the driver callback returns.
When the device driver wants to populate a range of virtual addresses, it can
-use either::
+use::
- long hmm_range_snapshot(struct hmm_range *range);
- long hmm_range_fault(struct hmm_range *range, bool block);
+ long hmm_range_fault(struct hmm_range *range, unsigned int flags);
-The first one (hmm_range_snapshot()) will only fetch present CPU page table
+With the HMM_RANGE_SNAPSHOT flag, it will only fetch present CPU page table
entries and will not trigger a page fault on missing or non-present entries.
-The second one does trigger a page fault on missing or read-only entries if
-write access is requested (see below). Page faults use the generic mm page
+Without that flag, it does trigger a page fault on missing or read-only entries
+if write access is requested (see below). Page faults use the generic mm page
fault code path just like a CPU page fault.
Both functions copy CPU page table entries into their pfns array argument. Each
range.flags = ...;
range.values = ...;
range.pfn_shift = ...;
- hmm_range_register(&range);
+ hmm_range_register(&range, mirror);
/*
* Just wait for range to be valid, safe to ignore return value as we
- * will use the return value of hmm_range_snapshot() below under the
+ * will use the return value of hmm_range_fault() below under the
* mmap_sem to ascertain the validity of the range.
*/
hmm_range_wait_until_valid(&range, TIMEOUT_IN_MSEC);
again:
down_read(&mm->mmap_sem);
- ret = hmm_range_snapshot(&range);
+ ret = hmm_range_fault(&range, HMM_RANGE_SNAPSHOT);
if (ret) {
up_read(&mm->mmap_sem);
if (ret == -EBUSY) {
/*
* No need to check hmm_range_wait_until_valid() return value
- * on retry we will get proper error with hmm_range_snapshot()
+ * on retry we will get proper error with hmm_range_fault()
*/
hmm_range_wait_until_valid(&range, TIMEOUT_IN_MSEC);
goto again;
===================================
Because the CPU cannot access device memory, migration must use the device DMA
-engine to perform copy from and to device memory. For this we need a new
-migration helper::
-
- int migrate_vma(const struct migrate_vma_ops *ops,
- struct vm_area_struct *vma,
- unsigned long mentries,
- unsigned long start,
- unsigned long end,
- unsigned long *src,
- unsigned long *dst,
- void *private);
-
-Unlike other migration functions it works on a range of virtual address, there
-are two reasons for that. First, device DMA copy has a high setup overhead cost
-and thus batching multiple pages is needed as otherwise the migration overhead
-makes the whole exercise pointless. The second reason is because the
-migration might be for a range of addresses the device is actively accessing.
-
-The migrate_vma_ops struct defines two callbacks. First one (alloc_and_copy())
-controls destination memory allocation and copy operation. Second one is there
-to allow the device driver to perform cleanup operations after migration::
-
- struct migrate_vma_ops {
- void (*alloc_and_copy)(struct vm_area_struct *vma,
- const unsigned long *src,
- unsigned long *dst,
- unsigned long start,
- unsigned long end,
- void *private);
- void (*finalize_and_map)(struct vm_area_struct *vma,
- const unsigned long *src,
- const unsigned long *dst,
- unsigned long start,
- unsigned long end,
- void *private);
- };
-
-It is important to stress that these migration helpers allow for holes in the
-virtual address range. Some pages in the range might not be migrated for all
-the usual reasons (page is pinned, page is locked, ...). This helper does not
-fail but just skips over those pages.
-
-The alloc_and_copy() might decide to not migrate all pages in the
-range (for reasons under the callback control). For those, the callback just
-has to leave the corresponding dst entry empty.
-
-Finally, the migration of the struct page might fail (for file backed page) for
-various reasons (failure to freeze reference, or update page cache, ...). If
-that happens, then the finalize_and_map() can catch any pages that were not
-migrated. Note those pages were still copied to a new page and thus we wasted
-bandwidth but this is considered as a rare event and a price that we are
-willing to pay to keep all the code simpler.
+engine to perform copy from and to device memory. For this we need to use
+migrate_vma_setup(), migrate_vma_pages(), and migrate_vma_finalize() helpers.
Memory cgroup (memcg) and rss accounting
config DRM_AMDGPU_USERPTR
bool "Always enable userptr write support"
depends on DRM_AMDGPU
- depends on HMM_MIRROR
+ depends on MMU
+ select HMM_MIRROR
+ select MMU_NOTIFIER
help
This option selects CONFIG_HMM and CONFIG_HMM_MIRROR if it
isn't already selected to enabled full userptr support.
#include <linux/pm_runtime.h>
#include <linux/vga_switcheroo.h>
#include <drm/drm_probe_helper.h>
+#include <linux/mmu_notifier.h>
#include "amdgpu.h"
#include "amdgpu_irq.h"
amdgpu_unregister_atpx_handler();
amdgpu_sync_fini();
amdgpu_fence_slab_fini();
+ mmu_notifier_synchronize();
}
module_init(amdgpu_init);
* Block for operations on BOs to finish and mark pages as accessed and
* potentially dirty.
*/
-static int amdgpu_mn_sync_pagetables_gfx(struct hmm_mirror *mirror,
- const struct hmm_update *update)
+static int
+amdgpu_mn_sync_pagetables_gfx(struct hmm_mirror *mirror,
+ const struct mmu_notifier_range *update)
{
struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror);
unsigned long start = update->start;
unsigned long end = update->end;
- bool blockable = update->blockable;
+ bool blockable = mmu_notifier_range_blockable(update);
struct interval_tree_node *it;
/* notification is exclusive, but interval is inclusive */
* necessitates evicting all user-mode queues of the process. The BOs
* are restorted in amdgpu_mn_invalidate_range_end_hsa.
*/
-static int amdgpu_mn_sync_pagetables_hsa(struct hmm_mirror *mirror,
- const struct hmm_update *update)
+static int
+amdgpu_mn_sync_pagetables_hsa(struct hmm_mirror *mirror,
+ const struct mmu_notifier_range *update)
{
struct amdgpu_mn *amn = container_of(mirror, struct amdgpu_mn, mirror);
unsigned long start = update->start;
unsigned long end = update->end;
- bool blockable = update->blockable;
+ bool blockable = mmu_notifier_range_blockable(update);
struct interval_tree_node *it;
/* notification is exclusive, but interval is inclusive */
range->flags = hmm_range_flags;
range->values = hmm_range_values;
range->pfn_shift = PAGE_SHIFT;
- INIT_LIST_HEAD(&range->list);
}
}
struct hmm_range *range;
unsigned long i;
uint64_t *pfns;
- int retry = 0;
int r = 0;
if (!mm) /* Happens during process shutdown */
0 : range->flags[HMM_PFN_WRITE];
range->pfn_flags_mask = 0;
range->pfns = pfns;
- hmm_range_register(range, mirror, start,
- start + ttm->num_pages * PAGE_SIZE, PAGE_SHIFT);
+ range->start = start;
+ range->end = start + ttm->num_pages * PAGE_SIZE;
+
+ hmm_range_register(range, mirror);
-retry:
/*
* Just wait for range to be valid, safe to ignore return value as we
* will use the return value of hmm_range_fault() below under the
hmm_range_wait_until_valid(range, HMM_RANGE_DEFAULT_TIMEOUT);
down_read(&mm->mmap_sem);
-
- r = hmm_range_fault(range, true);
- if (unlikely(r < 0)) {
- if (likely(r == -EAGAIN)) {
- /*
- * return -EAGAIN, mmap_sem is dropped
- */
- if (retry++ < MAX_RETRY_HMM_RANGE_FAULT)
- goto retry;
- else
- pr_err("Retry hmm fault too many times\n");
- }
-
- goto out_up_read;
- }
-
+ r = hmm_range_fault(range, 0);
up_read(&mm->mmap_sem);
+ if (unlikely(r < 0))
+ goto out_free_pfns;
+
for (i = 0; i < ttm->num_pages; i++) {
pages[i] = hmm_device_entry_to_page(range, pfns[i]);
if (unlikely(!pages[i])) {
return 0;
-out_up_read:
- if (likely(r != -EAGAIN))
- up_read(&mm->mmap_sem);
out_free_pfns:
hmm_range_unregister(range);
kvfree(pfns);
/* We want to receive a notification when the mm_struct is destroyed */
struct mmu_notifier mmu_notifier;
- /* Use for delayed freeing of kfd_process structure */
- struct rcu_head rcu;
-
unsigned int pasid;
unsigned int doorbell_index;
static struct kfd_process *find_process(const struct task_struct *thread);
static void kfd_process_ref_release(struct kref *ref);
-static struct kfd_process *create_process(const struct task_struct *thread,
- struct file *filep);
+static struct kfd_process *create_process(const struct task_struct *thread);
+static int kfd_process_init_cwsr_apu(struct kfd_process *p, struct file *filep);
static void evict_process_worker(struct work_struct *work);
static void restore_process_worker(struct work_struct *work);
if (process) {
pr_debug("Process already found\n");
} else {
- process = create_process(thread, filep);
+ process = create_process(thread);
+ if (IS_ERR(process))
+ goto out;
+
+ ret = kfd_process_init_cwsr_apu(process, filep);
+ if (ret) {
+ process = ERR_PTR(ret);
+ goto out;
+ }
if (!procfs.kobj)
goto out;
queue_work(kfd_process_wq, &p->release_work);
}
-static void kfd_process_destroy_delayed(struct rcu_head *rcu)
+static void kfd_process_free_notifier(struct mmu_notifier *mn)
{
- struct kfd_process *p = container_of(rcu, struct kfd_process, rcu);
-
- kfd_unref_process(p);
+ kfd_unref_process(container_of(mn, struct kfd_process, mmu_notifier));
}
static void kfd_process_notifier_release(struct mmu_notifier *mn,
mutex_unlock(&p->mutex);
- mmu_notifier_unregister_no_release(&p->mmu_notifier, mm);
- mmu_notifier_call_srcu(&p->rcu, &kfd_process_destroy_delayed);
+ mmu_notifier_put(&p->mmu_notifier);
}
static const struct mmu_notifier_ops kfd_process_mmu_notifier_ops = {
.release = kfd_process_notifier_release,
+ .free_notifier = kfd_process_free_notifier,
};
static int kfd_process_init_cwsr_apu(struct kfd_process *p, struct file *filep)
return 0;
}
-static struct kfd_process *create_process(const struct task_struct *thread,
- struct file *filep)
+/*
+ * On return the kfd_process is fully operational and will be freed when the
+ * mm is released
+ */
+static struct kfd_process *create_process(const struct task_struct *thread)
{
struct kfd_process *process;
int err = -ENOMEM;
process = kzalloc(sizeof(*process), GFP_KERNEL);
-
if (!process)
goto err_alloc_process;
- process->pasid = kfd_pasid_alloc();
- if (process->pasid == 0)
- goto err_alloc_pasid;
-
- if (kfd_alloc_process_doorbells(process) < 0)
- goto err_alloc_doorbells;
-
kref_init(&process->ref);
-
mutex_init(&process->mutex);
-
process->mm = thread->mm;
-
- /* register notifier */
- process->mmu_notifier.ops = &kfd_process_mmu_notifier_ops;
- err = mmu_notifier_register(&process->mmu_notifier, process->mm);
- if (err)
- goto err_mmu_notifier;
-
- hash_add_rcu(kfd_processes_table, &process->kfd_processes,
- (uintptr_t)process->mm);
-
process->lead_thread = thread->group_leader;
- get_task_struct(process->lead_thread);
-
INIT_LIST_HEAD(&process->per_device_data);
-
+ INIT_DELAYED_WORK(&process->eviction_work, evict_process_worker);
+ INIT_DELAYED_WORK(&process->restore_work, restore_process_worker);
+ process->last_restore_timestamp = get_jiffies_64();
kfd_event_init_process(process);
+ process->is_32bit_user_mode = in_compat_syscall();
+
+ process->pasid = kfd_pasid_alloc();
+ if (process->pasid == 0)
+ goto err_alloc_pasid;
+
+ if (kfd_alloc_process_doorbells(process) < 0)
+ goto err_alloc_doorbells;
err = pqm_init(&process->pqm, process);
if (err != 0)
goto err_process_pqm_init;
/* init process apertures*/
- process->is_32bit_user_mode = in_compat_syscall();
err = kfd_init_apertures(process);
if (err != 0)
goto err_init_apertures;
- INIT_DELAYED_WORK(&process->eviction_work, evict_process_worker);
- INIT_DELAYED_WORK(&process->restore_work, restore_process_worker);
- process->last_restore_timestamp = get_jiffies_64();
-
- err = kfd_process_init_cwsr_apu(process, filep);
+ /* Must be last, have to use release destruction after this */
+ process->mmu_notifier.ops = &kfd_process_mmu_notifier_ops;
+ err = mmu_notifier_register(&process->mmu_notifier, process->mm);
if (err)
- goto err_init_cwsr;
+ goto err_register_notifier;
+
+ get_task_struct(process->lead_thread);
+ hash_add_rcu(kfd_processes_table, &process->kfd_processes,
+ (uintptr_t)process->mm);
return process;
-err_init_cwsr:
+err_register_notifier:
kfd_process_free_outstanding_kfd_bos(process);
kfd_process_destroy_pdds(process);
err_init_apertures:
pqm_uninit(&process->pqm);
err_process_pqm_init:
- hash_del_rcu(&process->kfd_processes);
- synchronize_rcu();
- mmu_notifier_unregister_no_release(&process->mmu_notifier, process->mm);
-err_mmu_notifier:
- mutex_destroy(&process->mutex);
kfd_free_process_doorbells(process);
err_alloc_doorbells:
kfd_pasid_free(process->pasid);
err_alloc_pasid:
+ mutex_destroy(&process->mutex);
kfree(process);
err_alloc_process:
return ERR_PTR(err);
bool "(EXPERIMENTAL) Enable SVM (Shared Virtual Memory) support"
depends on DEVICE_PRIVATE
depends on DRM_NOUVEAU
- depends on HMM_MIRROR
+ depends on MMU
depends on STAGING
- select MIGRATE_VMA_HELPER
+ select HMM_MIRROR
+ select MMU_NOTIFIER
default n
help
Say Y here if you want to enable experimental support for
#define DMEM_CHUNK_SIZE (2UL << 20)
#define DMEM_CHUNK_NPAGES (DMEM_CHUNK_SIZE >> PAGE_SHIFT)
-struct nouveau_migrate;
-
enum nouveau_aper {
NOUVEAU_APER_VIRT,
NOUVEAU_APER_VRAM,
return container_of(page->pgmap, struct nouveau_dmem, pagemap);
}
-struct nouveau_dmem_fault {
- struct nouveau_drm *drm;
- struct nouveau_fence *fence;
- dma_addr_t *dma;
- unsigned long npages;
-};
+static unsigned long nouveau_dmem_page_addr(struct page *page)
+{
+ struct nouveau_dmem_chunk *chunk = page->zone_device_data;
+ unsigned long idx = page_to_pfn(page) - chunk->pfn_first;
-struct nouveau_migrate {
- struct vm_area_struct *vma;
- struct nouveau_drm *drm;
- struct nouveau_fence *fence;
- unsigned long npages;
- dma_addr_t *dma;
- unsigned long dma_nr;
-};
+ return (idx << PAGE_SHIFT) + chunk->bo->bo.offset;
+}
static void nouveau_dmem_page_free(struct page *page)
{
spin_unlock(&chunk->lock);
}
-static void
-nouveau_dmem_fault_alloc_and_copy(struct vm_area_struct *vma,
- const unsigned long *src_pfns,
- unsigned long *dst_pfns,
- unsigned long start,
- unsigned long end,
- void *private)
+static void nouveau_dmem_fence_done(struct nouveau_fence **fence)
{
- struct nouveau_dmem_fault *fault = private;
- struct nouveau_drm *drm = fault->drm;
- struct device *dev = drm->dev->dev;
- unsigned long addr, i, npages = 0;
- nouveau_migrate_copy_t copy;
- int ret;
-
-
- /* First allocate new memory */
- for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, i++) {
- struct page *dpage, *spage;
-
- dst_pfns[i] = 0;
- spage = migrate_pfn_to_page(src_pfns[i]);
- if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE))
- continue;
-
- dpage = alloc_page_vma(GFP_HIGHUSER, vma, addr);
- if (!dpage) {
- dst_pfns[i] = MIGRATE_PFN_ERROR;
- continue;
- }
- lock_page(dpage);
-
- dst_pfns[i] = migrate_pfn(page_to_pfn(dpage)) |
- MIGRATE_PFN_LOCKED;
- npages++;
- }
-
- /* Allocate storage for DMA addresses, so we can unmap later. */
- fault->dma = kmalloc(sizeof(*fault->dma) * npages, GFP_KERNEL);
- if (!fault->dma)
- goto error;
-
- /* Copy things over */
- copy = drm->dmem->migrate.copy_func;
- for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, i++) {
- struct nouveau_dmem_chunk *chunk;
- struct page *spage, *dpage;
- u64 src_addr, dst_addr;
-
- dpage = migrate_pfn_to_page(dst_pfns[i]);
- if (!dpage || dst_pfns[i] == MIGRATE_PFN_ERROR)
- continue;
-
- spage = migrate_pfn_to_page(src_pfns[i]);
- if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE)) {
- dst_pfns[i] = MIGRATE_PFN_ERROR;
- __free_page(dpage);
- continue;
- }
-
- fault->dma[fault->npages] =
- dma_map_page_attrs(dev, dpage, 0, PAGE_SIZE,
- PCI_DMA_BIDIRECTIONAL,
- DMA_ATTR_SKIP_CPU_SYNC);
- if (dma_mapping_error(dev, fault->dma[fault->npages])) {
- dst_pfns[i] = MIGRATE_PFN_ERROR;
- __free_page(dpage);
- continue;
- }
-
- dst_addr = fault->dma[fault->npages++];
-
- chunk = spage->zone_device_data;
- src_addr = page_to_pfn(spage) - chunk->pfn_first;
- src_addr = (src_addr << PAGE_SHIFT) + chunk->bo->bo.offset;
-
- ret = copy(drm, 1, NOUVEAU_APER_HOST, dst_addr,
- NOUVEAU_APER_VRAM, src_addr);
- if (ret) {
- dst_pfns[i] = MIGRATE_PFN_ERROR;
- __free_page(dpage);
- continue;
- }
+ if (fence) {
+ nouveau_fence_wait(*fence, true, false);
+ nouveau_fence_unref(fence);
+ } else {
+ /*
+ * FIXME wait for channel to be IDLE before calling finalizing
+ * the hmem object.
