diff mbox series

[v2,2/2] device-dax: "Hotremove" persistent memory that is used like normal RAM

Message ID 20190421014429.31206-3-pasha.tatashin@soleen.com (mailing list archive)
State Superseded
Headers show
Series "Hotremove" persistent memory | expand

Commit Message

Pasha Tatashin April 21, 2019, 1:44 a.m. UTC
It is now allowed to use persistent memory like a regular RAM, but
currently there is no way to remove this memory until machine is
rebooted.

This work expands the functionality to also allows hotremoving
previously hotplugged persistent memory, and recover the device for use
for other purposes.

To hotremove persistent memory, the management software must first
offline all memory blocks of dax region, and than unbind it from
device-dax/kmem driver. So, operations should look like this:

echo offline > echo offline > /sys/devices/system/memory/memoryN/state
...
echo dax0.0 > /sys/bus/dax/drivers/kmem/unbind

Note: if unbind is done without offlining memory beforehand, it won't be
possible to do dax0.0 hotremove, and dax's memory is going to be part of
System RAM until reboot.

Signed-off-by: Pavel Tatashin <pasha.tatashin@soleen.com>
---
 drivers/dax/dax-private.h |  2 +
 drivers/dax/kmem.c        | 91 +++++++++++++++++++++++++++++++++++++--
 2 files changed, 89 insertions(+), 4 deletions(-)

Comments

David Hildenbrand April 24, 2019, 8:55 p.m. UTC | #1
On 21.04.19 03:44, Pavel Tatashin wrote:
> It is now allowed to use persistent memory like a regular RAM, but
> currently there is no way to remove this memory until machine is
> rebooted.
> 
> This work expands the functionality to also allows hotremoving
> previously hotplugged persistent memory, and recover the device for use
> for other purposes.
> 
> To hotremove persistent memory, the management software must first
> offline all memory blocks of dax region, and than unbind it from
> device-dax/kmem driver. So, operations should look like this:
> 
> echo offline > echo offline > /sys/devices/system/memory/memoryN/state
> ...
> echo dax0.0 > /sys/bus/dax/drivers/kmem/unbind
> 
> Note: if unbind is done without offlining memory beforehand, it won't be
> possible to do dax0.0 hotremove, and dax's memory is going to be part of
> System RAM until reboot.
> 
> Signed-off-by: Pavel Tatashin <pasha.tatashin@soleen.com>
> ---
>  drivers/dax/dax-private.h |  2 +
>  drivers/dax/kmem.c        | 91 +++++++++++++++++++++++++++++++++++++--
>  2 files changed, 89 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/dax/dax-private.h b/drivers/dax/dax-private.h
> index a45612148ca0..999aaf3a29b3 100644
> --- a/drivers/dax/dax-private.h
> +++ b/drivers/dax/dax-private.h
> @@ -53,6 +53,7 @@ struct dax_region {
>   * @pgmap - pgmap for memmap setup / lifetime (driver owned)
>   * @ref: pgmap reference count (driver owned)
>   * @cmp: @ref final put completion (driver owned)
> + * @dax_mem_res: physical address range of hotadded DAX memory
>   */
>  struct dev_dax {
>  	struct dax_region *region;
> @@ -62,6 +63,7 @@ struct dev_dax {
>  	struct dev_pagemap pgmap;
>  	struct percpu_ref ref;
>  	struct completion cmp;
> +	struct resource *dax_kmem_res;
>  };
>  
>  static inline struct dev_dax *to_dev_dax(struct device *dev)
> diff --git a/drivers/dax/kmem.c b/drivers/dax/kmem.c
> index 4c0131857133..d4896b281036 100644
> --- a/drivers/dax/kmem.c
> +++ b/drivers/dax/kmem.c
> @@ -71,21 +71,104 @@ int dev_dax_kmem_probe(struct device *dev)
>  		kfree(new_res);
>  		return rc;
>  	}
> +	dev_dax->dax_kmem_res = new_res;
>  
>  	return 0;
>  }
>  
> +#ifdef CONFIG_MEMORY_HOTREMOVE
> +/*
> + * Check that device-dax's memory_blocks are offline. If a memory_block is not
> + * offline a warning is printed and an error is returned. dax hotremove can
> + * succeed only when every memory_block is offlined beforehand.
> + */
> +static int
> +offline_memblock_cb(struct memory_block *mem, void *arg)

Function name suggests that you are actually trying to offline memory
here. Maybe check_memblocks_offline_cb(), just like we have in
mm/memory_hotplug.c.

