On Sun, Feb 13, 2022 at 4:58 AM Shiyang Ruan <ruansy.fnst@xxxxxxxxxxx> wrote:
>
> v10.1 update:
> - Fix build error reported by kernel test robot
> - Add return code to dax_register_holder()
> - Add write lock to prevent concurrent registrations
> - Rename dax_get_holder() to dax_holder()
> - Add kernel doc for new added functions
>
> To easily track filesystem from a pmem device, we introduce a holder for
> dax_device structure, and also its operation. This holder is used to
> remember who is using this dax_device:
> - When it is the backend of a filesystem, the holder will be the
> instance of this filesystem.
> - When this pmem device is one of the targets in a mapped device, the
> holder will be this mapped device. In this case, the mapped device
> has its own dax_device and it will follow the first rule. So that we
> can finally track to the filesystem we needed.
>
> The holder and holder_ops will be set when filesystem is being mounted,
> or an target device is being activated.
>
> Signed-off-by: Shiyang Ruan <ruansy.fnst@xxxxxxxxxxx>
> ---
> drivers/dax/super.c | 95 +++++++++++++++++++++++++++++++++++++++++++++
> include/linux/dax.h | 30 ++++++++++++++
> 2 files changed, 125 insertions(+)
>
> diff --git a/drivers/dax/super.c b/drivers/dax/super.c
> index e3029389d809..d7fb4c36bf16 100644
> --- a/drivers/dax/super.c
> +++ b/drivers/dax/super.c
> @@ -21,6 +21,9 @@
> * @cdev: optional character interface for "device dax"
> * @private: dax driver private data
> * @flags: state and boolean properties
> + * @ops: operations for dax_device
> + * @holder_data: holder of a dax_device: could be filesystem or mapped device
> + * @holder_ops: operations for the inner holder
> */
> struct dax_device {
> struct inode inode;
> @@ -28,6 +31,9 @@ struct dax_device {
> void *private;
> unsigned long flags;
> const struct dax_operations *ops;
> + void *holder_data;
> + struct percpu_rw_semaphore holder_rwsem;
This lock is not necessary, see below...
> + const struct dax_holder_operations *holder_ops;
> };
>
> static dev_t dax_devt;
> @@ -193,6 +199,29 @@ int dax_zero_page_range(struct dax_device *dax_dev, pgoff_t pgoff,
> }
> EXPORT_SYMBOL_GPL(dax_zero_page_range);
>
> +int dax_holder_notify_failure(struct dax_device *dax_dev, u64 off,
> + u64 len, int mf_flags)
> +{
> + int rc, id;
> +
> + id = dax_read_lock();
> + if (!dax_alive(dax_dev)) {
> + rc = -ENXIO;
> + goto out;
> + }
> +
> + if (!dax_dev->holder_ops) {
> + rc = -EOPNOTSUPP;
> + goto out;
> + }
> +
> + rc = dax_dev->holder_ops->notify_failure(dax_dev, off, len, mf_flags);
> +out:
> + dax_read_unlock(id);
> + return rc;
> +}
> +EXPORT_SYMBOL_GPL(dax_holder_notify_failure);
> +
> #ifdef CONFIG_ARCH_HAS_PMEM_API
> void arch_wb_cache_pmem(void *addr, size_t size);
> void dax_flush(struct dax_device *dax_dev, void *addr, size_t size)
> @@ -268,6 +297,13 @@ void kill_dax(struct dax_device *dax_dev)
>
> clear_bit(DAXDEV_ALIVE, &dax_dev->flags);
> synchronize_srcu(&dax_srcu);
> +
> + /* Lock to prevent concurrent registrations. */
> + percpu_down_write(&dax_dev->holder_rwsem);
> + /* clear holder data */
> + dax_dev->holder_ops = NULL;
> + dax_dev->holder_data = NULL;
> + percpu_up_write(&dax_dev->holder_rwsem);
> }
> EXPORT_SYMBOL_GPL(kill_dax);
>
> @@ -393,6 +429,7 @@ struct dax_device *alloc_dax(void *private, const struct dax_operations *ops)
>
> dax_dev->ops = ops;
> dax_dev->private = private;
> + percpu_init_rwsem(&dax_dev->holder_rwsem);
> return dax_dev;
>
> err_dev:
> @@ -409,6 +446,64 @@ void put_dax(struct dax_device *dax_dev)
> }
> EXPORT_SYMBOL_GPL(put_dax);
>
> +/**
> + * dax_holder() - obtain the holder of a dax device
> + * @dax_dev: a dax_device instance
> +
> + * Return: the holder's data which represents the holder if registered,
> + * otherwize NULL.
