Re: [PATCH v4 12/22] iommu: Add I/O ASID allocator

[Jacob Pan]

Hi Jonathan,
Thanks for the review, comments inline below. I saw Jean already took
in changes based on your review in his sva/api tree. This is just some
additions.

On Tue, 18 Jun 2019 17:50:59 +0100
Jonathan Cameron <jonathan.cameron@xxxxxxxxxx> wrote:

> On Sun, 9 Jun 2019 06:44:12 -0700
> Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx> wrote:
> 
> > From: Jean-Philippe Brucker <jean-philippe.brucker@xxxxxxx>
> > 
> > Some devices might support multiple DMA address spaces, in
> > particular those that have the PCI PASID feature. PASID (Process
> > Address Space ID) allows to share process address spaces with
> > devices (SVA), partition a device into VM-assignable entities (VFIO
> > mdev) or simply provide multiple DMA address space to kernel
> > drivers. Add a global PASID allocator usable by different drivers
> > at the same time. Name it I/O ASID to avoid confusion with ASIDs
> > allocated by arch code, which are usually a separate ID space.
> > 
> > The IOASID space is global. Each device can have its own PASID
> > space, but by convention the IOMMU ended up having a global PASID
> > space, so that with SVA, each mm_struct is associated to a single
> > PASID.
> > 
> > The allocator is primarily used by IOMMU subsystem but in rare
> > occasions drivers would like to allocate PASIDs for devices that
> > aren't managed by an IOMMU, using the same ID space as IOMMU.
> > 
> > There are two types of allocators:
> > 1. default allocator - Always available, uses an XArray to track
> > 2. custom allocators - Can be registered at runtime, take precedence
> > over the default allocator.
> > 
> > Custom allocators have these attributes:
> > - provides platform specific alloc()/free() functions with private
> > data.
> > - allocation results lookup are not provided by the allocator,
> > lookup request must be done by the IOASID framework by its own
> > XArray.
> > - allocators can be unregistered at runtime, either fallback to the
> > next custom allocator or to the default allocator.  
> 
> What is the usecase for having a 'stack' of custom allocators?
> 
mainly for hotplug. If a vIOMMU provides custom allocation service to a
guest and got hot removed we can use other custom allocators from the
stack.
> > - custom allocators can share the same set of alloc()/free()
> > helpers, in this case they also share the same IOASID space, thus
> > the same XArray.
> > - switching between allocators requires all outstanding IOASIDs to
> > be freed unless the two allocators share the same alloc()/free()
> > helpers.  
> In general this approach has a lot of features where the
> justification is missing from this particular patch.  It may be
> useful to add some more background to this intro?
> 
Good point. Here are the use cases as complexity grow.
1. native (host kernel) PASID/IOASID allocation, this will only use
default allocator, users are IOMMU driver or any device drivers.
2. guest with a single vIOMMU. vIOMMU driver will register a custom
allocator which eventually calls into the host to allocate system-wide
PASID.
3. guest with two or more vIOMMUs but share the same backend to
allocate system-wide PASID. IOASID code detects the sharing based on
alloc/free ops.
4. guest with two or more vIOMMUs each allocate guest-wide PASIDs.
5. In case of hot plug of vIOMMU, the active custom allocator is
chosen from a list established during registration.

