[PATCH OLK-6.6 0/7] mm/numa_remote: fix NUMA.remote and DRAM in the same memory tier
Ho-Ren (Jack) Chuang (3): memory tier: dax/kmem: introduce an abstract layer for finding, allocating, and putting memory types memory tier: create CPUless memory tiers after obtaining HMAT info memory tier: consolidate the initialization of memory tiers Jinjiang Tu (1): mm/numa_remote: fix NUMA.remote and DRAM in the same memory tier Li Zhijian (2): acpi,mm: fix typo sibiling -> sibling mm/demotion: print demotion targets Yanfei Xu (1): memory tier: fix deadlock warning while onlining pages drivers/acpi/acpi_pad.c | 2 +- drivers/acpi/numa/hmat.c | 5 +- drivers/base/numa_remote.c | 23 +++++ drivers/dax/kmem.c | 30 +----- include/linux/memory-tiers.h | 17 +++- mm/memory-tiers.c | 180 ++++++++++++++++++++++++++--------- 6 files changed, 178 insertions(+), 79 deletions(-) -- 2.43.0
From: Li Zhijian <lizhijian@cn.fujitsu.com> mainline inclusion from mainline-v6.7-rc1 commit 51a23b1be92046f0cc52384d30cf060700f1a54e category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9812 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... ------------------------------------------- First found this typo as reviewing memory tier code. Fix it by sed like: $ sed -i 's/sibiling/sibling/g' $(git grep -l sibiling) so the acpi one will be corrected as well. Link: https://lkml.kernel.org/r/20230802092856.819328-1-lizhijian@cn.fujitsu.com Signed-off-by: Li Zhijian <lizhijian@cn.fujitsu.com> Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Cc: Huang, Ying <ying.huang@intel.com> Cc: Len Brown <lenb@kernel.org> Cc: "Rafael J. Wysocki" <rafael@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jinjiang Tu <tujinjiang@huawei.com> --- drivers/acpi/acpi_pad.c | 2 +- include/linux/memory-tiers.h | 2 +- mm/memory-tiers.c | 10 +++++----- 3 files changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/acpi/acpi_pad.c b/drivers/acpi/acpi_pad.c index 71e25c798976..b130acb39723 100644 --- a/drivers/acpi/acpi_pad.c +++ b/drivers/acpi/acpi_pad.c @@ -100,7 +100,7 @@ static void round_robin_cpu(unsigned int tsk_index) for_each_cpu(cpu, pad_busy_cpus) cpumask_or(tmp, tmp, topology_sibling_cpumask(cpu)); cpumask_andnot(tmp, cpu_online_mask, tmp); - /* avoid HT sibilings if possible */ + /* avoid HT siblings if possible */ if (cpumask_empty(tmp)) cpumask_andnot(tmp, cpu_online_mask, pad_busy_cpus); if (cpumask_empty(tmp)) { diff --git a/include/linux/memory-tiers.h b/include/linux/memory-tiers.h index c358c215f9cd..6debda505b51 100644 --- a/include/linux/memory-tiers.h +++ b/include/linux/memory-tiers.h @@ -23,7 +23,7 @@ struct memory_tier; struct memory_dev_type { /* list of memory types that are part of same tier as this type */ - struct list_head tier_sibiling; + struct list_head tier_sibling; /* list of memory types that are managed by one driver */ struct list_head list; /* abstract distance for this specific memory type */ diff --git a/mm/memory-tiers.c b/mm/memory-tiers.c index dadc56d2b1a2..0ec9ea04375a 100644 --- a/mm/memory-tiers.c +++ b/mm/memory-tiers.c @@ -142,7 +142,7 @@ static __always_inline nodemask_t get_memtier_nodemask(struct memory_tier *memti nodemask_t nodes = NODE_MASK_NONE; struct memory_dev_type *memtype; - list_for_each_entry(memtype, &memtier->memory_types, tier_sibiling) + list_for_each_entry(memtype, &memtier->memory_types, tier_sibling) nodes_or(nodes, nodes, memtype->nodes); return nodes; @@ -201,7 +201,7 @@ static struct memory_tier *find_create_memory_tier(struct memory_dev_type *memty * If the memtype is already part of a memory tier, * just return that. */ - if (!list_empty(&memtype->tier_sibiling)) { + if (!list_empty(&memtype->tier_sibling)) { list_for_each_entry(memtier, &memory_tiers, list) { if (adistance == memtier->adistance_start) return memtier; @@ -245,7 +245,7 @@ static struct memory_tier *find_create_memory_tier(struct memory_dev_type *memty memtier = new_memtier; link_memtype: - list_add(&memtype->tier_sibiling, &memtier->memory_types); + list_add(&memtype->tier_sibling, &memtier->memory_types); return memtier; } @@ -554,7 +554,7 @@ static bool clear_node_memory_tier(int node) memtype = node_memory_types[node].memtype; node_clear(node, memtype->nodes); if (nodes_empty(memtype->nodes)) { - list_del_init(&memtype->tier_sibiling); + list_del_init(&memtype->tier_sibling); if (list_empty(&memtier->memory_types)) destroy_memory_tier(memtier); } @@ -580,7 +580,7 @@ struct memory_dev_type *alloc_memory_type(int adistance) return ERR_PTR(-ENOMEM); memtype->adistance = adistance; - INIT_LIST_HEAD(&memtype->tier_sibiling); + INIT_LIST_HEAD(&memtype->tier_sibling); memtype->nodes = NODE_MASK_NONE; kref_init(&memtype->kref); return memtype; -- 2.43.0
From: Li Zhijian <lizhijian@fujitsu.com> mainline inclusion from mainline-v6.9-rc1 commit 601e793a749d6798764cd2eb7f459452f971ee61 category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9812 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... ------------------------------------------- Currently, when a demotion occurs, it will prioritize selecting a node from the preferred targets as the destination node for the demotion. If the preferred node does not meet the requirements, it will try from all the lower memory tier nodes until it finds a suitable demotion destination node or ultimately fails. However, the demotion target information isn't exposed to the users, especially the preferred target information, which relies on more factors. This makes it hard for users to understand the exact demotion behavior. Rather than having a new sysfs interface to expose this information, printing directly to kernel messages, just like the current page allocation fallback order does. A dmesg example with this patch is as follows: [ 0.704860] Demotion targets for Node 0: null [ 0.705456] Demotion targets for