> +{
> +	struct device *mem_dev = &mem->dev;
> +	bool is_offline;
> +
> +	device_lock(mem_dev);
> +	is_offline = mem_dev->offline;
> +	device_unlock(mem_dev);
> +
> +	if (!is_offline) {
> +		struct device *dev = (struct device *)arg;
> +		unsigned long spfn = section_nr_to_pfn(mem->start_section_nr);
> +		unsigned long epfn = section_nr_to_pfn(mem->end_section_nr);
> +		phys_addr_t spa = spfn << PAGE_SHIFT;
> +		phys_addr_t epa = epfn << PAGE_SHIFT;
> +
> +		dev_warn(dev, "memory block [%pa-%pa] is not offline\n",
> +			 &spa, &epa);
> +
> +		return -EBUSY;
> +	}
> +
> +	return 0;
> +}
> +
> +static int dev_dax_kmem_remove(struct device *dev)
> +{
> +	struct dev_dax *dev_dax = to_dev_dax(dev);
> +	struct resource *res = dev_dax->dax_kmem_res;
> +	resource_size_t kmem_start;
> +	resource_size_t kmem_size;
> +	unsigned long start_pfn;
> +	unsigned long end_pfn;
> +	int rc;
> +
> +	/*
> +	 * dax kmem resource does not exist, means memory was never hotplugged.
> +	 * So, nothing to do here.
> +	 */
> +	if (!res)
> +		return 0;
> +
> +	kmem_start = res->start;
> +	kmem_size = resource_size(res);
> +	start_pfn = kmem_start >> PAGE_SHIFT;
> +	end_pfn = start_pfn + (kmem_size >> PAGE_SHIFT) - 1;
> +
> +	/*
> +	 * Walk and check that every singe memory_block of dax region is
> +	 * offline
> +	 */
> +	lock_device_hotplug();
> +	rc = walk_memory_range(start_pfn, end_pfn, dev, offline_memblock_cb);
> +	unlock_device_hotplug();
> +
> +	/*
> +	 * If admin has not offlined memory beforehand, we cannot hotremove dax.
> +	 * Unfortunately, because unbind will still succeed there is no way for
> +	 * user to hotremove dax after this.
> +	 */
> +	if (rc)
> +		return rc;

Can't it happen that there is a race between you checking if memory is
offline and an admin onlining memory again? maybe pull the
remove_memory() into the locked region, using __remove_memory() instead.

