uprobes: Revert ref_ctr_offset in uprobe_unregister error path
Subsystems (3)
Jiri Olsa
reviewerPATCH v2 1/1 uprobes: Revert ref_ctr_offset in uprobe_unregister error path
From: Jiri Olsa <[email protected]>
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
---
v2 changes:
- adding proper Fixes tag and acks
kernel/events/uprobes.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/events/uprobes.c +2 -2
@@ -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 )
Signature · 1 line
2.49.0
Quoted 29 lines · Jiri Olsa
From: Jiri Olsa <[email protected]>
There's error path that could lead to inactive uprobe:
1)
uprobe_registersucceeds - updates instruction to int3 and changesref_ctrfrom 0 to 1
2)uprobe_unregisterfails - int3 stays in place, butref_ctris changed to 0 (it's not restored to 1 in the fail path) uprobe is leaked
3) anotheruprobe_registercomes and re-uses the leaked uprobe and succeds - but int3 is already in place, soref_ctrupdate is skipped and it stays 0 - uprobe CAN NOT be triggered now
4)uprobe_unregisterfails becauseref_ctrvalue is unexpectedFixing this by reverting the updated
ref_ctrvalue back to 1 in step 2), which is the case whenuprobe_unregisterfails (int3 stays in place), but we have already updated refctr.The new scenario will go as follows:
1)
uprobe_registersucceeds - updates instruction to int3 and changesref_ctrfrom 0 to 1
2)uprobe_unregisterfails - int3 stays in place andref_ctris reverted to 1.. uprobe is leaked
3) anotheruprobe_registercomes and re-uses the leaked uprobe and succeds - but int3 is already in place, soref_ctrupdate is skipped and it stays 1 - uprobe CAN be triggered now
4)uprobe_unregistersucceeds
Quoted 5 lines
Fixes 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")Acked-by Oleg Nesterov <[email protected]>Suggested-by Oleg Nesterov <[email protected]>Signed-off-by Jiri Olsa <[email protected]>
hi,
I can't find this in any related tree, was this pulled in?
thanks,
jirka
Quoted 25 lines
---
v2 changes:
- adding proper Fixes tag and ackskernel/events/uprobes.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)kernel/events/uprobes.c +2
-2@@ -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 )Signature · 1 line
2.49.0
hi, ping ;-)
Quoted 68 lines · Jiri Olsa
Quoted 29 lines · Jiri Olsa
From: Jiri Olsa <[email protected]>
There's error path that could lead to inactive uprobe:
1)
uprobe_registersucceeds - updates instruction to int3 and changesref_ctrfrom 0 to 1
2)uprobe_unregisterfails - int3 stays in place, butref_ctris changed to 0 (it's not restored to 1 in the fail path) uprobe is leaked
3) anotheruprobe_registercomes and re-uses the leaked uprobe and succeds - but int3 is already in place, soref_ctrupdate is skipped and it stays 0 - uprobe CAN NOT be triggered now
4)uprobe_unregisterfails becauseref_ctrvalue is unexpectedFixing this by reverting the updated
ref_ctrvalue back to 1 in step 2), which is the case whenuprobe_unregisterfails (int3 stays in place), but we have already updated refctr.The new scenario will go as follows:
1)
uprobe_registersucceeds - updates instruction to int3 and changesref_ctrfrom 0 to 1
2)uprobe_unregisterfails - int3 stays in place andref_ctris reverted to 1.. uprobe is leaked
3) anotheruprobe_registercomes and re-uses the leaked uprobe and succeds - but int3 is already in place, soref_ctrupdate is skipped and it stays 1 - uprobe CAN be triggered now
4)uprobe_unregistersucceedsQuoted 5 lines
Fixes 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")Acked-by Oleg Nesterov <[email protected]>Suggested-by Oleg Nesterov <[email protected]>Signed-off-by Jiri Olsa <[email protected]>hi,
I can't find this in any related tree, was this pulled in?thanks,