> > 
> > Signed-off-by: Jean-Philippe Brucker <jean-philippe.brucker@xxxxxxx>
> > Signed-off-by: Jacob Pan <jacob.jun.pan@xxxxxxxxxxxxxxx>
> > Link: https://lkml.org/lkml/2019/4/26/462  
> 
> Various comments inline.  Given the several cups of coffee this took
> to review I may well have misunderstood everything ;)
> 
> Jonathan
> > ---
> >  drivers/iommu/Kconfig  |   8 +
> >  drivers/iommu/Makefile |   1 +
> >  drivers/iommu/ioasid.c | 427
> > +++++++++++++++++++++++++++++++++++++++++++++++++
> > include/linux/ioasid.h |  74 +++++++++ 4 files changed, 510
> > insertions(+) create mode 100644 drivers/iommu/ioasid.c
> >  create mode 100644 include/linux/ioasid.h
> > 
> > diff --git a/drivers/iommu/Kconfig b/drivers/iommu/Kconfig
> > index 83664db..c40c4b5 100644
> > --- a/drivers/iommu/Kconfig
> > +++ b/drivers/iommu/Kconfig
> > @@ -3,6 +3,13 @@
> >  config IOMMU_IOVA
> >  	tristate
> >  
> > +# The IOASID allocator may also be used by non-IOMMU_API users
> > +config IOASID
> > +	tristate
> > +	help
> > +	  Enable the I/O Address Space ID allocator. A single ID
> > space shared
> > +	  between different users.
> > +
> >  # IOMMU_API always gets selected by whoever wants it.
> >  config IOMMU_API
> >  	bool
> > @@ -207,6 +214,7 @@ config INTEL_IOMMU_SVM
> >  	depends on INTEL_IOMMU && X86
> >  	select PCI_PASID
> >  	select MMU_NOTIFIER
> > +	select IOASID
> >  	help
> >  	  Shared Virtual Memory (SVM) provides a facility for
> > devices to access DMA resources through process address space by
> > diff --git a/drivers/iommu/Makefile b/drivers/iommu/Makefile
> > index 8c71a15..0efac6f 100644
> > --- a/drivers/iommu/Makefile
> > +++ b/drivers/iommu/Makefile
> > @@ -7,6 +7,7 @@ obj-$(CONFIG_IOMMU_DMA) += dma-iommu.o
> >  obj-$(CONFIG_IOMMU_IO_PGTABLE) += io-pgtable.o
> >  obj-$(CONFIG_IOMMU_IO_PGTABLE_ARMV7S) += io-pgtable-arm-v7s.o
> >  obj-$(CONFIG_IOMMU_IO_PGTABLE_LPAE) += io-pgtable-arm.o
> > +obj-$(CONFIG_IOASID) += ioasid.o
> >  obj-$(CONFIG_IOMMU_IOVA) += iova.o
> >  obj-$(CONFIG_OF_IOMMU)	+= of_iommu.o
> >  obj-$(CONFIG_MSM_IOMMU) += msm_iommu.o
> > diff --git a/drivers/iommu/ioasid.c b/drivers/iommu/ioasid.c
> > new file mode 100644
> > index 0000000..0919b70
> > --- /dev/null
> > +++ b/drivers/iommu/ioasid.c
> > @@ -0,0 +1,427 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * I/O Address Space ID allocator. There is one global IOASID
> > space, split into
> > + * subsets. Users create a subset with DECLARE_IOASID_SET, then
> > allocate and
> > + * free IOASIDs with ioasid_alloc and ioasid_free.
> > + */
> > +#define pr_fmt(fmt)	KBUILD_MODNAME ": " fmt
> > +
> > +#include <linux/xarray.h>
> > +#include <linux/ioasid.h>
> > +#include <linux/slab.h>
> > +#include <linux/module.h>
> > +#include <linux/spinlock.h>
> > +
> > +struct ioasid_data {
> > +	ioasid_t id;
> > +	struct ioasid_set *set;
> > +	void *private;
> > +	struct rcu_head rcu;
> > +};
> > +
> > +/*
> > + * struct ioasid_allocator_data - Internal data structure to hold
> > information
> > + * about an allocator. There are two types of allocators:
> > + *
> > + * - Default allocator always has its own XArray to track the
> > IOASIDs allocated.
> > + * - Custom allocators may share allocation helpers with different
> > private data.
> > + *   Custom allocators share the same helper functions also share
> > the same
> > + *   XArray.
> > + * Rules:
> > + * 1. Default allocator is always available, not dynamically
> > registered. This is
> > + *    to prevent race conditions with early boot code that want to
> > register
> > + *    custom allocators or allocate IOASIDs.
> > + * 2. Custom allocators take precedence over the default allocator.
> > + * 3. When all custom allocators sharing the same helper functions
> > are
> > + *    unregistered (e.g. due to hotplug), all outstanding IOASIDs
> > must be
> > + *    freed.
> > + * 4. When switching between custom allocators sharing the same
> > helper
> > + *    functions, outstanding IOASIDs are preserved.
> > + * 5. When switching between custom allocator and default
> > allocator, all IOASIDs
> > + *    must be freed to ensure unadulterated space for the new
> > allocator.
> > + *
> > + * @ops:	allocator helper functions and its data
> > + * @list:	registered custom allocators
> > + * @slist:	allocators share the same ops but different data
> > + * @flags:	attributes of the allocator
> > + * @users	number of allocators sharing the same ops and
> > XArray
> > + * @xa		xarray holds the IOASID space  
> Doc ordering should match the struct.
> 
Fixed in Jean's tree.
> > + */
> > +struct ioasid_allocator_data {
> > +	struct ioasid_allocator_ops *ops;
> > +	struct list_head list;
> > +	struct list_head slist;
> > +#define IOASID_ALLOCATOR_CUSTOM BIT(0) /* Needs framework to track
> > results */
> > +	unsigned long flags;