> +
> +	/* Hotremove memory, cannot fail because memory is already offlined */
> +	remove_memory(dev_dax->target_node, kmem_start, kmem_size);
> +
> +	/* Release and free dax resources */
> +	release_resource(res);
> +	kfree(res);
> +	dev_dax->dax_kmem_res = NULL;
> +
> +	return 0;
> +}
> +#else
>  static int dev_dax_kmem_remove(struct device *dev)
>  {
>  	/*
> -	 * Purposely leak the request_mem_region() for the device-dax
> -	 * range and return '0' to ->remove() attempts. The removal of
> -	 * the device from the driver always succeeds, but the region
> -	 * is permanently pinned as reserved by the unreleased
> +	 * Without hotremove purposely leak the request_mem_region() for the
> +	 * device-dax range and return '0' to ->remove() attempts. The removal
> +	 * of the device from the driver always succeeds, but the region is
> +	 * permanently pinned as reserved by the unreleased
>  	 * request_mem_region().
>  	 */
>  	return 0;
>  }
> +#endif /* CONFIG_MEMORY_HOTREMOVE */
>  
>  static struct dax_device_driver device_dax_kmem_driver = {
>  	.drv = {
>
Dan Williams April 24, 2019, 9:02 p.m. UTC | #2
On Wed, Apr 24, 2019 at 1:55 PM David Hildenbrand <david@redhat.com> wrote:
>
> On 21.04.19 03:44, Pavel Tatashin wrote:
> > It is now allowed to use persistent memory like a regular RAM, but
> > currently there is no way to remove this memory until machine is
> > rebooted.
> >
> > This work expands the functionality to also allows hotremoving
> > previously hotplugged persistent memory, and recover the device for use
> > for other purposes.
> >
> > To hotremove persistent memory, the management software must first
> > offline all memory blocks of dax region, and than unbind it from
> > device-dax/kmem driver. So, operations should look like this:
> >
> > echo offline > echo offline > /sys/devices/system/memory/memoryN/state
> > ...
> > echo dax0.0 > /sys/bus/dax/drivers/kmem/unbind
> >
> > Note: if unbind is done without offlining memory beforehand, it won't be
> > possible to do dax0.0 hotremove, and dax's memory is going to be part of
> > System RAM until reboot.
> >
> > Signed-off-by: Pavel Tatashin <pasha.tatashin@soleen.com>
> > ---
> >  drivers/dax/dax-private.h |  2 +
> >  drivers/dax/kmem.c        | 91 +++++++++++++++++++++++++++++++++++++--
> >  2 files changed, 89 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/dax/dax-private.h b/drivers/dax/dax-private.h
> > index a45612148ca0..999aaf3a29b3 100644
> > --- a/drivers/dax/dax-private.h
> > +++ b/drivers/dax/dax-private.h
> > @@ -53,6 +53,7 @@ struct dax_region {
> >   * @pgmap - pgmap for memmap setup / lifetime (driver owned)
> >   * @ref: pgmap reference count (driver owned)
> >   * @cmp: @ref final put completion (driver owned)
> > + * @dax_mem_res: physical address range of hotadded DAX memory
> >   */
> >  struct dev_dax {
> >       struct dax_region *region;
> > @@ -62,6 +63,7 @@ struct dev_dax {
> >       struct dev_pagemap pgmap;
> >       struct percpu_ref ref;
> >       struct completion cmp;
> > +     struct resource *dax_kmem_res;
> >  };
> >
> >  static inline struct dev_dax *to_dev_dax(struct device *dev)
> > diff --git a/drivers/dax/kmem.c b/drivers/dax/kmem.c
> > index 4c0131857133..d4896b281036 100644
> > --- a/drivers/dax/kmem.c
> > +++ b/drivers/dax/kmem.c
> > @@ -71,21 +71,104 @@ int dev_dax_kmem_probe(struct device *dev)
> >               kfree(new_res);
> >               return rc;
> >       }
> > +     dev_dax->dax_kmem_res = new_res;
> >
> >       return 0;
> >  }
> >
> > +#ifdef CONFIG_MEMORY_HOTREMOVE
> > +/*
> > + * Check that device-dax's memory_blocks are offline. If a memory_block is not
> > + * offline a warning is printed and an error is returned. dax hotremove can
> > + * succeed only when every memory_block is offlined beforehand.
> > + */
> > +static int
> > +offline_memblock_cb(struct memory_block *mem, void *arg)
>
> Function name suggests that you are actually trying to offline memory
> here. Maybe check_memblocks_offline_cb(), just like we have in
> mm/memory_hotplug.c.
>
> > +{
> > +     struct device *mem_dev = &mem->dev;
> > +     bool is_offline;
> > +
> > +     device_lock(mem_dev);
> > +     is_offline = mem_dev->offline;
> > +     device_unlock(mem_dev);
> > +
> > +     if (!is_offline) {
> > +             struct device *dev = (struct device *)arg;
> > +             unsigned long spfn = section_nr_to_pfn(mem->start_section_nr);
> > +             unsigned long epfn = section_nr_to_pfn(mem->end_section_nr);
> > +             phys_addr_t spa = spfn << PAGE_SHIFT;
> > +             phys_addr_t epa = epfn << PAGE_SHIFT;
> > +
> > +             dev_warn(dev, "memory block [%pa-%pa] is not offline\n",
> > +                      &spa, &epa);
> > +
> > +             return -EBUSY;
> > +     }
> > +
> > +     return 0;
> > +}
> > +
> > +static int dev_dax_kmem_remove(struct device *dev)
> > +{
> > +     struct dev_dax *dev_dax = to_dev_dax(dev);
> > +     struct resource *res = dev_dax->dax_kmem_res;
> > +     resource_size_t kmem_start;
> > +     resource_size_t kmem_size;
> > +     unsigned long start_pfn;
> > +     unsigned long end_pfn;
> > +     int rc;
> > +
> > +     /*
> > +      * dax kmem resource does not exist, means memory was never hotplugged.
> > +      * So, nothing to do here.
> > +      */
> > +     if (!res)
> > +             return 0;
> > +
> > +     kmem_start = res->start;
> > +     kmem_size = resource_size(res);
> > +     start_pfn = kmem_start >> PAGE_SHIFT;
> > +     end_pfn = start_pfn + (kmem_size >> PAGE_SHIFT) - 1;
> > +
> > +     /*
> > +      * Walk and check that every singe memory_block of dax region is
> > +      * offline
> > +      */
> > +     lock_device_hotplug();
> > +     rc = walk_memory_range(start_pfn, end_pfn, dev, offline_memblock_cb);
> > +     unlock_device_hotplug();
> > +
> > +     /*
> > +      * If admin has not offlined memory beforehand, we cannot hotremove dax.
> > +      * Unfortunately, because unbind will still succeed there is no way for
> > +      * user to hotremove dax after this.
> > +      */
> > +     if (rc)
> > +             return rc;
>
> Can't it happen that there is a race between you checking if memory is
> offline and an admin onlining memory again? maybe pull the
> remove_memory() into the locked region, using __remove_memory() instead.