Node 1: null // node 2 is onlined [ 32.259775] Demotion targets for Node 0: perferred: 2, fallback: 2 [ 32.261290] Demotion targets for Node 1: perferred: 2, fallback: 2 [ 32.262726] Demotion targets for Node 2: null // node 3 is onlined [ 42.448809] Demotion targets for Node 0: perferred: 2, fallback: 2-3 [ 42.450704] Demotion targets for Node 1: perferred: 2, fallback: 2-3 [ 42.452556] Demotion targets for Node 2: perferred: 3, fallback: 3 [ 42.454136] Demotion targets for Node 3: null // node 4 is onlined [ 52.676833] Demotion targets for Node 0: perferred: 2, fallback: 2-4 [ 52.678735] Demotion targets for Node 1: perferred: 2, fallback: 2-4 [ 52.680493] Demotion targets for Node 2: perferred: 4, fallback: 3-4 [ 52.682154] Demotion targets for Node 3: null [ 52.683405] Demotion targets for Node 4: null // node 5 is onlined [ 62.931902] Demotion targets for Node 0: perferred: 2, fallback: 2-5 [ 62.938266] Demotion targets for Node 1: perferred: 5, fallback: 2-5 [ 62.943515] Demotion targets for Node 2: perferred: 4, fallback: 3-4 [ 62.947471] Demotion targets for Node 3: null [ 62.949908] Demotion targets for Node 4: null [ 62.952137] Demotion targets for Node 5: perferred: 3, fallback: 3-4 Regarding this requirement, we have previously discussed [1]. The initial proposal involved introducing a new sysfs interface. However, due to concerns about potential changes and compatibility issues with the interface in the future, a consensus was not reached with the community. Therefore, this time, we are directly printing out the information. [1] https://lore.kernel.org/all/d1d5add8-8f4a-4578-8bf0-2cbe79b09989@fujitsu.com... Link: https://lkml.kernel.org/r/20240206020151.605516-1-lizhijian@fujitsu.com Signed-off-by: Li Zhijian <lizhijian@fujitsu.com> Reviewed-by: "Huang, Ying" <ying.huang@intel.com> Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jinjiang Tu <tujinjiang@huawei.com> --- mm/memory-tiers.c | 24 +++++++++++++++++++++++- 1 file changed, 23 insertions(+), 1 deletion(-) diff --git a/mm/memory-tiers.c b/mm/memory-tiers.c index 0ec9ea04375a..393f91e665f3 100644 --- a/mm/memory-tiers.c +++ b/mm/memory-tiers.c @@ -378,6 +378,26 @@ static void disable_all_demotion_targets(void) synchronize_rcu(); } +static void dump_demotion_targets(void) +{ + int node; + + for_each_node_state(node, N_MEMORY) { + struct memory_tier *memtier = __node_get_memory_tier(node); + nodemask_t preferred = node_demotion[node].preferred; + + if (!memtier) + continue; + + if (nodes_empty(preferred)) + pr_info("Demotion targets for Node %d: null\n", node); + else + pr_info("Demotion targets for Node %d: preferred: %*pbl, fallback: %*pbl\n", + node, nodemask_pr_args(&preferred), + nodemask_pr_args(&memtier->lower_tier_mask)); + } +} + /* * Find an automatic demotion target for all memory * nodes. Failing here is OK. It might just indicate @@ -462,7 +482,7 @@ static void establish_demotion_targets(void) * Now build the lower_tier mask for each node collecting node mask from * all memory tier below it. This allows us to fallback demotion page * allocation to a set of nodes that is closer the above selected - * perferred node. + * preferred node. */ lower_tier = node_states[N_MEMORY]; list_for_each_entry(memtier, &memory_tiers, list) { @@ -475,6 +495,8 @@ static void establish_demotion_targets(void) nodes_andnot(lower_tier, lower_tier, tier_nodes); memtier->lower_tier_mask = lower_tier; } + + dump_demotion_targets(); } #else -- 2.43.0
From: "Ho-Ren (Jack) Chuang" <horenchuang@bytedance.com> mainline inclusion from mainline-v6.10-rc1 commit a72a30af550c08423de1b9feecb6ceeddc434889 category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9812 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... ------------------------------------------- Patch series "Improved Memory Tier Creation for CPUless NUMA Nodes", v11. When a memory device, such as CXL1.1 type3 memory, is emulated as normal memory (E820_TYPE_RAM), the memory device is indistinguishable from normal DRAM in terms of memory tiering with the current implementation. The current memory tiering assigns all detected normal memory nodes to the same DRAM tier. This results in normal memory devices with different attributions being unable to be assigned to the correct memory tier, leading to the inability to migrate pages between different types of memory. https://lore.kernel.org/linux-mm/PH0PR08MB7955E9F08CCB64F23963B5C3A860A@PH0P... This patchset automatically resolves the issues. It delays the initialization of memory tiers for CPUless NUMA nodes until they obtain HMAT information and after all devices are initialized at boot time, eliminating the need for user intervention. If no HMAT is specified, it falls back to using `default_dram_type`. Example usecase: We have CXL memory on the host, and we create VMs with a new system memory device backed by host CXL memory. We inject CXL memory performance attributes through QEMU, and the guest now sees memory nodes with performance attributes in HMAT. With this change, we enable the guest kernel to construct the correct memory tiering for the memory nodes. This patch (of 2): Since different memory devices require finding, allocating, and putting memory types, these common steps are abstracted in this patch, enhancing the scalability and conciseness of the code. Link: https://lkml.kernel.org/r/20240405000707.2670063-1-horenchuang@bytedance.com Link: https://lkml.kernel.org/r/20240405000707.2670063-2-horenchuang@bytedance.com Signed-off-by: Ho-Ren (Jack) Chuang <horenchuang@bytedance.com> Reviewed-by: "Huang, Ying" <ying.huang@intel.com> Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawie.com> Cc: Alistair Popple <apopple@nvidia.com> Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Dave Jiang <dave.jiang@intel.com> Cc: Gregory Price <gourry.memverge@gmail.com> Cc: Hao Xiang <hao.xiang@bytedance.com> Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com> Cc: Michal Hocko <mhocko@suse.com> Cc: Ravi Jonnalagadda <ravis.opensrc@micron.com> Cc: SeongJae Park <sj@kernel.org> Cc: Tejun Heo <tj@kernel.org> Cc: Vishal Verma <vishal.l.verma@intel.