From d84f2f5a755208da3f93e17714631485cb3da11c Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Mon, 23 Sep 2019 15:35:40 -0700 Subject: drivers/base/node.c: simplify unregister_memory_block_under_nodes() We don't allow to offline memory block devices that belong to multiple numa nodes. Therefore, such devices can never get removed. It is sufficient to process a single node when removing the memory block. No need to iterate over each and every PFN. We already have the nid stored for each memory block. Make sure that the nid always has a sane value. Please note that checking for node_online(nid) is not required. If we would have a memory block belonging to a node that is no longer offline, then we would have a BUG in the node offlining code. Link: http://lkml.kernel.org/r/20190719135244.15242-1-david@redhat.com Signed-off-by: David Hildenbrand Cc: Greg Kroah-Hartman Cc: "Rafael J. Wysocki" Cc: David Hildenbrand Cc: Stephen Rothwell Cc: Pavel Tatashin Cc: Michal Hocko Cc: Oscar Salvador Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/base/memory.c | 1 + 1 file changed, 1 insertion(+) (limited to 'drivers/base/memory.c') diff --git a/drivers/base/memory.c b/drivers/base/memory.c index 20c39d1bcef8..154d5d4a0779 100644 --- a/drivers/base/memory.c +++ b/drivers/base/memory.c @@ -674,6 +674,7 @@ static int init_memory_block(struct memory_block **memory, mem->state = state; start_pfn = section_nr_to_pfn(mem->start_section_nr); mem->phys_device = arch_get_memory_phys_device(start_pfn); + mem->nid = NUMA_NO_NODE; ret = register_memory(mem); -- cgit v1.2.3-70-g09d2 From f915fb7fb2c1c57be05880012b46d2edd5124797 Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Mon, 23 Sep 2019 15:35:43 -0700 Subject: drivers/base/memory.c: fixup documentation of removable/phys_index/block_size_bytes Let's rephrase to memory block terminology and add some further clarifications. Link: http://lkml.kernel.org/r/20190806080826.5963-1-david@redhat.com Signed-off-by: David Hildenbrand Reviewed-by: Andrew Morton Cc: Greg Kroah-Hartman Cc: "Rafael J. Wysocki" Cc: Michal Hocko Cc: Oscar Salvador Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/base/memory.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) (limited to 'drivers/base/memory.c') diff --git a/drivers/base/memory.c b/drivers/base/memory.c index 154d5d4a0779..63fc5aa51c21 100644 --- a/drivers/base/memory.c +++ b/drivers/base/memory.c @@ -116,10 +116,8 @@ static unsigned long get_memory_block_size(void) } /* - * use this as the physical section index that this memsection - * uses. + * Show the first physical section index (number) of this memory block. */ - static ssize_t phys_index_show(struct device *dev, struct device_attribute *attr, char *buf) { @@ -131,7 +129,10 @@ static ssize_t phys_index_show(struct device *dev, } /* - * Show whether the section of memory is likely to be hot-removable + * Show whether the memory block is likely to be offlineable (or is already + * offline). Once offline, the memory block could be removed. The return + * value does, however, not indicate that there is a way to remove the + * memory block. */ static ssize_t removable_show(struct device *dev, struct device_attribute *attr, char *buf) @@ -455,7 +456,7 @@ static DEVICE_ATTR_RO(phys_device); static DEVICE_ATTR_RO(removable); /* - * Block size attribute stuff + * Show the memory block size (shared by all memory blocks). */ static ssize_t block_size_bytes_show(struct device *dev, struct device_attribute *attr, char *buf) -- cgit v1.2.3-70-g09d2 From 902ce63b337381092ff865f542e854ff3d0ebe2b Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Mon, 23 Sep 2019 15:35:46 -0700 Subject: driver/base/memory.c: validate memory block size early Let's validate the memory block size early, when initializing the memory device infrastructure. Fail hard in