I think the race is ok. The admin gets to keep the pieces of allowing
racing updates to the state and the kernel will keep the range active
until the next reboot.
Pasha Tatashin April 24, 2019, 9:34 p.m. UTC | #3
> > > +static int
> > > +offline_memblock_cb(struct memory_block *mem, void *arg)
> >
> > Function name suggests that you are actually trying to offline memory
> > here. Maybe check_memblocks_offline_cb(), just like we have in
> > mm/memory_hotplug.c.

Makes sense, I will rename to check_memblocks_offline_cb()

> > > +     lock_device_hotplug();
> > > +     rc = walk_memory_range(start_pfn, end_pfn, dev, offline_memblock_cb);
> > > +     unlock_device_hotplug();
> > > +
> > > +     /*
> > > +      * If admin has not offlined memory beforehand, we cannot hotremove dax.
> > > +      * Unfortunately, because unbind will still succeed there is no way for
> > > +      * user to hotremove dax after this.
> > > +      */
> > > +     if (rc)
> > > +             return rc;
> >
> > Can't it happen that there is a race between you checking if memory is
> > offline and an admin onlining memory again? maybe pull the
> > remove_memory() into the locked region, using __remove_memory() instead.
>
> I think the race is ok. The admin gets to keep the pieces of allowing
> racing updates to the state and the kernel will keep the range active
> until the next reboot.

Thank you for noticing this. I will pull it into locking region.
Because, __remove_memory() has this code:

1868   ret = walk_memory_range(PFN_DOWN(start), PFN_UP(start + size - 1), NULL,
1869   check_memblock_offlined_cb);
1870   if (ret)
1871      BUG();

Basically, panic if at least something is not offlined.