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jinjiang Tu <tujinjiang@huawei.com> --- drivers/dax/kmem.c | 30 ++++-------------------------- include/linux/memory-tiers.h | 13 +++++++++++++ mm/memory-tiers.c | 29 +++++++++++++++++++++++++++++ 3 files changed, 46 insertions(+), 26 deletions(-) diff --git a/drivers/dax/kmem.c b/drivers/dax/kmem.c index 2fb74acc0703..ca0f482ea7cb 100644 --- a/drivers/dax/kmem.c +++ b/drivers/dax/kmem.c @@ -54,36 +54,14 @@ static LIST_HEAD(kmem_memory_types); static struct memory_dev_type *kmem_find_alloc_memory_type(int adist) { - bool found = false; - struct memory_dev_type *mtype; - - mutex_lock(&kmem_memory_type_lock); - list_for_each_entry(mtype, &kmem_memory_types, list) { - if (mtype->adistance == adist) { - found = true; - break; - } - } - if (!found) { - mtype = alloc_memory_type(adist); - if (!IS_ERR(mtype)) - list_add(&mtype->list, &kmem_memory_types); - } - mutex_unlock(&kmem_memory_type_lock); - - return mtype; + guard(mutex)(&kmem_memory_type_lock); + return mt_find_alloc_memory_type(adist, &kmem_memory_types); } static void kmem_put_memory_types(void) { - struct memory_dev_type *mtype, *mtn; - - mutex_lock(&kmem_memory_type_lock); - list_for_each_entry_safe(mtype, mtn, &kmem_memory_types, list) { - list_del(&mtype->list); - put_memory_type(mtype); - } - mutex_unlock(&kmem_memory_type_lock); + guard(mutex)(&kmem_memory_type_lock); + mt_put_memory_types(&kmem_memory_types); } static int dev_dax_kmem_probe(struct dev_dax *dev_dax) diff --git a/include/linux/memory-tiers.h b/include/linux/memory-tiers.h index 6debda505b51..f0bbd272f152 100644 --- a/include/linux/memory-tiers.h +++ b/include/linux/memory-tiers.h @@ -48,6 +48,9 @@ int mt_calc_adistance(int node, int *adist); int mt_set_default_dram_perf(int nid, struct access_coordinate *perf, const char *source); int mt_perf_to_adistance(struct access_coordinate *perf, int *adist); +struct memory_dev_type *mt_find_alloc_memory_type(int adist, + struct list_head *memory_types); +void mt_put_memory_types(struct list_head *memory_types); #ifdef CONFIG_MIGRATION int next_demotion_node(int node); void node_get_allowed_targets(pg_data_t *pgdat, nodemask_t *targets); @@ -136,5 +139,15 @@ static inline int mt_perf_to_adistance(struct access_coordinate *perf, int *adis { return -EIO; } + +static inline struct memory_dev_type *mt_find_alloc_memory_type(int adist, + struct list_head *memory_types) +{ + return NULL; +} + +static inline void mt_put_memory_types(struct list_head *memory_types) +{ +} #endif /* CONFIG_NUMA */ #endif /* _LINUX_MEMORY_TIERS_H */ diff --git a/mm/memory-tiers.c b/mm/memory-tiers.c index 393f91e665f3..b3bfc93b529c 100644 --- a/mm/memory-tiers.c +++ b/mm/memory-tiers.c @@ -642,6 +642,35 @@ void clear_node_memory_type(int node, struct memory_dev_type *memtype) } EXPORT_SYMBOL_GPL(clear_node_memory_type); +struct memory_dev_type *mt_find_alloc_memory_type(int adist, struct list_head *memory_types) +{ + struct memory_dev_type *mtype; + + list_for_each_entry(mtype, memory_types, list) + if (mtype->adistance == adist) + return mtype; + + mtype = alloc_memory_type(adist); + if (IS_ERR(mtype)) + return mtype; + + list_add(&mtype->list, memory_types); + + return mtype; +} +EXPORT_SYMBOL_GPL(mt_find_alloc_memory_type); + +void mt_put_memory_types(struct list_head *memory_types) +{ + struct memory_dev_type *mtype, *mtn; + + list_for_each_entry_safe(mtype, mtn, memory_types, list) { + list_del(&mtype->list); + put_memory_type(mtype); + } +} +EXPORT_SYMBOL_GPL(mt_put_memory_types); + static void dump_hmem_attrs(struct access_coordinate *coord, const char *prefix) { pr_info( -- 2.43.0
From: "Ho-Ren (Jack) Chuang" <horenchuang@bytedance.com> mainline inclusion from mainline-v6.10-rc1 commit cf93be18fa1bb837fa1e3015a9919953fff6ef22 category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9812 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... ------------------------------------------- The current implementation treats emulated memory devices, such as CXL1.1 type3 memory, as normal DRAM when they are emulated as normal memory (E820_TYPE_RAM). However, these emulated devices have different characteristics than traditional DRAM, making it important to distinguish them. Thus, we modify the tiered memory initialization process to introduce a delay specifically for CPUless NUMA nodes. This delay ensures that the memory tier initialization for these nodes is deferred until HMAT information is obtained during the boot process. Finally, demotion tables are recalculated at the end. * late_initcall(memory_tier_late_init); Some device drivers may have initialized memory tiers between `memory_tier_init()` and `memory_tier_late_init()`, potentially bringing online memory nodes and configuring memory tiers. They should be excluded in the late init. * Handle cases where there is no HMAT when creating memory tiers There is a scenario where a CPUless node does not provide HMAT information. If no HMAT is specified, it falls back to using the default DRAM tier. * Introduce another new lock `default_dram_perf_lock` for adist calculation In the current implementation, iterating through CPUlist nodes requires holding the `memory_tier_lock`. However, `mt_calc_adistance()` will end up trying to acquire the same lock, leading to a potential deadlock. Therefore, we propose introducing a standalone `default_dram_perf_lock` to protect `default_dram_perf_*`. This approach not only avoids deadlock but also prevents holding a large lock simultaneously. * Upgrade `set_node_memory_tier` to