kernel/linux.git/drivers/dax/kmem.c, branch v6.12.80

dax: add missing MODULE_DESCRIPTION() macros

2024-06-17T23:42:16+00:00

make allmodconfig && make W=1 C=1 reports: WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/dax/hmem/dax_hmem.o WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/dax/device_dax.o WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/dax/kmem.o WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/dax/dax_pmem.o WARNING: modpost: missing MODULE_DESCRIPTION() in drivers/dax/dax_cxl.o Add all missing invocations of the MODULE_DESCRIPTION() macro. [iweiny: edit descriptions] Signed-off-by: Jeff Johnson Link: https://patch.msgid.link/r/20240605-md-drivers-dax-v1-1-3d448f3368b4@quicinc.com Signed-off-by: Ira Weiny

memory tier: dax/kmem: introduce an abstract layer for finding, allocating, and putting memory types

2024-05-06T00:53:25+00:00

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@PH0PR08MB7955.namprd08.prod.outlook.com/T/ 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 Reviewed-by: "Huang, Ying" Reviewed-by: Jonathan Cameron Cc: Alistair Popple Cc: Aneesh Kumar K.V Cc: Dan Williams Cc: Dave Jiang Cc: Gregory Price Cc: Hao Xiang Cc: Jonathan Cameron Cc: Michal Hocko Cc: Ravi Jonnalagadda Cc: SeongJae Park Cc: Tejun Heo Cc: Vishal Verma Signed-off-by: Andrew Morton

dax/kmem: allow kmem to add memory with memmap_on_memory

2023-12-11T00:51:35+00:00

Large amounts of memory managed by the kmem driver may come in via CXL, and it is often desirable to have the memmap for this memory on the new memory itself. Enroll kmem-managed memory for memmap_on_memory semantics if the dax region originates via CXL. For non-CXL dax regions, retain the existing default behavior of hot adding without memmap_on_memory semantics. Link: https://lkml.kernel.org/r/20231107-vv-kmem_memmap-v10-3-1253ec050ed0@intel.com Signed-off-by: Vishal Verma Reviewed-by: Jonathan Cameron Reviewed-by: David Hildenbrand Reviewed-by: "Huang, Ying" Tested-by: Li Zhijian [cxl.kmem and nvdimm.kmem] Cc: Michal Hocko Cc: Oscar Salvador Cc: Dan Williams Cc: Dave Jiang Cc: Dave Hansen Cc: Aneesh Kumar K.V Cc: Fan Ni Cc: Jeff Moyer Signed-off-by: Andrew Morton

dax, kmem: calculate abstract distance with general interface

2023-10-16T22:44:39+00:00

Previously, a fixed abstract distance MEMTIER_DEFAULT_DAX_ADISTANCE is used for slow memory type in kmem driver. This limits the usage of kmem driver, for example, it cannot be used for HBM (high bandwidth memory). So, we use the general abstract distance calculation mechanism in kmem drivers to get more accurate abstract distance on systems with proper support. The original MEMTIER_DEFAULT_DAX_ADISTANCE is used as fallback only. Now, multiple memory types may be managed by kmem. These memory types are put into the "kmem_memory_types" list and protected by kmem_memory_type_lock. Link: https://lkml.kernel.org/r/20230926060628.265989-5-ying.huang@intel.com Signed-off-by: "Huang, Ying" Tested-by: Bharata B Rao Reviewed-by: Dave Jiang Reviewed-by: Alistair Popple Cc: Aneesh Kumar K.V Cc: Wei Xu Cc: Dan Williams Cc: Dave Hansen Cc: Davidlohr Bueso Cc: Johannes Weiner Cc: Jonathan Cameron Cc: Michal Hocko Cc: Yang Shi Cc: Rafael J Wysocki Signed-off-by: Andrew Morton

memory tier: rename destroy_memory_type() to put_memory_type()

2023-08-18T17:12:11+00:00

It appears that destroy_memory_type() isn't a very good name because we usually will not free the memory_type here. So rename it to a more appropriate name i.e. put_memory_type(). Link: https://lkml.kernel.org/r/20230706063905.543800-1-linmiaohe@huawei.com Signed-off-by: Miaohe Lin Suggested-by: Huang, Ying Reviewed-by: "Huang, Ying" Reviewed-by: Xiao Yang Cc: Dan Williams Cc: Dave Jiang Cc: Vishal Verma Signed-off-by: Andrew Morton