jirkaQuoted 25 lines
---
v2 changes:
- adding proper Fixes tag and ackskernel/events/uprobes.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)kernel/events/uprobes.c +2
-2@@ -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 )Signature · 1 line
2.49.0
ping
Quoted 71 lines · Jiri Olsa
hi, ping ;-)
Quoted 68 lines · Jiri Olsa
Quoted 29 lines · Jiri Olsa
From: Jiri Olsa <[email protected]>
There's error path that could lead to inactive uprobe:
1)
uprobe_registersucceeds - updates instruction to int3 and changesref_ctrfrom 0 to 1
2)uprobe_unregisterfails - int3 stays in place, butref_ctris changed to 0 (it's not restored to 1 in the fail path) uprobe is leaked
3) anotheruprobe_registercomes and re-uses the leaked uprobe and succeds - but int3 is already in place, soref_ctrupdate is skipped and it stays 0 - uprobe CAN NOT be triggered now
4)uprobe_unregisterfails becauseref_ctrvalue is unexpectedFixing this by reverting the updated
ref_ctrvalue back to 1 in step 2), which is the case whenuprobe_unregisterfails (int3 stays in place), but we have already updated refctr.The new scenario will go as follows:
1)
uprobe_registersucceeds - updates instruction to int3 and changesref_ctrfrom 0 to 1
2)uprobe_unregisterfails - int3 stays in place andref_ctris reverted to 1.. uprobe is leaked
3) anotheruprobe_registercomes and re-uses the leaked uprobe and succeds - but int3 is already in place, soref_ctrupdate is skipped and it stays 1 - uprobe CAN be triggered now
4)uprobe_unregistersucceedsQuoted 5 lines
Fixes 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")Acked-by Oleg Nesterov <[email protected]>Suggested-by Oleg Nesterov <[email protected]>Signed-off-by Jiri Olsa <[email protected]>hi,
I can't find this in any related tree, was this pulled in?thanks,
jirkaQuoted 25 lines
---
v2 changes:
- adding proper Fixes tag and ackskernel/events/uprobes.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)kernel/events/uprobes.c +2
-2@@ -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 )Signature · 1 line
2.49.0
Quoted 44 lines · Jiri Olsa
ping
Quoted 41 lines · Jiri Olsa
hi, ping ;-)
Quoted 38 lines · Jiri Olsa
Quoted 29 lines · Jiri Olsa
From: Jiri Olsa <[email protected]>
There's error path that could lead to inactive uprobe:
1)
uprobe_registersucceeds - updates instruction to int3 and changesref_ctrfrom 0 to 1
2)uprobe_unregisterfails - int3 stays in place, butref_ctris changed to 0 (it's not restored to 1 in the fail path) uprobe is leaked
3) anotheruprobe_registercomes and re-uses the leaked uprobe and succeds - but int3 is already in place, soref_ctrupdate is skipped and it stays 0 - uprobe CAN NOT be triggered now
4)uprobe_unregisterfails becauseref_ctrvalue is unexpectedFixing this by reverting the updated
ref_ctrvalue back to 1 in step 2), which is the case whenuprobe_unregisterfails (int3 stays in place), but we have already updated refctr.The new scenario will go as follows:
1)
uprobe_registersucceeds - updates instruction to int3 and changesref_ctrfrom 0 to 1
2)uprobe_unregisterfails - int3 stays in place andref_ctris reverted to 1.. uprobe is leaked
3) anotheruprobe_registercomes and re-uses the leaked uprobe and succeds - but int3 is already in place, soref_ctrupdate is skipped and it stays 1 - uprobe CAN be triggered now
4)uprobe_unregistersucceedsQuoted 5 lines
Fixes 1cc33161a83d ("uprobes: Support SDT markers having reference count (semaphore)")Acked-by Oleg Nesterov <[email protected]>Suggested-by Oleg Nesterov <[email protected]>Signed-off-by Jiri Olsa <[email protected]>hi,
I can't find this in any related tree, was this pulled in?
@Andrew, I assume you should pick this.
Signature · 3 lines
Cheers, David / dhildenb