support additional cases, including default DRAM, late CPUless, and hot-plugged initializations. To cover hot-plugged memory nodes, `mt_calc_adistance()` and `mt_find_alloc_memory_type()` are moved into `set_node_memory_tier()` to handle cases where memtype is not initialized and where HMAT information is available. * Introduce `default_memory_types` for those memory types that are not initialized by device drivers. Because late initialized memory and default DRAM memory need to be managed, a default memory type is created for storing all memory types that are not initialized by device drivers and as a fallback. Link: https://lkml.kernel.org/r/20240405000707.2670063-3-horenchuang@bytedance.com Signed-off-by: Ho-Ren (Jack) Chuang <horenchuang@bytedance.com> Signed-off-by: Hao Xiang <hao.xiang@bytedance.com> Reviewed-by: "Huang, Ying" <ying.huang@intel.com> Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Cc: Alistair Popple <apopple@nvidia.com> Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Dave Jiang <dave.jiang@intel.com> Cc: Gregory Price <gourry.memverge@gmail.com> Cc: Michal Hocko <mhocko@suse.com> Cc: Ravi Jonnalagadda <ravis.opensrc@micron.com> Cc: SeongJae Park <sj@kernel.org> Cc: Tejun Heo <tj@kernel.org> Cc: Vishal Verma <vishal.l.verma@intel.com> Cc: Jonathan Cameron <Jonathan.Cameron@huawie.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jinjiang Tu <tujinjiang@huawei.com> --- mm/memory-tiers.c | 94 +++++++++++++++++++++++++++++++++++------------ 1 file changed, 70 insertions(+), 24 deletions(-) diff --git a/mm/memory-tiers.c b/mm/memory-tiers.c index b3bfc93b529c..0dcca1ef4941 100644 --- a/mm/memory-tiers.c +++ b/mm/memory-tiers.c @@ -37,6 +37,11 @@ struct node_memory_type_map { static DEFINE_MUTEX(memory_tier_lock); static LIST_HEAD(memory_tiers); +/* + * The list is used to store all memory types that are not created + * by a device driver. + */ +static LIST_HEAD(default_memory_types); static struct node_memory_type_map node_memory_types[MAX_NUMNODES]; struct memory_dev_type *default_dram_type; @@ -127,6 +132,8 @@ static struct demotion_nodes *node_demotion __read_mostly; static BLOCKING_NOTIFIER_HEAD(mt_adistance_algorithms); +/* The lock is used to protect `default_dram_perf*` info and nid. */ +static DEFINE_MUTEX(default_dram_perf_lock); static bool default_dram_perf_error; static struct access_coordinate default_dram_perf; static int default_dram_perf_ref_nid = NUMA_NO_NODE; @@ -524,7 +531,8 @@ static inline void __init_node_memory_type(int node, struct memory_dev_type *mem static struct memory_tier *set_node_memory_tier(int node) { struct memory_tier *memtier; - struct memory_dev_type *memtype; + struct memory_dev_type *memtype = default_dram_type; + int adist = MEMTIER_ADISTANCE_DRAM; pg_data_t *pgdat = NODE_DATA(node); @@ -533,7 +541,16 @@ static struct memory_tier *set_node_memory_tier(int node) if (!node_state(node, N_MEMORY)) return ERR_PTR(-EINVAL); - __init_node_memory_type(node, default_dram_type); + mt_calc_adistance(node, &adist); + if (!node_memory_types[node].memtype) { + memtype = mt_find_alloc_memory_type(adist, &default_memory_types); + if (IS_ERR(memtype)) { + memtype = default_dram_type; + pr_info("Failed to allocate a memory type. Fall back.\n"); + } + } + + __init_node_memory_type(node, memtype); memtype = node_memory_types[node].memtype; node_set(node, memtype->nodes); @@ -671,6 +688,35 @@ void mt_put_memory_types(struct list_head *memory_types) } EXPORT_SYMBOL_GPL(mt_put_memory_types); +/* + * This is invoked via `late_initcall()` to initialize memory tiers for + * CPU-less memory nodes after driver initialization, which is + * expected to provide `adistance` algorithms. + */ +static int __init memory_tier_late_init(void) +{ + int nid; + + guard(mutex)(&memory_tier_lock); + for_each_node_state(nid, N_MEMORY) { + /* + * Some device drivers may have initialized memory tiers + * between `memory_tier_init()` and `memory_tier_late_init()`, + * potentially bringing online memory nodes and + * configuring memory tiers. Exclude them here. + */ + if (node_memory_types[nid].memtype) + continue; + + set_node_memory_tier(nid); + } + + establish_demotion_targets(); + + return 0; +} +late_initcall(memory_tier_late_init); + static void dump_hmem_attrs(struct access_coordinate *coord, const char *prefix) { pr_info( @@ -682,25 +728,19 @@ static void dump_hmem_attrs(struct access_coordinate *coord, const char *prefix) int mt_set_default_dram_perf(int nid, struct access_coordinate *perf, const char *source) { - int rc = 0; - - mutex_lock(&memory_tier_lock); - if (default_dram_perf_error) { - rc = -EIO; - goto out; - } + guard(mutex)(&default_dram_perf_lock); + if (default_dram_perf_error) + return -EIO; if (perf->read_latency + perf->write_latency == 0 || - perf->read_bandwidth + perf->write_bandwidth == 0) { - rc = -EINVAL; - goto out; - } + perf->read_bandwidth + perf->write_bandwidth == 0) + return -EINVAL; if (default_dram_perf_ref_nid == NUMA_NO_NODE) { default_dram_perf = *perf; default_dram_perf_ref_nid = nid; default_dram_perf_ref_source = kstrdup(source, GFP_KERNEL); - goto out; + return 0; } /* @@ -728,27 +768,25 @@ int mt_set_default_dram_perf(int nid, struct access_coordinate *perf, pr_info( " disable default DRAM node performance based abstract distance algorithm.