dax/kmem: Pass valid argument to memory_group_register_static

2023-06-23T07:32:06+00:00

memory_group_register_static takes maximum number of pages as the argument while dev_dax_kmem_probe passes total_len (in bytes) as the argument. IIUC, I don't see any crash/panic impact as such. As, memory_group_register_static just set the max_pages limit which is used in auto_movable_zone_for_pfn to determine the zone. which might cause these condition to behave differently, This will be true always so jump will happen to kernel_zone ... if (!auto_movable_can_online_movable(NUMA_NO_NODE, group, nr_pages)) goto kernel_zone; ... kernel_zone: return default_kernel_zone_for_pfn(nid, pfn, nr_pages); Here, In below, zone_intersects compare range will be larger as nr_pages will be higher (derived from total_len passed in dev_dax_kmem_probe). ... static struct zone *default_kernel_zone_for_pfn(int nid, unsigned long start_pfn, unsigned long nr_pages) { struct pglist_data *pgdat = NODE_DATA(nid); int zid; for (zid = 0; zid < ZONE_NORMAL; zid++) { struct zone *zone = &pgdat->node_zones[zid]; if (zone_intersects(zone, start_pfn, nr_pages)) return zone; } return &pgdat->node_zones[ZONE_NORMAL]; } Incorrect zone will be returned here, which in later time might cause bigger problem. Fixes: eedf634aac3b ("dax/kmem: use a single static memory group for a single probed unit") Signed-off-by: Tarun Sahu Link: https://lore.kernel.org/r/20230621155025.370672-1-tsahu@linux.ibm.com Reviewed-by: Vishal Verma Signed-off-by: Vishal Verma

dax/kmem: Fix leak of memory-hotplug resources

2023-02-17T22:58:01+00:00

While experimenting with CXL region removal the following corruption of /proc/iomem appeared. Before: f010000000-f04fffffff : CXL Window 0 f010000000-f02fffffff : region4 f010000000-f02fffffff : dax4.0 f010000000-f02fffffff : System RAM (kmem) After (modprobe -r cxl_test): f010000000-f02fffffff : **redacted binary garbage** f010000000-f02fffffff : System RAM (kmem) ...and testing further the same is visible with persistent memory assigned to kmem: Before: 480000000-243fffffff : Persistent Memory 480000000-57e1fffff : namespace3.0 580000000-243fffffff : dax3.0 580000000-243fffffff : System RAM (kmem) After (ndctl disable-region all): 480000000-243fffffff : Persistent Memory 580000000-243fffffff : ***redacted binary garbage*** 580000000-243fffffff : System RAM (kmem) The corrupted data is from a use-after-free of the "dax4.0" and "dax3.0" resources, and it also shows that the "System RAM (kmem)" resource is not being removed. The bug does not appear after "modprobe -r kmem", it requires the parent of "dax4.0" and "dax3.0" to be removed which re-parents the leaked "System RAM (kmem)" instances. Those in turn reference the freed resource as a parent. First up for the fix is release_mem_region_adjustable() needs to reliably delete the resource inserted by add_memory_driver_managed(). That is thwarted by a check for IORESOURCE_SYSRAM that predates the dax/kmem driver, from commit: 65c78784135f ("kernel, resource: check for IORESOURCE_SYSRAM in release_mem_region_adjustable") That appears to be working around the behavior of HMM's "MEMORY_DEVICE_PUBLIC" facility that has since been deleted. With that check removed the "System RAM (kmem)" resource gets removed, but corruption still occurs occasionally because the "dax" resource is not reliably removed. The dax range information is freed before the device is unregistered, so the driver can not reliably recall (another use after free) what it is meant to release. Lastly if that use after free got lucky, the driver was covering up the leak of "System RAM (kmem)" due to its use of release_resource() which detaches, but does not free, child resources. The switch to remove_resource() forces remove_memory() to be responsible for the deletion of the resource added by add_memory_driver_managed(). Fixes: c2f3011ee697 ("device-dax: add an allocation interface for device-dax instances") Cc: Cc: Oscar Salvador Cc: David Hildenbrand Cc: Pavel Tatashin Reviewed-by: Vishal Verma Reviewed-by: Pasha Tatashin Reviewed-by: Dave Jiang Link: https://lore.kernel.org/r/167653656244.3147810.5705900882794040229.stgit@dwillia2-xfh.jf.intel.com Signed-off-by: Dan Williams