> + */
> +void *dax_holder(struct dax_device *dax_dev)
> +{
> + if (!dax_alive(dax_dev))
> + return NULL;
> +
> + return dax_dev->holder_data;
> +}
> +EXPORT_SYMBOL_GPL(dax_holder);
> +
> +/**
> + * dax_register_holder() - register a holder to a dax device
> + * @dax_dev: a dax_device instance
> + * @holder: a pointer to a holder's data which represents the holder
> + * @ops: operations of this holder
> +
> + * Return: negative errno if an error occurs, otherwise 0.
> + */
> +int dax_register_holder(struct dax_device *dax_dev, void *holder,
> + const struct dax_holder_operations *ops)
> +{
> + if (!dax_alive(dax_dev))
> + return -ENXIO;
> +
> + /* Already registered */
> + if (dax_holder(dax_dev))
> + return -EBUSY;
Delete this...
> +
> + /* Lock to prevent concurrent registrations. */
> + percpu_down_write(&dax_dev->holder_rwsem);
...and just use cmpxchg:
if (cmpxchg(&dax_dev->holder_data, NULL, holder))
return -EBUSY;
dax_dev->holder_ops = ops;
...and then on the release side you can require that the caller
specify @holder before clearing to make the unregistration symmetric:
if (cmpxchg(&dax_dev->holder_data, holder, NULL) != holder))
return -EBUSY;
dax_dev->holder_ops = NULL;
> + dax_dev->holder_data = holder;
> + dax_dev->holder_ops = ops;
> + percpu_up_write(&dax_dev->holder_rwsem);
> +
> + return 0;
> +}
> +EXPORT_SYMBOL_GPL(dax_register_holder);
> +
> +/**
> + * dax_unregister_holder() - unregister the holder for a dax device
> + * @dax_dev: a dax_device instance
> + */
> +void dax_unregister_holder(struct dax_device *dax_dev)
Per above, require the holder to pass in their holder_data again.
> +{
> + if (!dax_alive(dax_dev))
> + return;
> +
> + dax_dev->holder_data = NULL;
> + dax_dev->holder_ops = NULL;
> +}
> +EXPORT_SYMBOL_GPL(dax_unregister_holder);
> +
> /**
> * inode_dax: convert a public inode into its dax_dev
> * @inode: An inode with i_cdev pointing to a dax_dev
> diff --git a/include/linux/dax.h b/include/linux/dax.h
> index 9fc5f99a0ae2..9800d84e5b7d 100644
> --- a/include/linux/dax.h
> +++ b/include/linux/dax.h
> @@ -32,8 +32,24 @@ struct dax_operations {
> int (*zero_page_range)(struct dax_device *, pgoff_t, size_t);
> };
>
> +struct dax_holder_operations {
> + /*
> + * notify_failure - notify memory failure into inner holder device
> + * @dax_dev: the dax device which contains the holder
> + * @offset: offset on this dax device where memory failure occurs
> + * @len: length of this memory failure event
> + * @flags: action flags for memory failure handler
> + */
> + int (*notify_failure)(struct dax_device *dax_dev, u64 offset,
> + u64 len, int mf_flags);
> +};
> +
> #if IS_ENABLED(CONFIG_DAX)
> struct dax_device *alloc_dax(void *private, const struct dax_operations *ops);
> +int dax_register_holder(struct dax_device *dax_dev, void *holder,
> + const struct dax_holder_operations *ops);
> +void dax_unregister_holder(struct dax_device *dax_dev);
> +void *dax_holder(struct dax_device *dax_dev);
> void put_dax(struct dax_device *dax_dev);
> void kill_dax(struct dax_device *dax_dev);
> void dax_write_cache(struct dax_device *dax_dev, bool wc);
> @@ -53,6 +69,18 @@ static inline bool daxdev_mapping_supported(struct vm_area_struct *vma,
> return dax_synchronous(dax_dev);
> }
> #else
> +static inline int dax_register_holder(struct dax_device *dax_dev, void *holder,
> + const struct dax_holder_operations *ops)
> +{
> + return 0;
> +}
> +static inline void dax_unregister_holder(struct dax_device *dax_dev)
> +{
> +}
> +static inline void *dax_holder(struct dax_device *dax_dev)
> +{
> + return NULL;
> +}
> static inline struct dax_device *alloc_dax(void *private,
> const struct dax_operations *ops)
> {
> @@ -185,6 +213,8 @@ size_t dax_copy_to_iter(struct dax_device *dax_dev, pgoff_t pgoff, void *addr,
> size_t bytes, struct iov_iter *i);
> int dax_zero_page_range(struct dax_device *dax_dev, pgoff_t pgoff,
> size_t nr_pages);
> +int dax_holder_notify_failure(struct dax_device *dax_dev, u64 off, u64 len,
> + int mf_flags);
> void dax_flush(struct dax_device *dax_dev, void *addr, size_t size);
>
> ssize_t dax_iomap_rw(struct kiocb *iocb, struct iov_iter *iter,
> --
> 2.34.1
>
>
>
