole wrote:
> I think that as a minimum, the following patch should be evaluted,
> but am unsure to whom I should submit it (after I test):
>>> []
 Agreed - the intention here is certainly to have no side effects. It
 looks like 'no_printk()' is used in quite a few other places that would
 benefit from this change. So we probably want a generic
 'really_no_printk()' macro.
>>>
>>> https://lkml.org/lkml/2012/6/17/231
>>
>> I don't see this in the tree.
> 
> It never got applied.
> 
>> Also maybe we should just convert
>> no_printk() to do what your 'eliminated_printk()'.
> 
> Some of them at least.
> 
>> So we can convert all users with this change?
> 
> I don't think so, I think there are some
> function evaluation/side effects that are
> required.  I believe some do hardware I/O.
> 
> It'd be good to at least isolate them.
> 
> I'm not sure how to find them via some
> automated tool/mechanism though.
> 
> I asked Julia Lawall about it once in this
> thread:  https://lkml.org/lkml/2014/12/3/696
> 

Seems rather fragile to have side effects that we rely
upon hidden in a printk().

Just convert them and see what breaks :)
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Vlad Yasevich]

On 12/03/2015 01:06 PM, Marcelo wrote:
> 
> 
> Em 3 de dezembro de 2015 15:59:10 BRST, Eric Dumazet  
> escreveu:
>> On Thu, 2015-12-03 at 15:43 -0200, Marcelo Ricardo Leitner wrote:
>>
>>> Vlad, others,
>>>
>>> It's been a long time but this was introduced by commit 914e1c8b6980
>>> ("sctp: Inherit all socket options from parent correctly."). This is
>> not
>>> very consistent with how other protocols work and it will be hard to
>>> keep tracking a negative mask of flags that we can't copy.
>>>
>>> I reviewed the list of options and I'm thinking that only
>>> SO_BINDTODEVICE is worth copying, leaving the others for the
>> application
>>> to re-set, as it is for other protocols. So I'm thinking on simply:
>>>
>>> -   newsk->sk_flags = sk->sk_flags;
>>> +   newsk->sk_flags = sk->sk_flags & SO_BINDTODEVICE;
>>>
>>> in the above.
>>>
>>> What do you think?
>>
>> I think SO_BINDTODEVICE is not a flag ;)
>>
>> #define SO_BINDTODEVICE25
> 
> Oops, indeed!
> Idea persists.
> Thx!
> 

Hmm...  sk_clone_lock() appears to copy the flags as well, so it would
appear the tcp accept() sockets would also have timestamping set.

I can see how we probably shouldn't being copying sk_flags as there isn't
much there that need to be set.

-vlad


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo]

Em 3 de dezembro de 2015 15:59:10 BRST, Eric Dumazet  
escreveu:
>On Thu, 2015-12-03 at 15:43 -0200, Marcelo Ricardo Leitner wrote:
>
>> Vlad, others,
>> 
>> It's been a long time but this was introduced by commit 914e1c8b6980
>> ("sctp: Inherit all socket options from parent correctly."). This is
>not
>> very consistent with how other protocols work and it will be hard to
>> keep tracking a negative mask of flags that we can't copy.
>> 
>> I reviewed the list of options and I'm thinking that only
>> SO_BINDTODEVICE is worth copying, leaving the others for the
>application
>> to re-set, as it is for other protocols. So I'm thinking on simply:
>> 
>> -   newsk->sk_flags = sk->sk_flags;
>> +   newsk->sk_flags = sk->sk_flags & SO_BINDTODEVICE;
>> 
>> in the above.
>> 
>> What do you think?
>
>I think SO_BINDTODEVICE is not a flag ;)
>
>#define SO_BINDTODEVICE25

Oops, indeed!
Idea persists.
Thx!
-- 
Sent from mobile. Please excuse my brevity.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Eric Dumazet]

On Thu, 2015-12-03 at 15:43 -0200, Marcelo Ricardo Leitner wrote:

> Vlad, others,
> 
> It's been a long time but this was introduced by commit 914e1c8b6980
> ("sctp: Inherit all socket options from parent correctly."). This is not
> very consistent with how other protocols work and it will be hard to
> keep tracking a negative mask of flags that we can't copy.
> 
> I reviewed the list of options and I'm thinking that only
> SO_BINDTODEVICE is worth copying, leaving the others for the application
> to re-set, as it is for other protocols. So I'm thinking on simply:
> 
> -   newsk->sk_flags = sk->sk_flags;
> +   newsk->sk_flags = sk->sk_flags & SO_BINDTODEVICE;
> 
> in the above.
> 
> What do you think?

