make idr_remove_all() do removal -before- free_layer()
Subsystems (2)
PATCH 1/1 make idr_remove_all() do removal -before- free_layer()
The following patch fixes a problem in the IDR system, where an idr_remove_all() hands a data element to call_rcu() (via free_layer()) before making that data element inaccessible to new readers. This is very bad, and results in readers still having a reference to this data element at the end of the grace period. Tests on large machines that concurrently map and unmap user-space memory within the same multithreaded process result in crashes within about five minutes. Applying this patch increases the kernel's longevity to the three-to-eight-hour range.
There appear to be other similar problems in idr_get_empty_slot() and sub_remove(), but I fixed the easy one in idr_remove_all() first. It is therefore no surprise that failures still occur.
(Yes, and I did look at the relevant patch last year without spotting this one. Goes to show the value of testing as well as code review, I guess...)
Nadia, Manfred, any thoughts?
---
lib/idr.c +1 -1
@@ -449,6 +449,7 @@ void idr_remove_all(struct idr *idp) n = idp->layers *IDR_BITS ; p = idp->top; +rcu_assign_pointer (idp->top,NULL ); max =1 << n; id =0 ;
@@ -467,7 +468,6 @@ void idr_remove_all(struct idr *idp) p = *--paa; } } -rcu_assign_pointer (idp->top,NULL ); idp->layers =0 ; }EXPORT_SYMBOL (idr_remove_all);
----- End forwarded message -----
Quoted 20 lines · Paul E. McKenney
The following patch fixes a problem in the IDR system, where an
idr_remove_all()hands a data element tocall_rcu()(viafree_layer()) before making that data element inaccessible to new readers. This is very bad, and results in readers still having a reference to this data element at the end of the grace period. Tests on large machines that concurrently map and unmap user-space memory within the same multithreaded process result in crashes within about five minutes. Applying this patch increases the kernel's longevity to the three-to-eight-hour range.There appear to be other similar problems in
idr_get_empty_slot()andsub_remove(), but I fixed the easy one inidr_remove_all()first. It is therefore no surprise that failures still occur.(Yes, and I did look at the relevant patch last year without spotting this one. Goes to show the value of testing as well as code review, I guess...)
Quoted 5 lines
Nadia, Manfred, any thoughts?
Located-by Milton Miller II <[email protected]>Tested-by Milton Miller II <[email protected]>Signed-off-by Paul E. McKenney <[email protected]>
Hm, looks like something we really want to see fixed in 2.6.29-final, right?
Ingo
Quoted 23 lines · Ingo Molnar
Quoted 20 lines · Paul E. McKenney
The following patch fixes a problem in the IDR system, where an
idr_remove_all()hands a data element tocall_rcu()(viafree_layer()) before making that data element inaccessible to new readers. This is very bad, and results in readers still having a reference to this data element at the end of the grace period. Tests on large machines that concurrently map and unmap user-space memory within the same multithreaded process result in crashes within about five minutes. Applying this patch increases the kernel's longevity to the three-to-eight-hour range.There appear to be other similar problems in
idr_get_empty_slot()andsub_remove(), but I fixed the easy one inidr_remove_all()first. It is therefore no surprise that failures still occur.(Yes, and I did look at the relevant patch last year without spotting this one. Goes to show the value of testing as well as code review, I guess...)
Quoted 8 lines
Quoted 5 lines
Nadia, Manfred, any thoughts?
Located-by Milton Miller II <[email protected]>Tested-by Milton Miller II <[email protected]>Signed-off-by Paul E. McKenney <[email protected]>Hm, looks like something we really want to see fixed in 2.6.29-final, right?
This was located in real testing, so I agree that it is pretty high priority. So this patch should go into 2.6.29.
The priority of the remaining yet-as-unknown fixes depends on their complexity and risk.
Thanx, Paul