[PATCHv2 mm-stable] uprobes: Revert ref_ctr_offset in uprobe_unregister error path

[PATCHv2 mm-stable] uprobes: Revert ref_ctr_offset in uprobe_unregister error path

[Jiri Olsa]

From: Jiri Olsa <olsajiri@gmail.com>

There's error path that could lead to inactive uprobe:

  1) uprobe_register succeeds - updates instruction to int3 and
     changes ref_ctr from 0 to 1
  2) uprobe_unregister fails  - int3 stays in place, but ref_ctr
     is changed to 0 (it's not restored to 1 in the fail path)
     uprobe is leaked
  3) another uprobe_register comes and re-uses the leaked uprobe
     and succeds - but int3 is already in place, so ref_ctr update
     is skipped and it stays 0 - uprobe CAN NOT be triggered now
  4) uprobe_unregister fails because ref_ctr value is unexpected

Fixing this by reverting the updated ref_ctr value back to 1 in step 2),
which is the case when uprobe_unregister fails (int3 stays in place),
but we have already updated refctr.

The new scenario will go as follows:

  1) uprobe_register succeeds - updates instruction to int3 and
     changes ref_ctr from 0 to 1
  2) uprobe_unregister fails  - int3 stays in place and ref_ctr
     is reverted to 1..  uprobe is leaked
  3) another uprobe_register comes and re-uses the leaked uprobe
     and succeds - but int3 is already in place, so ref_ctr update
     is skipped and it stays 1 - uprobe CAN be triggered now
  4) uprobe_unregister succeeds

Fixes: 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")
Acked-by: David Hildenbrand <david@redhat.com>
Acked-by: Oleg Nesterov <oleg@redhat.com>
Suggested-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
---
v2 changes:
- adding proper Fixes tag and acks

 kernel/events/uprobes.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
index 4c965ba77f9f..84ee7b590861 100644
--- a/kernel/events/uprobes.c
+++ b/kernel/events/uprobes.c
@@ -581,8 +581,8 @@ int uprobe_write_opcode(struct arch_uprobe *auprobe, struct vm_area_struct *vma,
 
 out:
 	/* Revert back reference counter if instruction update failed. */
-	if (ret < 0 && is_register && ref_ctr_updated)
-		update_ref_ctr(uprobe, mm, -1);
+	if (ret < 0 && ref_ctr_updated)
+		update_ref_ctr(uprobe, mm, is_register ? -1 : 1);
 
 	/* try collapse pmd for compound page */
 	if (ret > 0)
-- 
2.49.0

Re: [PATCHv2 mm-stable] uprobes: Revert ref_ctr_offset in uprobe_unregister error path

[Jiri Olsa]

On Wed, May 14, 2025 at 12:18:09PM +0200, Jiri Olsa wrote:
> From: Jiri Olsa <olsajiri@gmail.com>
> 
> There's error path that could lead to inactive uprobe:
> 
>   1) uprobe_register succeeds - updates instruction to int3 and
>      changes ref_ctr from 0 to 1
>   2) uprobe_unregister fails  - int3 stays in place, but ref_ctr
>      is changed to 0 (it's not restored to 1 in the fail path)
>      uprobe is leaked
>   3) another uprobe_register comes and re-uses the leaked uprobe
>      and succeds - but int3 is already in place, so ref_ctr update
>      is skipped and it stays 0 - uprobe CAN NOT be triggered now
>   4) uprobe_unregister fails because ref_ctr value is unexpected
> 
> Fixing this by reverting the updated ref_ctr value back to 1 in step 2),
> which is the case when uprobe_unregister fails (int3 stays in place),
> but we have already updated refctr.
> 
> The new scenario will go as follows:
> 
>   1) uprobe_register succeeds - updates instruction to int3 and
>      changes ref_ctr from 0 to 1
>   2) uprobe_unregister fails  - int3 stays in place and ref_ctr
>      is reverted to 1..  uprobe is leaked
>   3) another uprobe_register comes and re-uses the leaked uprobe
>      and succeds - but int3 is already in place, so ref_ctr update
>      is skipped and it stays 1 - uprobe CAN be triggered now
>   4) uprobe_unregister succeeds
> 
> Fixes: 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")
> Acked-by: David Hildenbrand <david@redhat.com>
> Acked-by: Oleg Nesterov <oleg@redhat.com>
> Suggested-by: Oleg Nesterov <oleg@redhat.com>
> Signed-off-by: Jiri Olsa <jolsa@kernel.org>

hi,
I can't find this in any related tree, was this pulled in?

thanks,
jirka


> ---
> v2 changes:
> - adding proper Fixes tag and acks
> 
>  kernel/events/uprobes.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
> index 4c965ba77f9f..84ee7b590861 100644
> --- a/kernel/events/uprobes.c
> +++ b/kernel/events/uprobes.c
> @@ -581,8 +581,8 @@ int uprobe_write_opcode(struct arch_uprobe *auprobe, struct vm_area_struct *vma,
>  
>  out:
>  	/* Revert back reference counter if instruction update failed. */
> -	if (ret < 0 && is_register && ref_ctr_updated)
> -		update_ref_ctr(uprobe, mm, -1);
> +	if (ret < 0 && ref_ctr_updated)
> +		update_ref_ctr(uprobe, mm, is_register ? -1 : 1);
>  
>  	/* try collapse pmd for compound page */
>  	if (ret > 0)
> -- 
> 2.49.0
> 

