Merge 'notifier_get_put' into hmm.git

Jason Gunthorpe says:

====================
Add mmu_notifier_get/put for managing mmu notifier registrations

This series introduces a new registration flow for mmu_notifiers based on
the idea that the user would like to get a single refcounted piece of
memory for a mm, keyed to its use.

For instance many users of mmu_notifiers use an interval tree or similar
to dispatch notifications to some object. There are many objects but only
one notifier subscription per mm holding the tree.

Of the 12 places that call mmu_notifier_register:
 - 7 are maintaining some kind of obvious mapping of mm_struct to
   mmu_notifier registration, ie in some linked list or hash table. Of
   the 7 this series converts 4 (gru, hmm, RDMA, radeon)

 - 3 (hfi1, gntdev, vhost) are registering multiple notifiers, but each
   one immediately does some VA range filtering, ie with an interval tree.
   These would be better with a global subsystem-wide range filter and
   could convert to this API.

 - 2 (kvm, amd_iommu) are deliberately using a single mm at a time, and
   really can't use this API. One of the intel-svm's modes is also in this
   list

The 3/7 unconverted drivers are:
 - intel-svm
   This driver tracks mm's in a global linked list 'global_svm_list'
   and would benefit from this API.

   Its flow is a bit complex, since it also wants a set of non-shared
   notifiers.

 - i915_gem_usrptr
   This driver tracks mm's in a per-device hash
   table (dev_priv->mm_structs), but only has an optional use of
   mmu_notifiers.  Since it still seems to need the hash table it is
   difficult to convert.

 - amdkfd/kfd_process
   This driver is using a global SRCU hash table to track mm's

   The control flow here is very complicated and the driver is relying on
   this hash table to be fast on the ioctl syscall path.

   It would definitely benefit, but only if the ioctl path didn't need to
   do the search so often.
====================

Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>

3 files changed, 36 insertions, 17 deletions

diff --git a/include/linux/hmm.h b/include/linux/hmm.h
index 51e18fbb8953..3fec513b9c00 100644
--- a/include/linux/hmm.h
+++ b/include/linux/hmm.h
@@ -84,15 +84,12 @@
  * @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;
 };
 
@@ -409,13 +406,6 @@ long hmm_range_dma_unmap(struct hmm_range *range,
  */
 #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 */
diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
index 3a37a89eb7a7..525d25d93330 100644
--- a/include/linux/mm_types.h
+++ b/include/linux/mm_types.h
@@ -25,7 +25,6 @@
 
 struct address_space;
 struct mem_cgroup;
-struct hmm;
 
 /*
  * Each physical page in the system has a struct page associated with
@@ -502,11 +501,6 @@ struct mm_struct {
 		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;
 
 	/*
diff --git a/include/linux/mmu_notifier.h b/include/linux/mmu_notifier.h
index b6c004bd9f6a..31aa971315a1 100644
--- a/include/linux/mmu_notifier.h
+++ b/include/linux/mmu_notifier.h
@@ -211,6 +211,19 @@ struct mmu_notifier_ops {
 	 */
 	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);
 };
 
 /*
@@ -227,6 +240,9 @@ struct mmu_notifier_ops {
 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)
@@ -234,6 +250,21 @@ 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,
@@ -581,6 +612,10 @@ static inline void mmu_notifier_mm_destroy(struct mm_struct *mm)
 #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 */