summaryrefslogtreecommitdiffstats
path: root/drivers/dax
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2017-07-07 09:44:06 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2017-07-07 09:44:06 -0700
commitb6ffe9ba46016f8351896ccee33bebcd0e5ea7c0 (patch)
tree839a5a070eabe8851797330ea77ca7eb7c93bcc1 /drivers/dax
parent9f45efb9286268e01d5022d34a58a68f53ca3072 (diff)
parent9d92573fff3ec70785ef1815cc80573f70e7a921 (diff)
downloadlinux-b6ffe9ba46016f8351896ccee33bebcd0e5ea7c0.tar.bz2
Merge tag 'libnvdimm-for-4.13' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm
Pull libnvdimm updates from Dan Williams: "libnvdimm updates for the latest ACPI and UEFI specifications. This pull request also includes new 'struct dax_operations' enabling to undo the abuse of copy_user_nocache() for copy operations to pmem. The dax work originally missed 4.12 to address concerns raised by Al. Summary: - Introduce the _flushcache() family of memory copy helpers and use them for persistent memory write operations on x86. The _flushcache() semantic indicates that the cache is either bypassed for the copy operation (movnt) or any lines dirtied by the copy operation are written back (clwb, clflushopt, or clflush). - Extend dax_operations with ->copy_from_iter() and ->flush() operations. These operations and other infrastructure updates allow all persistent memory specific dax functionality to be pushed into libnvdimm and the pmem driver directly. It also allows dax-specific sysfs attributes to be linked to a host device, for example: /sys/block/pmem0/dax/write_cache - Add support for the new NVDIMM platform/firmware mechanisms introduced in ACPI 6.2 and UEFI 2.7. This support includes the v1.2 namespace label format, extensions to the address-range-scrub command set, new error injection commands, and a new BTT (block-translation-table) layout. These updates support inter-OS and pre-OS compatibility. - Fix a longstanding memory corruption bug in nfit_test. - Make the pmem and nvdimm-region 'badblocks' sysfs files poll(2) capable. - Miscellaneous fixes and small updates across libnvdimm and the nfit driver. Acknowledgements that came after the branch was pushed: commit 6aa734a2f38e ("libnvdimm, region, pmem: fix 'badblocks' sysfs_get_dirent() reference lifetime") was reviewed by Toshi Kani <toshi.kani@hpe.com>" * tag 'libnvdimm-for-4.13' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm: (42 commits) libnvdimm, namespace: record 'lbasize' for pmem namespaces acpi/nfit: Issue Start ARS to retrieve existing records libnvdimm: New ACPI 6.2 DSM functions acpi, nfit: Show bus_dsm_mask in sysfs libnvdimm, acpi, nfit: Add bus level dsm mask for pass thru. acpi, nfit: Enable DSM pass thru for root functions. libnvdimm: passthru functions clear to send libnvdimm, btt: convert some info messages to warn/err libnvdimm, region, pmem: fix 'badblocks' sysfs_get_dirent() reference lifetime libnvdimm: fix the clear-error check in nsio_rw_bytes libnvdimm, btt: fix btt_rw_page not returning errors acpi, nfit: quiet invalid block-aperture-region warnings libnvdimm, btt: BTT updates for UEFI 2.7 format acpi, nfit: constify *_attribute_group libnvdimm, pmem: disable dax flushing when pmem is fronting a volatile region libnvdimm, pmem, dax: export a cache control attribute dax: convert to bitmask for flags dax: remove default copy_from_iter fallback libnvdimm, nfit: enable support for volatile ranges libnvdimm, pmem: fix persistence warning ...