\n"); default_dram_perf_error = true; - rc = -EINVAL; + return -EINVAL; } -out: - mutex_unlock(&memory_tier_lock); - return rc; + return 0; } int mt_perf_to_adistance(struct access_coordinate *perf, int *adist) { + guard(mutex)(&default_dram_perf_lock); if (default_dram_perf_error) return -EIO; - if (default_dram_perf_ref_nid == NUMA_NO_NODE) - return -ENOENT; - if (perf->read_latency + perf->write_latency == 0 || perf->read_bandwidth + perf->write_bandwidth == 0) return -EINVAL; - mutex_lock(&memory_tier_lock); + if (default_dram_perf_ref_nid == NUMA_NO_NODE) + return -ENOENT; + /* * The abstract distance of a memory node is in direct proportion to * its memory latency (read + write) and inversely proportional to its @@ -761,7 +799,6 @@ int mt_perf_to_adistance(struct access_coordinate *perf, int *adist) (default_dram_perf.read_latency + default_dram_perf.write_latency) * (default_dram_perf.read_bandwidth + default_dram_perf.write_bandwidth) / (perf->read_bandwidth + perf->write_bandwidth); - mutex_unlock(&memory_tier_lock); return 0; } @@ -874,7 +911,8 @@ static int __init memory_tier_init(void) * For now we can have 4 faster memory tiers with smaller adistance * than default DRAM tier. */ - default_dram_type = alloc_memory_type(MEMTIER_ADISTANCE_DRAM); + default_dram_type = mt_find_alloc_memory_type(MEMTIER_ADISTANCE_DRAM, + &default_memory_types); if (IS_ERR(default_dram_type)) panic("%s() failed to allocate default DRAM tier\n", __func__); @@ -884,6 +922,14 @@ static int __init memory_tier_init(void) * types assigned. */ for_each_node_state(node, N_MEMORY) { + if (!node_state(node, N_CPU)) + /* + * Defer memory tier initialization on + * CPUless numa nodes. These will be initialized + * after firmware and devices are initialized. + */ + continue; + memtier = set_node_memory_tier(node); if (IS_ERR(memtier)) /* -- 2.43.0
From: "Ho-Ren (Jack) Chuang" <horen.chuang@linux.dev> mainline inclusion from mainline-v6.11-rc1 commit 823430c8e9d98c5865af518c782d0493b76aa511 category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9812 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... ------------------------------------------- The current memory tier initialization process is distributed across two different functions, memory_tier_init() and memory_tier_late_init(). This design is hard to maintain. Thus, this patch is proposed to reduce the possible code paths by consolidating different initialization patches into one. The earlier discussion with Jonathan and Ying is listed here: https://lore.kernel.org/lkml/20240405150244.00004b49@Huawei.com/ If we want to put these two initializations together, they must be placed together in the later function. Because only at that time, the HMAT information will be ready, adist between nodes can be calculated, and memory tiering can be established based on the adist. So we position the initialization at memory_tier_init() to the memory_tier_late_init() call. Moreover, it's natural to keep memory_tier initialization in drivers at device_initcall() level. If we simply move the set_node_memory_tier() from memory_tier_init() to late_initcall(), it will result in HMAT not registering the mt_adistance_algorithm callback function, because set_node_memory_tier() is not performed during the memory tiering initialization phase, leading to a lack of correct default_dram information. Therefore, we introduced a nodemask to pass the information of the default DRAM nodes. The reason for not choosing to reuse default_dram_type->nodes is that it is not clean enough. So in the end, we use a __initdata variable, which is a variable that is released once initialization is complete, including both CPU and memory nodes for HMAT to iterate through. Link: https://lkml.kernel.org/r/20240704072646.437579-1-horen.chuang@linux.dev Signed-off-by: Ho-Ren (Jack) Chuang <horenchuang@bytedance.com> Suggested-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Reviewed-by: "Huang, Ying" <ying.huang@intel.com> Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com> Cc: Alistair Popple <apopple@nvidia.com> Cc: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com> Cc: Dan Williams <dan.j.williams@intel.com> Cc: Dave Jiang <dave.jiang@intel.com> Cc: Gregory Price <gourry.memverge@gmail.com> Cc: Len Brown <lenb@kernel.org> Cc: Michal Hocko <mhocko@suse.com> Cc: Rafael J. Wysocki <rafael@kernel.org> Cc: Ravi Jonnalagadda <ravis.opensrc@micron.com> Cc: SeongJae Park <sj@kernel.org> Cc: Tejun Heo <tj@kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jinjiang Tu <tujinjiang@huawei.com> --- drivers/acpi/numa/hmat.c | 5 +--- include/linux/memory-tiers.h | 2 ++ mm/memory-tiers.c | 54 ++++++++++++++---------------------- 3 files changed, 24 insertions(+), 37 deletions(-) diff --git a/drivers/acpi/numa/hmat.c b/drivers/acpi/numa/hmat.c index 1d6450c13922..c1fd44933db0 100644 --- a/drivers/acpi/numa/hmat.c +++ b/drivers/acpi/numa/hmat.c @@ -860,10 +860,7 @@ static int hmat_set_default_dram_perf(void) struct memory_target *target; struct access_coordinate *attrs; - if (!default_dram_type) - return -EIO; - - for_each_node_mask(nid, default_dram_type->nodes) { + for_each_node_mask(nid, default_dram_nodes) { pxm = node_to_pxm(nid); target = find_mem_target(pxm); if (!target) diff --git a/include/linux/memory-tiers.h b/include/linux/memory-tiers.h index f0bbd272f152..7a805796fcfd 100644 --- a/include/linux/memory-tiers.h +++ b/include/linux/memory-tiers.h @@ -38,6 +38,7 @@ struct access_coordinate; #ifdef CONFIG_NUMA extern bool numa_demotion_enabled; extern struct memory_dev_type *default_dram_type; +extern nodemask_t default_dram_nodes; struct memory_dev_type *alloc_memory_type(int adistance); void put_memory_type(struct memory_dev_type *memtype); void init_node_memory_type(int node, struct memory_dev_type *default_type); @@ -76,6 +77,7 @@ static inline bool node_is_toptier(int node) #define numa_demotion_enabled false #define default_dram_type NULL +#define default_dram_nodes NODE_MASK_NONE /* * CONFIG_NUMA implementation returns non NULL error. */ diff --git a/mm/memory-tiers.c b/mm/memory-tiers.c index 0dcca1ef4941..a317145b138d 100644 --- a/mm/memory-tiers.c +++ b/mm/memory-tiers.c @@ -44,6 +44,7 @@ static LIST_HEAD(memory_tiers); static LIST_HEAD(default_memory_types); static struct node_memory_type_map node_memory_types[MAX_NUMNODES]; struct memory_dev_type *default_dram_type; +nodemask_t default_dram_nodes __initdata = NODE_MASK_NONE; static struct bus_type memory_tier_subsys = { .name = "memory_tiering", @@ -690,28 +691,35 @@ EXPORT_SYMBOL_GPL(mt_put_memory_types); /* * This is invoked via `late_initcall()` to initialize memory tiers for - * CPU-less memory nodes after driver initialization, which is - * expected to provide `adistance` algorithms. + * memory nodes, both with and without CPUs. After the initialization of + * firmware and devices, adistance algorithms are expected to be provided. */ static int __init memory_tier_late_init(void) { int nid; + struct memory_tier *memtier; + get_online_mems(); guard(mutex)(&memory_tier_lock); + + /* Assign each uninitialized N_MEMORY node to a memory tier. */ for_each_node_state(nid, N_MEMORY) { /* - * Some device drivers may have initialized memory tiers - * between `memory_tier_init()` and `memory_tier_late_init()`, - * potentially bringing online memory nodes and - * configuring memory tiers. Exclude them here. + * Some device drivers may have initialized + * memory tiers, potentially bringing memory nodes + * online and configuring memory tiers. + * Exclude them here. */ if (node_memory_types[nid].memtype) continue; - set_node_memory_tier(nid); + memtier = set_node_memory_tier(nid); + if (IS_ERR(memtier)) + continue; } establish_demotion_targets(); + put_online_mems(); return 0; } @@ -894,8 +902,7 @@ static int __meminit memtier_hotplug_callback(struct notifier_block *self, static int __init memory_tier_init(void) { - int ret, node; - struct memory_tier *memtier; + int ret; ret = subsys_virtual_register(&memory_tier_subsys, NULL); if (ret) @@ -906,7 +913,8 @@ static int __init memory_tier_init(void) GFP_KERNEL); WARN_ON(!node_demotion); #endif - mutex_lock(&memory_tier_lock); + + guard(mutex)(&memory_tier_lock); /* * For now we can have 4 faster memory tiers with smaller adistance * than default DRAM tier. @@ -916,29 +924,9 @@ static int __init memory_tier_init(void) if (IS_ERR(default_dram_type)) panic("%s() failed to allocate default DRAM tier\n", __func__); - /* - * Look at all the existing N_MEMORY nodes and add them to - * default memory tier or to a tier if we already have memory - * types assigned. - */ - for_each_node_state(node, N_MEMORY) { - if (!node_state(node, N_CPU)) - /* - * Defer memory tier initialization on - * CPUless numa nodes. These will be initialized - * after firmware and devices are initialized. - */ - continue; - - memtier = set_node_memory_tier(node); - if (IS_ERR(memtier)) - /* - * Continue with memtiers we are able to setup - */ - break; - } - establish_demotion_targets(); - mutex_unlock(&memory_tier_lock); + /* Record nodes with memory and CPU to set default DRAM performance. */ + nodes_and(default_dram_nodes, node_states[N_MEMORY], + node_states[N_CPU]); hotplug_memory_notifier(memtier_hotplug_callback, MEMTIER_HOTPLUG_PRI); return 0; -- 2.43.0
From: Yanfei Xu <yanfei.xu@intel.com> mainline inclusion from mainline-v6.12-rc1 commit 073c78edf5bbd32046a117fc7f71312dc3d2c607 category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9812 Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?i... ------------------------------------------- commit 823430c8e9d9 ("memory tier: consolidate the initialization of memory tiers") introduces a locking change that use guard(mutex) to instead of mutex_lock/unlock() for memory_tier_lock. It unexpectedly expanded the locked region to include the hotplug_memory_notifier(), as a result, it triggers an locking dependency detected of ABBA deadlock. Exclude hotplug_memory_notifier() from the locked region to fixing it. The deadlock scenario is that when a memory online event occurs, the execution of memory notifier will access the read lock of the memory_chain.rwsem, then the reigistration of the memory notifier in memory_tier_init() acquires the write lock of the memory_chain.rwsem while holding memory_tier_lock. Then the memory online event continues to invoke the memory hotplug callback registered by memory_tier_init(). Since this callback tries to acquire the memory_tier_lock, a deadlock occurs. In fact, this deadlock can't happen because memory_tier_init() always executes before memory online events happen due to the subsys_initcall() has an higher priority than module_init(). [ 133.491106] WARNING: possible circular locking dependency detected [ 133.493656] 6.11.0-rc2+ #146 Tainted: G O N [ 133.504290] ------------------------------------------------------ [ 133.515194] (udev-worker)/1133 is trying to acquire lock: [ 133.525715] ffffffff87044e28 (memory_tier_lock){+.+.}-{3:3}, at: memtier_hotplug_callback+0x383/0x4b0 [ 133.536449] [ 133.536449] but task is already holding lock: [ 133.549847] ffffffff875d3310 ((memory_chain).rwsem){++++}-{3:3}, at: blocking_notifier_call_chain+0x60/0xb0 [ 133.556781] [ 133.556781] which lock already depends on the new lock. [ 133.556781] [ 133.569957] [ 133.569957] the existing dependency chain (in reverse order) is: [ 133.577618] [ 133.577618] -> #1 ((memory_chain).rwsem){++++}-{3:3}: [ 133.584997] down_write+0x97/0x210 [ 133.588647] blocking_notifier_chain_register+0x71/0xd0 [ 133.592537] register_memory_notifier+0x26/0x30 [ 133.596314] memory_tier_init+0x187/0x300 [ 133.599864] do_one_initcall+0x117/0x5d0 [ 133.603399] kernel_init_freeable+0xab0/0xeb0 [ 133.606986] kernel_init+0x28/0x2f0 [ 133.610312] ret_from_fork+0x59/0x90 [ 133.613652] ret_from_fork_asm+0x1a/0x30 [ 133.617012] [ 133.617012] -> #0 (memory_tier_lock){+.+.