+ */
}
+}
- nouveau_fence_new(drm->dmem->migrate.chan, false, &fault->fence);
-
- return;
-
-error:
- for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, ++i) {
- struct page *page;
+static vm_fault_t nouveau_dmem_fault_copy_one(struct nouveau_drm *drm,
+ struct vm_fault *vmf, struct migrate_vma *args,
+ dma_addr_t *dma_addr)
+{
+ struct device *dev = drm->dev->dev;
+ struct page *dpage, *spage;
- if (!dst_pfns[i] || dst_pfns[i] == MIGRATE_PFN_ERROR)
- continue;
+ spage = migrate_pfn_to_page(args->src[0]);
+ if (!spage || !(args->src[0] & MIGRATE_PFN_MIGRATE))
+ return 0;
- page = migrate_pfn_to_page(dst_pfns[i]);
- dst_pfns[i] = MIGRATE_PFN_ERROR;
- if (page == NULL)
- continue;
+ dpage = alloc_page_vma(GFP_HIGHUSER, vmf->vma, vmf->address);
+ if (!dpage)
+ return VM_FAULT_SIGBUS;
+ lock_page(dpage);
- __free_page(page);
- }
-}
+ *dma_addr = dma_map_page(dev, dpage, 0, PAGE_SIZE, DMA_BIDIRECTIONAL);
+ if (dma_mapping_error(dev, *dma_addr))
+ goto error_free_page;
-void nouveau_dmem_fault_finalize_and_map(struct vm_area_struct *vma,
- const unsigned long *src_pfns,
- const unsigned long *dst_pfns,
- unsigned long start,
- unsigned long end,
- void *private)
-{
- struct nouveau_dmem_fault *fault = private;
- struct nouveau_drm *drm = fault->drm;
+ if (drm->dmem->migrate.copy_func(drm, 1, NOUVEAU_APER_HOST, *dma_addr,
+ NOUVEAU_APER_VRAM, nouveau_dmem_page_addr(spage)))
+ goto error_dma_unmap;
- if (fault->fence) {
- nouveau_fence_wait(fault->fence, true, false);
- nouveau_fence_unref(&fault->fence);
- } else {
- /*
- * FIXME wait for channel to be IDLE before calling finalizing
- * the hmem object below (nouveau_migrate_hmem_fini()).
- */
- }
+ args->dst[0] = migrate_pfn(page_to_pfn(dpage)) | MIGRATE_PFN_LOCKED;
+ return 0;
- while (fault->npages--) {
- dma_unmap_page(drm->dev->dev, fault->dma[fault->npages],
- PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
- }
- kfree(fault->dma);
+error_dma_unmap:
+ dma_unmap_page(dev, *dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL);
+error_free_page:
+ __free_page(dpage);
+ return VM_FAULT_SIGBUS;
}
-static const struct migrate_vma_ops nouveau_dmem_fault_migrate_ops = {
- .alloc_and_copy = nouveau_dmem_fault_alloc_and_copy,
- .finalize_and_map = nouveau_dmem_fault_finalize_and_map,
-};
-
static vm_fault_t nouveau_dmem_migrate_to_ram(struct vm_fault *vmf)
{
struct nouveau_dmem *dmem = page_to_dmem(vmf->page);
- unsigned long src[1] = {0}, dst[1] = {0};
- struct nouveau_dmem_fault fault = { .drm = dmem->drm };
- int ret;
+ struct nouveau_drm *drm = dmem->drm;
+ struct nouveau_fence *fence;
+ unsigned long src = 0, dst = 0;
+ dma_addr_t dma_addr = 0;
+ vm_fault_t ret;
+ struct migrate_vma args = {
+ .vma = vmf->vma,
+ .start = vmf->address,
+ .end = vmf->address + PAGE_SIZE,
+ .src = &src,
+ .dst = &dst,
+ };
/*
* FIXME what we really want is to find some heuristic to migrate more
* than just one page on CPU fault. When such fault happens it is very
* likely that more surrounding page will CPU fault too.
*/
- ret = migrate_vma(&nouveau_dmem_fault_migrate_ops, vmf->vma,
- vmf->address, vmf->address + PAGE_SIZE,
- src, dst, &fault);
- if (ret)
+ if (migrate_vma_setup(&args) < 0)
return VM_FAULT_SIGBUS;
+ if (!args.cpages)
+ return 0;
- if (dst[0] == MIGRATE_PFN_ERROR)
- return VM_FAULT_SIGBUS;
+ ret = nouveau_dmem_fault_copy_one(drm, vmf, &args, &dma_addr);
+ if (ret || dst == 0)
+ goto done;
- return 0;
+ nouveau_fence_new(dmem->migrate.chan, false, &fence);
+ migrate_vma_pages(&args);
+ nouveau_dmem_fence_done(&fence);
+ dma_unmap_page(drm->dev->dev, dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL);
+done:
+ migrate_vma_finalize(&args);
+ return ret;
}
static const struct dev_pagemap_ops nouveau_dmem_pagemap_ops = {
drm->dmem = NULL;
}
-static void
-nouveau_dmem_migrate_alloc_and_copy(struct vm_area_struct *vma,
- const unsigned long *src_pfns,
- unsigned long *dst_pfns,
- unsigned long start,
- unsigned long end,
- void *private)
+static unsigned long nouveau_dmem_migrate_copy_one(struct nouveau_drm *drm,
+ unsigned long src, dma_addr_t *dma_addr)
{
- struct nouveau_migrate *migrate = private;
- struct nouveau_drm *drm = migrate->drm;
struct device *dev = drm->dev->dev;
- unsigned long addr, i, npages = 0;
- nouveau_migrate_copy_t copy;
- int ret;
-
- /* First allocate new memory */
- for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, i++) {
- struct page *dpage, *spage;
-
- dst_pfns[i] = 0;
- spage = migrate_pfn_to_page(src_pfns[i]);
- if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE))
- continue;
-
- dpage = nouveau_dmem_page_alloc_locked(drm);
- if (!dpage)
- continue;
-
- dst_pfns[i] = migrate_pfn(page_to_pfn(dpage)) |
- MIGRATE_PFN_LOCKED |
- MIGRATE_PFN_DEVICE;
- npages++;
- }
-
- if (!npages)
- return;
-
- /* Allocate storage for DMA addresses, so we can unmap later. */
- migrate->dma = kmalloc(sizeof(*migrate->dma) * npages, GFP_KERNEL);
- if (!migrate->dma)
- goto error;
-
- /* Copy things over */
- copy = drm->dmem->migrate.copy_func;
- for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, i++) {
- struct nouveau_dmem_chunk *chunk;
- struct page *spage, *dpage;
- u64 src_addr, dst_addr;
-
- dpage = migrate_pfn_to_page(dst_pfns[i]);
- if (!dpage || dst_pfns[i] == MIGRATE_PFN_ERROR)
- continue;
-
- chunk = dpage->zone_device_data;
- dst_addr = page_to_pfn(dpage) - chunk->pfn_first;
- dst_addr = (dst_addr << PAGE_SHIFT) + chunk->bo->bo.offset;
-
- spage = migrate_pfn_to_page(src_pfns[i]);
- if (!spage || !(src_pfns[i] & MIGRATE_PFN_MIGRATE)) {
- nouveau_dmem_page_free_locked(drm, dpage);
- dst_pfns[i] = 0;
- continue;
- }
-
- migrate->dma[migrate->dma_nr] =
- dma_map_page_attrs(dev, spage, 0, PAGE_SIZE,
- PCI_DMA_BIDIRECTIONAL,
- DMA_ATTR_SKIP_CPU_SYNC);
- if (dma_mapping_error(dev, migrate->dma[migrate->dma_nr])) {
- nouveau_dmem_page_free_locked(drm, dpage);
- dst_pfns[i] = 0;
- continue;
- }
-
- src_addr = migrate->dma[migrate->dma_nr++];
+ struct page *dpage, *spage;
- ret = copy(drm, 1, NOUVEAU_APER_VRAM, dst_addr,
- NOUVEAU_APER_HOST, src_addr);
- if (ret) {
- nouveau_dmem_page_free_locked(drm, dpage);
- dst_pfns[i] = 0;
- continue;
- }
- }
-
- nouveau_fence_new(drm->dmem->migrate.chan, false, &migrate->fence);
+ spage = migrate_pfn_to_page(src);
+ if (!spage || !(src & MIGRATE_PFN_MIGRATE))
+ goto out;
- return;
+ dpage = nouveau_dmem_page_alloc_locked(drm);
+ if (!dpage)
+ return 0;
-error:
- for (addr = start, i = 0; addr < end; addr += PAGE_SIZE, ++i) {
- struct page *page;
+ *dma_addr = dma_map_page(dev, spage, 0, PAGE_SIZE, DMA_BIDIRECTIONAL);
+ if (dma_mapping_error(dev, *dma_addr))
+ goto out_free_page;
- if (!dst_pfns[i] || dst_pfns[i] == MIGRATE_PFN_ERROR)
- continue;
+ if (drm->dmem->migrate.copy_func(drm, 1, NOUVEAU_APER_VRAM,
+ nouveau_dmem_page_addr(dpage), NOUVEAU_APER_HOST,
+ *dma_addr))
+ goto out_dma_unmap;
- page = migrate_pfn_to_page(dst_pfns[i]);
- dst_pfns[i] = MIGRATE_PFN_ERROR;
- if (page == NULL)
- continue;
+ return migrate_pfn(page_to_pfn(dpage)) | MIGRATE_PFN_LOCKED;
- __free_page(page);
- }
+out_dma_unmap:
+ dma_unmap_page(dev, *dma_addr, PAGE_SIZE, DMA_BIDIRECTIONAL);
+out_free_page:
+ nouveau_dmem_page_free_locked(drm, dpage);
+out:
+ return 0;
}
-void nouveau_dmem_migrate_finalize_and_map(struct vm_area_struct *vma,
- const unsigned long *src_pfns,
- const unsigned long *dst_pfns,
- unsigned long start,
- unsigned long end,
- void *private)
+static void nouveau_dmem_migrate_chunk(struct nouveau_drm *drm,
+ struct migrate_vma *args, dma_addr_t *dma_addrs)
{
- struct nouveau_migrate *migrate = private;
- struct nouveau_drm *drm = migrate->drm;
-
- if (migrate->fence) {
- nouveau_fence_wait(migrate->fence, true, false);
- nouveau_fence_unref(&migrate->fence);
- } else {
- /*
- * FIXME wait for channel to be IDLE before finalizing
- * the hmem object below (nouveau_migrate_hmem_fini()) ?
- */
+ struct nouveau_fence *fence;
+ unsigned long addr = args->start, nr_dma = 0, i;
+
+ for (i = 0; addr < args->end; i++) {
+ args->dst[i] = nouveau_dmem_migrate_copy_one(drm, args->src[i],
+ dma_addrs + nr_dma);
+ if (args->dst[i])
+ nr_dma++;
+ addr += PAGE_SIZE;
}
- while (migrate->dma_nr--) {
- dma_unmap_page(drm->dev->dev, migrate->dma[migrate->dma_nr],
- PAGE_SIZE, PCI_DMA_BIDIRECTIONAL);
- }
- kfree(migrate->dma);
+ nouveau_fence_new(drm->dmem->migrate.chan, false, &fence);
+ migrate_vma_pages(args);
+ nouveau_dmem_fence_done(&fence);
+ while (nr_dma--) {
+ dma_unmap_page(drm->dev->dev, dma_addrs[nr_dma], PAGE_SIZE,
+ DMA_BIDIRECTIONAL);
+ }
/*
- * FIXME optimization: update GPU page table to point to newly
- * migrated memory.
+ * FIXME optimization: update GPU page table to point to newly migrated
+ * memory.