Diffstat (limited to 'drivers/dax')
-rw-r--r--drivers/dax/super.c118
1 files changed, 112 insertions, 6 deletions
diff --git a/drivers/dax/super.c b/drivers/dax/super.c
index 922d0823f8ec..ce9e563e6e1d 100644
--- a/drivers/dax/super.c
+++ b/drivers/dax/super.c
@@ -18,6 +18,7 @@
#include <linux/cdev.h>
#include <linux/hash.h>
#include <linux/slab.h>
+#include <linux/uio.h>
#include <linux/dax.h>
#include <linux/fs.h>
@@ -115,13 +116,20 @@ int __bdev_dax_supported(struct super_block *sb, int blocksize)
EXPORT_SYMBOL_GPL(__bdev_dax_supported);
#endif
+enum dax_device_flags {
+ /* !alive + rcu grace period == no new operations / mappings */
+ DAXDEV_ALIVE,
+ /* gate whether dax_flush() calls the low level flush routine */
+ DAXDEV_WRITE_CACHE,
+};
+
/**
* struct dax_device - anchor object for dax services
* @inode: core vfs
* @cdev: optional character interface for "device dax"
* @host: optional name for lookups where the device path is not available
* @private: dax driver private data
- * @alive: !alive + rcu grace period == no new operations / mappings
+ * @flags: state and boolean properties
*/
struct dax_device {
struct hlist_node list;
@@ -129,10 +137,75 @@ struct dax_device {
struct cdev cdev;
const char *host;
void *private;
- bool alive;
+ unsigned long flags;
const struct dax_operations *ops;
};
+static ssize_t write_cache_show(struct device *dev,
+ struct device_attribute *attr, char *buf)
+{
+ struct dax_device *dax_dev = dax_get_by_host(dev_name(dev));
+ ssize_t rc;
+
+ WARN_ON_ONCE(!dax_dev);
+ if (!dax_dev)
+ return -ENXIO;
+
+ rc = sprintf(buf, "%d\n", !!test_bit(DAXDEV_WRITE_CACHE,
+ &dax_dev->flags));
+ put_dax(dax_dev);
+ return rc;
+}
+
+static ssize_t write_cache_store(struct device *dev,
+ struct device_attribute *attr, const char *buf, size_t len)
+{
+ bool write_cache;
+ int rc = strtobool(buf, &write_cache);
+ struct dax_device *dax_dev = dax_get_by_host(dev_name(dev));
+
+ WARN_ON_ONCE(!dax_dev);
+ if (!dax_dev)
+ return -ENXIO;
+
+ if (rc)
+ len = rc;
+ else if (write_cache)
+ set_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+ else
+ clear_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+
+ put_dax(dax_dev);
+ return len;
+}
+static DEVICE_ATTR_RW(write_cache);
+
+static umode_t dax_visible(struct kobject *kobj, struct attribute *a, int n)
+{
+ struct device *dev = container_of(kobj, typeof(*dev), kobj);
+ struct dax_device *dax_dev = dax_get_by_host(dev_name(dev));
+
+ WARN_ON_ONCE(!dax_dev);
+ if (!dax_dev)
+ return 0;
+
+ if (a == &dev_attr_write_cache.attr && !dax_dev->ops->flush)
+ return 0;
+ return a->mode;
+}
+
+static struct attribute *dax_attributes[] = {
+ &dev_attr_write_cache.attr,
+ NULL,
+};
+
+struct attribute_group dax_attribute_group = {
+ .name = "dax",
+ .attrs = dax_attributes,
+ .is_visible = dax_visible,
+};
+EXPORT_SYMBOL_GPL(dax_attribute_group);
+
/**
* dax_direct_access() - translate a device pgoff to an absolute pfn
* @dax_dev: a dax_device instance representing the logical memory range
@@ -172,10 +245,43 @@ long dax_direct_access(struct dax_device *dax_dev, pgoff_t pgoff, long nr_pages,
}
EXPORT_SYMBOL_GPL(dax_direct_access);
+size_t dax_copy_from_iter(struct dax_device *dax_dev, pgoff_t pgoff, void *addr,
+ size_t bytes, struct iov_iter *i)
+{
+ if (!dax_alive(dax_dev))
+ return 0;
+
+ return dax_dev->ops->copy_from_iter(dax_dev, pgoff, addr, bytes, i);
+}
+EXPORT_SYMBOL_GPL(dax_copy_from_iter);
+
+void dax_flush(struct dax_device *dax_dev, pgoff_t pgoff, void *addr,
+ size_t size)
+{
+ if (!dax_alive(dax_dev))
+ return;
+
+ if (!test_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags))
+ return;
+
+ if (dax_dev->ops->flush)
+ dax_dev->ops->flush(dax_dev, pgoff, addr, size);
+}
+EXPORT_SYMBOL_GPL(dax_flush);
+
+void dax_write_cache(struct dax_device *dax_dev, bool wc)
+{
+ if (wc)
+ set_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+ else
+ clear_bit(DAXDEV_WRITE_CACHE, &dax_dev->flags);
+}
+EXPORT_SYMBOL_GPL(dax_write_cache);
+
bool dax_alive(struct dax_device *dax_dev)
{
lockdep_assert_held(&dax_srcu);
- return dax_dev->alive;
+ return test_bit(DAXDEV_ALIVE, &dax_dev->flags);
}
EXPORT_SYMBOL_GPL(dax_alive);
@@ -195,7 +301,7 @@ void kill_dax(struct dax_device *dax_dev)
if (!dax_dev)
return;
- dax_dev->alive = false;
+ clear_bit(DAXDEV_ALIVE, &dax_dev->flags);
synchronize_srcu(&dax_srcu);
@@ -239,7 +345,7 @@ static void dax_destroy_inode(struct inode *inode)
{
struct dax_device *dax_dev = to_dax_dev(inode);
- WARN_ONCE(dax_dev->alive,
+ WARN_ONCE(test_bit(DAXDEV_ALIVE, &dax_dev->flags),
"kill_dax() must be called before final iput()\n");
call_rcu(&inode->i_rcu, dax_i_callback);
}
@@ -291,7 +397,7 @@ static struct dax_device *dax_dev_get(dev_t devt)
dax_dev = to_dax_dev(inode);
if (inode->i_state & I_NEW) {
- dax_dev->alive = true;
+ set_bit(DAXDEV_ALIVE, &dax_dev->flags);
inode->i_cdev = &dax_dev->cdev;
inode->i_mode = S_IFCHR;
inode->i_flags = S_DAX;