}-{3:3}: [ 133.623390] __lock_acquire+0x2efd/0x5c60 [ 133.626730] lock_acquire+0x1ce/0x580 [ 133.629757] __mutex_lock+0x15c/0x1490 [ 133.632731] mutex_lock_nested+0x1f/0x30 [ 133.635717] memtier_hotplug_callback+0x383/0x4b0 [ 133.638748] notifier_call_chain+0xbf/0x370 [ 133.641647] blocking_notifier_call_chain+0x76/0xb0 [ 133.644636] memory_notify+0x2e/0x40 [ 133.647427] online_pages+0x597/0x720 [ 133.650246] memory_subsys_online+0x4f6/0x7f0 [ 133.653107] device_online+0x141/0x1d0 [ 133.655831] online_memory_block+0x4d/0x60 [ 133.658616] walk_memory_blocks+0xc0/0x120 [ 133.661419] add_memory_resource+0x51d/0x6c0 [ 133.664202] add_memory_driver_managed+0xf5/0x180 [ 133.667060] dev_dax_kmem_probe+0x7f7/0xb40 [kmem] [ 133.669949] dax_bus_probe+0x147/0x230 [ 133.672687] really_probe+0x27f/0xac0 [ 133.675463] __driver_probe_device+0x1f3/0x460 [ 133.678493] driver_probe_device+0x56/0x1b0 [ 133.681366] __driver_attach+0x277/0x570 [ 133.684149] bus_for_each_dev+0x145/0x1e0 [ 133.686937] driver_attach+0x49/0x60 [ 133.689673] bus_add_driver+0x2f3/0x6b0 [ 133.692421] driver_register+0x170/0x4b0 [ 133.695118] __dax_driver_register+0x141/0x1b0 [ 133.697910] dax_kmem_init+0x54/0xff0 [kmem] [ 133.700794] do_one_initcall+0x117/0x5d0 [ 133.703455] do_init_module+0x277/0x750 [ 133.706054] load_module+0x5d1d/0x74f0 [ 133.708602] init_module_from_file+0x12c/0x1a0 [ 133.711234] idempotent_init_module+0x3f1/0x690 [ 133.713937] __x64_sys_finit_module+0x10e/0x1a0 [ 133.716492] x64_sys_call+0x184d/0x20d0 [ 133.719053] do_syscall_64+0x6d/0x140 [ 133.721537] entry_SYSCALL_64_after_hwframe+0x76/0x7e [ 133.724239] [ 133.724239] other info that might help us debug this: [ 133.724239] [ 133.730832] Possible unsafe locking scenario: [ 133.730832] [ 133.735298] CPU0 CPU1 [ 133.737759] ---- ---- [ 133.740165] rlock((memory_chain).rwsem); [ 133.742623] lock(memory_tier_lock); [ 133.745357] lock((memory_chain).rwsem); [ 133.748141] lock(memory_tier_lock); [ 133.750489] [ 133.750489] *** DEADLOCK *** [ 133.750489] [ 133.756742] 6 locks held by (udev-worker)/1133: [ 133.759179] #0: ffff888207be6158 (&dev->mutex){....}-{3:3}, at: __driver_attach+0x26c/0x570 [ 133.762299] #1: ffffffff875b5868 (device_hotplug_lock){+.+.}-{3:3}, at: lock_device_hotplug+0x20/0x30 [ 133.765565] #2: ffff88820cf6a108 (&dev->mutex){....}-{3:3}, at: device_online+0x2f/0x1d0 [ 133.768978] #3: ffffffff86d08ff0 (cpu_hotplug_lock){++++}-{0:0}, at: mem_hotplug_begin+0x17/0x30 [ 133.772312] #4: ffffffff8702dfb0 (mem_hotplug_lock){++++}-{0:0}, at: mem_hotplug_begin+0x23/0x30 [ 133.775544] #5: ffffffff875d3310 ((memory_chain).rwsem){++++}-{3:3}, at: blocking_notifier_call_chain+0x60/0xb0 [ 133.779113] [ 133.779113] stack backtrace: [ 133.783728] CPU: 5 UID: 0 PID: 1133 Comm: (udev-worker) Tainted: G O N 6.11.0-rc2+ #146 [ 133.787220] Tainted: [O]=OOT_MODULE, [N]=TEST [ 133.789948] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015 [ 133.793291] Call Trace: [ 133.795826] <TASK> [ 133.798284] dump_stack_lvl+0xea/0x150 [ 133.801025] dump_stack+0x19/0x20 [ 133.803609] print_circular_bug+0x477/0x740 [ 133.806341] check_noncircular+0x2f4/0x3e0 [ 133.809056] ? __pfx_check_noncircular+0x10/0x10 [ 133.811866] ? __pfx_lockdep_lock+0x10/0x10 [ 133.814670] ? __sanitizer_cov_trace_const_cmp8+0x1c/0x30 [ 133.817610] __lock_acquire+0x2efd/0x5c60 [ 133.820339] ? __pfx___lock_acquire+0x10/0x10 [ 133.823128] ? __dax_driver_register+0x141/0x1b0 [ 133.825926] ? do_one_initcall+0x117/0x5d0 [ 133.828648] lock_acquire+0x1ce/0x580 [ 133.831349] ? memtier_hotplug_callback+0x383/0x4b0 [ 133.834293] ? __pfx_lock_acquire+0x10/0x10 [ 133.837134] __mutex_lock+0x15c/0x1490 [ 133.839829] ? memtier_hotplug_callback+0x383/0x4b0 [ 133.842753] ? memtier_hotplug_callback+0x383/0x4b0 [ 133.845602] ? __this_cpu_preempt_check+0x21/0x30 [ 133.848438] ? __pfx___mutex_lock+0x10/0x10 [ 133.851200] ? __pfx_lock_acquire+0x10/0x10 [ 133.853935] ? global_dirty_limits+0xc0/0x160 [ 133.856699] ? __sanitizer_cov_trace_switch+0x58/0xa0 [ 133.859564] mutex_lock_nested+0x1f/0x30 [ 133.862251] ? mutex_lock_nested+0x1f/0x30 [ 133.864964] memtier_hotplug_callback+0x383/0x4b0 [ 133.867752] notifier_call_chain+0xbf/0x370 [ 133.870550] ? writeback_set_ratelimit+0xe8/0x160 [ 133.873372] blocking_notifier_call_chain+0x76/0xb0 [ 133.876311] memory_notify+0x2e/0x40 [ 133.879013] online_pages+0x597/0x720 [ 133.881686] ? irqentry_exit+0x3e/0xa0 [ 133.884397] ? __pfx_online_pages+0x10/0x10 [ 133.887244] ? __sanitizer_cov_trace_const_cmp8+0x1c/0x30 [ 133.890299] ? mhp_init_memmap_on_memory+0x7a/0x1c0 [ 133.893203] memory_subsys_online+0x4f6/0x7f0 [ 133.896099] ? __pfx_memory_subsys_online+0x10/0x10 [ 133.899039] ? xa_load+0x16d/0x2e0 [ 133.901667] ? __pfx_xa_load+0x10/0x10 [ 133.904366] ? __pfx_memory_subsys_online+0x10/0x10 [ 133.907218] device_online+0x141/0x1d0 [ 133.909845] online_memory_block+0x4d/0x60 [ 133.912494] walk_memory_blocks+0xc0/0x120 [ 133.915104] ? __pfx_online_memory_block+0x10/0x10 [ 133.917776] add_memory_resource+0x51d/0x6c0 [ 133.920404] ? __pfx_add_memory_resource+0x10/0x10 [ 133.923104] ? _raw_write_unlock+0x31/0x60 [ 133.925781] ? register_memory_resource+0x119/0x180 [ 133.928450] add_memory_driver_managed+0xf5/0x180 [ 133.931036] dev_dax_kmem_probe+0x7f7/0xb40 [kmem] [ 133.933665] ? __pfx_dev_dax_kmem_probe+0x10/0x10 [kmem] [ 133.936332] ? __pfx___up_read+0x10/0x10 [ 133.938878] dax_bus_probe+0x147/0x230 [ 133.941332] ? __pfx_dax_bus_probe+0x10/0x10 [ 133.943954] really_probe+0x27f/0xac0 [ 133.946387] ? __sanitizer_cov_trace_const_cmp1+0x1e/0x30 [ 133.949106] __driver_probe_device+0x1f3/0x460 [ 133.951704] ? parse_option_str+0x149/0x190 [ 133.954241] driver_probe_device+0x56/0x1b0 [ 133.956749] __driver_attach+0x277/0x570 [ 133.959228] ? __pfx___driver_attach+0x10/0x10 [ 133.961776] bus_for_each_dev+0x145/0x1e0 [ 133.964367] ? __pfx_bus_for_each_dev+0x10/0x10 [ 133.967019] ? __kasan_check_read+0x15/0x20 [ 133.969543] ? _raw_spin_unlock+0x31/0x60 [ 133.972132] driver_attach+0x49/0x60 [ 133.974536] bus_add_driver+0x2f3/0x6b0 [ 133.977044] driver_register+0x170/0x4b0 [ 133.979480] __dax_driver_register+0x141/0x1b0 [ 133.982126] ? __pfx_dax_kmem_init+0x10/0x10 [kmem] [ 