Re: [PATCHv2 mm-stable] uprobes: Revert ref_ctr_offset in uprobe_unregister error path

[Jiri Olsa]

hi, ping ;-)

On Wed, Jun 04, 2025 at 10:28:42PM +0200, Jiri Olsa wrote:
> On Wed, May 14, 2025 at 12:18:09PM +0200, Jiri Olsa wrote:
> > From: Jiri Olsa <olsajiri@gmail.com>
> > 
> > There's error path that could lead to inactive uprobe:
> > 
> >   1) uprobe_register succeeds - updates instruction to int3 and
> >      changes ref_ctr from 0 to 1
> >   2) uprobe_unregister fails  - int3 stays in place, but ref_ctr
> >      is changed to 0 (it's not restored to 1 in the fail path)
> >      uprobe is leaked
> >   3) another uprobe_register comes and re-uses the leaked uprobe
> >      and succeds - but int3 is already in place, so ref_ctr update
> >      is skipped and it stays 0 - uprobe CAN NOT be triggered now
> >   4) uprobe_unregister fails because ref_ctr value is unexpected
> > 
> > Fixing this by reverting the updated ref_ctr value back to 1 in step 2),
> > which is the case when uprobe_unregister fails (int3 stays in place),
> > but we have already updated refctr.
> > 
> > The new scenario will go as follows:
> > 
> >   1) uprobe_register succeeds - updates instruction to int3 and
> >      changes ref_ctr from 0 to 1
> >   2) uprobe_unregister fails  - int3 stays in place and ref_ctr
> >      is reverted to 1..  uprobe is leaked
> >   3) another uprobe_register comes and re-uses the leaked uprobe
> >      and succeds - but int3 is already in place, so ref_ctr update
> >      is skipped and it stays 1 - uprobe CAN be triggered now
> >   4) uprobe_unregister succeeds
> > 
> > Fixes: 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")
> > Acked-by: David Hildenbrand <david@redhat.com>
> > Acked-by: Oleg Nesterov <oleg@redhat.com>
> > Suggested-by: Oleg Nesterov <oleg@redhat.com>
> > Signed-off-by: Jiri Olsa <jolsa@kernel.org>
> 
> hi,
> I can't find this in any related tree, was this pulled in?
> 
> thanks,
> jirka
> 
> 
> > ---
> > v2 changes:
> > - adding proper Fixes tag and acks
> > 
> >  kernel/events/uprobes.c | 4 ++--
> >  1 file changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
> > index 4c965ba77f9f..84ee7b590861 100644
> > --- a/kernel/events/uprobes.c
> > +++ b/kernel/events/uprobes.c
> > @@ -581,8 +581,8 @@ int uprobe_write_opcode(struct arch_uprobe *auprobe, struct vm_area_struct *vma,
> >  
> >  out:
> >  	/* Revert back reference counter if instruction update failed. */
> > -	if (ret < 0 && is_register && ref_ctr_updated)
> > -		update_ref_ctr(uprobe, mm, -1);
> > +	if (ret < 0 && ref_ctr_updated)
> > +		update_ref_ctr(uprobe, mm, is_register ? -1 : 1);
> >  
> >  	/* try collapse pmd for compound page */
> >  	if (ret > 0)
> > -- 
> > 2.49.0
> > 

Re: [PATCHv2 mm-stable] uprobes: Revert ref_ctr_offset in uprobe_unregister error path

[Jiri Olsa]