*/
+ migrate_vma_finalize(args);
}
-static const struct migrate_vma_ops nouveau_dmem_migrate_ops = {
- .alloc_and_copy = nouveau_dmem_migrate_alloc_and_copy,
- .finalize_and_map = nouveau_dmem_migrate_finalize_and_map,
-};
-
int
nouveau_dmem_migrate_vma(struct nouveau_drm *drm,
struct vm_area_struct *vma,
unsigned long start,
unsigned long end)
{
- unsigned long *src_pfns, *dst_pfns, npages;
- struct nouveau_migrate migrate = {0};
- unsigned long i, c, max;
- int ret = 0;
-
- npages = (end - start) >> PAGE_SHIFT;
- max = min(SG_MAX_SINGLE_ALLOC, npages);
- src_pfns = kzalloc(sizeof(long) * max, GFP_KERNEL);
- if (src_pfns == NULL)
- return -ENOMEM;
- dst_pfns = kzalloc(sizeof(long) * max, GFP_KERNEL);
- if (dst_pfns == NULL) {
- kfree(src_pfns);
- return -ENOMEM;
- }
+ unsigned long npages = (end - start) >> PAGE_SHIFT;
+ unsigned long max = min(SG_MAX_SINGLE_ALLOC, npages);
+ dma_addr_t *dma_addrs;
+ struct migrate_vma args = {
+ .vma = vma,
+ .start = start,
+ };
+ unsigned long c, i;
+ int ret = -ENOMEM;
+
+ args.src = kcalloc(max, sizeof(args.src), GFP_KERNEL);
+ if (!args.src)
+ goto out;
+ args.dst = kcalloc(max, sizeof(args.dst), GFP_KERNEL);
+ if (!args.dst)
+ goto out_free_src;
- migrate.drm = drm;
- migrate.vma = vma;
- migrate.npages = npages;
- for (i = 0; i < npages; i += c) {
- unsigned long next;
+ dma_addrs = kmalloc_array(max, sizeof(*dma_addrs), GFP_KERNEL);
+ if (!dma_addrs)
+ goto out_free_dst;
+ for (i = 0; i < npages; i += c) {
c = min(SG_MAX_SINGLE_ALLOC, npages);
- next = start + (c << PAGE_SHIFT);
- ret = migrate_vma(&nouveau_dmem_migrate_ops, vma, start,
- next, src_pfns, dst_pfns, &migrate);
+ args.end = start + (c << PAGE_SHIFT);
+ ret = migrate_vma_setup(&args);
if (ret)
- goto out;
- start = next;
+ goto out_free_dma;
+
+ if (args.cpages)
+ nouveau_dmem_migrate_chunk(drm, &args, dma_addrs);
+ args.start = args.end;
}
+ ret = 0;
+out_free_dma:
+ kfree(dma_addrs);
+out_free_dst:
+ kfree(args.dst);
+out_free_src:
+ kfree(args.src);
out:
- kfree(dst_pfns);
- kfree(src_pfns);
return ret;
}
npages = (range->end - range->start) >> PAGE_SHIFT;
for (i = 0; i < npages; ++i) {
- struct nouveau_dmem_chunk *chunk;
struct page *page;
uint64_t addr;
- page = hmm_pfn_to_page(range, range->pfns[i]);
+ page = hmm_device_entry_to_page(range, range->pfns[i]);
if (page == NULL)
continue;
continue;
}
- chunk = page->zone_device_data;
- addr = page_to_pfn(page) - chunk->pfn_first;
- addr = (addr + chunk->bo->bo.mem.start) << PAGE_SHIFT;
-
+ addr = nouveau_dmem_page_addr(page);
range->pfns[i] &= ((1UL << range->pfn_shift) - 1);
range->pfns[i] |= (addr >> PAGE_SHIFT) << range->pfn_shift;
}
static inline void nouveau_dmem_fini(struct nouveau_drm *drm) {}
static inline void nouveau_dmem_suspend(struct nouveau_drm *drm) {}
static inline void nouveau_dmem_resume(struct nouveau_drm *drm) {}
-
-static inline int nouveau_dmem_migrate_vma(struct nouveau_drm *drm,
- struct vm_area_struct *vma,
- unsigned long start,
- unsigned long end)
-{
- return 0;
-}
-
-static inline void nouveau_dmem_convert_pfn(struct nouveau_drm *drm,
- struct hmm_range *range) {}
#endif /* IS_ENABLED(CONFIG_DRM_NOUVEAU_SVM) */
#endif
#include <linux/pci.h>
#include <linux/pm_runtime.h>
#include <linux/vga_switcheroo.h>
+#include <linux/mmu_notifier.h>
#include <drm/drmP.h>
#include <drm/drm_crtc_helper.h>
#ifdef CONFIG_NOUVEAU_PLATFORM_DRIVER
platform_driver_unregister(&nouveau_platform_driver);
#endif
+ if (IS_ENABLED(CONFIG_DRM_NOUVEAU_SVM))
+ mmu_notifier_synchronize();
}
module_init(nouveau_drm_init);
static int
nouveau_svmm_sync_cpu_device_pagetables(struct hmm_mirror *mirror,
- const struct hmm_update *update)
+ const struct mmu_notifier_range *update)
{
struct nouveau_svmm *svmm = container_of(mirror, typeof(*svmm), mirror);
unsigned long start = update->start;
unsigned long limit = update->end;
- if (!update->blockable)
+ if (!mmu_notifier_range_blockable(update))
return -EAGAIN;
SVMM_DBG(svmm, "invalidate %016lx-%016lx", start, limit);
}
static int
-nouveau_range_fault(struct hmm_mirror *mirror, struct hmm_range *range)
+nouveau_range_fault(struct nouveau_svmm *svmm, struct hmm_range *range)
{
long ret;
range->default_flags = 0;
range->pfn_flags_mask = -1UL;
- ret = hmm_range_register(range, mirror,
- range->start, range->end,
- PAGE_SHIFT);
+ ret = hmm_range_register(range, &svmm->mirror);
if (ret) {
- up_read(&range->vma->vm_mm->mmap_sem);
+ up_read(&svmm->mm->mmap_sem);
return (int)ret;
}
if (!hmm_range_wait_until_valid(range, HMM_RANGE_DEFAULT_TIMEOUT)) {
- up_read(&range->vma->vm_mm->mmap_sem);
- return -EAGAIN;
+ up_read(&svmm->mm->mmap_sem);
+ return -EBUSY;
}
- ret = hmm_range_fault(range, true);
+ ret = hmm_range_fault(range, 0);
if (ret <= 0) {
if (ret == 0)
ret = -EBUSY;
- up_read(&range->vma->vm_mm->mmap_sem);
+ up_read(&svmm->mm->mmap_sem);
hmm_range_unregister(range);
return ret;
}
args.i.p.addr + args.i.p.size, fn - fi);
/* Have HMM fault pages within the fault window to the GPU. */
- range.vma = vma;
range.start = args.i.p.addr;
range.end = args.i.p.addr + args.i.p.size;
range.pfns = args.phys;
range.values = nouveau_svm_pfn_values;
range.pfn_shift = NVIF_VMM_PFNMAP_V0_ADDR_SHIFT;
again:
- ret = nouveau_range_fault(&svmm->mirror, &range);
+ ret = nouveau_range_fault(svmm, &range);
if (ret == 0) {
mutex_lock(&svmm->mutex);
if (!nouveau_range_done(&range)) {
/* tracking pinned memory */
u64 vram_pin_size;
u64 gart_pin_size;
-
- struct mutex mn_lock;
- DECLARE_HASHTABLE(mn_hash, 7);
};
bool radeon_is_px(struct drm_device *dev);
init_rwsem(&rdev->pm.mclk_lock);
init_rwsem(&rdev->exclusive_lock);
init_waitqueue_head(&rdev->irq.vblank_queue);
- mutex_init(&rdev->mn_lock);
- hash_init(rdev->mn_hash);
r = radeon_gem_init(rdev);
if (r)
return r;
#include <linux/module.h>
#include <linux/pm_runtime.h>
#include <linux/vga_switcheroo.h>
+#include <linux/mmu_notifier.h>
#include <drm/drm_crtc_helper.h>
#include <drm/drm_drv.h>
{
pci_unregister_driver(pdriver);
radeon_unregister_atpx_handler();
+ mmu_notifier_synchronize();
}
module_init(radeon_init);
#include "radeon.h"
struct radeon_mn {
- /* constant after initialisation */
- struct radeon_device *rdev;
- struct mm_struct *mm;
struct mmu_notifier mn;
- /* only used on destruction */
- struct work_struct work;
-
- /* protected by rdev->mn_lock */
- struct hlist_node node;
-
/* objects protected by lock */
struct mutex lock;
struct rb_root_cached objects;
struct list_head bos;
};
-/**
- * radeon_mn_destroy - destroy the rmn
- *
- * @work: previously sheduled work item
- *
- * Lazy destroys the notifier from a work item
- */
-static void radeon_mn_destroy(struct work_struct *work)
-{
- struct radeon_mn *rmn = container_of(work, struct radeon_mn, work);
- struct radeon_device *rdev = rmn->rdev;
- struct radeon_mn_node *node, *next_node;
- struct radeon_bo *bo, *next_bo;
-
- mutex_lock(&rdev->mn_lock);
- mutex_lock(&rmn->lock);
- hash_del(&rmn->node);
- rbtree_postorder_for_each_entry_safe(node, next_node,
- &rmn->objects.rb_root, it.rb) {
-
- interval_tree_remove(&node->it, &rmn->objects);
- list_for_each_entry_safe(bo, next_bo, &node->bos, mn_list) {
- bo->mn = NULL;
- list_del_init(&bo->mn_list);
- }
- kfree(node);
- }
- mutex_unlock(&rmn->lock);
- mutex_unlock(&rdev->mn_lock);
- mmu_notifier_unregister(&rmn->mn, rmn->mm);
- kfree(rmn);
-}
-
-/**
- * radeon_mn_release - callback to notify about mm destruction
- *
- * @mn: our notifier
- * @mn: the mm this callback is about
- *
- * Shedule a work item to lazy destroy our notifier.
- */
-static void radeon_mn_release(struct mmu_notifier *mn,
- struct mm_struct *mm)
-{
- struct radeon_mn *rmn = container_of(mn, struct radeon_mn, mn);
- INIT_WORK(&rmn->work, radeon_mn_destroy);
- schedule_work(&rmn->work);
-}
-
/**
* radeon_mn_invalidate_range_start - callback to notify about mm change
*
return ret;
}
-static const struct mmu_notifier_ops radeon_mn_ops = {
- .release = radeon_mn_release,
- .invalidate_range_start = radeon_mn_invalidate_range_start,
-};
+static void radeon_mn_release(struct mmu_notifier *mn, struct mm_struct *mm)
+{
+ struct mmu_notifier_range range = {
+ .mm = mm,
+ .start = 0,
+ .end = ULONG_MAX,
+ .flags = 0,
+ .event = MMU_NOTIFY_UNMAP,
+ };
+
+ radeon_mn_invalidate_range_start(mn, &range);
+}
-/**
- * radeon_mn_get - create notifier context
- *
- * @rdev: radeon device pointer
- *
- * Creates a notifier context for current->mm.
- */
-static struct radeon_mn *radeon_mn_get(struct radeon_device *rdev)
+static struct mmu_notifier *radeon_mn_alloc_notifier(struct mm_struct *mm)
{
- struct mm_struct *mm = current->mm;
struct radeon_mn *rmn;
- int r;
-
- if (down_write_killable(&mm->mmap_sem))
- return ERR_PTR(-EINTR);
-
- mutex_lock(&rdev->mn_lock);
-
- hash_for_each_possible(rdev->mn_hash, rmn, node, (unsigned long)mm)
- if (rmn->mm == mm)
- goto release_locks;
rmn = kzalloc(sizeof(*rmn), GFP_KERNEL);
- if (!rmn) {
- rmn = ERR_PTR(-ENOMEM);
- goto release_locks;
- }
+ if (!rmn)
+ return ERR_PTR(-ENOMEM);
- rmn->rdev = rdev;
- rmn->mm = mm;
- rmn->mn.ops = &radeon_mn_ops;
mutex_init(&rmn->lock);
rmn->objects = RB_ROOT_CACHED;
-
- r = __mmu_notifier_register(&rmn->mn, mm);
- if (r)
- goto free_rmn;
-
- hash_add(rdev->mn_hash, &rmn->node, (unsigned long)mm);
-
-release_locks:
- mutex_unlock(&rdev->mn_lock);
- up_write(&mm->mmap_sem);
-
- return rmn;
-
-free_rmn:
- mutex_unlock(&rdev->mn_lock);
- up_write(&mm->mmap_sem);
- kfree(rmn);
+ return &rmn->mn;
+}
- return ERR_PTR(r);
+static void radeon_mn_free_notifier(struct mmu_notifier *mn)
+{
+ kfree(container_of(mn, struct radeon_mn, mn));
}
+static const struct mmu_notifier_ops radeon_mn_ops = {
+ .release = radeon_mn_release,
+ .invalidate_range_start = radeon_mn_invalidate_range_start,
+ .alloc_notifier = radeon_mn_alloc_notifier,
+ .free_notifier = radeon_mn_free_notifier,
+};
+
/**
* radeon_mn_register - register a BO for notifier updates
*
int radeon_mn_register(struct radeon_bo *bo, unsigned long addr)
{
unsigned long end = addr + radeon_bo_size(bo) - 1;
- struct radeon_device *rdev = bo->rdev;
+ struct mmu_notifier *mn;
struct radeon_mn *rmn;
struct radeon_mn_node *node = NULL;
struct list_head bos;
struct interval_tree_node *it;
- rmn = radeon_mn_get(rdev);
- if (IS_ERR(rmn))
- return PTR_ERR(rmn);
+ mn = mmu_notifier_get(&radeon_mn_ops, current->mm);
+ if (IS_ERR(mn))
+ return PTR_ERR(mn);
+ rmn = container_of(mn, struct radeon_mn, mn);
INIT_LIST_HEAD(&bos);
*/
void radeon_mn_unregister(struct radeon_bo *bo)
{
- struct radeon_device *rdev = bo->rdev;
- struct radeon_mn *rmn;
+ struct radeon_mn *rmn = bo->mn;
struct list_head *head;
- mutex_lock(&rdev->mn_lock);
- rmn = bo->mn;
- if (rmn == NULL) {
- mutex_unlock(&rdev->mn_lock);
- return;
- }
-
mutex_lock(&rmn->lock);
/* save the next list entry for later */
head = bo->mn_list.next;
- bo->mn = NULL;
list_del(&bo->mn_list);
if (list_empty(head)) {
}
mutex_unlock(&rmn->lock);
- mutex_unlock(&rdev->mn_lock);
+
+ mmu_notifier_put(&rmn->mn);
+ bo->mn = NULL;
}
gru_free_tables();
misc_deregister(&gru_miscdev);
gru_proc_exit();
+ mmu_notifier_synchronize();
}
static const struct file_operations gru_fops = {
struct gru_mm_struct {
struct mmu_notifier ms_notifier;
- atomic_t ms_refcnt;
spinlock_t ms_asid_lock; /* protects ASID assignment */
atomic_t ms_range_active;/* num range_invals active */
- char ms_released;
wait_queue_head_t ms_wait_queue;
DECLARE_BITMAP(ms_asidmap, GRU_MAX_GRUS);
struct gru_mm_tracker ms_asids[GRU_MAX_GRUS];
gms, range->start, range->end);
}
-static void gru_release(struct mmu_notifier *mn, struct mm_struct *mm)
+static struct mmu_notifier *gru_alloc_notifier(struct mm_struct *mm)
{
- struct gru_mm_struct *gms = container_of(mn, struct gru_mm_struct,
- ms_notifier);
+ struct gru_mm_struct *gms;
+
+ gms = kzalloc(sizeof(*gms), GFP_KERNEL);
+ if (!gms)
+ return ERR_PTR(-ENOMEM);
+ STAT(gms_alloc);
+ spin_lock_init(&gms->ms_asid_lock);
+ init_waitqueue_head(&gms->ms_wait_queue);
- gms->ms_released = 1;
- gru_dbg(grudev, "gms %p\n", gms);
+ return &gms->ms_notifier;
}
+static void gru_free_notifier(struct mmu_notifier *mn)
+{
+ kfree(container_of(mn, struct gru_mm_struct, ms_notifier));
+ STAT(gms_free);
+}
static const struct mmu_notifier_ops gru_mmuops = {
.invalidate_range_start = gru_invalidate_range_start,
.invalidate_range_end = gru_invalidate_range_end,
- .release = gru_release,
+ .alloc_notifier = gru_alloc_notifier,
+ .free_notifier = gru_free_notifier,
};
-/* Move this to the basic mmu_notifier file. But for now... */
-static struct mmu_notifier *mmu_find_ops(struct mm_struct *mm,
- const struct mmu_notifier_ops *ops)
-{
- struct mmu_notifier *mn, *gru_mn = NULL;
-
- if (mm->mmu_notifier_mm) {
- rcu_read_lock();
- hlist_for_each_entry_rcu(mn, &mm->mmu_notifier_mm->list,
- hlist)
- if (mn->ops == ops) {
- gru_mn = mn;
- break;
- }
- rcu_read_unlock();
- }
- return gru_mn;
-}
-
struct gru_mm_struct *gru_register_mmu_notifier(void)
{
- struct gru_mm_struct *gms;
struct mmu_notifier *mn;
- int err;
-
- mn = mmu_find_ops(current->mm, &gru_mmuops);
- if (mn) {
- gms = container_of(mn, struct gru_mm_struct, ms_notifier);
- atomic_inc(&gms->ms_refcnt);
- } else {
- gms = kzalloc(sizeof(*gms), GFP_KERNEL);
- if (!gms)
- return ERR_PTR(-ENOMEM);
- STAT(gms_alloc);
- spin_lock_init(&gms->ms_asid_lock);
- gms->ms_notifier.ops = &gru_mmuops;
- atomic_set(&gms->ms_refcnt, 1);
- init_waitqueue_head(&gms->ms_wait_queue);
- err = __mmu_notifier_register(&gms->ms_notifier, current->mm);
- if (err)
- goto error;
- }
- if (gms)
- gru_dbg(grudev, "gms %p, refcnt %d\n", gms,
- atomic_read(&gms->ms_refcnt));
- return gms;
-error:
- kfree(gms);
- return ERR_PTR(err);
+
+ mn = mmu_notifier_get_locked(&gru_mmuops, current->mm);
+ if (IS_ERR(mn))
+ return ERR_CAST(mn);
+
+ return container_of(mn, struct gru_mm_struct, ms_notifier);
}
void gru_drop_mmu_notifier(struct gru_mm_struct *gms)
{
- gru_dbg(grudev, "gms %p, refcnt %d, released %d\n", gms,
- atomic_read(&gms->ms_refcnt), gms->ms_released);
- if (atomic_dec_return(&gms->ms_refcnt) == 0) {
- if (!gms->ms_released)
- mmu_notifier_unregister(&gms->ms_notifier, current->mm);
- kfree(gms);
- STAT(gms_free);
- }
+ mmu_notifier_put(&gms->ms_notifier);
}
/*
* @notifiers: count of active mmu notifiers
*/
struct hmm {
- struct mm_struct *mm;
- struct kref kref;
+ struct mmu_notifier mmu_notifier;
spinlock_t ranges_lock;
struct list_head ranges;
struct list_head mirrors;
- struct mmu_notifier mmu_notifier;
struct rw_semaphore mirrors_sem;
wait_queue_head_t wq;
- struct rcu_head rcu;
long notifiers;
};
* @values: pfn value for some special case (none, special, error, ...)