dax: Assign RAM regions to memory-hotplug by default

2023-02-11T01:33:40+00:00

The default mode for device-dax instances is backwards for RAM-regions as evidenced by the fact that it tends to catch end users by surprise. "Where is my memory?". Recall that platforms are increasingly shipping with performance-differentiated memory pools beyond typical DRAM and NUMA effects. This includes HBM (high-bandwidth-memory) and CXL (dynamic interleave, varied media types, and future fabric attached possibilities). For this reason the EFI_MEMORY_SP (EFI Special Purpose Memory => Linux 'Soft Reserved') attribute is expected to be applied to all memory-pools that are not the general purpose pool. This designation gives an Operating System a chance to defer usage of a memory pool until later in the boot process where its performance properties can be interrogated and administrator policy can be applied. 'Soft Reserved' memory can be anything from too limited and precious to be part of the general purpose pool (HBM), too slow to host hot kernel data structures (some PMEM media), or anything in between. However, in the absence of an explicit policy, the memory should at least be made usable by default. The current device-dax default hides all non-general-purpose memory behind a device interface. The expectation is that the distribution of users that want the memory online by default vs device-dedicated-access by default follows the Pareto principle. A small number of enlightened users may want to do userspace memory management through a device, but general users just want the kernel to make the memory available with an option to get more advanced later. Arrange for all device-dax instances not backed by PMEM to default to attaching to the dax_kmem driver. From there the baseline memory hotplug policy (CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE / memhp_default_state=) gates whether the memory comes online or stays offline. Where, if it stays offline, it can be reliably converted back to device-mode where it can be partitioned, or fronted by a userspace allocator. So, if someone wants device-dax instances for their 'Soft Reserved' memory: 1/ Build a kernel with CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE=n or boot with memhp_default_state=offline, or roll the dice and hope that the kernel has not pinned a page in that memory before step 2. 2/ Write a udev rule to convert the target dax device(s) from 'system-ram' mode to 'devdax' mode: daxctl reconfigure-device $dax -m devdax -f Cc: Michal Hocko Cc: David Hildenbrand Cc: Dave Hansen Reviewed-by: Gregory Price Tested-by: Fan Ni Reviewed-by: Dave Jiang Link: https://lore.kernel.org/r/167602003336.1924368.6809503401422267885.stgit@dwillia2-xfh.jf.intel.com Signed-off-by: Dan Williams

mm/demotion/dax/kmem: set node's abstract distance to MEMTIER_DEFAULT_DAX_ADISTANCE

2022-09-27T02:46:11+00:00

By default, all nodes are assigned to the default memory tier which is the memory tier designated for nodes with DRAM Set dax kmem device node's tier to slower memory tier by assigning abstract distance to MEMTIER_DEFAULT_DAX_ADISTANCE. Low-level drivers like papr_scm or ACPI NFIT can initialize memory device type to a more accurate value based on device tree details or HMAT. If the kernel doesn't find the memory type initialized, a default slower memory type is assigned by the kmem driver. [aneesh.kumar@linux.ibm.com: assign correct memory type for multiple dax devices with the same node affinity] Link: https://lkml.kernel.org/r/20220826100224.542312-1-aneesh.kumar@linux.ibm.com Link: https://lkml.kernel.org/r/20220818131042.113280-5-aneesh.kumar@linux.ibm.com Signed-off-by: Aneesh Kumar K.V Reviewed-by: "Huang, Ying" Acked-by: Wei Xu Cc: Alistair Popple Cc: Bharata B Rao Cc: Dan Williams Cc: Dave Hansen Cc: Davidlohr Bueso Cc: Hesham Almatary Cc: Jagdish Gediya Cc: Johannes Weiner Cc: Jonathan Cameron Cc: Michal Hocko Cc: Tim Chen Cc: Yang Shi Cc: SeongJae Park Signed-off-by: Andrew Morton

dax/kmem: use a single static memory group for a single probed unit

2021-09-08T18:50:23+00:00

Although dax/kmem users often disable auto-onlining and instead online memory manually (usually to ZONE_MOVABLE), there is still value in having auto-onlining be aware of the relationship of memory blocks. Let's treat one probed unit as a single static memory device, similar to a single ACPI memory device. Link: https://lkml.kernel.org/r/20210806124715.17090-7-david@redhat.com Signed-off-by: David Hildenbrand Cc: Anshuman Khandual Cc: Dan Williams Cc: Dave Hansen Cc: Greg Kroah-Hartman Cc: Hui Zhu Cc: Jason Wang Cc: Len Brown Cc: Marek Kedzierski Cc: "Michael S. Tsirkin" Cc: Michal Hocko Cc: Mike Rapoport Cc: Oscar Salvador Cc: Pankaj Gupta Cc: Pavel Tatashin Cc: Rafael J. Wysocki Cc: "Rafael J. Wysocki" Cc: Vitaly Kuznetsov Cc: Vlastimil Babka Cc: Wei Yang Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds