From: Dan Williams <dan.j.williams@intel.com>
To: Shiyang Ruan <ruansy.fnst@fujitsu.com>
Cc: Christoph Hellwig <hch@infradead.org>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
linux-xfs <linux-xfs@vger.kernel.org>,
Linux NVDIMM <nvdimm@lists.linux.dev>,
Linux MM <linux-mm@kvack.org>,
linux-fsdevel <linux-fsdevel@vger.kernel.org>,
"Darrick J. Wong" <djwong@kernel.org>,
david <david@fromorbit.com>, Jane Chu <jane.chu@oracle.com>
Subject: Re: [PATCH v10.1 1/9] dax: Introduce holder for dax_device
Date: Tue, 15 Feb 2022 14:06:26 -0800 [thread overview]
Message-ID: <CAPcyv4j9VTGxsNv4QFMONTjh_NWPkQ3dkG-uKvTyRp+X0bBvew@mail.gmail.com> (raw)
In-Reply-To: <20220213125832.2722009-1-ruansy.fnst@fujitsu.com>
On Sun, Feb 13, 2022 at 4:58 AM Shiyang Ruan <ruansy.fnst@fujitsu.com> 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@fujitsu.com>
> ---
> 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
>
>
>
next prev parent reply other threads:[~2022-02-15 22:06 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-01-27 12:40 [PATCH v10 0/9] fsdax: introduce fs query to support reflink Shiyang Ruan
2022-01-27 12:40 ` [PATCH v10 1/9] dax: Introduce holder for dax_device Shiyang Ruan
2022-01-27 16:13 ` kernel test robot
2022-01-27 16:44 ` kernel test robot
2022-02-02 13:03 ` Christoph Hellwig
2022-02-13 12:58 ` [PATCH v10.1 " Shiyang Ruan
2022-02-15 22:06 ` Dan Williams [this message]
2022-01-27 12:40 ` [PATCH v10 2/9] mm: factor helpers for memory_failure_dev_pagemap Shiyang Ruan
2022-02-01 21:03 ` Matthew Wilcox
2022-02-15 22:11 ` Dan Williams
2022-01-27 12:40 ` [PATCH v10 3/9] pagemap,pmem: Introduce ->memory_failure() Shiyang Ruan
2022-02-15 22:38 ` Dan Williams
2022-01-27 12:40 ` [PATCH v10 4/9] fsdax: fix function description Shiyang Ruan
2022-02-02 13:04 ` Christoph Hellwig
2022-02-15 23:51 ` Dan Williams
2022-01-27 12:40 ` [PATCH v10 5/9] fsdax: Introduce dax_load_page() Shiyang Ruan
2022-02-16 1:34 ` Dan Williams
2022-02-16 3:02 ` Shiyang Ruan
2022-02-16 3:07 ` Dan Williams
2022-01-27 12:40 ` [PATCH v10 6/9] mm: move pgoff_address() to vma_pgoff_address() Shiyang Ruan
2022-02-16 1:37 ` Dan Williams
2022-01-27 12:40 ` [PATCH v10 7/9] mm: Introduce mf_dax_kill_procs() for fsdax case Shiyang Ruan
2022-02-16 1:47 ` Dan Williams
2022-02-16 1:49 ` Dan Williams
2022-01-27 12:40 ` [PATCH v10 8/9] xfs: Implement ->notify_failure() for XFS Shiyang Ruan
2022-01-27 17:56 ` kernel test robot
2022-01-27 19:39 ` kernel test robot
2022-02-01 20:41 ` Darrick J. Wong
2022-02-13 13:02 ` [PATCH v10.1 " Shiyang Ruan
2022-02-15 1:46 ` Darrick J. Wong
2022-02-15 9:42 ` Shiyang Ruan
2022-02-16 1:56 ` [PATCH v10 " Dan Williams
2022-01-27 12:40 ` [PATCH v10 9/9] fsdax: set a CoW flag when associate reflink mappings Shiyang Ruan
2022-02-16 2:09 ` Dan Williams
2022-02-16 2:55 ` Shiyang Ruan
2022-02-16 3:09 ` Dan Williams
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAPcyv4j9VTGxsNv4QFMONTjh_NWPkQ3dkG-uKvTyRp+X0bBvew@mail.gmail.com \
--to=dan.j.williams@intel.com \
--cc=david@fromorbit.com \
--cc=djwong@kernel.org \
--cc=hch@infradead.org \
--cc=jane.chu@oracle.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-xfs@vger.kernel.org \
--cc=nvdimm@lists.linux.dev \
--cc=ruansy.fnst@fujitsu.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).