* @default_flags: default flags for the range (write, read, ... see hmm doc)
* @pfn_flags_mask: allows to mask pfn flags so that only default_flags matter
- * @page_shift: device virtual address shift value (should be >= PAGE_SHIFT)
* @pfn_shifts: pfn shift value (should be <= PAGE_SHIFT)
* @valid: pfns array did not change since it has been fill by an HMM function
*/
struct hmm_range {
struct hmm *hmm;
- struct vm_area_struct *vma;
struct list_head list;
unsigned long start;
unsigned long end;
const uint64_t *values;
uint64_t default_flags;
uint64_t pfn_flags_mask;
- uint8_t page_shift;
uint8_t pfn_shift;
bool valid;
};
-/*
- * hmm_range_page_shift() - return the page shift for the range
- * @range: range being queried
- * Return: page shift (page size = 1 << page shift) for the range
- */
-static inline unsigned hmm_range_page_shift(const struct hmm_range *range)
-{
- return range->page_shift;
-}
-
-/*
- * hmm_range_page_size() - return the page size for the range
- * @range: range being queried
- * Return: page size for the range in bytes
- */
-static inline unsigned long hmm_range_page_size(const struct hmm_range *range)
-{
- return 1UL << hmm_range_page_shift(range);
-}
-
/*
* hmm_range_wait_until_valid() - wait for range to be valid
* @range: range affected by invalidation to wait on
range->flags[HMM_PFN_VALID];
}
-/*
- * Old API:
- * hmm_pfn_to_page()
- * hmm_pfn_to_pfn()
- * hmm_pfn_from_page()
- * hmm_pfn_from_pfn()
- *
- * This are the OLD API please use new API, it is here to avoid cross-tree
- * merge painfullness ie we convert things to new API in stages.
- */
-static inline struct page *hmm_pfn_to_page(const struct hmm_range *range,
- uint64_t pfn)
-{
- return hmm_device_entry_to_page(range, pfn);
-}
-
-static inline unsigned long hmm_pfn_to_pfn(const struct hmm_range *range,
- uint64_t pfn)
-{
- return hmm_device_entry_to_pfn(range, pfn);
-}
-
-static inline uint64_t hmm_pfn_from_page(const struct hmm_range *range,
- struct page *page)
-{
- return hmm_device_entry_from_page(range, page);
-}
-
-static inline uint64_t hmm_pfn_from_pfn(const struct hmm_range *range,
- unsigned long pfn)
-{
- return hmm_device_entry_from_pfn(range, pfn);
-}
-
/*
* Mirroring: how to synchronize device page table with CPU page table.
*
struct hmm_mirror;
-/*
- * enum hmm_update_event - type of update
- * @HMM_UPDATE_INVALIDATE: invalidate range (no indication as to why)
- */
-enum hmm_update_event {
- HMM_UPDATE_INVALIDATE,
-};
-
-/*
- * struct hmm_update - HMM update information for callback
- *
- * @start: virtual start address of the range to update
- * @end: virtual end address of the range to update
- * @event: event triggering the update (what is happening)
- * @blockable: can the callback block/sleep ?
- */
-struct hmm_update {
- unsigned long start;
- unsigned long end;
- enum hmm_update_event event;
- bool blockable;
-};
-
/*
* struct hmm_mirror_ops - HMM mirror device operations callback
*
/* sync_cpu_device_pagetables() - synchronize page tables
*
* @mirror: pointer to struct hmm_mirror
- * @update: update information (see struct hmm_update)
- * Return: -EAGAIN if update.blockable false and callback need to
- * block, 0 otherwise.
+ * @update: update information (see struct mmu_notifier_range)
+ * Return: -EAGAIN if mmu_notifier_range_blockable(update) is false
+ * and callback needs to block, 0 otherwise.
*
* This callback ultimately originates from mmu_notifiers when the CPU
* page table is updated. The device driver must update its page table
* page tables are completely updated (TLBs flushed, etc); this is a
* synchronous call.
*/
- int (*sync_cpu_device_pagetables)(struct hmm_mirror *mirror,
- const struct hmm_update *update);
+ int (*sync_cpu_device_pagetables)(
+ struct hmm_mirror *mirror,
+ const struct mmu_notifier_range *update);
};
/*
/*
* Please see Documentation/vm/hmm.rst for how to use the range API.
*/
-int hmm_range_register(struct hmm_range *range,
- struct hmm_mirror *mirror,
- unsigned long start,
- unsigned long end,
- unsigned page_shift);
+int hmm_range_register(struct hmm_range *range, struct hmm_mirror *mirror);
void hmm_range_unregister(struct hmm_range *range);
-long hmm_range_snapshot(struct hmm_range *range);
-long hmm_range_fault(struct hmm_range *range, bool block);
+
+/*
+ * Retry fault if non-blocking, drop mmap_sem and return -EAGAIN in that case.
+ */
+#define HMM_FAULT_ALLOW_RETRY (1 << 0)
+
+/* Don't fault in missing PTEs, just snapshot the current state. */
+#define HMM_FAULT_SNAPSHOT (1 << 1)
+
+long hmm_range_fault(struct hmm_range *range, unsigned int flags);
+
long hmm_range_dma_map(struct hmm_range *range,
struct device *device,
dma_addr_t *daddrs,
- bool block);
+ unsigned int flags);
long hmm_range_dma_unmap(struct hmm_range *range,
- struct vm_area_struct *vma,
struct device *device,
dma_addr_t *daddrs,
bool dirty);
*/
#define HMM_RANGE_DEFAULT_TIMEOUT 1000
-/* Below are for HMM internal use only! Not to be used by device driver! */
-static inline void hmm_mm_init(struct mm_struct *mm)
-{
- mm->hmm = NULL;
-}
-#else /* IS_ENABLED(CONFIG_HMM_MIRROR) */
-static inline void hmm_mm_init(struct mm_struct *mm) {}
#endif /* IS_ENABLED(CONFIG_HMM_MIRROR) */
#endif /* LINUX_HMM_H */
struct resource *devm_request_free_mem_region(struct device *dev,
struct resource *base, unsigned long size);
+struct resource *request_free_mem_region(struct resource *base,
+ unsigned long size, const char *name);
#endif /* __ASSEMBLY__ */
#endif /* _LINUX_IOPORT_H */
struct percpu_ref *ref;
struct percpu_ref internal_ref;
struct completion done;
- struct device *dev;
enum memory_type type;
unsigned int flags;
u64 pci_p2pdma_bus_offset;
}
#ifdef CONFIG_ZONE_DEVICE
+void *memremap_pages(struct dev_pagemap *pgmap, int nid);
+void memunmap_pages(struct dev_pagemap *pgmap);
void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap);
void devm_memunmap_pages(struct device *dev, struct dev_pagemap *pgmap);
struct dev_pagemap *get_dev_pagemap(unsigned long pfn,
#define MIGRATE_PFN_MIGRATE (1UL << 1)
#define MIGRATE_PFN_LOCKED (1UL << 2)
#define MIGRATE_PFN_WRITE (1UL << 3)
-#define MIGRATE_PFN_DEVICE (1UL << 4)
-#define MIGRATE_PFN_ERROR (1UL << 5)
#define MIGRATE_PFN_SHIFT 6
static inline struct page *migrate_pfn_to_page(unsigned long mpfn)
return (pfn << MIGRATE_PFN_SHIFT) | MIGRATE_PFN_VALID;
}
-/*
- * struct migrate_vma_ops - migrate operation callback
- *
- * @alloc_and_copy: alloc destination memory and copy source memory to it
- * @finalize_and_map: allow caller to map the successfully migrated pages
- *
- *
- * The alloc_and_copy() callback happens once all source pages have been locked,
- * unmapped and checked (checked whether pinned or not). All pages that can be
- * migrated will have an entry in the src array set with the pfn value of the
- * page and with the MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE flag set (other
- * flags might be set but should be ignored by the callback).
- *
- * The alloc_and_copy() callback can then allocate destination memory and copy
- * source memory to it for all those entries (ie with MIGRATE_PFN_VALID and
- * MIGRATE_PFN_MIGRATE flag set). Once these are allocated and copied, the
- * callback must update each corresponding entry in the dst array with the pfn
- * value of the destination page and with the MIGRATE_PFN_VALID and
- * MIGRATE_PFN_LOCKED flags set (destination pages must have their struct pages
- * locked, via lock_page()).
- *
- * At this point the alloc_and_copy() callback is done and returns.
- *
- * Note that the callback does not have to migrate all the pages that are
- * marked with MIGRATE_PFN_MIGRATE flag in src array unless this is a migration
- * from device memory to system memory (ie the MIGRATE_PFN_DEVICE flag is also
- * set in the src array entry). If the device driver cannot migrate a device
- * page back to system memory, then it must set the corresponding dst array
- * entry to MIGRATE_PFN_ERROR. This will trigger a SIGBUS if CPU tries to
- * access any of the virtual addresses originally backed by this page. Because
- * a SIGBUS is such a severe result for the userspace process, the device
- * driver should avoid setting MIGRATE_PFN_ERROR unless it is really in an
- * unrecoverable state.
- *
- * For empty entry inside CPU page table (pte_none() or pmd_none() is true) we
- * do set MIGRATE_PFN_MIGRATE flag inside the corresponding source array thus
- * allowing device driver to allocate device memory for those unback virtual
- * address. For this the device driver simply have to allocate device memory
- * and properly set the destination entry like for regular migration. Note that
- * this can still fails and thus inside the device driver must check if the
- * migration was successful for those entry inside the finalize_and_map()
- * callback just like for regular migration.
- *
- * THE alloc_and_copy() CALLBACK MUST NOT CHANGE ANY OF THE SRC ARRAY ENTRIES
- * OR BAD THINGS WILL HAPPEN !
- *
- *
- * The finalize_and_map() callback happens after struct page migration from
- * source to destination (destination struct pages are the struct pages for the
- * memory allocated by the alloc_and_copy() callback). Migration can fail, and
- * thus the finalize_and_map() allows the driver to inspect which pages were
- * successfully migrated, and which were not. Successfully migrated pages will
- * have the MIGRATE_PFN_MIGRATE flag set for their src array entry.
- *
- * It is safe to update device page table from within the finalize_and_map()
- * callback because both destination and source page are still locked, and the
- * mmap_sem is held in read mode (hence no one can unmap the range being
- * migrated).
- *
- * Once callback is done cleaning up things and updating its page table (if it
- * chose to do so, this is not an obligation) then it returns. At this point,
- * the HMM core will finish up the final steps, and the migration is complete.
- *
- * THE finalize_and_map() CALLBACK MUST NOT CHANGE ANY OF THE SRC OR DST ARRAY
- * ENTRIES OR BAD THINGS WILL HAPPEN !
- */
-struct migrate_vma_ops {
- void (*alloc_and_copy)(struct vm_area_struct *vma,
- const unsigned long *src,
- unsigned long *dst,
- unsigned long start,
- unsigned long end,
- void *private);
- void (*finalize_and_map)(struct vm_area_struct *vma,
- const unsigned long *src,
- const unsigned long *dst,
- unsigned long start,
- unsigned long end,
- void *private);
+struct migrate_vma {
+ struct vm_area_struct *vma;
+ /*
+ * Both src and dst array must be big enough for
+ * (end - start) >> PAGE_SHIFT entries.
+ *
+ * The src array must not be modified by the caller after
+ * migrate_vma_setup(), and must not change the dst array after
+ * migrate_vma_pages() returns.
+ */
+ unsigned long *dst;
+ unsigned long *src;
+ unsigned long cpages;
+ unsigned long npages;
+ unsigned long start;
+ unsigned long end;
};
-#if defined(CONFIG_MIGRATE_VMA_HELPER)
-int migrate_vma(const struct migrate_vma_ops *ops,
- struct vm_area_struct *vma,
- unsigned long start,
- unsigned long end,
- unsigned long *src,
- unsigned long *dst,
- void *private);
-#else
-static inline int migrate_vma(const struct migrate_vma_ops *ops,
- struct vm_area_struct *vma,
- unsigned long start,
- unsigned long end,
- unsigned long *src,
- unsigned long *dst,
- void *private)
-{
- return -EINVAL;
-}
-#endif /* IS_ENABLED(CONFIG_MIGRATE_VMA_HELPER) */
+int migrate_vma_setup(struct migrate_vma *args);
+void migrate_vma_pages(struct migrate_vma *migrate);
+void migrate_vma_finalize(struct migrate_vma *migrate);
#endif /* CONFIG_MIGRATION */
struct address_space;
struct mem_cgroup;
-struct hmm;
/*
* Each physical page in the system has a struct page associated with
atomic_long_t hugetlb_usage;
#endif
struct work_struct async_put_work;
-
-#ifdef CONFIG_HMM_MIRROR
- /* HMM needs to track a few things per mm */
- struct hmm *hmm;
-#endif
} __randomize_layout;
/*
*/
void (*invalidate_range)(struct mmu_notifier *mn, struct mm_struct *mm,
unsigned long start, unsigned long end);
+
+ /*
+ * These callbacks are used with the get/put interface to manage the
+ * lifetime of the mmu_notifier memory. alloc_notifier() returns a new
+ * notifier for use with the mm.
+ *
+ * free_notifier() is only called after the mmu_notifier has been
+ * fully put, calls to any ops callback are prevented and no ops
+ * callbacks are currently running. It is called from a SRCU callback
+ * and cannot sleep.
+ */
+ struct mmu_notifier *(*alloc_notifier)(struct mm_struct *mm);
+ void (*free_notifier)(struct mmu_notifier *mn);
};
/*
struct mmu_notifier {
struct hlist_node hlist;
const struct mmu_notifier_ops *ops;
+ struct mm_struct *mm;
+ struct rcu_head rcu;
+ unsigned int users;
};
static inline int mm_has_notifiers(struct mm_struct *mm)
return unlikely(mm->mmu_notifier_mm);
}
+struct mmu_notifier *mmu_notifier_get_locked(const struct mmu_notifier_ops *ops,
+ struct mm_struct *mm);
+static inline struct mmu_notifier *
+mmu_notifier_get(const struct mmu_notifier_ops *ops, struct mm_struct *mm)
+{
+ struct mmu_notifier *ret;
+
+ down_write(&mm->mmap_sem);
+ ret = mmu_notifier_get_locked(ops, mm);
+ up_write(&mm->mmap_sem);
+ return ret;
+}
+void mmu_notifier_put(struct mmu_notifier *mn);
+void mmu_notifier_synchronize(void);
+
extern int mmu_notifier_register(struct mmu_notifier *mn,
struct mm_struct *mm);
extern int __mmu_notifier_register(struct mmu_notifier *mn,
#define pudp_huge_clear_flush_notify pudp_huge_clear_flush
#define set_pte_at_notify set_pte_at
+static inline void mmu_notifier_synchronize(void)
+{
+}
+
#endif /* CONFIG_MMU_NOTIFIER */
#endif /* _LINUX_MMU_NOTIFIER_H */
mm_init_owner(mm, p);
RCU_INIT_POINTER(mm->exe_file, NULL);
mmu_notifier_mm_init(mm);
- hmm_mm_init(mm);
init_tlb_flush_pending(mm);
#if defined(CONFIG_TRANSPARENT_HUGEPAGE) && !USE_SPLIT_PMD_PTLOCKS
mm->pmd_huge_pte = NULL;
EXPORT_SYMBOL(resource_list_free);
#ifdef CONFIG_DEVICE_PRIVATE
-/**
- * devm_request_free_mem_region - find free region for device private memory
- *
- * @dev: device struct to bind the resource to
- * @size: size in bytes of the device memory to add
- * @base: resource tree to look in
- *
- * This function tries to find an empty range of physical address big enough to
- * contain the new resource, so that it can later be hotplugged as ZONE_DEVICE
- * memory, which in turn allocates struct pages.
- */
-struct resource *devm_request_free_mem_region(struct device *dev,
- struct resource *base, unsigned long size)
+static struct resource *__request_free_mem_region(struct device *dev,
+ struct resource *base, unsigned long size, const char *name)
{
resource_size_t end, addr;
struct resource *res;
REGION_DISJOINT)
continue;
- res = devm_request_mem_region(dev, addr, size, dev_name(dev));
+ if (dev)
+ res = devm_request_mem_region(dev, addr, size, name);
+ else
+ res = request_mem_region(addr, size, name);
if (!res)
return ERR_PTR(-ENOMEM);
res->desc = IORES_DESC_DEVICE_PRIVATE_MEMORY;
return ERR_PTR(-ERANGE);
}
+
+/**
+ * devm_request_free_mem_region - find free region for device private memory
+ *
+ * @dev: device struct to bind the resource to
+ * @size: size in bytes of the device memory to add
+ * @base: resource tree to look in
+ *
+ * This function tries to find an empty range of physical address big enough to
+ * contain the new resource, so that it can later be hotplugged as ZONE_DEVICE
+ * memory, which in turn allocates struct pages.
+ */
+struct resource *devm_request_free_mem_region(struct device *dev,
+ struct resource *base, unsigned long size)
+{
+ return __request_free_mem_region(dev, base, size, dev_name(dev));
+}
EXPORT_SYMBOL_GPL(devm_request_free_mem_region);
+
+struct resource *request_free_mem_region(struct resource *base,
+ unsigned long size, const char *name)
+{
+ return __request_free_mem_region(NULL, base, size, name);
+}
+EXPORT_SYMBOL_GPL(request_free_mem_region);
+
#endif /* CONFIG_DEVICE_PRIVATE */
static int __init strict_iomem(char *str)
If FS_DAX is enabled, then say Y.
-config MIGRATE_VMA_HELPER
- bool
-
config DEV_PAGEMAP_OPS
bool
+#
+# Helpers to mirror range of the CPU page tables of a process into device page
+# tables.
+#
config HMM_MIRROR
- bool "HMM mirror CPU page table into a device page table"
- depends on (X86_64 || PPC64)
- depends on MMU && 64BIT
- select MMU_NOTIFIER
- help
- Select HMM_MIRROR if you want to mirror range of the CPU page table of a
- process into a device page table. Here, mirror means "keep synchronized".
- Prerequisites: the device must provide the ability to write-protect its
- page tables (at PAGE_SIZE granularity), and must be able to recover from
- the resulting potential page faults.
+ bool
+ depends on MMU
+ depends on MMU_NOTIFIER
config DEVICE_PRIVATE
bool "Unaddressable device memory (GPU memory, ...)"
#include <linux/mmu_notifier.h>
#include <linux/memory_hotplug.h>
-static const struct mmu_notifier_ops hmm_mmu_notifier_ops;
-
-/**
- * hmm_get_or_create - register HMM against an mm (HMM internal)
- *
- * @mm: mm struct to attach to
- * Returns: returns an HMM object, either by referencing the existing
- * (per-process) object, or by creating a new one.
- *
- * This is not intended to be used directly by device drivers. If mm already
- * has an HMM struct then it get a reference on it and returns it. Otherwise
- * it allocates an HMM struct, initializes it, associate it with the mm and
- * returns it.
- */
-static struct hmm *hmm_get_or_create(struct mm_struct *mm)
+static struct mmu_notifier *hmm_alloc_notifier(struct mm_struct *mm)
{
struct hmm *hmm;
- lockdep_assert_held_write(&mm->mmap_sem);
-
- /* Abuse the page_table_lock to also protect mm->hmm. */
- spin_lock(&mm->page_table_lock);
- hmm = mm->hmm;
- if (mm->hmm && kref_get_unless_zero(&mm->hmm->kref))
- goto out_unlock;
- spin_unlock(&mm->page_table_lock);
-
- hmm = kmalloc(sizeof(*hmm), GFP_KERNEL);
+ hmm = kzalloc(sizeof(*hmm), GFP_KERNEL);
if (!hmm)
- return NULL;
+ return ERR_PTR(-ENOMEM);
+
init_waitqueue_head(&hmm->wq);
INIT_LIST_HEAD(&hmm->mirrors);
init_rwsem(&hmm->mirrors_sem);
- hmm->mmu_notifier.ops = NULL;
INIT_LIST_HEAD(&hmm->ranges);
spin_lock_init(&hmm->ranges_lock);
- kref_init(&hmm->kref);
hmm->notifiers = 0;
- hmm->mm = mm;
-
- hmm->mmu_notifier.ops = &hmm_mmu_notifier_ops;
- if (__mmu_notifier_register(&hmm->mmu_notifier, mm)) {
- kfree(hmm);
- return NULL;
- }
-
- mmgrab(hmm->mm);
-
- /*
- * We hold the exclusive mmap_sem here so we know that mm->hmm is
- * still NULL or 0 kref, and is safe to update.
- */
- spin_lock(&mm->page_table_lock);
- mm->hmm = hmm;
-
-out_unlock:
- spin_unlock(&mm->page_table_lock);
- return hmm;
+ return &hmm->mmu_notifier;
}
-static void hmm_free_rcu(struct rcu_head *rcu)
+static void hmm_free_notifier(struct mmu_notifier *mn)
{
- struct hmm *hmm = container_of(rcu, struct hmm, rcu);
+ struct hmm *hmm = container_of(mn, struct hmm, mmu_notifier);
- mmdrop(hmm->mm);
+ WARN_ON(!list_empty(&hmm->ranges));
+ WARN_ON(!list_empty(&hmm->mirrors));
kfree(hmm);
}
-static void hmm_free(struct kref *kref)
-{
- struct hmm *hmm = container_of(kref, struct hmm, kref);
-
- spin_lock(&hmm->mm->page_table_lock);
- if (hmm->mm->hmm == hmm)
- hmm->mm->hmm = NULL;
- spin_unlock(&hmm->mm->page_table_lock);
-
- mmu_notifier_unregister_no_release(&hmm->mmu_notifier, hmm->mm);
- mmu_notifier_call_srcu(&hmm->rcu, hmm_free_rcu);
-}
-
-static inline void hmm_put(struct hmm *hmm)
-{
- kref_put(&hmm->kref, hmm_free);
-}
-
static void hmm_release(struct mmu_notifier *mn, struct mm_struct *mm)
{
struct hmm *hmm = container_of(mn, struct hmm, mmu_notifier);
struct hmm_mirror *mirror;
- /* Bail out if hmm is in the process of being freed */
- if (!kref_get_unless_zero(&hmm->kref))
- return;
-
/*
* Since hmm_range_register() holds the mmget() lock hmm_release() is
* prevented as long as a range exists.
mirror->ops->release(mirror);
}
up_read(&hmm->mirrors_sem);
-
- hmm_put(hmm);
}
static void notifiers_decrement(struct hmm *hmm)
{
struct hmm *hmm = container_of(mn, struct hmm, mmu_notifier);
struct hmm_mirror *mirror;
- struct hmm_update update;
struct hmm_range *range;
unsigned long flags;
int ret = 0;
- if (!kref_get_unless_zero(&hmm->kref))
- return 0;
-
- update.start = nrange->start;
- update.end = nrange->end;
- update.event = HMM_UPDATE_INVALIDATE;
- update.blockable = mmu_notifier_range_blockable(nrange);
-
spin_lock_irqsave(&hmm->ranges_lock, flags);
hmm->notifiers++;
list_for_each_entry(range, &hmm->ranges, list) {
- if (update.end < range->start || update.start >= range->end)
+ if (nrange->end < range->start || nrange->start >= range->end)
continue;
range->valid = false;
list_for_each_entry(mirror, &hmm->mirrors, list) {
int rc;
- rc = mirror->ops->sync_cpu_device_pagetables(mirror, &update);
+ rc = mirror->ops->sync_cpu_device_pagetables(mirror, nrange);
if (rc) {
- if (WARN_ON(update.blockable || rc != -EAGAIN))
+ if (WARN_ON(mmu_notifier_range_blockable(nrange) ||
+ rc != -EAGAIN))
continue;
ret = -EAGAIN;
break;
out:
if (ret)
notifiers_decrement(hmm);
- hmm_put(hmm);
return ret;
}
{
struct hmm *hmm = container_of(mn, struct hmm, mmu_notifier);
- if (!kref_get_unless_zero(&hmm->kref))
- return;
-
notifiers_decrement(hmm);
- hmm_put(hmm);
}
static const struct mmu_notifier_ops hmm_mmu_notifier_ops = {
.release = hmm_release,
.invalidate_range_start = hmm_invalidate_range_start,
.invalidate_range_end = hmm_invalidate_range_end,
+ .alloc_notifier = hmm_alloc_notifier,
+ .free_notifier = hmm_free_notifier,
};
/*
*
* To start mirroring a process address space, the device driver must register
* an HMM mirror struct.
+ *
+ * The caller cannot unregister the hmm_mirror while any ranges are
+ * registered.
+ *
+ * Callers using this function must put a call to mmu_notifier_synchronize()
+ * in their module exit functions.
*/
int hmm_mirror_register(struct hmm_mirror *mirror, struct mm_struct *mm)
{
+ struct mmu_notifier *mn;
+
lockdep_assert_held_write(&mm->mmap_sem);
/* Sanity check */
if (!mm || !mirror || !mirror->ops)
return -EINVAL;
- mirror->hmm = hmm_get_or_create(mm);
- if (!mirror->hmm)
- return -ENOMEM;
+ mn = mmu_notifier_get_locked(&hmm_mmu_notifier_ops, mm);
+ if (IS_ERR(mn))
+ return PTR_ERR(mn);
+ mirror->hmm = container_of(mn, struct hmm, mmu_notifier);
down_write(&mirror->hmm->mirrors_sem);
list_add(&mirror->list, &mirror->hmm->mirrors);
down_write(&hmm->mirrors_sem);
list_del(&mirror->list);
up_write(&hmm->mirrors_sem);
- hmm_put(hmm);
+ mmu_notifier_put(&hmm->mmu_notifier);
}
EXPORT_SYMBOL(hmm_mirror_unregister);
struct hmm_range *range;
struct dev_pagemap *pgmap;
unsigned long last;
- bool fault;
- bool block;
+ unsigned int flags;
};
static int hmm_vma_do_fault(struct mm_walk *walk, unsigned long addr,
struct vm_area_struct *vma = walk->vma;
vm_fault_t ret;
- flags |= hmm_vma_walk->block ? 0 : FAULT_FLAG_ALLOW_RETRY;
- flags |= write_fault ? FAULT_FLAG_WRITE : 0;
+ if (hmm_vma_walk->flags & HMM_FAULT_ALLOW_RETRY)
+ flags |= FAULT_FLAG_ALLOW_RETRY;
+ if (write_fault)
+ flags |= FAULT_FLAG_WRITE;
+
ret = handle_mm_fault(vma, addr, flags);
- if (ret & VM_FAULT_RETRY)
+ if (ret & VM_FAULT_RETRY) {
+ /* Note, handle_mm_fault did up_read(&mm->mmap_sem)) */
return -EAGAIN;
+ }
if (ret & VM_FAULT_ERROR) {
*pfn = range->values[HMM_PFN_ERROR];
return -EFAULT;
}
/*
- * hmm_vma_walk_hole() - handle a range lacking valid pmd or pte(s)
- * @start: range virtual start address (inclusive)
+ * hmm_vma_walk_hole_() - handle a range lacking valid pmd or pte(s)
+ * @addr: range virtual start address (inclusive)
* @end: range virtual end address (exclusive)
* @fault: should we fault or not ?
* @write_fault: write fault ?
struct hmm_vma_walk *hmm_vma_walk = walk->private;
struct hmm_range *range = hmm_vma_walk->range;
uint64_t *pfns = range->pfns;
- unsigned long i, page_size;
+ unsigned long i;
hmm_vma_walk->last = addr;
- page_size = hmm_range_page_size(range);
- i = (addr - range->start) >> range->page_shift;
+ i = (addr - range->start) >> PAGE_SHIFT;
- for (; addr < end; addr += page_size, i++) {
+ for (; addr < end; addr += PAGE_SIZE, i++) {
pfns[i] = range->values[HMM_PFN_NONE];
if (fault || write_fault) {
int ret;
{
struct hmm_range *range = hmm_vma_walk->range;
- if (!hmm_vma_walk->fault)
+ if (hmm_vma_walk->flags & HMM_FAULT_SNAPSHOT)
return;
/*
* So we not only consider the individual per page request we also
* consider the default flags requested for the range. The API can
- * be use in 2 fashions. The first one where the HMM user coalesce
- * multiple page fault into one request and set flags per pfns for
- * of those faults. The second one where the HMM user want to pre-
+ * be used 2 ways. The first one where the HMM user coalesces
+ * multiple page faults into one request and sets flags per pfn for
+ * those faults. The second one where the HMM user wants to pre-
* fault a range with specific flags. For the latter one it is a
* waste to have the user pre-fill the pfn arrays with a default
* flags value.
/* We aren't ask to do anything ... */
if (!(pfns & range->flags[HMM_PFN_VALID]))
return;
- /* If this is device memory than only fault if explicitly requested */
+ /* If this is device memory then only fault if explicitly requested */
if ((cpu_flags & range->flags[HMM_PFN_DEVICE_PRIVATE])) {
/* Do we fault on device memory ? */
if (pfns & range->flags[HMM_PFN_DEVICE_PRIVATE]) {
{
unsigned long i;
- if (!hmm_vma_walk->fault) {
+ if (hmm_vma_walk->flags & HMM_FAULT_SNAPSHOT) {
*fault = *write_fault = false;
return;
}
range->flags[HMM_PFN_VALID];
}
-static inline uint64_t pud_to_hmm_pfn_flags(struct hmm_range *range, pud_t pud)
-{
- if (!pud_present(pud))
- return 0;
- return pud_write(pud) ? range->flags[HMM_PFN_VALID] |
- range->flags[HMM_PFN_WRITE] :
- range->flags[HMM_PFN_VALID];
-}
-
-static int hmm_vma_handle_pmd(struct mm_walk *walk,
- unsigned long addr,
- unsigned long end,
- uint64_t *pfns,
- pmd_t pmd)
-{
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
+static int hmm_vma_handle_pmd(struct mm_walk *walk, unsigned long addr,
+ unsigned long end, uint64_t *pfns, pmd_t pmd)
+{
struct hmm_vma_walk *hmm_vma_walk = walk->private;
struct hmm_range *range = hmm_vma_walk->range;
unsigned long pfn, npages, i;
if (pmd_protnone(pmd) || fault || write_fault)
return hmm_vma_walk_hole_(addr, end, fault, write_fault, walk);
- pfn = pmd_pfn(pmd) + pte_index(addr);
+ pfn = pmd_pfn(pmd) + ((addr & ~PMD_MASK) >> PAGE_SHIFT);
for (i = 0; addr < end; addr += PAGE_SIZE, i++, pfn++) {
if (pmd_devmap(pmd)) {
hmm_vma_walk->pgmap = get_dev_pagemap(pfn,
}
hmm_vma_walk->last = end;
return 0;
-#else
- /* If THP is not enabled then we should never reach that code ! */
- return -EINVAL;
-#endif
}
+#else /* CONFIG_TRANSPARENT_HUGEPAGE */
+/* stub to allow the code below to compile */
+int hmm_vma_handle_pmd(struct mm_walk *walk, unsigned long addr,
+ unsigned long end, uint64_t *pfns, pmd_t pmd);
+#endif /* CONFIG_TRANSPARENT_HUGEPAGE */
static inline uint64_t pte_to_hmm_pfn_flags(struct hmm_range *range, pte_t pte)
{
{
struct hmm_vma_walk *hmm_vma_walk = walk->private;
struct hmm_range *range = hmm_vma_walk->range;
- struct vm_area_struct *vma = walk->vma;
bool fault, write_fault;
uint64_t cpu_flags;
pte_t pte = *ptep;
if (fault || write_fault) {
pte_unmap(ptep);
hmm_vma_walk->last = addr;
- migration_entry_wait(vma->vm_mm,
- pmdp, addr);
+ migration_entry_wait(walk->mm, pmdp, addr);
return -EBUSY;
}
return 0;
{
struct hmm_vma_walk *hmm_vma_walk = walk->private;
struct hmm_range *range = hmm_vma_walk->range;
- struct vm_area_struct *vma = walk->vma;
uint64_t *pfns = range->pfns;
unsigned long addr = start, i;
pte_t *ptep;
pmd_t pmd;
-
again:
pmd = READ_ONCE(*pmdp);
if (pmd_none(pmd))
return hmm_vma_walk_hole(start, end, walk);
- if (pmd_huge(pmd) && (range->vma->vm_flags & VM_HUGETLB))
- return hmm_pfns_bad(start, end, walk);
-
if (thp_migration_supported() && is_pmd_migration_entry(pmd)) {
bool fault, write_fault;
unsigned long npages;
0, &fault, &write_fault);
if (fault || write_fault) {
hmm_vma_walk->last = addr;
- pmd_migration_entry_wait(vma->vm_mm, pmdp);
+ pmd_migration_entry_wait(walk->mm, pmdp);
return -EBUSY;
}
return 0;
if (pmd_devmap(pmd) || pmd_trans_huge(pmd)) {
/*
- * No need to take pmd_lock here, even if some other threads
+ * No need to take pmd_lock here, even if some other thread
* is splitting the huge pmd we will get that event through
* mmu_notifier callback.
*
- * So just read pmd value and check again its a transparent
+ * So just read pmd value and check again it's a transparent
* huge or device mapping one and compute corresponding pfn
* values.
*/
}
/*
- * We have handled all the valid case above ie either none, migration,
+ * We have handled all the valid cases above ie either none, migration,
* huge or transparent huge. At this point either it is a valid pmd
* entry pointing to pte directory or it is a bad pmd that will not
* recover.
return 0;
}
-static int hmm_vma_walk_pud(pud_t *pudp,
- unsigned long start,
- unsigned long end,
- struct mm_walk *walk)
+#if defined(CONFIG_ARCH_HAS_PTE_DEVMAP) && \
+ defined(CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD)
+static inline uint64_t pud_to_hmm_pfn_flags(struct hmm_range *range, pud_t pud)
+{
+ if (!pud_present(pud))
+ return 0;
+ return pud_write(pud) ? range->flags[HMM_PFN_VALID] |
+ range->flags[HMM_PFN_WRITE] :
+ range->flags[HMM_PFN_VALID];
+}
+
+static int hmm_vma_walk_pud(pud_t *pudp, unsigned long start, unsigned long end,
+ struct mm_walk *walk)
{
struct hmm_vma_walk *hmm_vma_walk = walk->private;
struct hmm_range *range = hmm_vma_walk->range;
return 0;
}
+#else
+#define hmm_vma_walk_pud NULL
+#endif
+#ifdef CONFIG_HUGETLB_PAGE
static int hmm_vma_walk_hugetlb_entry(pte_t *pte, unsigned long hmask,
unsigned long start, unsigned long end,
struct mm_walk *walk)
{
-#ifdef CONFIG_HUGETLB_PAGE
- unsigned long addr = start, i, pfn, mask, size, pfn_inc;
+ unsigned long addr = start, i, pfn;
struct hmm_vma_walk *hmm_vma_walk = walk->private;
struct hmm_range *range = hmm_vma_walk->range;
struct vm_area_struct *vma = walk->vma;
- struct hstate *h = hstate_vma(vma);
uint64_t orig_pfn, cpu_flags;
bool fault, write_fault;
spinlock_t *ptl;
pte_t entry;
int ret = 0;
- size = 1UL << huge_page_shift(h);
- mask = size - 1;
- if (range->page_shift != PAGE_SHIFT) {
- /* Make sure we are looking at full page. */
- if (start & mask)
- return -EINVAL;
- if (end < (start + size))
- return -EINVAL;
- pfn_inc = size >> PAGE_SHIFT;
- } else {
- pfn_inc = 1;
- size = PAGE_SIZE;
- }
-
-
- ptl = huge_pte_lock(hstate_vma(walk->vma), walk->mm, pte);
+ ptl = huge_pte_lock(hstate_vma(vma), walk->mm, pte);
entry = huge_ptep_get(pte);
- i = (start - range->start) >> range->page_shift;
+ i = (start - range->start) >> PAGE_SHIFT;
orig_pfn = range->pfns[i];
range->pfns[i] = range->values[HMM_PFN_NONE];
cpu_flags = pte_to_hmm_pfn_flags(range, entry);
goto unlock;
}
- pfn = pte_pfn(entry) + ((start & mask) >> range->page_shift);
- for (; addr < end; addr += size, i++, pfn += pfn_inc)
+ pfn = pte_pfn(entry) + ((start & ~hmask) >> PAGE_SHIFT);
+ for (; addr < end; addr += PAGE_SIZE, i++, pfn++)
range->pfns[i] = hmm_device_entry_from_pfn(range, pfn) |
cpu_flags;
hmm_vma_walk->last = end;
return hmm_vma_walk_hole_(addr, end, fault, write_fault, walk);
return ret;
-#else /* CONFIG_HUGETLB_PAGE */
- return -EINVAL;
-#endif
}
+#else
+#define hmm_vma_walk_hugetlb_entry NULL
+#endif /* CONFIG_HUGETLB_PAGE */
static void hmm_pfns_clear(struct hmm_range *range,
uint64_t *pfns,
* hmm_range_register() - start tracking change to CPU page table over a range
* @range: range
* @mm: the mm struct for the range of virtual address
- * @start: start virtual address (inclusive)
- * @end: end virtual address (exclusive)
- * @page_shift: expect page shift for the range
- * Returns 0 on success, -EFAULT if the address space is no longer valid
+ *
+ * Return: 0 on success, -EFAULT if the address space is no longer valid
*
* Track updates to the CPU page table see include/linux/hmm.h
*/
-int hmm_range_register(struct hmm_range *range,
- struct hmm_mirror *mirror,
- unsigned long start,
- unsigned long end,
- unsigned page_shift)
+int hmm_range_register(struct hmm_range *range, struct hmm_mirror *mirror)
{
- unsigned long mask = ((1UL << page_shift) - 1UL);
struct hmm *hmm = mirror->hmm;
unsigned long flags;
range->valid = false;
range->hmm = NULL;
- if ((start & mask) || (end & mask))
+ if ((range->start & (PAGE_SIZE - 1)) || (range->end & (PAGE_SIZE - 1)))
return -EINVAL;
- if (start >= end)
+ if (range->start >= range->end)
return -EINVAL;
- range->page_shift = page_shift;
- range->start = start;
- range->end = end;
-
/* Prevent hmm_release() from running while the range is valid */
- if (!mmget_not_zero(hmm->mm))
+ if (!mmget_not_zero(hmm->mmu_notifier.mm))
return -EFAULT;
/* Initialize range to track CPU page table updates. */
spin_lock_irqsave(&hmm->ranges_lock, flags);
range->hmm = hmm;
- kref_get(&hmm->kref);
list_add(&range->list, &hmm->ranges);
/*
spin_unlock_irqrestore(&hmm->ranges_lock, flags);
/* Drop reference taken by hmm_range_register() */
- mmput(hmm->mm);
- hmm_put(hmm);
+ mmput(hmm->mmu_notifier.mm);
/*
* The range is now invalid and the ref on the hmm is dropped, so
}
EXPORT_SYMBOL(hmm_range_unregister);
-/*
- * hmm_range_snapshot() - snapshot CPU page table for a range
- * @range: range
- * Return: -EINVAL if invalid argument, -ENOMEM out of memory, -EPERM invalid
- * permission (for instance asking for write and range is read only),
- * -EBUSY if you need to retry, -EFAULT invalid (ie either no valid
- * vma or it is illegal to access that range), number of valid pages
- * in range->pfns[] (from range start address).
+/**
+ * hmm_range_fault - try to fault some address in a virtual address range
+ * @range: range being faulted
+ * @flags: HMM_FAULT_* flags
*
- * This snapshots the CPU page table for a range of virtual addresses. Snapshot
- * validity is tracked by range struct. See in include/linux/hmm.h for example
- * on how to use.
- */
-long hmm_range_snapshot(struct hmm_range *range)
-{
- const unsigned long device_vma = VM_IO | VM_PFNMAP | VM_MIXEDMAP;
- unsigned long start = range->start, end;
- struct hmm_vma_walk hmm_vma_walk;
- struct hmm *hmm = range->hmm;
- struct vm_area_struct *vma;
- struct mm_walk mm_walk;
-
- lockdep_assert_held(&hmm->mm->mmap_sem);
- do {
- /* If range is no longer valid force retry. */
- if (!range->valid)
- return -EBUSY;
-
- vma = find_vma(hmm->mm, start);
- if (vma == NULL || (vma->vm_flags & device_vma))
- return -EFAULT;
-
- if (is_vm_hugetlb_page(vma)) {
- if (huge_page_shift(hstate_vma(vma)) !=
- range->page_shift &&
- range->page_shift != PAGE_SHIFT)
- return -EINVAL;
- } else {
- if (range->page_shift != PAGE_SHIFT)
- return -EINVAL;
- }
-
- if (!(vma->vm_flags & VM_READ)) {
- /*
- * If vma do not allow read access, then assume that it
- * does not allow write access, either. HMM does not
- * support architecture that allow write without read.
- */
- hmm_pfns_clear(range, range->pfns,
- range->start, range->end);
- return -EPERM;
- }
-
- range->vma = vma;
- hmm_vma_walk.pgmap = NULL;
- hmm_vma_walk.last = start;
- hmm_vma_walk.fault = false;
- hmm_vma_walk.range = range;
- mm_walk.private = &hmm_vma_walk;
- end = min(range->end, vma->vm_end);
-
- mm_walk.vma = vma;
- mm_walk.mm = vma->vm_mm;
- mm_walk.pte_entry = NULL;
- mm_walk.test_walk = NULL;
- mm_walk.hugetlb_entry = NULL;
- mm_walk.pud_entry = hmm_vma_walk_pud;
- mm_walk.pmd_entry = hmm_vma_walk_pmd;
- mm_walk.pte_hole = hmm_vma_walk_hole;
- mm_walk.hugetlb_entry = hmm_vma_walk_hugetlb_entry;
-
- walk_page_range(start, end, &mm_walk);
- start = end;
- } while (start < range->end);
-
- return (hmm_vma_walk.last - range->start) >> PAGE_SHIFT;
-}
-EXPORT_SYMBOL(hmm_range_snapshot);
-
-/*
- * hmm_range_fault() - try to fault some address in a virtual address range
- * @range: range being faulted
- * @block: allow blocking on fault (if true it sleeps and do not drop mmap_sem)
- * Return: number of valid pages in range->pfns[] (from range start
- * address). This may be zero. If the return value is negative,
- * then one of the following values may be returned:
+ * Return: the number of valid pages in range->pfns[] (from range start
+ * address), which may be zero. On error one of the following status codes
+ * can be returned:
*
- * -EINVAL invalid arguments or mm or virtual address are in an
- * invalid vma (for instance device file vma).
- * -ENOMEM: Out of memory.
- * -EPERM: Invalid permission (for instance asking for write and
- * range is read only).
- * -EAGAIN: If you need to retry and mmap_sem was drop. This can only
- * happens if block argument is false.
- * -EBUSY: If the the range is being invalidated and you should wait
- * for invalidation to finish.
- * -EFAULT: Invalid (ie either no valid vma or it is illegal to access
- * that range), number of valid pages in range->pfns[] (from
- * range start address).
+ * -EINVAL: Invalid arguments or mm or virtual address is in an invalid vma
+ * (e.g., device file vma).
+ * -ENOMEM: Out of memory.
+ * -EPERM: Invalid permission (e.g., asking for write and range is read
+ * only).
+ * -EAGAIN: A page fault needs to be retried and mmap_sem was dropped.
+ * -EBUSY: The range has been invalidated and the caller needs to wait for
+ * the invalidation to finish.
+ * -EFAULT: Invalid (i.e., either no valid vma or it is illegal to access
+ * that range) number of valid pages in range->pfns[] (from
+ * range start address).
*
* This is similar to a regular CPU page fault except that it will not trigger
* any memory migration if the memory being faulted is not accessible by CPUs
* On error, for one virtual address in the range, the function will mark the
* corresponding HMM pfn entry with an error flag.
*/
-long hmm_range_fault(struct hmm_range *range, bool block)
+long hmm_range_fault(struct hmm_range *range, unsigned int flags)
{
const unsigned long device_vma = VM_IO | VM_PFNMAP | VM_MIXEDMAP;
unsigned long start = range->start, end;
struct mm_walk mm_walk;
int ret;
- lockdep_assert_held(&hmm->mm->mmap_sem);
+ lockdep_assert_held(&hmm->mmu_notifier.mm->mmap_sem);
do {
/* If range is no longer valid force retry. */
if (!range->valid)
return -EBUSY;
- vma = find_vma(hmm->mm, start);
+ vma = find_vma(hmm->mmu_notifier.mm, start);
if (vma == NULL || (vma->vm_flags & device_vma))
return -EFAULT;
- if (is_vm_hugetlb_page(vma)) {
- if (huge_page_shift(hstate_vma(vma)) !=
- range->page_shift &&
- range->page_shift != PAGE_SHIFT)
- return -EINVAL;
- } else {
- if (range->page_shift != PAGE_SHIFT)
- return -EINVAL;
- }
-
if (!(vma->vm_flags & VM_READ)) {
/*
* If vma do not allow read access, then assume that it
return -EPERM;
}
- range->vma = vma;
hmm_vma_walk.pgmap = NULL;
hmm_vma_walk.last = start;
- hmm_vma_walk.fault = true;
- hmm_vma_walk.block = block;
+ hmm_vma_walk.flags = flags;
hmm_vma_walk.range = range;
mm_walk.private = &hmm_vma_walk;
end = min(range->end, vma->vm_end);
EXPORT_SYMBOL(hmm_range_fault);
/**
- * hmm_range_dma_map() - hmm_range_fault() and dma map page all in one.
- * @range: range being faulted
- * @device: device against to dma map page to
- * @daddrs: dma address of mapped pages
- * @block: allow blocking on fault (if true it sleeps and do not drop mmap_sem)
- * Return: number of pages mapped on success, -EAGAIN if mmap_sem have been
- * drop and you need to try again, some other error value otherwise
+ * hmm_range_dma_map - hmm_range_fault() and dma map page all in one.
+ * @range: range being faulted
+ * @device: device to map page to
+ * @daddrs: array of dma addresses for the mapped pages
+ * @flags: HMM_FAULT_*
*
- * Note same usage pattern as hmm_range_fault().
+ * Return: the number of pages mapped on success (including zero), or any
+ * status return from hmm_range_fault() otherwise.
*/
-long hmm_range_dma_map(struct hmm_range *range,
- struct device *device,
- dma_addr_t *daddrs,
- bool block)
+long hmm_range_dma_map(struct hmm_range *range, struct device *device,
+ dma_addr_t *daddrs, unsigned int flags)
{
unsigned long i, npages, mapped;
long ret;
- ret = hmm_range_fault(range, block);
+ ret = hmm_range_fault(range, flags);
if (ret <= 0)
return ret ? ret : -EBUSY;
/**
* hmm_range_dma_unmap() - unmap range of that was map with hmm_range_dma_map()
* @range: range being unmapped
- * @vma: the vma against which the range (optional)
* @device: device against which dma map was done
* @daddrs: dma address of mapped pages
* @dirty: dirty page if it had the write flag set
* concurrent mmu notifier or sync_cpu_device_pagetables() to make progress.
*/
long hmm_range_dma_unmap(struct hmm_range *range,
- struct vm_area_struct *vma,
struct device *device,
dma_addr_t *daddrs,
bool dirty)
EXPORT_SYMBOL(devmap_managed_key);
static atomic_t devmap_managed_enable;
-static void devmap_managed_enable_put(void *data)
+static void devmap_managed_enable_put(void)
{
if (atomic_dec_and_test(&devmap_managed_enable))
static_branch_disable(&devmap_managed_key);
}
-static int devmap_managed_enable_get(struct device *dev, struct dev_pagemap *pgmap)
+static int devmap_managed_enable_get(struct dev_pagemap *pgmap)
{
if (!pgmap->ops || !pgmap->ops->page_free) {
WARN(1, "Missing page_free method\n");
if (atomic_inc_return(&devmap_managed_enable) == 1)
static_branch_enable(&devmap_managed_key);
- return devm_add_action_or_reset(dev, devmap_managed_enable_put, NULL);
+ return 0;
}
#else
-static int devmap_managed_enable_get(struct device *dev, struct dev_pagemap *pgmap)
+static int devmap_managed_enable_get(struct dev_pagemap *pgmap)
{
return -EINVAL;
}
+static void devmap_managed_enable_put(void)
+{
+}
#endif /* CONFIG_DEV_PAGEMAP_OPS */
static void pgmap_array_delete(struct resource *res)
pgmap->ref = NULL;
}
-static void devm_memremap_pages_release(void *data)
+void memunmap_pages(struct dev_pagemap *pgmap)
{
- struct dev_pagemap *pgmap = data;
- struct device *dev = pgmap->dev;
struct resource *res = &pgmap->res;
unsigned long pfn;
int nid;
untrack_pfn(NULL, PHYS_PFN(res->start), resource_size(res));
pgmap_array_delete(res);
- dev_WARN_ONCE(dev, pgmap->altmap.alloc,
- "%s: failed to free all reserved pages\n", __func__);
+ WARN_ONCE(pgmap->altmap.alloc, "failed to free all reserved pages\n");
+ devmap_managed_enable_put();
+}
+EXPORT_SYMBOL_GPL(memunmap_pages);
+
+static void devm_memremap_pages_release(void *data)
+{
+ memunmap_pages(data);
}
static void dev_pagemap_percpu_release(struct percpu_ref *ref)
complete(&pgmap->done);
}
-/**
- * devm_memremap_pages - remap and provide memmap backing for the given resource
- * @dev: hosting device for @res
- * @pgmap: pointer to a struct dev_pagemap
- *
- * Notes:
- * 1/ At a minimum the res and type members of @pgmap must be initialized
- * by the caller before passing it to this function
- *
- * 2/ The altmap field may optionally be initialized, in which case
- * PGMAP_ALTMAP_VALID must be set in pgmap->flags.
- *
- * 3/ The ref field may optionally be provided, in which pgmap->ref must be
- * 'live' on entry and will be killed and reaped at
- * devm_memremap_pages_release() time, or if this routine fails.
- *
- * 4/ res is expected to be a host memory range that could feasibly be
- * treated as a "System RAM" range, i.e. not a device mmio range, but
- * this is not enforced.
+/*
+ * Not device managed version of dev_memremap_pages, undone by
+ * memunmap_pages(). Please use dev_memremap_pages if you have a struct
+ * device available.
*/
-void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap)
+void *memremap_pages(struct dev_pagemap *pgmap, int nid)
{
struct resource *res = &pgmap->res;
struct dev_pagemap *conflict_pgmap;
.altmap = pgmap_altmap(pgmap),
};
pgprot_t pgprot = PAGE_KERNEL;
- int error, nid, is_ram;
+ int error, is_ram;
bool need_devmap_managed = true;
switch (pgmap->type) {
}
if (need_devmap_managed) {
- error = devmap_managed_enable_get(dev, pgmap);
+ error = devmap_managed_enable_get(pgmap);
if (error)
return ERR_PTR(error);
}
conflict_pgmap = get_dev_pagemap(PHYS_PFN(res->start), NULL);
if (conflict_pgmap) {
- dev_WARN(dev, "Conflicting mapping in same section\n");
+ WARN(1, "Conflicting mapping in same section\n");
put_dev_pagemap(conflict_pgmap);
error = -ENOMEM;
goto err_array;
conflict_pgmap = get_dev_pagemap(PHYS_PFN(res->end), NULL);
if (conflict_pgmap) {
- dev_WARN(dev, "Conflicting mapping in same section\n");
+ WARN(1, "Conflicting mapping in same section\n");
put_dev_pagemap(conflict_pgmap);
error = -ENOMEM;
goto err_array;
goto err_array;
}
- pgmap->dev = dev;
-
error = xa_err(xa_store_range(&pgmap_array, PHYS_PFN(res->start),
PHYS_PFN(res->end), pgmap, GFP_KERNEL));
if (error)
goto err_array;
- nid = dev_to_node(dev);
if (nid < 0)
nid = numa_mem_id();
PHYS_PFN(res->start),
PHYS_PFN(resource_size(res)), pgmap);
percpu_ref_get_many(pgmap->ref, pfn_end(pgmap) - pfn_first(pgmap));
-
- error = devm_add_action_or_reset(dev, devm_memremap_pages_release,
- pgmap);
- if (error)
- return ERR_PTR(error);
-
return __va(res->start);
err_add_memory:
err_array:
dev_pagemap_kill(pgmap);
dev_pagemap_cleanup(pgmap);
+ devmap_managed_enable_put();
return ERR_PTR(error);
}
+EXPORT_SYMBOL_GPL(memremap_pages);
+
+/**
+ * devm_memremap_pages - remap and provide memmap backing for the given resource
+ * @dev: hosting device for @res
+ * @pgmap: pointer to a struct dev_pagemap
+ *
+ * Notes:
+ * 1/ At a minimum the res and type members of @pgmap must be initialized
+ * by the caller before passing it to this function
+ *
+ * 2/ The altmap field may optionally be initialized, in which case
+ * PGMAP_ALTMAP_VALID must be set in pgmap->flags.
+ *
+ * 3/ The ref field may optionally be provided, in which pgmap->ref must be
+ * 'live' on entry and will be killed and reaped at
+ * devm_memremap_pages_release() time, or if this routine fails.
+ *
+ * 4/ res is expected to be a host memory range that could feasibly be
+ * treated as a "System RAM" range, i.e. not a device mmio range, but
+ * this is not enforced.
+ */
+void *devm_memremap_pages(struct device *dev, struct dev_pagemap *pgmap)
+{
+ int error;
+ void *ret;
+
+ ret = memremap_pages(pgmap, dev_to_node(dev));
+ if (IS_ERR(ret))
+ return ret;
+
+ error = devm_add_action_or_reset(dev, devm_memremap_pages_release,
+ pgmap);
+ if (error)
+ return ERR_PTR(error);
+ return ret;
+}
EXPORT_SYMBOL_GPL(devm_memremap_pages);
void devm_memunmap_pages(struct device *dev, struct dev_pagemap *pgmap)
#endif /* CONFIG_NUMA */
-#if defined(CONFIG_MIGRATE_VMA_HELPER)
-struct migrate_vma {
- struct vm_area_struct *vma;
- unsigned long *dst;
- unsigned long *src;
- unsigned long cpages;
- unsigned long npages;
- unsigned long start;
- unsigned long end;
-};
-
+#ifdef CONFIG_DEVICE_PRIVATE
static int migrate_vma_collect_hole(unsigned long start,
unsigned long end,
struct mm_walk *walk)
goto next;
page = device_private_entry_to_page(entry);
- mpfn = migrate_pfn(page_to_pfn(page))|
- MIGRATE_PFN_DEVICE | MIGRATE_PFN_MIGRATE;
+ mpfn = migrate_pfn(page_to_pfn(page)) |
+ MIGRATE_PFN_MIGRATE;
if (is_write_device_private_entry(entry))
mpfn |= MIGRATE_PFN_WRITE;
} else {
}
}
+/**
+ * migrate_vma_setup() - prepare to migrate a range of memory
+ * @args: contains the vma, start, and and pfns arrays for the migration
+ *
+ * Returns: negative errno on failures, 0 when 0 or more pages were migrated
+ * without an error.
+ *
+ * Prepare to migrate a range of memory virtual address range by collecting all
+ * the pages backing each virtual address in the range, saving them inside the
+ * src array. Then lock those pages and unmap them. Once the pages are locked
+ * and unmapped, check whether each page is pinned or not. Pages that aren't
+ * pinned have the MIGRATE_PFN_MIGRATE flag set (by this function) in the
+ * corresponding src array entry. Then restores any pages that are pinned, by
+ * remapping and unlocking those pages.
+ *
+ * The caller should then allocate destination memory and copy source memory to
+ * it for all those entries (ie with MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE
+ * flag set). Once these are allocated and copied, the caller must update each
+ * corresponding entry in the dst array with the pfn value of the destination
+ * page and with the MIGRATE_PFN_VALID and MIGRATE_PFN_LOCKED flags set
+ * (destination pages must have their struct pages locked, via lock_page()).
+ *
+ * Note that the caller does not have to migrate all the pages that are marked
+ * with MIGRATE_PFN_MIGRATE flag in src array unless this is a migration from
+ * device memory to system memory. If the caller cannot migrate a device page
+ * back to system memory, then it must return VM_FAULT_SIGBUS, which has severe
+ * consequences for the userspace process, so it must be avoided if at all
+ * possible.
+ *
+ * For empty entries inside CPU page table (pte_none() or pmd_none() is true) we
+ * do set MIGRATE_PFN_MIGRATE flag inside the corresponding source array thus
+ * allowing the caller to allocate device memory for those unback virtual
+ * address. For this the caller simply has to allocate device memory and
+ * properly set the destination entry like for regular migration. Note that
+ * this can still fails and thus inside the device driver must check if the
+ * migration was successful for those entries after calling migrate_vma_pages()
+ * just like for regular migration.
+ *
+ * After that, the callers must call migrate_vma_pages() to go over each entry
+ * in the src array that has the MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE flag
+ * set. If the corresponding entry in dst array has MIGRATE_PFN_VALID flag set,
+ * then migrate_vma_pages() to migrate struct page information from the source
+ * struct page to the destination struct page. If it fails to migrate the
+ * struct page information, then it clears the MIGRATE_PFN_MIGRATE flag in the
+ * src array.
+ *
+ * At this point all successfully migrated pages have an entry in the src
+ * array with MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE flag set and the dst
+ * array entry with MIGRATE_PFN_VALID flag set.
+ *
+ * Once migrate_vma_pages() returns the caller may inspect which pages were
+ * successfully migrated, and which were not. Successfully migrated pages will
+ * have the MIGRATE_PFN_MIGRATE flag set for their src array entry.
+ *
+ * It is safe to update device page table after migrate_vma_pages() because
+ * both destination and source page are still locked, and the mmap_sem is held
+ * in read mode (hence no one can unmap the range being migrated).
+ *
+ * Once the caller is done cleaning up things and updating its page table (if it
+ * chose to do so, this is not an obligation) it finally calls
+ * migrate_vma_finalize() to update the CPU page table to point to new pages
+ * for successfully migrated pages or otherwise restore the CPU page table to
+ * point to the original source pages.
+ */
+int migrate_vma_setup(struct migrate_vma *args)
+{
+ long nr_pages = (args->end - args->start) >> PAGE_SHIFT;
+
+ args->start &= PAGE_MASK;
+ args->end &= PAGE_MASK;
+ if (!args->vma || is_vm_hugetlb_page(args->vma) ||
+ (args->vma->vm_flags & VM_SPECIAL) || vma_is_dax(args->vma))
+ return -EINVAL;
+ if (nr_pages <= 0)
+ return -EINVAL;
+ if (args->start < args->vma->vm_start ||
+ args->start >= args->vma->vm_end)
+ return -EINVAL;
+ if (args->end <= args->vma->vm_start || args->end > args->vma->vm_end)
+ return -EINVAL;
+ if (!args->src || !args->dst)
+ return -EINVAL;
+
+ memset(args->src, 0, sizeof(*args->src) * nr_pages);
+ args->cpages = 0;
+ args->npages = 0;
+
+ migrate_vma_collect(args);
+
+ if (args->cpages)
+ migrate_vma_prepare(args);
+ if (args->cpages)
+ migrate_vma_unmap(args);
+
+ /*
+ * At this point pages are locked and unmapped, and thus they have
+ * stable content and can safely be copied to destination memory that
+ * is allocated by the drivers.
+ */
+ return 0;
+
+}
+EXPORT_SYMBOL(migrate_vma_setup);
+
static void migrate_vma_insert_page(struct migrate_vma *migrate,
unsigned long addr,
struct page *page,
*src &= ~MIGRATE_PFN_MIGRATE;
}
-/*
+/**
* migrate_vma_pages() - migrate meta-data from src page to dst page
* @migrate: migrate struct containing all migration information
*
* struct page. This effectively finishes the migration from source page to the
* destination page.
*/
-static void migrate_vma_pages(struct migrate_vma *migrate)
+void migrate_vma_pages(struct migrate_vma *migrate)
{
const unsigned long npages = migrate->npages;
const unsigned long start = migrate->start;
if (notified)
mmu_notifier_invalidate_range_only_end(&range);
}
+EXPORT_SYMBOL(migrate_vma_pages);
-/*
+/**
* migrate_vma_finalize() - restore CPU page table entry
* @migrate: migrate struct containing all migration information
*
* This also unlocks the pages and puts them back on the lru, or drops the extra
* refcount, for device pages.