I think SO_BINDTODEVICE is not a flag ;)

#define SO_BINDTODEVICE25


--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Marcelo]

On Thu, Dec 03, 2015 at 01:35:37PM -0500, Vlad Yasevich wrote:
> On 12/03/2015 01:06 PM, Marcelo wrote:
> > 
> > 
> > Em 3 de dezembro de 2015 15:59:10 BRST, Eric Dumazet 
> >  escreveu:
> >> On Thu, 2015-12-03 at 15:43 -0200, Marcelo Ricardo Leitner wrote:
> >>
> >>> Vlad, others,
> >>>
> >>> It's been a long time but this was introduced by commit 914e1c8b6980
> >>> ("sctp: Inherit all socket options from parent correctly."). This is
> >> not
> >>> very consistent with how other protocols work and it will be hard to
> >>> keep tracking a negative mask of flags that we can't copy.
> >>>
> >>> I reviewed the list of options and I'm thinking that only
> >>> SO_BINDTODEVICE is worth copying, leaving the others for the
> >> application
> >>> to re-set, as it is for other protocols. So I'm thinking on simply:
> >>>
> >>> -   newsk->sk_flags = sk->sk_flags;
> >>> +   newsk->sk_flags = sk->sk_flags & SO_BINDTODEVICE;
> >>>
> >>> in the above.
> >>>
> >>> What do you think?
> >>
> >> I think SO_BINDTODEVICE is not a flag ;)
> >>
> >> #define SO_BINDTODEVICE25
> > 
> > Oops, indeed!
> > Idea persists.
> > Thx!
> > 
> 
> Hmm...  sk_clone_lock() appears to copy the flags as well, so it would
> appear the tcp accept() sockets would also have timestamping set.

Ahh right, through a memcpy. I completely missed that.

And later on it does:
if (sock_needs_netstamp(sk) &&
newsk->sk_flags & SK_FLAGS_TIMESTAMP)
net_enable_timestamp();

> I can see how we probably shouldn't being copying sk_flags as there isn't
> much there that need to be set.

I take that back then, we can enable timestamp like the above instead.
I'll test and post a patch soon.

Thanks,
Marcelo

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Wed, Nov 25, 2015 at 4:12 PM, Vlad Yasevich <vyasev...@gmail.com> wrote:
> On 11/24/2015 03:45 PM, Neil Horman wrote:
>> On Tue, Nov 24, 2015 at 11:10:32AM +0100, Dmitry Vyukov wrote:
>>> On Tue, Nov 24, 2015 at 10:31 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
>>>> On Tue, Nov 24, 2015 at 10:15 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
>>>>> Hello,
>>>>>
>>>>> The following program triggers use-after-free in sctp_do_sm:
>>>>>
>>>>> // autogenerated by syzkaller (http://github.com/google/syzkaller)
>>>>> #include 
>>>>> #include 
>>>>> #include 
>>>>>
>>>>> int main()
>>>>> {
>>>>> long r0 = syscall(SYS_socket, 0xaul, 0x80805ul, 0x0ul, 0, 0, 0);
>>>>> long r1 = syscall(SYS_mmap, 0x2000ul, 0x1ul, 0x3ul,
>>>>> 0x32ul, 0xul, 0x0ul);
>>>>> memcpy((void*)0x20002fe4,
>>>>> "\x0a\x00\x33\xe7\xeb\x9d\xcf\x61\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xc5\xc8\x88\x64",
>>>>> 28);
>>>>> long r3 = syscall(SYS_bind, r0, 0x20002fe4ul, 0x1cul, 0, 0, 0);
>>>>> memcpy((void*)0x2faa,
>>>>> "\x9b\x01\x7d\xcd\xb8\x6a\xc7\x3d\x09\x3a\x07\x00\xa7\xc4\xe9\xee\x0a\xd6\xec\xde\x26\x75\x5f\x22\xae\x4e\x33\x00\xb0\x76\x10\x70\xd6\xca\x19\xbc\x15\x83\xcf\x2e\xbc\x99\x0c\x5e\x83\x89\xc1\x44\x9c\x6e\x74\xd8\x5d\x5d\xd0\xf0\xdf\x47\xc0\x00\x71\x0b\x55\x4c\xab\xf0\xd8\x90\xd5\x92\x8c\x6e\x33\x22\x15\x5b\x19\xfb\xed\xdd\xa6\xac\xcb\x60\xcf\xe2\xde\xed\xdb\x95\x5c\xaa\x20\xa3",
>>>>> 94);
>>>>> memcpy((void*)0x233a,
>>>>> "\x02\x00\x33\xe2\x7f\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00",
>>>>> 128);
>>>>> long r6 = syscall(SYS_sendto, r0, 0x2faaul, 0x5eul,
>>>>> 0x81ul, 0x233aul, 0x80ul);
>>>>> return 0;
>>>>> }
>>>>>
>>>>>
>>>>> ==
>>>>> BUG: KASAN: use-after-free in sctp_do_sm+0x42f6/0x4f60 at addr 
>>>>> 880036fa80a8
>>>>> Read of size 4 by task a.out/5664
>>>>> =
>>>>> BUG kmalloc-4096 (Tainted: GB  ): kasan: bad access detected
>>>>> -
>>>>>
>>>>> INFO: Allocated in sctp_association_new+0x6f/0x1ea0 age=8 cpu=1 pid=5664
>>>>> [<  none  >] kmem_cache_alloc_trace+0x1cf/0x220 ./mm/slab.c:3707
>>>>> [<  none  >] sctp_association_new+0x6f/0x1ea0
>>>>> [<  none  >] sctp_sendmsg+0x1954/0x28e0
>>>>> [<  none  >] inet_sendmsg+0x316/0x4f0 ./net/ipv4/af_inet.c:802
>>>>> [< inline >] __sock_sendmsg_nosec ./net/socket.c:641
>>>>> [< inline >] __sock_sendmsg ./net/socket.c:651
>>>>> [<  none  >] sock_sendmsg+0xca/0x110 ./net/socket.c:662
>>>>> [<  none  >] SYSC_sendto+0x208/0x350 ./net/socket.c:1841
>>>>> [<  none  >] SyS_sendto+0x40/0x50 ./net/socket.c:1862
>>>>> [<  none  >] entry_SYSCALL_64_fastpath+0x16/0x7a
>>>>>
>>>>> INFO: Freed in sctp_association_put+0x150/0x250 age=14 cpu=1 pid=5664
>>>>> [<  none  >] kfree+0x199/0x1b0 ./mm/slab.c:1211
>>>>> [<  none  >] sctp_association_put+0x150/0x250
>>>>> [<  none  >] sctp_association_free+0x498/0x630
>>>>> [<  none  >] sctp_do_sm+0xd8b/0x4f60
>>>>> [<  none  >] sctp_primitive_SHUTDOWN+0xa9/0xd0
>>>>> [<  none  >] sctp_close+0x616/0x790
>>>>> [<  none  >] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
>>>>> [< 