case the value is not suitable. As nobody checks the return value of memory_dev_init(), turn it into a void function and fail with a panic in all scenarios instead. Otherwise, we'll crash later during boot when core/drivers expect that the memory device infrastructure (including memory_block_size_bytes()) works as expected. I think long term, we should move the whole memory block size configuration (set_memory_block_size_order() and memory_block_size_bytes()) into drivers/base/memory.c. Link: http://lkml.kernel.org/r/20190806090142.22709-1-david@redhat.com Signed-off-by: David Hildenbrand Cc: Greg Kroah-Hartman Cc: "Rafael J. Wysocki" Cc: Pavel Tatashin Cc: Michal Hocko Cc: Dan Williams Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/base/memory.c | 31 +++++++++---------------------- include/linux/memory.h | 6 +++--- 2 files changed, 12 insertions(+), 25 deletions(-) (limited to 'drivers/base/memory.c') diff --git a/drivers/base/memory.c b/drivers/base/memory.c index 63fc5aa51c21..763154c82eb3 100644 --- a/drivers/base/memory.c +++ b/drivers/base/memory.c @@ -100,21 +100,6 @@ unsigned long __weak memory_block_size_bytes(void) } EXPORT_SYMBOL_GPL(memory_block_size_bytes); -static unsigned long get_memory_block_size(void) -{ - unsigned long block_sz; - - block_sz = memory_block_size_bytes(); - - /* Validate blk_sz is a power of 2 and not less than section size */ - if ((block_sz & (block_sz - 1)) || (block_sz < MIN_MEMORY_BLOCK_SIZE)) { - WARN_ON(1); - block_sz = MIN_MEMORY_BLOCK_SIZE; - } - - return block_sz; -} - /* * Show the first physical section index (number) of this memory block. */ @@ -461,7 +446,7 @@ static DEVICE_ATTR_RO(removable); static ssize_t block_size_bytes_show(struct device *dev, struct device_attribute *attr, char *buf) { - return sprintf(buf, "%lx\n", get_memory_block_size()); + return sprintf(buf, "%lx\n", memory_block_size_bytes()); } static DEVICE_ATTR_RO(block_size_bytes); @@ -812,19 +797,22 @@ static const struct attribute_group *memory_root_attr_groups[] = { /* * Initialize the sysfs support for memory devices... */ -int __init memory_dev_init(void) +void __init memory_dev_init(void) { int ret; int err; unsigned long block_sz, nr; + /* Validate the configured memory block size */ + block_sz = memory_block_size_bytes(); + if (!is_power_of_2(block_sz) || block_sz < MIN_MEMORY_BLOCK_SIZE) + panic("Memory block size not suitable: 0x%lx\n", block_sz); + sections_per_block = block_sz / MIN_MEMORY_BLOCK_SIZE; + ret = subsys_system_register(&memory_subsys, memory_root_attr_groups); if (ret) goto out; - block_sz = get_memory_block_size(); - sections_per_block = block_sz / MIN_MEMORY_BLOCK_SIZE; - /* * Create entries for memory sections that were found * during boot and have been initialized @@ -840,8 +828,7 @@ int __init memory_dev_init(void) out: if (ret) - printk(KERN_ERR "%s() failed: %d\n", __func__, ret); - return ret; + panic("%s() failed: %d\n", __func__, ret); } /** diff --git a/include/linux/memory.h b/include/linux/memory.h index 02e633f3ede0..b3d9df060626 100644 --- a/include/linux/memory.h +++ b/include/linux/memory.h @@ -80,9 +80,9 @@ struct mem_section; #define IPC_CALLBACK_PRI 10 #ifndef CONFIG_MEMORY_HOTPLUG_SPARSE -static inline int memory_dev_init(void) +static inline void memory_dev_init(void) { - return 0; + return; } static inline int register_memory_notifier(struct notifier_block *nb) { @@ -113,7 +113,7 @@ extern int register_memory_isolate_notifier(struct notifier_block *nb); extern void unregister_memory_isolate_notifier(struct notifier_block *nb); int create_memory_block_devices(unsigned long start, unsigned long size); void remove_memory_block_devices(unsigned long start, unsigned long size); -extern int memory_dev_init(void); +extern void memory_dev_init(void); extern int memory_notify(unsigned long val, void *v); extern