x86/resctrl: Fix arch_mbm_* array overrun on SNC
Subsystems (2)
PATCH 1/1 x86/resctrl: Fix arch_mbm_* array overrun on SNC
When using resctrl on systems with Sub-NUMA Clustering enabled, monitoring groups may be allocated RMID values which would overrun the arch_mbm_{local,total} arrays.
This is due to inconsistencies in whether the SNC-adjusted num_rmid value or the unadjusted value in resctrl_arch_system_num_rmid_idx() is used. The num_rmid value for the L3 resource is currently:
resctrl_arch_system_num_rmid_idx() / snc_nodes_per_l3_cache
As a simple fix, make resctrl_arch_system_num_rmid_idx() return the SNC-adjusted, L3 num_rmid value on x86.
--- arch/x86/include/asm/resctrl.h | 6 ------ arch/x86/kernel/cpu/resctrl/core.c | 8 ++++++++ include/linux/resctrl.h | 3 +++ 3 files changed, 11 insertions(+), 6 deletions(-)
arch/x86/include/asm/resctrl.h +0 -6
@@ -156,12 +156,6 @@ static inline void resctrl_sched_in(struct task_struct *tsk)__resctrl_sched_in (tsk); } -static inline u32 resctrl_arch_system_num_rmid_idx (void ) -{ -/* RMID are independent numbers for x86. num_rmid_idx == num_rmid */ -return boot_cpu_data.x86_cache_max_rmid +1 ; -} -static inline void resctrl_arch_rmid_idx_decode (u32 idx,u32 *closid,u32 *rmid) { *rmid = idx;
arch/x86/kernel/cpu/resctrl/core.c +8 -0
@@ -119,6 +119,14 @@ struct rdt_hw_resource rdt_resources_all[] = { }, }; +u32 resctrl_arch_system_num_rmid_idx (void ) +{ +struct rdt_resource *r = &rdt_resources_all[RDT_RESOURCE_L3 ].r_resctrl; + +/* RMID are independent numbers for x86. num_rmid_idx == num_rmid */ +return r->num_rmid; +} +/* * cache_alloc_hsw_probe() - Have to probe for Intel haswell server CPUs * as they do not have CPUID enumeration support for Cache allocation.
include/linux/resctrl.h +3 -0
@@ -248,6 +248,9 @@ struct resctrl_schema {/* The number of closid supported by this resource regardless of CDP */ u32 resctrl_arch_get_num_closid (struct rdt_resource *r); + +u32 resctrl_arch_system_num_rmid_idx (void ); +int resctrl_arch_update_domains (struct rdt_resource *r,u32 closid);/*
Signature · 1 line
2.45.2.1089.g2a221341d9-goog
Reinette Chatre
maintainerHi Peter,
Quoted 4 lines · Peter Newman
When using resctrl on systems with Sub-NUMA Clustering enabled, monitoring groups may be allocated RMID values which would overrun the arch_mbm_{local,total} arrays.
Quoted 9 lines
This is due to inconsistencies in whether the SNC-adjusted
num_rmidvalue or the unadjusted value inresctrl_arch_system_num_rmid_idx()is used. Thenum_rmidvalue for the L3 resource is currently:
resctrl_arch_system_num_rmid_idx()/snc_nodes_per_l3_cacheAs a simple fix, make
resctrl_arch_system_num_rmid_idx()return the SNC-adjusted, L3num_rmidvalue on x86.
Thank you very much for finding, root-causing, and providing a fix for the issue.
Quoted 49 lines
Fixes e13db55b5a0d ("x86/resctrl: Introduce snc_nodes_per_l3_cache")Signed-off-by Peter Newman <[email protected]>--- arch/x86/include/asm/resctrl.h | 6 ------ arch/x86/kernel/cpu/resctrl/core.c | 8 ++++++++ include/linux/resctrl.h | 3 +++ 3 files changed, 11 insertions(+), 6 deletions(-)arch/x86/include/asm/resctrl.h +0
-6@@ -156,12 +156,6 @@ static inline void resctrl_sched_in(struct task_struct *tsk)__resctrl_sched_in (tsk); } -static inline u32 resctrl_arch_system_num_rmid_idx (void ) -{ -/* RMID are independent numbers for x86. num_rmid_idx == num_rmid */ -return boot_cpu_data.x86_cache_max_rmid +1 ; -} -static inline void resctrl_arch_rmid_idx_decode (u32 idx,u32 *closid,u32 *rmid) { *rmid = idx;arch/x86/kernel/cpu/resctrl/core.c +8
-0@@ -119,6 +119,14 @@ struct rdt_hw_resource rdt_resources_all[] = { }, }; +u32 resctrl_arch_system_num_rmid_idx (void ) +{ +struct rdt_resource *r = &rdt_resources_all[RDT_RESOURCE_L3 ].r_resctrl; + +/* RMID are independent numbers for x86. num_rmid_idx == num_rmid */ +return r->num_rmid; +} +/* * cache_alloc_hsw_probe() - Have to probe for Intel haswell server CPUs * as they do not have CPUID enumeration support for Cache allocation. include/linux/resctrl.h +0
-0
include/linux/resctrl.h struct resctrl_schema { · line 253
Quoted 6 lines
include/linux/resctrl.h +3
-0@@ -248,6 +248,9 @@ struct resctrl_schema {/* The number of closid supported by this resource regardless of CDP */ u32 resctrl_arch_get_num_closid (struct rdt_resource *r); + +u32 resctrl_arch_system_num_rmid_idx (void ); +
nit: the additional empty lines are unnecessary.