Re: use-after-free in sctp_do_sm

[Vlad Yasevich]

On 11/24/2015 03:45 PM, Neil Horman wrote:
> On Tue, Nov 24, 2015 at 11:10:32AM +0100, Dmitry Vyukov wrote:
>> On Tue, Nov 24, 2015 at 10:31 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
>>> On Tue, Nov 24, 2015 at 10:15 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
>>>> Hello,
>>>>
>>>> The following program triggers use-after-free in sctp_do_sm:
>>>>
>>>> // autogenerated by syzkaller (http://github.com/google/syzkaller)
>>>> #include 
>>>> #include 
>>>> #include 
>>>>
>>>> int main()
>>>> {
>>>> long r0 = syscall(SYS_socket, 0xaul, 0x80805ul, 0x0ul, 0, 0, 0);
>>>> long r1 = syscall(SYS_mmap, 0x2000ul, 0x1ul, 0x3ul,
>>>> 0x32ul, 0xul, 0x0ul);
>>>> memcpy((void*)0x20002fe4,
>>>> "\x0a\x00\x33\xe7\xeb\x9d\xcf\x61\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xc5\xc8\x88\x64",
>>>> 28);
>>>> long r3 = syscall(SYS_bind, r0, 0x20002fe4ul, 0x1cul, 0, 0, 0);
>>>> memcpy((void*)0x2faa,
>>>> "\x9b\x01\x7d\xcd\xb8\x6a\xc7\x3d\x09\x3a\x07\x00\xa7\xc4\xe9\xee\x0a\xd6\xec\xde\x26\x75\x5f\x22\xae\x4e\x33\x00\xb0\x76\x10\x70\xd6\xca\x19\xbc\x15\x83\xcf\x2e\xbc\x99\x0c\x5e\x83\x89\xc1\x44\x9c\x6e\x74\xd8\x5d\x5d\xd0\xf0\xdf\x47\xc0\x00\x71\x0b\x55\x4c\xab\xf0\xd8\x90\xd5\x92\x8c\x6e\x33\x22\x15\x5b\x19\xfb\xed\xdd\xa6\xac\xcb\x60\xcf\xe2\xde\xed\xdb\x95\x5c\xaa\x20\xa3",
>>>> 94);
>>>> memcpy((void*)0x233a,
>>>> "\x02\x00\x33\xe2\x7f\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00",
>>>> 128);
>>>> long r6 = syscall(SYS_sendto, r0, 0x2faaul, 0x5eul,
>>>> 0x81ul, 0x233aul, 0x80ul);
>>>> return 0;
>>>> }
>>>>
>>>>
>>>> ==
>>>> BUG: KASAN: use-after-free in sctp_do_sm+0x42f6/0x4f60 at addr 
>>>> 880036fa80a8
>>>> Read of size 4 by task a.out/5664
>>>> =
>>>> BUG kmalloc-4096 (Tainted: GB  ): kasan: bad access detected
>>>> -
>>>>
>>>> INFO: Allocated in sctp_association_new+0x6f/0x1ea0 age=8 cpu=1 pid=5664
>>>> [<  none  >] kmem_cache_alloc_trace+0x1cf/0x220 ./mm/slab.c:3707
>>>> [<  none  >] sctp_association_new+0x6f/0x1ea0
>>>> [<  none  >] sctp_sendmsg+0x1954/0x28e0
>>>> [<  none  >] inet_sendmsg+0x316/0x4f0 ./net/ipv4/af_inet.c:802
>>>> [< inline >] __sock_sendmsg_nosec ./net/socket.c:641
>>>> [< inline >] __sock_sendmsg ./net/socket.c:651
>>>> [<  none  >] sock_sendmsg+0xca/0x110 ./net/socket.c:662
>>>> [<  none  >] SYSC_sendto+0x208/0x350 ./net/socket.c:1841
>>>> [<  none  >] SyS_sendto+0x40/0x50 ./net/socket.c:1862
>>>> [<  none  >] entry_SYSCALL_64_fastpath+0x16/0x7a
>>>>
>>>> INFO: Freed in sctp_association_put+0x150/0x250 age=14 cpu=1 pid=5664
>>>> [<  none  >] kfree+0x199/0x1b0 ./mm/slab.c:1211
>>>> [<  none  >] sctp_association_put+0x150/0x250
>>>> [<  none  >] sctp_association_free+0x498/0x630
>>>> [<  none  >] sctp_do_sm+0xd8b/0x4f60
>>>> [<  none  >] sctp_primitive_SHUTDOWN+0xa9/0xd0
>>>> [<  none  >] sctp_close+0x616/0x790
>>>> [<  none  >] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
>>>> [<  none  >] inet6_release+0x50/0x70 ./net/ipv6/af_inet6.c:416
>>>> [< inline >] constant_test_bit 
>>>> ././arch/x86/include/asm/bitops.h:321
>>>> [<  none  >] sock_release+0x8d/0x200 ./net/socket.c:601
>>>> [<  none  >] sock_close+0x16/0x20 ./