int memory_isolate_notify(unsigned long val, void *v); extern struct memory_block *find_memory_block(struct mem_section *); -- cgit v1.2.3-70-g09d2 From b6c88d3b9d38f9448e0fcf44847a075ea81d5ca2 Mon Sep 17 00:00:00 2001 From: David Hildenbrand Date: Mon, 23 Sep 2019 15:35:49 -0700 Subject: drivers/base/memory.c: don't store end_section_nr in memory blocks Each memory block spans the same amount of sections/pages/bytes. The size is determined before the first memory block is created. No need to store what we can easily calculate - and the calculations even look simpler now. Michal brought up the idea of variable-sized memory blocks. However, if we ever implement something like this, we will need an API compatibility switch and reworks at various places (most code assumes a fixed memory block size). So let's cleanup what we have right now. While at it, fix the variable naming in register_mem_sect_under_node() - we no longer talk about a single section. Link: http://lkml.kernel.org/r/20190809110200.2746-1-david@redhat.com Signed-off-by: David Hildenbrand Cc: Greg Kroah-Hartman Cc: "Rafael J. Wysocki" Cc: Pavel Tatashin Cc: Michal Hocko Cc: Dan Williams Cc: Oscar Salvador Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/base/memory.c | 1 - drivers/base/node.c | 10 +++++----- include/linux/memory.h | 1 - mm/memory_hotplug.c | 2 +- 4 files changed, 6 insertions(+), 8 deletions(-) (limited to 'drivers/base/memory.c') diff --git a/drivers/base/memory.c b/drivers/base/memory.c index 763154c82eb3..6bea4f3f8040 100644 --- a/drivers/base/memory.c +++ b/drivers/base/memory.c @@ -656,7 +656,6 @@ static int init_memory_block(struct memory_block **memory, return -ENOMEM; mem->start_section_nr = block_id * sections_per_block; - mem->end_section_nr = mem->start_section_nr + sections_per_block - 1; mem->state = state; start_pfn = section_nr_to_pfn(mem->start_section_nr); mem->phys_device = arch_get_memory_phys_device(start_pfn); diff --git a/drivers/base/node.c b/drivers/base/node.c index 840c95baa1d8..257449cf061f 100644 --- a/drivers/base/node.c +++ b/drivers/base/node.c @@ -756,13 +756,13 @@ static int __ref get_nid_for_pfn(unsigned long pfn) static int register_mem_sect_under_node(struct memory_block *mem_blk, void *arg) { + unsigned long memory_block_pfns = memory_block_size_bytes() / PAGE_SIZE; + unsigned long start_pfn = section_nr_to_pfn(mem_blk->start_section_nr); + unsigned long end_pfn = start_pfn + memory_block_pfns - 1; int ret, nid = *(int *)arg; - unsigned long pfn, sect_start_pfn, sect_end_pfn; + unsigned long pfn; - sect_start_pfn = section_nr_to_pfn(mem_blk->start_section_nr); - sect_end_pfn = section_nr_to_pfn(mem_blk->end_section_nr); - sect_end_pfn += PAGES_PER_SECTION - 1; - for (pfn = sect_start_pfn; pfn <= sect_end_pfn; pfn++) { + for (pfn = start_pfn; pfn <= end_pfn; pfn++) { int page_nid; /* diff --git a/include/linux/memory.h b/include/linux/memory.h index b3d9df060626..0ebb105eb261 100644 --- a/include/linux/memory.h +++ b/include/linux/memory.h @@ -25,7 +25,6 @@ struct memory_block { unsigned long start_section_nr; - unsigned long end_section_nr; unsigned long state; /* serialized by the dev->lock */ int section_count; /* serialized by mem_sysfs_mutex */ int online_type; /* for passing data to online routine */ diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 5b8811945bbb..3706a137d880 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -1654,7 +1654,7 @@ static int check_memblock_offlined_cb(struct memory_block *mem, void *arg) phys_addr_t beginpa, endpa; beginpa = PFN_PHYS(section_nr_to_pfn(mem->start_section_nr)); - endpa = PFN_PHYS(section_nr_to_pfn(mem->end_section_nr + 1))-1; + endpa = beginpa + memory_block_size_bytes() - 1; pr_warn("removing memory fails, because memory [%pa-%pa] is onlined\n", &beginpa, &endpa); -- cgit v1.2.3-70-g09d2