133.984724] dax_kmem_init+0x54/0xff0 [kmem] [ 133.987284] ? __pfx_dax_kmem_init+0x10/0x10 [kmem] [ 133.989965] do_one_initcall+0x117/0x5d0 [ 133.992506] ? __pfx_do_one_initcall+0x10/0x10 [ 133.995185] ? __kasan_kmalloc+0x88/0xa0 [ 133.997748] ? kasan_poison+0x3e/0x60 [ 134.000288] ? kasan_unpoison+0x2c/0x60 [ 134.002762] ? kasan_poison+0x3e/0x60 [ 134.005202] ? __asan_register_globals+0x62/0x80 [ 134.007753] ? __pfx_dax_kmem_init+0x10/0x10 [kmem] [ 134.010439] do_init_module+0x277/0x750 [ 134.012953] load_module+0x5d1d/0x74f0 [ 134.015406] ? __pfx_load_module+0x10/0x10 [ 134.017887] ? __pfx_ima_post_read_file+0x10/0x10 [ 134.020470] ? __sanitizer_cov_trace_const_cmp8+0x1c/0x30 [ 134.023127] ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20 [ 134.025767] ? security_kernel_post_read_file+0xa2/0xd0 [ 134.028429] ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20 [ 134.031162] ? kernel_read_file+0x503/0x820 [ 134.033645] ? __pfx_kernel_read_file+0x10/0x10 [ 134.036232] ? __pfx___lock_acquire+0x10/0x10 [ 134.038766] init_module_from_file+0x12c/0x1a0 [ 134.041291] ? init_module_from_file+0x12c/0x1a0 [ 134.043936] ? __pfx_init_module_from_file+0x10/0x10 [ 134.046516] ? __this_cpu_preempt_check+0x21/0x30 [ 134.049091] ? __kasan_check_read+0x15/0x20 [ 134.051551] ? do_raw_spin_unlock+0x60/0x210 [ 134.054077] idempotent_init_module+0x3f1/0x690 [ 134.056643] ? __pfx_idempotent_init_module+0x10/0x10 [ 134.059318] ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20 [ 134.061995] ? __fget_light+0x17d/0x210 [ 134.064428] __x64_sys_finit_module+0x10e/0x1a0 [ 134.066976] x64_sys_call+0x184d/0x20d0 [ 134.069405] do_syscall_64+0x6d/0x140 [ 134.071926] entry_SYSCALL_64_after_hwframe+0x76/0x7e [yanfei.xu@intel.com: add mutex_lock/unlock() pair back] Link: https://lkml.kernel.org/r/20240830102447.1445296-1-yanfei.xu@intel.com Link: https://lkml.kernel.org/r/20240827113614.1343049-1-yanfei.xu@intel.com Fixes: 823430c8e9d9 ("memory tier: consolidate the initialization of memory tiers") Signed-off-by: Yanfei Xu <yanfei.xu@intel.com> Reviewed-by: "Huang, Ying" <ying.huang@intel.com> Cc: Ho-Ren (Jack) Chuang <horen.chuang@linux.dev> Cc: Jonathan Cameron <Jonathan.Cameron@huawei.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jinjiang Tu <tujinjiang@huawei.com> --- mm/memory-tiers.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/mm/memory-tiers.c b/mm/memory-tiers.c index a317145b138d..02eb903a9077 100644 --- a/mm/memory-tiers.c +++ b/mm/memory-tiers.c @@ -914,13 +914,14 @@ static int __init memory_tier_init(void) WARN_ON(!node_demotion); #endif - guard(mutex)(&memory_tier_lock); + mutex_lock(&memory_tier_lock); /* * For now we can have 4 faster memory tiers with smaller adistance * than default DRAM tier. */ default_dram_type = mt_find_alloc_memory_type(MEMTIER_ADISTANCE_DRAM, &default_memory_types); + mutex_unlock(&memory_tier_lock); if (IS_ERR(default_dram_type)) panic("%s() failed to allocate default DRAM tier\n", __func__); -- 2.43.0
hulk inclusion category: bugfix bugzilla: https://atomgit.com/openeuler/kernel/issues/9812 ---------------------------------------- NUMA.remote's memory tier sits at the same level as DRAM, which breaks both NUMA demote and NUMA balancing promote: cold pages can no longer be demoted to the remote node, and hot pages cannot be promoted back to local DRAM. Because there is currently no bandwidth/latency data for NUMA.remote, the adist cannot be derived through the normal path. Hardcode the NUMA.remote adist to MEMTIER_ADISTANCE_DRAM * 5, so that it lands below DRAM in the memory tier hierarchy, making demote/promote work. Fixes: 5d0bd1321546 ("mm/numa_remote: support to hotplug/hotremove remote memory") Signed-off-by: Jinjiang Tu <tujinjiang@huawei.com> --- drivers/base/numa_remote.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/drivers/base/numa_remote.c b/drivers/base/numa_remote.c index f924d7867394..b1ed72020482 100644 --- a/drivers/base/numa_remote.c +++ b/drivers/base/numa_remote.c @@ -12,6 +12,7 @@ #include <linux/memory.h> #include <linux/numa_remote.h> #include <linux/oom.h> +#include <linux/memory-tiers.h> #include "../../mm/hugetlb_vmemmap.h" #include "../../mm/internal.h" @@ -20,6 +21,8 @@ /* The default distance between two remote node */ #define REMOTE_TO_REMOTE_DISTANCE 255 +#define MEMTIER_DEFAULT_UB_ADISTANCE (MEMTIER_ADISTANCE_DRAM * 5) + bool numa_remote_enabled __ro_after_init; static bool numa_remote_nofallback_mode __ro_after_init; static bool numa_remote_preonline_mode __ro_after_init; @@ -680,6 +683,24 @@ int numa_remote_report_node_meminfo(char *buf, int len, int nid) nid, K(atomic_long_read(&pre_online_pages_node[nid]))); } +static int ub_calculate_adistance(struct notifier_block *self, + unsigned long nid, void *data) +{ + int *adist = data; + + if (!numa_is_remote_node(nid)) + return NOTIFY_OK; + + *adist = MEMTIER_DEFAULT_UB_ADISTANCE; + + return NOTIFY_STOP; +} + +static struct notifier_block ub_adist_nb __meminitdata = { + .notifier_call = ub_calculate_adistance, + .priority = 100, +}; + static int __init numa_remote_init(void) { int ret; @@ -700,6 +721,8 @@ static int __init numa_remote_init(void) if (ret) goto err_register_notifier; + register_mt_adistance_algorithm(&ub_adist_nb); + return 0; err_register_notifier: -- 2.43.0
反馈： 您发送到kernel@openeuler.org的补丁/补丁集，已成功转换为PR！ PR链接地址： https://atomgit.com/openeuler/kernel/merge_requests/26908 邮件列表地址：https://mailweb.openeuler.org/archives/list/kernel@openeuler.org/message/2SW... FeedBack: The patch(es) which you have sent to kernel@openeuler.org mailing list has been converted to a pull request successfully! Pull request link: https://atomgit.com/openeuler/kernel/merge_requests/26908 Mailing list address: https://mailweb.openeuler.org/archives/list/kernel@openeuler.org/message/2SW...
participants (2)
-
Jinjiang Tu -
patchwork bot