*/
-static void migrate_vma_finalize(struct migrate_vma *migrate)
+void migrate_vma_finalize(struct migrate_vma *migrate)
{
const unsigned long npages = migrate->npages;
unsigned long i;
}
}
}
-
-/*
- * migrate_vma() - migrate a range of memory inside vma
- *
- * @ops: migration callback for allocating destination memory and copying
- * @vma: virtual memory area containing the range to be migrated
- * @start: start address of the range to migrate (inclusive)
- * @end: end address of the range to migrate (exclusive)
- * @src: array of hmm_pfn_t containing source pfns
- * @dst: array of hmm_pfn_t containing destination pfns
- * @private: pointer passed back to each of the callback
- * Returns: 0 on success, error code otherwise
- *
- * This function tries to migrate a range of memory virtual address range, using
- * callbacks to allocate and copy memory from source to destination. First it
- * collects all the pages backing each virtual address in the range, saving this
- * inside the src array. Then it locks those pages and unmaps them. Once the pages
- * are locked and unmapped, it checks whether each page is pinned or not. Pages
- * that aren't pinned have the MIGRATE_PFN_MIGRATE flag set (by this function)
- * in the corresponding src array entry. It then restores any pages that are
- * pinned, by remapping and unlocking those pages.
- *
- * At this point it calls the alloc_and_copy() callback. For documentation on
- * what is expected from that callback, see struct migrate_vma_ops comments in
- * include/linux/migrate.h
- *
- * After the alloc_and_copy() callback, this function goes over each entry in
- * the src array that has the MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE flag
- * set. If the corresponding entry in dst array has MIGRATE_PFN_VALID flag set,
- * then the function tries to migrate struct page information from the source
- * struct page to the destination struct page. If it fails to migrate the struct
- * page information, then it clears the MIGRATE_PFN_MIGRATE flag in the src
- * array.
- *
- * At this point all successfully migrated pages have an entry in the src
- * array with MIGRATE_PFN_VALID and MIGRATE_PFN_MIGRATE flag set and the dst
- * array entry with MIGRATE_PFN_VALID flag set.
- *
- * It then calls the finalize_and_map() callback. See comments for "struct
- * migrate_vma_ops", in include/linux/migrate.h for details about
- * finalize_and_map() behavior.
- *
- * After the finalize_and_map() callback, for successfully migrated pages, this
- * function updates the CPU page table to point to new pages, otherwise it
- * restores the CPU page table to point to the original source pages.
- *
- * Function returns 0 after the above steps, even if no pages were migrated
- * (The function only returns an error if any of the arguments are invalid.)
- *
- * Both src and dst array must be big enough for (end - start) >> PAGE_SHIFT
- * unsigned long entries.
- */
-int migrate_vma(const struct migrate_vma_ops *ops,
- struct vm_area_struct *vma,
- unsigned long start,
- unsigned long end,
- unsigned long *src,
- unsigned long *dst,
- void *private)
-{
- struct migrate_vma migrate;
-
- /* Sanity check the arguments */
- start &= PAGE_MASK;
- end &= PAGE_MASK;
- if (!vma || is_vm_hugetlb_page(vma) || (vma->vm_flags & VM_SPECIAL) ||
- vma_is_dax(vma))
- return -EINVAL;
- if (start < vma->vm_start || start >= vma->vm_end)
- return -EINVAL;
- if (end <= vma->vm_start || end > vma->vm_end)
- return -EINVAL;
- if (!ops || !src || !dst || start >= end)
- return -EINVAL;
-
- memset(src, 0, sizeof(*src) * ((end - start) >> PAGE_SHIFT));
- migrate.src = src;
- migrate.dst = dst;
- migrate.start = start;
- migrate.npages = 0;
- migrate.cpages = 0;
- migrate.end = end;
- migrate.vma = vma;
-
- /* Collect, and try to unmap source pages */
- migrate_vma_collect(&migrate);
- if (!migrate.cpages)
- return 0;
-
- /* Lock and isolate page */
- migrate_vma_prepare(&migrate);
- if (!migrate.cpages)
- return 0;
-
- /* Unmap pages */
- migrate_vma_unmap(&migrate);
- if (!migrate.cpages)
- return 0;
-
- /*
- * At this point pages are locked and unmapped, and thus they have
- * stable content and can safely be copied to destination memory that
- * is allocated by the callback.
- *
- * Note that migration can fail in migrate_vma_struct_page() for each
- * individual page.
- */
- ops->alloc_and_copy(vma, src, dst, start, end, private);
-
- /* This does the real migration of struct page */
- migrate_vma_pages(&migrate);
-
- ops->finalize_and_map(vma, src, dst, start, end, private);
-
- /* Unlock and remap pages */
- migrate_vma_finalize(&migrate);
-
- return 0;
-}
-EXPORT_SYMBOL(migrate_vma);
-#endif /* defined(MIGRATE_VMA_HELPER) */
+EXPORT_SYMBOL(migrate_vma_finalize);
+#endif /* CONFIG_DEVICE_PRIVATE */
pr_info("%pS callback failed with %d in %sblockable context.\n",
mn->ops->invalidate_range_start, _ret,
!mmu_notifier_range_blockable(range) ? "non-" : "");
+ WARN_ON(mmu_notifier_range_blockable(range) ||
+ ret != -EAGAIN);
ret = _ret;
}
}
}
EXPORT_SYMBOL_GPL(__mmu_notifier_invalidate_range);
-static int do_mmu_notifier_register(struct mmu_notifier *mn,
- struct mm_struct *mm,
- int take_mmap_sem)
+/*
+ * Same as mmu_notifier_register but here the caller must hold the
+ * mmap_sem in write mode.
+ */
+int __mmu_notifier_register(struct mmu_notifier *mn, struct mm_struct *mm)
{
- struct mmu_notifier_mm *mmu_notifier_mm;
+ struct mmu_notifier_mm *mmu_notifier_mm = NULL;
int ret;
+ lockdep_assert_held_write(&mm->mmap_sem);
BUG_ON(atomic_read(&mm->mm_users) <= 0);
- ret = -ENOMEM;
- mmu_notifier_mm = kmalloc(sizeof(struct mmu_notifier_mm), GFP_KERNEL);
- if (unlikely(!mmu_notifier_mm))
- goto out;
+ mn->mm = mm;
+ mn->users = 1;
- if (take_mmap_sem)
- down_write(&mm->mmap_sem);
- ret = mm_take_all_locks(mm);
- if (unlikely(ret))
- goto out_clean;
+ if (!mm->mmu_notifier_mm) {
+ /*
+ * kmalloc cannot be called under mm_take_all_locks(), but we
+ * know that mm->mmu_notifier_mm can't change while we hold
+ * the write side of the mmap_sem.
+ */
+ mmu_notifier_mm =
+ kmalloc(sizeof(struct mmu_notifier_mm), GFP_KERNEL);
+ if (!mmu_notifier_mm)
+ return -ENOMEM;
- if (!mm_has_notifiers(mm)) {
INIT_HLIST_HEAD(&mmu_notifier_mm->list);
spin_lock_init(&mmu_notifier_mm->lock);
-
- mm->mmu_notifier_mm = mmu_notifier_mm;
- mmu_notifier_mm = NULL;
}
+
+ ret = mm_take_all_locks(mm);
+ if (unlikely(ret))
+ goto out_clean;
+
+ /* Pairs with the mmdrop in mmu_notifier_unregister_* */
mmgrab(mm);
/*
* We can't race against any other mmu notifier method either
* thanks to mm_take_all_locks().
*/
+ if (mmu_notifier_mm)
+ mm->mmu_notifier_mm = mmu_notifier_mm;
+
spin_lock(&mm->mmu_notifier_mm->lock);
hlist_add_head_rcu(&mn->hlist, &mm->mmu_notifier_mm->list);
spin_unlock(&mm->mmu_notifier_mm->lock);
mm_drop_all_locks(mm);
+ BUG_ON(atomic_read(&mm->mm_users) <= 0);
+ return 0;
+
out_clean:
- if (take_mmap_sem)
- up_write(&mm->mmap_sem);
kfree(mmu_notifier_mm);
-out:
- BUG_ON(atomic_read(&mm->mm_users) <= 0);
return ret;
}
+EXPORT_SYMBOL_GPL(__mmu_notifier_register);
-/*
+/**
+ * mmu_notifier_register - Register a notifier on a mm
+ * @mn: The notifier to attach
+ * @mm: The mm to attach the notifier to
+ *
* Must not hold mmap_sem nor any other VM related lock when calling
* this registration function. Must also ensure mm_users can't go down
* to zero while this runs to avoid races with mmu_notifier_release,
* so mm has to be current->mm or the mm should be pinned safely such
* as with get_task_mm(). If the mm is not current->mm, the mm_users
* pin should be released by calling mmput after mmu_notifier_register
- * returns. mmu_notifier_unregister must be always called to
- * unregister the notifier. mm_count is automatically pinned to allow
- * mmu_notifier_unregister to safely run at any time later, before or
- * after exit_mmap. ->release will always be called before exit_mmap
- * frees the pages.
+ * returns.
+ *
+ * mmu_notifier_unregister() or mmu_notifier_put() must be always called to
+ * unregister the notifier.
+ *
+ * While the caller has a mmu_notifier get the mn->mm pointer will remain
+ * valid, and can be converted to an active mm pointer via mmget_not_zero().
*/
int mmu_notifier_register(struct mmu_notifier *mn, struct mm_struct *mm)
{
- return do_mmu_notifier_register(mn, mm, 1);
+ int ret;
+
+ down_write(&mm->mmap_sem);
+ ret = __mmu_notifier_register(mn, mm);
+ up_write(&mm->mmap_sem);
+ return ret;
}
EXPORT_SYMBOL_GPL(mmu_notifier_register);
-/*
- * Same as mmu_notifier_register but here the caller must hold the
- * mmap_sem in write mode.
+static struct mmu_notifier *
+find_get_mmu_notifier(struct mm_struct *mm, const struct mmu_notifier_ops *ops)
+{
+ struct mmu_notifier *mn;
+
+ spin_lock(&mm->mmu_notifier_mm->lock);
+ hlist_for_each_entry_rcu (mn, &mm->mmu_notifier_mm->list, hlist) {
+ if (mn->ops != ops)
+ continue;
+
+ if (likely(mn->users != UINT_MAX))
+ mn->users++;
+ else
+ mn = ERR_PTR(-EOVERFLOW);
+ spin_unlock(&mm->mmu_notifier_mm->lock);
+ return mn;
+ }
+ spin_unlock(&mm->mmu_notifier_mm->lock);
+ return NULL;
+}
+
+/**
+ * mmu_notifier_get_locked - Return the single struct mmu_notifier for
+ * the mm & ops
+ * @ops: The operations struct being subscribe with
+ * @mm : The mm to attach notifiers too
+ *
+ * This function either allocates a new mmu_notifier via
+ * ops->alloc_notifier(), or returns an already existing notifier on the
+ * list. The value of the ops pointer is used to determine when two notifiers
+ * are the same.
+ *
+ * Each call to mmu_notifier_get() must be paired with a call to
+ * mmu_notifier_put(). The caller must hold the write side of mm->mmap_sem.
+ *
+ * While the caller has a mmu_notifier get the mm pointer will remain valid,
+ * and can be converted to an active mm pointer via mmget_not_zero().
*/
-int __mmu_notifier_register(struct mmu_notifier *mn, struct mm_struct *mm)
+struct mmu_notifier *mmu_notifier_get_locked(const struct mmu_notifier_ops *ops,
+ struct mm_struct *mm)
{
- return do_mmu_notifier_register(mn, mm, 0);
+ struct mmu_notifier *mn;
+ int ret;
+
+ lockdep_assert_held_write(&mm->mmap_sem);
+
+ if (mm->mmu_notifier_mm) {
+ mn = find_get_mmu_notifier(mm, ops);
+ if (mn)
+ return mn;
+ }
+
+ mn = ops->alloc_notifier(mm);
+ if (IS_ERR(mn))
+ return mn;
+ mn->ops = ops;
+ ret = __mmu_notifier_register(mn, mm);
+ if (ret)
+ goto out_free;
+ return mn;
+out_free:
+ mn->ops->free_notifier(mn);
+ return ERR_PTR(ret);
}
-EXPORT_SYMBOL_GPL(__mmu_notifier_register);
+EXPORT_SYMBOL_GPL(mmu_notifier_get_locked);
/* this is called after the last mmu_notifier_unregister() returned */
void __mmu_notifier_mm_destroy(struct mm_struct *mm)
}
EXPORT_SYMBOL_GPL(mmu_notifier_unregister_no_release);
+static void mmu_notifier_free_rcu(struct rcu_head *rcu)
+{
+ struct mmu_notifier *mn = container_of(rcu, struct mmu_notifier, rcu);
+ struct mm_struct *mm = mn->mm;
+
+ mn->ops->free_notifier(mn);
+ /* Pairs with the get in __mmu_notifier_register() */
+ mmdrop(mm);
+}
+
+/**
+ * mmu_notifier_put - Release the reference on the notifier
+ * @mn: The notifier to act on
+ *
+ * This function must be paired with each mmu_notifier_get(), it releases the
+ * reference obtained by the get. If this is the last reference then process
+ * to free the notifier will be run asynchronously.
+ *
+ * Unlike mmu_notifier_unregister() the get/put flow only calls ops->release
+ * when the mm_struct is destroyed. Instead free_notifier is always called to
+ * release any resources held by the user.
+ *
+ * As ops->release is not guaranteed to be called, the user must ensure that
+ * all sptes are dropped, and no new sptes can be established before
+ * mmu_notifier_put() is called.
+ *
+ * This function can be called from the ops->release callback, however the
+ * caller must still ensure it is called pairwise with mmu_notifier_get().
+ *
+ * Modules calling this function must call mmu_notifier_synchronize() in
+ * their __exit functions to ensure the async work is completed.
+ */
+void mmu_notifier_put(struct mmu_notifier *mn)
+{
+ struct mm_struct *mm = mn->mm;
+
+ spin_lock(&mm->mmu_notifier_mm->lock);
+ if (WARN_ON(!mn->users) || --mn->users)
+ goto out_unlock;
+ hlist_del_init_rcu(&mn->hlist);
+ spin_unlock(&mm->mmu_notifier_mm->lock);
+
+ call_srcu(&srcu, &mn->rcu, mmu_notifier_free_rcu);
+ return;
+
+out_unlock:
+ spin_unlock(&mm->mmu_notifier_mm->lock);
+}
+EXPORT_SYMBOL_GPL(mmu_notifier_put);
+
+/**
+ * mmu_notifier_synchronize - Ensure all mmu_notifiers are freed
+ *
+ * This function ensures that all outstanding async SRU work from
+ * mmu_notifier_put() is completed. After it returns any mmu_notifier_ops
+ * associated with an unused mmu_notifier will no longer be called.
+ *
+ * Before using the caller must ensure that all of its mmu_notifiers have been
+ * fully released via mmu_notifier_put().
+ *
+ * Modules using the mmu_notifier_put() API should call this in their __exit
+ * function to avoid module unloading races.
+ */
+void mmu_notifier_synchronize(void)
+{
+ synchronize_srcu(&srcu);
+}
+EXPORT_SYMBOL_GPL(mmu_notifier_synchronize);
+
bool
mmu_notifier_range_update_to_read_only(const struct mmu_notifier_range *range)
{
}
}
- pr_info("%s initialised, %lu pages in %ums\n", dev_name(pgmap->dev),
+ pr_info("%s initialised %lu pages in %ums\n", __func__,
size, jiffies_to_msecs(jiffies - start));
}
if (!nfit_res)
return devm_memremap_pages(dev, pgmap);
- pgmap->dev = dev;
if (!pgmap->ref) {
if (pgmap->ops && (pgmap->ops->kill || pgmap->ops->cleanup))
return ERR_PTR(-EINVAL);
projects / linux-2.6-microblaze.git / commitdiff