Pasha
David Hildenbrand April 25, 2019, 7:41 a.m. UTC | #4
On 24.04.19 23:34, Pavel Tatashin wrote:
>>>> +static int
>>>> +offline_memblock_cb(struct memory_block *mem, void *arg)
>>>
>>> Function name suggests that you are actually trying to offline memory
>>> here. Maybe check_memblocks_offline_cb(), just like we have in
>>> mm/memory_hotplug.c.
> 
> Makes sense, I will rename to check_memblocks_offline_cb()
> 
>>>> +     lock_device_hotplug();
>>>> +     rc = walk_memory_range(start_pfn, end_pfn, dev, offline_memblock_cb);
>>>> +     unlock_device_hotplug();
>>>> +
>>>> +     /*
>>>> +      * If admin has not offlined memory beforehand, we cannot hotremove dax.
>>>> +      * Unfortunately, because unbind will still succeed there is no way for
>>>> +      * user to hotremove dax after this.
>>>> +      */
>>>> +     if (rc)
>>>> +             return rc;
>>>
>>> Can't it happen that there is a race between you checking if memory is
>>> offline and an admin onlining memory again? maybe pull the
>>> remove_memory() into the locked region, using __remove_memory() instead.
>>
>> I think the race is ok. The admin gets to keep the pieces of allowing
>> racing updates to the state and the kernel will keep the range active
>> until the next reboot.
> 
> Thank you for noticing this. I will pull it into locking region.
> Because, __remove_memory() has this code:
> 
> 1868   ret = walk_memory_range(PFN_DOWN(start), PFN_UP(start + size - 1), NULL,
> 1869   check_memblock_offlined_cb);
> 1870   if (ret)
> 1871      BUG();
> 

Yes, also I think you can let go of the device_lock in
check_memblocks_offline_cb, lock_device_hotplug() should take care of
this (see Documentation/core-api/memory-hotplug.rst - "locking internals")

Cheers!
Pasha Tatashin April 25, 2019, 12:30 p.m. UTC | #5
>
> Yes, also I think you can let go of the device_lock in
> check_memblocks_offline_cb, lock_device_hotplug() should take care of
> this (see Documentation/core-api/memory-hotplug.rst - "locking internals")
>
Hi David,

Thank you for your comments. I went through memory-hotplug.rst, and I
still think that device_lock() is needed here. In this particular case
it can be replaced with something like READ_ONCE(), but for simplicity
it is better to have device_lock()/device_unlock() as this is not a
performance critical code.

I do not see any lock ordering issues with this code, as we are
holding lock_device_hotplug() first that prevents userland from
adding/removing memory during this check.

https://soleen.com/source/xref/linux/arch/powerpc/platforms/powernv/memtrace.c?r=98fa15f3#248

Here we have a similar code:
lock_device_hotplug();
   online_mem_block();
    device_online()
     device_lock(dev);

Pasha
David Hildenbrand April 25, 2019, 12:38 p.m. UTC | #6
On 25.04.19 14:30, Pavel Tatashin wrote:
>>
>> Yes, also I think you can let go of the device_lock in
>> check_memblocks_offline_cb, lock_device_hotplug() should take care of
>> this (see Documentation/core-api/memory-hotplug.rst - "locking internals")
>>
> Hi David,
> 
> Thank you for your comments. I went through memory-hotplug.rst, and I
> still think that device_lock() is needed here. In this particular case
> it can be replaced with something like READ_ONCE(), but for simplicity
> it is better to have device_lock()/device_unlock() as this is not a
> performance critical code.
> 
> I do not see any lock ordering issues with this code, as we are
> holding lock_device_hotplug() first that prevents userland from
> adding/removing memory during this check.

Yes, lock ordering is not an issue, I rather think that the device
hotplug lock will guard us in all situations. E.g. remove_memory() also
does not use it when checking if all blocks are offline. But you can
leave it in if you think it is needed.

> 
> https://soleen.com/source/xref/linux/arch/powerpc/platforms/powernv/memtrace.c?r=98fa15f3#248
> 
> Here we have a similar code:
> lock_device_hotplug();
>    online_mem_block();
>     device_online()
>      device_lock(dev);
> 
> Pasha
>
diff mbox series

Patch

diff --git a/drivers/dax/dax-private.h b/drivers/dax/dax-private.h
index a45612148ca0..999aaf3a29b3 100644
--- a/drivers/dax/dax-private.h
+++ b/drivers/dax/dax-private.h
@@ -53,6 +53,7 @@  struct dax_region {
  * @pgmap - pgmap for memmap setup / lifetime (driver owned)
  * @ref: pgmap reference count (driver owned)
  * @cmp: @ref final put completion (driver owned)
+ * @dax_mem_res: physical address range of hotadded DAX memory
  */
 struct dev_dax {
 	struct dax_region *region;
@@ -62,6 +63,7 @@  struct dev_dax {
 	struct dev_pagemap pgmap;
 	struct percpu_ref ref;
 	struct completion cmp;
+	struct resource *dax_kmem_res;
 };
 