include/linux/resctrl.h struct resctrl_schema { · line 256
Quoted 3 lines
include/linux/resctrl.h +0
-0int resctrl_arch_update_domains (struct rdt_resource *r,u32 closid);/*
| Reviewed-by: Reinette Chatre <[email protected]>
Reinette
Reinette Chatre
maintainerHi Peter,
Quoted 73 lines · Reinette Chatre
Quoted 4 lines · Peter Newman
When using resctrl on systems with Sub-NUMA Clustering enabled, monitoring groups may be allocated RMID values which would overrun the arch_mbm_{local,total} arrays.
Quoted 9 lines
This is due to inconsistencies in whether the SNC-adjusted
num_rmidvalue or the unadjusted value inresctrl_arch_system_num_rmid_idx()is used. Thenum_rmidvalue for the L3 resource is currently:
resctrl_arch_system_num_rmid_idx()/snc_nodes_per_l3_cacheAs a simple fix, make
resctrl_arch_system_num_rmid_idx()return the SNC-adjusted, L3num_rmidvalue on x86.Thank you very much for finding, root-causing, and providing a fix for the issue.
Quoted 47 lines
Fixes e13db55b5a0d ("x86/resctrl: Introduce snc_nodes_per_l3_cache")Signed-off-by Peter Newman <[email protected]>---
arch/x86/include/asm/resctrl.h | 6 ------
arch/x86/kernel/cpu/resctrl/core.c| 8 ++++++++
include/linux/resctrl.h | 3 +++3 files changed, 11 insertions(+), 6 deletions(-)arch/x86/include/asm/resctrl.h +0
-6@@ -156,12 +156,6 @@ static inline void resctrl_sched_in(struct task_struct *tsk) __resctrl_sched_in(tsk); } -static inline u32 resctrl_arch_system_num_rmid_idx (void ) -{ - /* RMID are independent numbers for x86. num_rmid_idx == num_rmid */ - return boot_cpu_data.x86_cache_max_rmid + 1; -} - static inline void resctrl_arch_rmid_idx_decode(u32 idx, u32 *closid, u32 *rmid) { *rmid = idx;arch/x86/kernel/cpu/resctrl/core.c +8
-0@@ -119,6 +119,14 @@ struct rdt_hw_resource rdt_resources_all[] = { }, }; +u32 resctrl_arch_system_num_rmid_idx (void ) +{ + struct rdt_resource *r = &rdt_resources_all[RDT_RESOURCE_L3].r_resctrl; + + /* RMID are independent numbers for x86. num_rmid_idx == num_rmid */ + return r->num_rmid; +} + /** cache_alloc_hsw_probe() - Have to probe for Intel haswell server CPUs * as they do not have CPUID enumeration support for Cache allocation. include/linux/resctrl.h +0
-0include/linux/resctrl.h struct resctrl_schema { · line 252
Quoted 5 lines
include/linux/resctrl.h +3
-0@@ -248,6 +248,9 @@ struct resctrl_schema { /* The number of closid supported by this resource regardless of CDP */ u32 resctrl_arch_get_num_closid(struct rdt_resource *r); + +u32 resctrl_arch_system_num_rmid_idx (void ); +nit: the additional empty lines are unnecessary.
Quoted 5 lines
include/linux/resctrl.h struct resctrl_schema { · line 256
Quoted 2 lines
include/linux/resctrl.h +0
-0int resctrl_arch_update_domains(struct rdt_resource *r, u32 closid); /*| Reviewed-by: Reinette Chatre <[email protected]>
It would be great if this fix can be included in the kernel release. Could you please send V2 with nit and tag applied to be ready for inclusion?
Thank you
Reinette