ping

On Wed, Jun 11, 2025 at 10:57:08AM +0200, Jiri Olsa wrote:
> hi, ping ;-)
> 
> On Wed, Jun 04, 2025 at 10:28:42PM +0200, Jiri Olsa wrote:
> > On Wed, May 14, 2025 at 12:18:09PM +0200, Jiri Olsa wrote:
> > > From: Jiri Olsa <olsajiri@gmail.com>
> > > 
> > > There's error path that could lead to inactive uprobe:
> > > 
> > >   1) uprobe_register succeeds - updates instruction to int3 and
> > >      changes ref_ctr from 0 to 1
> > >   2) uprobe_unregister fails  - int3 stays in place, but ref_ctr
> > >      is changed to 0 (it's not restored to 1 in the fail path)
> > >      uprobe is leaked
> > >   3) another uprobe_register comes and re-uses the leaked uprobe
> > >      and succeds - but int3 is already in place, so ref_ctr update
> > >      is skipped and it stays 0 - uprobe CAN NOT be triggered now
> > >   4) uprobe_unregister fails because ref_ctr value is unexpected
> > > 
> > > Fixing this by reverting the updated ref_ctr value back to 1 in step 2),
> > > which is the case when uprobe_unregister fails (int3 stays in place),
> > > but we have already updated refctr.
> > > 
> > > The new scenario will go as follows:
> > > 
> > >   1) uprobe_register succeeds - updates instruction to int3 and
> > >      changes ref_ctr from 0 to 1
> > >   2) uprobe_unregister fails  - int3 stays in place and ref_ctr
> > >      is reverted to 1..  uprobe is leaked
> > >   3) another uprobe_register comes and re-uses the leaked uprobe
> > >      and succeds - but int3 is already in place, so ref_ctr update
> > >      is skipped and it stays 1 - uprobe CAN be triggered now
> > >   4) uprobe_unregister succeeds
> > > 
> > > Fixes: 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")
> > > Acked-by: David Hildenbrand <david@redhat.com>
> > > Acked-by: Oleg Nesterov <oleg@redhat.com>
> > > Suggested-by: Oleg Nesterov <oleg@redhat.com>
> > > Signed-off-by: Jiri Olsa <jolsa@kernel.org>
> > 
> > hi,
> > I can't find this in any related tree, was this pulled in?
> > 
> > thanks,
> > jirka
> > 
> > 
> > > ---
> > > v2 changes:
> > > - adding proper Fixes tag and acks
> > > 
> > >  kernel/events/uprobes.c | 4 ++--
> > >  1 file changed, 2 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/kernel/events/uprobes.c b/kernel/events/uprobes.c
> > > index 4c965ba77f9f..84ee7b590861 100644
> > > --- a/kernel/events/uprobes.c
> > > +++ b/kernel/events/uprobes.c
> > > @@ -581,8 +581,8 @@ int uprobe_write_opcode(struct arch_uprobe *auprobe, struct vm_area_struct *vma,
> > >  
> > >  out:
> > >  	/* Revert back reference counter if instruction update failed. */
> > > -	if (ret < 0 && is_register && ref_ctr_updated)
> > > -		update_ref_ctr(uprobe, mm, -1);
> > > +	if (ret < 0 && ref_ctr_updated)
> > > +		update_ref_ctr(uprobe, mm, is_register ? -1 : 1);
> > >  
> > >  	/* try collapse pmd for compound page */
> > >  	if (ret > 0)
> > > -- 
> > > 2.49.0
> > > 

Re: [PATCHv2 mm-stable] uprobes: Revert ref_ctr_offset in uprobe_unregister error path

[David Hildenbrand]

On 17.06.25 13:58, Jiri Olsa wrote:
> ping
> 
> On Wed, Jun 11, 2025 at 10:57:08AM +0200, Jiri Olsa wrote:
>> hi, ping ;-)
>>
>> On Wed, Jun 04, 2025 at 10:28:42PM +0200, Jiri Olsa wrote:
>>> On Wed, May 14, 2025 at 12:18:09PM +0200, Jiri Olsa wrote:
>>>> From: Jiri Olsa <olsajiri@gmail.com>
>>>>
>>>> There's error path that could lead to inactive uprobe:
>>>>
>>>>    1) uprobe_register succeeds - updates instruction to int3 and
>>>>       changes ref_ctr from 0 to 1
>>>>    2) uprobe_unregister fails  - int3 stays in place, but ref_ctr
>>>>       is changed to 0 (it's not restored to 1 in the fail path)
>>>>       uprobe is leaked
>>>>    3) another uprobe_register comes and re-uses the leaked uprobe
>>>>       and succeds - but int3 is already in place, so ref_ctr update
>>>>       is skipped and it stays 0 - uprobe CAN NOT be triggered now
>>>>    4) uprobe_unregister fails because ref_ctr value is unexpected
>>>>
>>>> Fixing this by reverting the updated ref_ctr value back to 1 in step 2),
>>>> which is the case when uprobe_unregister fails (int3 stays in place),
>>>> but we have already updated refctr.
>>>>
>>>> The new scenario will go as follows:
>>>>
>>>>    1) uprobe_register succeeds - updates instruction to int3 and
>>>>       changes ref_ctr from 0 to 1
>>>>    2) uprobe_unregister fails  - int3 stays in place and ref_ctr
>>>>       is reverted to 1..  uprobe is leaked
>>>>    3) another uprobe_register comes and re-uses the leaked uprobe
>>>>       and succeds - but int3 is already in place, so ref_ctr update
>>>>       is skipped and it stays 1 - uprobe CAN be triggered now
>>>>    4) uprobe_unregister succeeds
>>>>
>>>> Fixes: 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")
>>>> Acked-by: David Hildenbrand <david@redhat.com>
>>>> Acked-by: Oleg Nesterov <oleg@redhat.com>
>>>> Suggested-by: Oleg Nesterov <oleg@redhat.com>
>>>> Signed-off-by: Jiri Olsa <jolsa@kernel.org>
>>>
>>> hi,
>>> I can't find this in any related tree, was this pulled in?

@Andrew, I assume you should pick this.

-- 
Cheers,

David / dhildenb