Re: use-after-free in sctp_do_sm

[Neil Horman]

On Tue, Nov 24, 2015 at 11:10:32AM +0100, Dmitry Vyukov wrote:
> On Tue, Nov 24, 2015 at 10:31 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
> > On Tue, Nov 24, 2015 at 10:15 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
> >> Hello,
> >>
> >> The following program triggers use-after-free in sctp_do_sm:
> >>
> >> // autogenerated by syzkaller (http://github.com/google/syzkaller)
> >> #include 
> >> #include 
> >> #include 
> >>
> >> int main()
> >> {
> >> long r0 = syscall(SYS_socket, 0xaul, 0x80805ul, 0x0ul, 0, 0, 0);
> >> long r1 = syscall(SYS_mmap, 0x2000ul, 0x1ul, 0x3ul,
> >> 0x32ul, 0xul, 0x0ul);
> >> memcpy((void*)0x20002fe4,
> >> "\x0a\x00\x33\xe7\xeb\x9d\xcf\x61\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xc5\xc8\x88\x64",
> >> 28);
> >> long r3 = syscall(SYS_bind, r0, 0x20002fe4ul, 0x1cul, 0, 0, 0);
> >> memcpy((void*)0x2faa,
> >> "\x9b\x01\x7d\xcd\xb8\x6a\xc7\x3d\x09\x3a\x07\x00\xa7\xc4\xe9\xee\x0a\xd6\xec\xde\x26\x75\x5f\x22\xae\x4e\x33\x00\xb0\x76\x10\x70\xd6\xca\x19\xbc\x15\x83\xcf\x2e\xbc\x99\x0c\x5e\x83\x89\xc1\x44\x9c\x6e\x74\xd8\x5d\x5d\xd0\xf0\xdf\x47\xc0\x00\x71\x0b\x55\x4c\xab\xf0\xd8\x90\xd5\x92\x8c\x6e\x33\x22\x15\x5b\x19\xfb\xed\xdd\xa6\xac\xcb\x60\xcf\xe2\xde\xed\xdb\x95\x5c\xaa\x20\xa3",
> >> 94);
> >> memcpy((void*)0x233a,
> >> "\x02\x00\x33\xe2\x7f\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00",
> >> 128);
> >> long r6 = syscall(SYS_sendto, r0, 0x2faaul, 0x5eul,
> >> 0x81ul, 0x233aul, 0x80ul);
> >> return 0;
> >> }
> >>
> >>
> >> ==
> >> BUG: KASAN: use-after-free in sctp_do_sm+0x42f6/0x4f60 at addr 
> >> 880036fa80a8
> >> Read of size 4 by task a.out/5664
> >> =
> >> BUG kmalloc-4096 (Tainted: GB  ): kasan: bad access detected
> >> -
> >>
> >> INFO: Allocated in sctp_association_new+0x6f/0x1ea0 age=8 cpu=1 pid=5664
> >> [<  none  >] kmem_cache_alloc_trace+0x1cf/0x220 ./mm/slab.c:3707
> >> [<  none  >] sctp_association_new+0x6f/0x1ea0
> >> [<  none  >] sctp_sendmsg+0x1954/0x28e0
> >> [<  none  >] inet_sendmsg+0x316/0x4f0 ./net/ipv4/af_inet.c:802
> >> [< inline >] __sock_sendmsg_nosec ./net/socket.c:641
> >> [< inline >] __sock_sendmsg ./net/socket.c:651
> >> [<  none  >] sock_sendmsg+0xca/0x110 ./net/socket.c:662
> >> [<  none  >] SYSC_sendto+0x208/0x350 ./net/socket.c:1841
> >> [<  none  >] SyS_sendto+0x40/0x50 ./net/socket.c:1862
> >> [<  none  >] entry_SYSCALL_64_fastpath+0x16/0x7a
> >>
> >> INFO: Freed in sctp_association_put+0x150/0x250 age=14 cpu=1 pid=5664
> >> [<  none  >] kfree+0x199/0x1b0 ./mm/slab.c:1211
> >> [<  none  >] sctp_association_put+0x150/0x250
> >> [<  none  >] sctp_association_free+0x498/0x630
> >> [<  none  >] sctp_do_sm+0xd8b/0x4f60
> >> [<  none  >] sctp_primitive_SHUTDOWN+0xa9/0xd0
> >> [<  none  >] sctp_close+0x616/0x790
> >> [<  none  >] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
> >> [<  none  >] inet6_release+0x50/0x70 ./net/ipv6/af_inet6.c:416
> >> [< inline >] constant_test_bit 
> >> ././arch/x86/include/asm/bitops.h:321
> >> [<  none  >] sock_release+0x8d/0x200 ./net/socket.c:601
> >> [<  none  >] sock_close+0x16/0x20 ./net/socket.c:1188
> >> [<  none  >] __fput+0x21d/0x6e0 ./fs/file_table.c:265
> >> [<  none  >] fput+0x15/0x20 ./fs/file_table.c:84
> >> [<  none  >] task_work_run+0x163/0