 static inline struct dev_dax *to_dev_dax(struct device *dev)
diff --git a/drivers/dax/kmem.c b/drivers/dax/kmem.c
index 4c0131857133..d4896b281036 100644
--- a/drivers/dax/kmem.c
+++ b/drivers/dax/kmem.c
@@ -71,21 +71,104 @@  int dev_dax_kmem_probe(struct device *dev)
 		kfree(new_res);
 		return rc;
 	}
+	dev_dax->dax_kmem_res = new_res;
 
 	return 0;
 }
 
+#ifdef CONFIG_MEMORY_HOTREMOVE
+/*
+ * Check that device-dax's memory_blocks are offline. If a memory_block is not
+ * offline a warning is printed and an error is returned. dax hotremove can
+ * succeed only when every memory_block is offlined beforehand.
+ */
+static int
+offline_memblock_cb(struct memory_block *mem, void *arg)
+{
+	struct device *mem_dev = &mem->dev;
+	bool is_offline;
+
+	device_lock(mem_dev);
+	is_offline = mem_dev->offline;
+	device_unlock(mem_dev);
+
+	if (!is_offline) {
+		struct device *dev = (struct device *)arg;
+		unsigned long spfn = section_nr_to_pfn(mem->start_section_nr);
+		unsigned long epfn = section_nr_to_pfn(mem->end_section_nr);
+		phys_addr_t spa = spfn << PAGE_SHIFT;
+		phys_addr_t epa = epfn << PAGE_SHIFT;
+
+		dev_warn(dev, "memory block [%pa-%pa] is not offline\n",
+			 &spa, &epa);
+
+		return -EBUSY;
+	}
+
+	return 0;
+}
+
+static int dev_dax_kmem_remove(struct device *dev)
+{
+	struct dev_dax *dev_dax = to_dev_dax(dev);
+	struct resource *res = dev_dax->dax_kmem_res;
+	resource_size_t kmem_start;
+	resource_size_t kmem_size;
+	unsigned long start_pfn;
+	unsigned long end_pfn;
+	int rc;
+
+	/*
+	 * dax kmem resource does not exist, means memory was never hotplugged.
+	 * So, nothing to do here.
+	 */
+	if (!res)
+		return 0;
+
+	kmem_start = res->start;
+	kmem_size = resource_size(res);
+	start_pfn = kmem_start >> PAGE_SHIFT;
+	end_pfn = start_pfn + (kmem_size >> PAGE_SHIFT) - 1;
+
+	/*
+	 * Walk and check that every singe memory_block of dax region is
+	 * offline
+	 */
+	lock_device_hotplug();
+	rc = walk_memory_range(start_pfn, end_pfn, dev, offline_memblock_cb);
+	unlock_device_hotplug();
+
+	/*
+	 * If admin has not offlined memory beforehand, we cannot hotremove dax.
+	 * Unfortunately, because unbind will still succeed there is no way for
+	 * user to hotremove dax after this.
+	 */
+	if (rc)
+		return rc;
+
+	/* Hotremove memory, cannot fail because memory is already offlined */
+	remove_memory(dev_dax->target_node, kmem_start, kmem_size);
+
+	/* Release and free dax resources */
+	release_resource(res);
+	kfree(res);
+	dev_dax->dax_kmem_res = NULL;
+
+	return 0;
+}
+#else
 static int dev_dax_kmem_remove(struct device *dev)
 {
 	/*
-	 * Purposely leak the request_mem_region() for the device-dax
-	 * range and return '0' to ->remove() attempts. The removal of
-	 * the device from the driver always succeeds, but the region
-	 * is permanently pinned as reserved by the unreleased
+	 * Without hotremove purposely leak the request_mem_region() for the
+	 * device-dax range and return '0' to ->remove() attempts. The removal
+	 * of the device from the driver always succeeds, but the region is
+	 * permanently pinned as reserved by the unreleased
 	 * request_mem_region().
 	 */
 	return 0;
 }
+#endif /* CONFIG_MEMORY_HOTREMOVE */
 
 static struct dax_device_driver device_dax_kmem_driver = {
 	.drv = {