This add documentation on how HMM works and a more in depth view of how it should be use by device driver writers. Signed-off-by: Jérôme Glisse <jglisse@xxxxxxxxxx> --- Documentation/vm/hmm.txt | 219 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 219 insertions(+) create mode 100644 Documentation/vm/hmm.txt diff --git a/Documentation/vm/hmm.txt b/Documentation/vm/hmm.txt new file mode 100644 index 0000000..febed50 --- /dev/null +++ b/Documentation/vm/hmm.txt @@ -0,0 +1,219 @@ +Heterogeneous Memory Management (HMM) +------------------------------------- + +The raison d'�e of HMM is to provide a common API for device driver that +wants to mirror a process address space on there device and/or migrate system +memory to device memory. Device driver can decide to only use one aspect of +HMM (mirroring or memory migration), for instance some device can directly +access process address space through hardware (for instance PCIe ATS/PASID), +but still want to benefit from memory migration capabilities that HMM offer. + +While HMM rely on existing kernel infrastructure (namely mmu_notifier) some +of its features (memory migration, atomic access) require integration with +core mm kernel code. Having HMM as the common intermediary is more appealing +than having each device driver hooking itself inside the common mm code. + +Moreover HMM as a layer allows integration with DMA API or page reclaimation. + + +Mirroring address space on the device: +-------------------------------------- + +Device that can't directly access transparently the process address space, need +to mirror the CPU page table into there own page table. HMM helps to keep the +device page table synchronize with the CPU page table. It is not expected that +the device will fully mirror the CPU page table but only mirror region that are +actively accessed by the device. For that reasons HMM only helps populating and +synchronizing device page table for range that the device driver explicitly ask +for. + +Mirroring address space inside the device page table is easy with HMM : + + /* Create a mirror for the current process for your device. */ + your_hmm_mirror->hmm_mirror.device = your_hmm_device; + hmm_mirror_register(&your_hmm_mirror->hmm_mirror); + + ... + + /* Mirror memory (in read mode) between addressA and addressB */ + your_hmm_event->hmm_event.start = addressA; + your_hmm_event->hmm_event.end = addressB; + your_hmm_event->hmm_event.etype = HMM_DEVICE_RFAULT; + hmm_mirror_fault(&your_hmm_mirror->hmm_mirror, &your_hmm_event->hmm_event); + /* HMM callback into your driver with the >update() callback. During the + * callback use the HMM page table to populate the device page table. You + * can only use the HMM page table to populate the device page table for + * the specified range during the >update() callback, at any other point in + * time the HMM page table content should be assume to be undefined. + */ + your_hmm_device->update(mirror, event); + + ... + + /* Process is quiting or device done stop the mirroring and cleanup. */ + hmm_mirror_unregister(&your_hmm_mirror->hmm_mirror); + /* Device driver can free your_hmm_mirror */ + + +HMM mirror page table: +---------------------- + +Each hmm_mirror object is associated with a mirror page table that HMM keeps +synchronize with the CPU page table by using the mmu_notifier API. HMM is using +its own generic page table format because it needs to store DMA address, which +are bigger than long on some architecture, and have more flags per entry than +radix tree allows. + +The HMM page table mostly mirror x86 page table layout. A page holds a global +directory and each entry points to a lower level directory. Unlike regular CPU +page table, directory level are more aggressively freed and remove from the HMM +mirror page table. This means device driver needs to use the HMM helpers and to +follow directive on when and how to access the mirror page table. HMM use the +per page spinlock of directory page to synchronize update of directory ie update +can happen on different directory concurently. + +As a rules the mirror page table can only be accessed by device driver from one +of the HMM device callback. Any access from outside a callback is illegal and +gives undertimed result. + +Accessing the mirror page table from a device callback needs to use the HMM +page table helpers. A loop to access entry for a range of address looks like : + + /* Initialize a HMM page table iterator. */ + struct hmm_pt_iter iter; + hmm_pt_iter_init(&iter, &mirror->pt) + + /* Get pointer to HMM page table entry for a given address. */ + dma_addr_t *hmm_pte; + hmm_pte = hmm_pt_iter_walk(&iter, &addr, &next); + +If there is no valid entry directory for given range address then hmm_pte is +NULL. If there is a valid entry directory then you can access the hmm_pte and +the pointer will stay valid as long as you do not call hmm_pt_iter_walk() with +the same iter struct for a different address or call hmm_pt_iter_fini(). + +While the HMM page table entry pointer stays valid you can only modify the +value it is pointing to by using one of HMM helpers (hmm_pte_*()) as other +threads might be updating the same entry concurrently. The device driver only +need to update an HMM page table entry to set the dirty bit, so driver should +only be using hmm_pte_set_dirty(). + +Similarly to extract information the device driver should use one of the helper +like hmm_pte_dma_addr() or hmm_pte_pfn() (if HMM is not doing DMA mapping which +is a device driver at initialization parameter). + + +Migrating system memory to device memory: +----------------------------------------- + +Device like discret GPU often have there own local memory which offer bigger +bandwidth and smaller latency than access to system memory for the GPU. This +local memory is not necessarily accessible by the CPU. Device local memory will +remain revealent for the foreseeable future as bandwidth of GPU