Re: use-after-free in sctp_do_sm

[Eric Dumazet]

On Tue, 2015-11-24 at 15:45 -0500, Neil Horman wrote:
> On Tue, Nov 24, 2015 at 11:10:32AM +0100, Dmitry Vyukov wrote:
> > On Tue, Nov 24, 2015 at 10:31 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
> > > On Tue, Nov 24, 2015 at 10:15 AM, Dmitry Vyukov <dvyu...@google.com> 
> > > wrote:
> > >> Hello,
> > >>
> > >> The following program triggers use-after-free in sctp_do_sm:
> > >>
> > >> // autogenerated by syzkaller (http://github.com/google/syzkaller)
> > >> #include 
> > >> #include 
> > >> #include 
> > >>
> > >> int main()
> > >> {
> > >> long r0 = syscall(SYS_socket, 0xaul, 0x80805ul, 0x0ul, 0, 0, 0);
> > >> long r1 = syscall(SYS_mmap, 0x2000ul, 0x1ul, 0x3ul,
> > >> 0x32ul, 0xul, 0x0ul);
> > >> memcpy((void*)0x20002fe4,
> > >> "\x0a\x00\x33\xe7\xeb\x9d\xcf\x61\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xc5\xc8\x88\x64",
> > >> 28);
> > >> long r3 = syscall(SYS_bind, r0, 0x20002fe4ul, 0x1cul, 0, 0, 0);
> > >> memcpy((void*)0x2faa,
> > >> "\x9b\x01\x7d\xcd\xb8\x6a\xc7\x3d\x09\x3a\x07\x00\xa7\xc4\xe9\xee\x0a\xd6\xec\xde\x26\x75\x5f\x22\xae\x4e\x33\x00\xb0\x76\x10\x70\xd6\xca\x19\xbc\x15\x83\xcf\x2e\xbc\x99\x0c\x5e\x83\x89\xc1\x44\x9c\x6e\x74\xd8\x5d\x5d\xd0\xf0\xdf\x47\xc0\x00\x71\x0b\x55\x4c\xab\xf0\xd8\x90\xd5\x92\x8c\x6e\x33\x22\x15\x5b\x19\xfb\xed\xdd\xa6\xac\xcb\x60\xcf\xe2\xde\xed\xdb\x95\x5c\xaa\x20\xa3",
> > >> 94);
> > >> memcpy((void*)0x233a,
> > >> "\x02\x00\x33\xe2\x7f\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00",
> > >> 128);
> > >> long r6 = syscall(SYS_sendto, r0, 0x2faaul, 0x5eul,
> > >> 0x81ul, 0x233aul, 0x80ul);
> > >> return 0;
> > >> }
> > >>
> > >>
> > >> ==
> > >> BUG: KASAN: use-after-free in sctp_do_sm+0x42f6/0x4f60 at addr 
> > >> 880036fa80a8
> > >> Read of size 4 by task a.out/5664
> > >> =
> > >> BUG kmalloc-4096 (Tainted: GB  ): kasan: bad access detected
> > >> -
> > >>
> > >> INFO: Allocated in sctp_association_new+0x6f/0x1ea0 age=8 cpu=1 pid=5664
> > >> [<  none  >] kmem_cache_alloc_trace+0x1cf/0x220 ./mm/slab.c:3707
> > >> [<  none  >] sctp_association_new+0x6f/0x1ea0
> > >> [<  none  >] sctp_sendmsg+0x1954/0x28e0
> > >> [<  none  >] inet_sendmsg+0x316/0x4f0 ./net/ipv4/af_inet.c:802
> > >> [< inline >] __sock_sendmsg_nosec ./net/socket.c:641
> > >> [< inline >] __sock_sendmsg ./net/socket.c:651
> > >> [<  none  >] sock_sendmsg+0xca/0x110 ./net/socket.c:662
> > >> [<  none  >] SYSC_sendto+0x208/0x350 ./net/socket.c:1841
> > >> [<  none  >] SyS_sendto+0x40/0x50 ./net/socket.c:1862
> > >> [<  none  >] entry_SYSCALL_64_fastpath+0x16/0x7a
> > >>
> > >> INFO: Freed in sctp_association_put+0x150/0x250 age=14 cpu=1 pid=5664
> > >> [<  none  >] kfree+0x199/0x1b0 ./mm/slab.c:1211
> > >> [<  none  >] sctp_association_put+0x150/0x250
> > >> [<  none  >] sctp_association_free+0x498/0x630
> > >> [<  none  >] sctp_do_sm+0xd8b/0x4f60
> > >> [<  none  >] sctp_primitive_SHUTDOWN+0xa9/0xd0
> > >> [<  none  >] sctp_close+0x616/0x790
> > >> [<  none  >] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
> > >> [<  none  >] inet6_release+0x50/0x70 ./net/ipv6/af_inet6.c:416
> > >> [< inline >] constant_test_bit 
> > >> ././arch/x86/include/asm/bitops.h:

Re: use-after-free in sctp_do_sm

[David Miller]

From: Neil Horman 
Date: Tue, 24 Nov 2015 15:45:54 -0500

>> The right commit is:
>> 
>> commit 7d267278a9ece963d77eefec61630223fce08c6c
>> Author: Rainer Weikusat
>> Date:   Fri Nov 20 22:07:23 2015 +
>> unix: avoid use-after-free in ep_remove_wait_queue
> This commit doesn't seem to exist