memory keep +increasing faster than bandwidth of system memory and as latency of PCIe does +not decrease. + +Thus to maximize use of device like GPU, program need to use the device memory. +Userspace API wants to make this as transparent as it can be, so that there is +no need for complex modification of applications. + +Transparent use of device memory for range of address of a process require core +mm code modifications. Adding a new memory zone for devices memory did not make +sense given that such memory is often only accessible by the device only. This +is why we decided to use a special kind of swap, migrated memory is mark as a +special swap entry inside the CPU page table. + +While HMM handles the migration process, it does not decide what range or when +to migrate memory. The decision to perform such migration is under the control +of the device driver. Migration back to system memory happens either because +the CPU try to access the memory or because device driver decided to migrate +the memory back. + + + /* Migrate system memory between addressA and addressB to device memory. */ + your_hmm_event->hmm_event.start = addressA; + your_hmm_event->hmm_event.end = addressB; + your_hmm_event->hmm_event.etype = HMM_COPY_TO_DEVICE; + hmm_mirror_fault(&your_hmm_mirror->hmm_mirror, &your_hmm_event->hmm_event); + /* HMM callback into your driver with the >copy_to_device() callback. + * Device driver must allocate device memory, DMA system memory to device + * memory, update the device page table to point to device memory and + * return. See hmm.h for details instructions and how failure are handled. + */ + your_hmm_device->copy_to_device(mirror, event, dst, addressA, addressB); + + +Right now HMM only support migrating anonymous private memory. Migration of +share memory and more generaly file mapped memory is on the road map. + + +Locking consideration and overall design: +----------------------------------------- + +As a rule HMM will handle proper locking on the behalf of the device driver, +as such device driver does not need to take any mm lock before calling into +the HMM code. + +HMM is also responsible of the hmm_device and hmm_mirror object lifetime. The +device driver can only free those after calling hmm_device_unregister() or +hmm_mirror_unregister() respectively. + +All the lock inside any of the HMM structure should never be use by the device +driver. They are intended to be use only and only by HMM code. Below is short +description of the 3 main locks that exist for HMM internal use. Educational +purpose only. + +Each process mm has one and only one struct hmm associated with it. Each hmm +struct can be use by several different mirror. There is one and only one mirror +per mm and device pair. So in essence the hmm struct is the core that dispatch +everything to every single mirror, each of them corresponding to a specific +device. The list of mirror for an hmm struct is protected by a semaphore as it +sees mostly read access. + +Each time a device fault a range of address it calls hmm_mirror_fault(), HMM +keeps track, inside the hmm struct, of each range currently being faulted. It +does that so it can synchronize with any CPU page table update. If there is a +CPU page table update then a callback through mmu_notifier will happen and HMM +will try to interrupt the device page fault that conflict (ie address range +overlap with the range being updated) and wait for them to back off. This +insure that at no point in time the device driver see transient page table +information. The list of active fault is protected by a spinlock, query on +that list should be short and quick (we haven't gather enough statistic on +that side yet to have a good idea of the average access pattern). + +Each device driver wanting to use HMM must register one and only one hmm_device +struct per physical device with HMM. The hmm_device struct have pointer to the +device driver call back and keeps track of active mirrors for a given device. +The active mirrors list is protected by a spinlock. + + +Future work: +------------ + +Improved atomic access by the device to system memory. Some platform bus (PCIe) +offer limited number of atomic memory operations, some platform do not even +have any kind of atomic memory operations by a device. In order to allow such +atomic operation we want to map page read only the CPU while the device perform +its operation. For this we need a new case inside the CPU write fault code path +to synchronize with the device. + +We want to allow program to lock a range of memory inside device memory and +forbid CPU access while the memory is lock inside the device. Any CPU access +to locked range would result in SIGBUS. We think that madvise() would be the +right syscall into which we could plug that feature. + +In order to minimize kernel memory consumption and overhead of DMA mapping, we +want to introduce new DMA API that allows to manage mapping on IOMMU directory +page basis. This would allow to map/unmap/update DMA mapping in bulk and +minimize IOMMU update and flushing overhead. Moreover this would allow to +improve IOMMU bad access reporting for DMA address inside those directory. + +Because update to the device page table might require "heavy" synchronization +with the device, the mmu_notifier callback might have to sleep while HMM is +waiting for the device driver to report device page table update completion. +This is especialy bad if this happens during page reclaimation, this might +bring the system to pause. We want to mitigate this, either by maintaining a +new intermediate lru level in which we put pages actively mirrored by a device +or by some other mecanism. For time being we advice that device driver that +use HMM explicitly explain this corner case so that user are aware that this +can happens if there is memory pressure. -- 2.4.3 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx"> email@xxxxxxxxx </a>