It's in the 'net' tree.  Which hasn't been pulled into 'net-next' for
a few days.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

use-after-free in sctp_do_sm

[Dmitry Vyukov]

Hello,

The following program triggers use-after-free in sctp_do_sm:

// autogenerated by syzkaller (http://github.com/google/syzkaller)
#include 
#include 
#include 

int main()
{
long r0 = syscall(SYS_socket, 0xaul, 0x80805ul, 0x0ul, 0, 0, 0);
long r1 = syscall(SYS_mmap, 0x2000ul, 0x1ul, 0x3ul,
0x32ul, 0xul, 0x0ul);
memcpy((void*)0x20002fe4,
"\x0a\x00\x33\xe7\xeb\x9d\xcf\x61\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xc5\xc8\x88\x64",
28);
long r3 = syscall(SYS_bind, r0, 0x20002fe4ul, 0x1cul, 0, 0, 0);
memcpy((void*)0x2faa,
"\x9b\x01\x7d\xcd\xb8\x6a\xc7\x3d\x09\x3a\x07\x00\xa7\xc4\xe9\xee\x0a\xd6\xec\xde\x26\x75\x5f\x22\xae\x4e\x33\x00\xb0\x76\x10\x70\xd6\xca\x19\xbc\x15\x83\xcf\x2e\xbc\x99\x0c\x5e\x83\x89\xc1\x44\x9c\x6e\x74\xd8\x5d\x5d\xd0\xf0\xdf\x47\xc0\x00\x71\x0b\x55\x4c\xab\xf0\xd8\x90\xd5\x92\x8c\x6e\x33\x22\x15\x5b\x19\xfb\xed\xdd\xa6\xac\xcb\x60\xcf\xe2\xde\xed\xdb\x95\x5c\xaa\x20\xa3",
94);
memcpy((void*)0x233a,
"\x02\x00\x33\xe2\x7f\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00",
128);
long r6 = syscall(SYS_sendto, r0, 0x2faaul, 0x5eul,
0x81ul, 0x233aul, 0x80ul);
return 0;
}


==========
BUG: KASAN: use-after-free in sctp_do_sm+0x42f6/0x4f60 at addr 880036fa80a8
Read of size 4 by task a.out/5664
=
BUG kmalloc-4096 (Tainted: GB  ): kasan: bad access detected
-

INFO: Allocated in sctp_association_new+0x6f/0x1ea0 age=8 cpu=1 pid=5664
[<  none  >] kmem_cache_alloc_trace+0x1cf/0x220 ./mm/slab.c:3707
[<  none  >] sctp_association_new+0x6f/0x1ea0
[<  none  >] sctp_sendmsg+0x1954/0x28e0
[<  none  >] inet_sendmsg+0x316/0x4f0 ./net/ipv4/af_inet.c:802
[< inline >] __sock_sendmsg_nosec ./net/socket.c:641
[< inline >] __sock_sendmsg ./net/socket.c:651
[<  none  >] sock_sendmsg+0xca/0x110 ./net/socket.c:662
[<  none  >] SYSC_sendto+0x208/0x350 ./net/socket.c:1841
[<  none  >] SyS_sendto+0x40/0x50 ./net/socket.c:1862
[<  none  >] entry_SYSCALL_64_fastpath+0x16/0x7a

INFO: Freed in sctp_association_put+0x150/0x250 age=14 cpu=1 pid=5664
[<  none  >] kfree+0x199/0x1b0 ./mm/slab.c:1211
[<  none  >] sctp_association_put+0x150/0x250
[<  none  >] sctp_association_free+0x498/0x630
[<  none  >] sctp_do_sm+0xd8b/0x4f60
[<  none  >] sctp_primitive_SHUTDOWN+0xa9/0xd0
[<  none  >] sctp_close+0x616/0x790
[<  none  >] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
[<  none  >] inet6_release+0x50/0x70 ./net/ipv6/af_inet6.c:416
[< inline >] constant_test_bit ././arch/x86/include/asm/bitops.h:321
[<  none  >] sock_release+0x8d/0x200 ./net/socket.c:601
[<  none  >] sock_close+0x16/0x20 ./net/socket.c:1188
[<  none  >] __fput+0x21d/0x6e0 ./fs/file_table.c:265
[<  none  >] fput+0x15/0x20 ./fs/file_table.c:84
[<  none  >] task_work_run+0x163/0x1f0 ./include/trace/events/rcu.h:20
[< inline >] __list_add ./include/linux/list.h:42
[< inline >] list_add_tail ./include/linux/list.h:76
[< inline >] list_move_tail ./include/linux/list.h:168
[< inline >] reparent_leader ./kernel/exit.c:618
[< inline >] forget_original_parent ./kernel/exit.c:669
[< inline >] exit_notify ./kernel/exit.c:697
[<  none  >] do_exit+0x809/0x2b90 ./kernel/exit.c:878
[<  none  >] do_group_exit+0x108/0x320 ./kernel/exit.c:985

INFO: Slab 0xeadbea00 objects=7 used=1 fp=0x880036fa8000
flags=0x1004080
INFO: Object 0x880036fa8000 @offset=0 fp=0x880036fad668
CPU: 1 PID: 5664 Comm: a.out Tainted: GB   4.4.0-rc1+ #81
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
  880061d6f700 825d3336 88003e806d00
 880036fa8000 880036fa8000 880061d6f730 81618784
 88003e806d00 eadbea00 880036fa8000 

Call Trace:
 [] __asan_report_load4_noabort+0x3e/0x40
 [] sctp_do_sm+0x4

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Tue, Nov 24, 2015 at 10:15 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
> Hello,
>
> The following program triggers use-after-free in sctp_do_sm:
>
> // autogenerated by syzkaller (http://github.com/google/syzkaller)
> #include 
> #include 
> #include 
>
> int main()
> {
> long r0 = syscall(SYS_socket, 0xaul, 0x80805ul, 0x0ul, 0, 0, 0);
> long r1 = syscall(SYS_mmap, 0x2000ul, 0x1ul, 0x3ul,
> 0x32ul, 0xul, 0x0ul);
> memcpy((void*)0x20002fe4,
> "\x0a\x00\x33\xe7\xeb\x9d\xcf\x61\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xc5\xc8\x88\x64",
> 28);
> long r3 = syscall(SYS_bind, r0, 0x20002fe4ul, 0x1cul, 0, 0, 0);
> memcpy((void*)0x2faa,
> "\x9b\x01\x7d\xcd\xb8\x6a\xc7\x3d\x09\x3a\x07\x00\xa7\xc4\xe9\xee\x0a\xd6\xec\xde\x26\x75\x5f\x22\xae\x4e\x33\x00\xb0\x76\x10\x70\xd6\xca\x19\xbc\x15\x83\xcf\x2e\xbc\x99\x0c\x5e\x83\x89\xc1\x44\x9c\x6e\x74\xd8\x5d\x5d\xd0\xf0\xdf\x47\xc0\x00\x71\x0b\x55\x4c\xab\xf0\xd8\x90\xd5\x92\x8c\x6e\x33\x22\x15\x5b\x19\xfb\xed\xdd\xa6\xac\xcb\x60\xcf\xe2\xde\xed\xdb\x95\x5c\xaa\x20\xa3",
> 94);
> memcpy((void*)0x233a,
> "\x02\x00\x33\xe2\x7f\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00",
> 128);
> long r6 = syscall(SYS_sendto, r0, 0x2faaul, 0x5eul,
> 0x81ul, 0x233aul, 0x80ul);
>     return 0;
> }
>
>
> ==
> BUG: KASAN: use-after-free in sctp_do_sm+0x42f6/0x4f60 at addr 
> 880036fa80a8
> Read of size 4 by task a.out/5664
> =
> BUG kmalloc-4096 (Tainted: GB  ): kasan: bad access detected
> -
>
> INFO: Allocated in sctp_association_new+0x6f/0x1ea0 age=8 cpu=1 pid=5664
> [<  none  >] kmem_cache_alloc_trace+0x1cf/0x220 ./mm/slab.c:3707
> [<  none  >] sctp_association_new+0x6f/0x1ea0
> [<  none  >] sctp_sendmsg+0x1954/0x28e0
> [<  none  >] inet_sendmsg+0x316/0x4f0 ./net/ipv4/af_inet.c:802
> [< inline >] __sock_sendmsg_nosec ./net/socket.c:641
> [< inline >] __sock_sendmsg ./net/socket.c:651
> [<  none  >] sock_sendmsg+0xca/0x110 ./net/socket.c:662
> [<  none  >] SYSC_sendto+0x208/0x350 ./net/socket.c:1841
> [<  none  >] SyS_sendto+0x40/0x50 ./net/socket.c:1862
> [<  none  >] entry_SYSCALL_64_fastpath+0x16/0x7a
>
> INFO: Freed in sctp_association_put+0x150/0x250 age=14 cpu=1 pid=5664
> [<  none  >] kfree+0x199/0x1b0 ./mm/slab.c:1211
> [<  none  >] sctp_association_put+0x150/0x250
> [<  none  >] sctp_association_free+0x498/0x630
> [<  none  >] sctp_do_sm+0xd8b/0x4f60
> [<  none  >] sctp_primitive_SHUTDOWN+0xa9/0xd0
> [<  none  >] sctp_close+0x616/0x790
> [<  none  >] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
> [<  none  >] inet6_release+0x50/0x70 ./net/ipv6/af_inet6.c:416
> [< inline >] constant_test_bit ././arch/x86/include/asm/bitops.h:321
> [<  none  >] sock_release+0x8d/0x200 ./net/socket.c:601
> [<  none  >] sock_close+0x16/0x20 ./net/socket.c:1188
> [<  none  >] __fput+0x21d/0x6e0 ./fs/file_table.c:265
> [<  none  >] fput+0x15/0x20 ./fs/file_table.c:84
> [<  none  >] task_work_run+0x163/0x1f0 ./include/trace/events/rcu.h:20
> [< inline >] __list_add ./include/linux/list.h:42
> [< inline >] list_add_tail ./include/linux/list.h:76
> [< inline >] list_move_tail ./include/linux/list.h:168
> [< inline >] reparent_leader ./kernel/exit.c:618
> [< inline >] forget_original_parent ./kernel/exit.c:669
> [< inline >] exit_notify ./kernel/exit.c:697
> [<  none  >] do_exit+0x809/0x2b90 ./kernel/exit.c:878
> [<  none  >] do_group_exit+0x108/0x320 ./kernel/exit.c:985
>
> INFO: Slab 0xeadbea00 objects=7 used=1 fp=0x880036fa8000
> flags=0x1004080
> INFO: Object 0x880036fa8

Re: use-after-free in sctp_do_sm

[Dmitry Vyukov]

On Tue, Nov 24, 2015 at 10:31 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
> On Tue, Nov 24, 2015 at 10:15 AM, Dmitry Vyukov <dvyu...@google.com> wrote:
>> Hello,
>>
>> The following program triggers use-after-free in sctp_do_sm:
>>
>> // autogenerated by syzkaller (http://github.com/google/syzkaller)
>> #include 
>> #include 
>> #include 
>>
>> int main()
>> {
>> long r0 = syscall(SYS_socket, 0xaul, 0x80805ul, 0x0ul, 0, 0, 0);
>> long r1 = syscall(SYS_mmap, 0x2000ul, 0x1ul, 0x3ul,
>> 0x32ul, 0xul, 0x0ul);
>> memcpy((void*)0x20002fe4,
>> "\x0a\x00\x33\xe7\xeb\x9d\xcf\x61\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x01\xc5\xc8\x88\x64",
>> 28);
>> long r3 = syscall(SYS_bind, r0, 0x20002fe4ul, 0x1cul, 0, 0, 0);
>> memcpy((void*)0x2faa,
>> "\x9b\x01\x7d\xcd\xb8\x6a\xc7\x3d\x09\x3a\x07\x00\xa7\xc4\xe9\xee\x0a\xd6\xec\xde\x26\x75\x5f\x22\xae\x4e\x33\x00\xb0\x76\x10\x70\xd6\xca\x19\xbc\x15\x83\xcf\x2e\xbc\x99\x0c\x5e\x83\x89\xc1\x44\x9c\x6e\x74\xd8\x5d\x5d\xd0\xf0\xdf\x47\xc0\x00\x71\x0b\x55\x4c\xab\xf0\xd8\x90\xd5\x92\x8c\x6e\x33\x22\x15\x5b\x19\xfb\xed\xdd\xa6\xac\xcb\x60\xcf\xe2\xde\xed\xdb\x95\x5c\xaa\x20\xa3",
>> 94);
>> memcpy((void*)0x233a,
>> "\x02\x00\x33\xe2\x7f\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00",
>> 128);
>> long r6 = syscall(SYS_sendto, r0, 0x20000faaul, 0x5eul,
>> 0x81ul, 0x233aul, 0x80ul);
>> return 0;
>> }
>>
>>
>> ==
>> BUG: KASAN: use-after-free in sctp_do_sm+0x42f6/0x4f60 at addr 
>> 880036fa80a8
>> Read of size 4 by task a.out/5664
>> =
>> BUG kmalloc-4096 (Tainted: GB  ): kasan: bad access detected
>> -
>>
>> INFO: Allocated in sctp_association_new+0x6f/0x1ea0 age=8 cpu=1 pid=5664
>> [<  none  >] kmem_cache_alloc_trace+0x1cf/0x220 ./mm/slab.c:3707
>> [<  none  >] sctp_association_new+0x6f/0x1ea0
>> [<  none  >] sctp_sendmsg+0x1954/0x28e0
>> [<  none  >] inet_sendmsg+0x316/0x4f0 ./net/ipv4/af_inet.c:802
>> [< inline >] __sock_sendmsg_nosec ./net/socket.c:641
>> [< inline >] __sock_sendmsg ./net/socket.c:651
>> [<  none  >] sock_sendmsg+0xca/0x110 ./net/socket.c:662
>> [<  none  >] SYSC_sendto+0x208/0x350 ./net/socket.c:1841
>> [<  none  >] SyS_sendto+0x40/0x50 ./net/socket.c:1862
>> [<  none  >] entry_SYSCALL_64_fastpath+0x16/0x7a
>>
>> INFO: Freed in sctp_association_put+0x150/0x250 age=14 cpu=1 pid=5664
>> [<  none  >] kfree+0x199/0x1b0 ./mm/slab.c:1211
>> [<  none  >] sctp_association_put+0x150/0x250
>> [<  none  >] sctp_association_free+0x498/0x630
>> [<  none  >] sctp_do_sm+0xd8b/0x4f60
>> [<  none  >] sctp_primitive_SHUTDOWN+0xa9/0xd0
>> [<  none  >] sctp_close+0x616/0x790
>> [<  none  >] inet_release+0xed/0x1c0 ./net/ipv4/af_inet.c:471
>> [<  none  >] inet6_release+0x50/0x70 ./net/ipv6/af_inet6.c:416
>> [< inline >] constant_test_bit ././arch/x86/include/asm/bitops.h:321
>> [<  none  >] sock_release+0x8d/0x200 ./net/socket.c:601
>> [<  none  >] sock_close+0x16/0x20 ./net/socket.c:1188
>> [<  none  >] __fput+0x21d/0x6e0 ./fs/file_table.c:265
>> [<  none  >] fput+0x15/0x20 ./fs/file_table.c:84
>> [<  none  >] task_work_run+0x163/0x1f0 
>> ./include/trace/events/rcu.h:20
>> [< inline >] __list_add ./include/linux/list.h:42
>> [< inline >] list_add_tail ./include/linux/list.h:76
>> [< inline >] list_move_tail ./include/linux/list.h:168
>> [< inline >] reparent_leader ./kernel/exit.c:618
>> [< inline >] forget_original_pare