> > +	struct xarray xa;
> > +	refcount_t users;
> > +};
> > +
> > +static DEFINE_MUTEX(ioasid_allocator_lock);
> > +static LIST_HEAD(allocators_list);
> > +static ioasid_t default_alloc(ioasid_t min, ioasid_t max, void
> > *opaque); +static void default_free(ioasid_t ioasid, void *opaque);
> > +
> > +static struct ioasid_allocator_ops default_ops = {
> > +	.alloc = default_alloc,
> > +	.free = default_free  
> 
> Pure churn prevention but doesn't seem implausible that we'll end up
> with more ops at somepoint, so add the commas on last elements?
> 
Sounds good, avoid a new line. Fixed in Jean's tree.
> > +};
> > +
> > +static struct ioasid_allocator_data default_allocator = {
> > +	.ops = &default_ops,
> > +	.flags = 0,
> > +	.xa = XARRAY_INIT(ioasid_xa, XA_FLAGS_ALLOC)
> > +};
> > +
> > +static struct ioasid_allocator_data *active_allocator =
> > &default_allocator; +
> > +static ioasid_t default_alloc(ioasid_t min, ioasid_t max, void
> > *opaque) +{  
> Why not reorder to avoid having the forward defs above?  It's only
> moving things a few lines in this case so I'm not seeing a strong
> readability argument.
> 
There is a circular dependency. default_alloc() depends on
default_allocator which in turn depends on default_ops.
> > +	ioasid_t id;
> > +
> > +	if (xa_alloc(&default_allocator.xa, &id, opaque,
> > XA_LIMIT(min, max), GFP_KERNEL)) {
> > +		pr_err("Failed to alloc ioasid from %d to %d\n",
> > min, max);  
> 
> It seems xa_alloc can return a few different error values. Perhaps
> worth printing out what we got as might help in debugging?
> 
yes, will do.
> > +		return INVALID_IOASID;
> > +	}
> > +
> > +	return id;
> > +}
> > +
> > +void default_free(ioasid_t ioasid, void *opaque)
> > +{
> > +	struct ioasid_data *ioasid_data;
> > +
> > +	ioasid_data = xa_erase(&default_allocator.xa, ioasid);
> > +	kfree_rcu(ioasid_data, rcu);
> > +}
> > +
> > +/* Allocate and initialize a new custom allocator with its helper
> > functions */ +static inline struct ioasid_allocator_data
> > *ioasid_alloc_allocator(struct ioasid_allocator_ops *ops)  
> 
> Why inline?
> 
Good catch, fixed by Jean in his tree.
> > +{
> > +	struct ioasid_allocator_data *ia_data;
> > +
> > +	ia_data = kzalloc(sizeof(*ia_data), GFP_KERNEL);
> > +	if (!ia_data)
> > +		return NULL;
> > +
> > +	xa_init_flags(&ia_data->xa, XA_FLAGS_ALLOC);
> > +	INIT_LIST_HEAD(&ia_data->slist);
> > +	ia_data->flags |= IOASID_ALLOCATOR_CUSTOM;
> > +	ia_data->ops = ops;
> > +
> > +	/* For tracking custom allocators that share the same ops
> > */
> > +	list_add_tail(&ops->list, &ia_data->slist);
> > +	refcount_set(&ia_data->users, 1);
> > +
> > +	return ia_data;
> > +}
> > +
> > +static inline bool use_same_ops(struct ioasid_allocator_ops *a,
> > struct ioasid_allocator_ops *b) +{
> > +	return (a->free == b->free) && (a->alloc == b->alloc);
> > +}
> > +
> > +/**
> > + * ioasid_register_allocator - register a custom allocator
> > + * @ops: the custom allocator ops to be registered
> > + *
> > + * Custom allocators take precedence over the default xarray based
> > allocator.
> > + * Private data associated with the IOASID allocated by the custom
> > allocators
> > + * are managed by IOASID framework similar to data stored in xa by
> > default
> > + * allocator.
> > + *
> > + * There can be multiple allocators registered but only one is
> > active. In case
> > + * of runtime removal of a custom allocator, the next one is
> > activated based
> > + * on the registration ordering.
> > + *
> > + * Multiple allocators can share the same alloc() function, in
> > this case the
> > + * IOASID space is shared.
> > + *  
> Nitpick: This extra blank line seems inconsistent.
> 
fixed by Jean in his tree.
> > + */
> > +int ioasid_register_allocator(struct ioasid_allocator_ops *ops)
> > +{
> > +	struct ioasid_allocator_data *ia_data;
> > +	struct ioasid_allocator_data *pallocator;
> > +	int ret = 0;
> > +
> > +	mutex_lock(&ioasid_allocator_lock);
> > +
> > +	ia_data = ioasid_alloc_allocator(ops);
> > +	if (!ia_data) {
> > +		ret = -ENOMEM;
> > +		goto out_unlock;
> > +	}
> > +
> > +	/*
> > +	 * No particular preference, we activate the first one and
> > keep
> > +	 * the later registered allocators in a list in case the
> > first one gets
> > +	 * removed due to hotplug.
> > +	 */
> > +	if (list_empty(&allocators_list)) {
> > +		WARN_ON(active_allocator != &default_allocator);
> > +		/* Use this new allocator if default is not active
> > */
> > +		if (xa_empty(&active_allocator->xa)) {
> > +			active_allocator = ia_data;
> > +			list_add_tail(&ia_data->list,
> > &allocators_list);
> > +			goto out_unlock;
> > +		}
> > +		pr_warn("Default allocator active with outstanding
> > IOASID\n");
> > +		ret = -EAGAIN;
> > +		goto out_free;
> > +	}
> > +
> > +	/* Check if the allocator is already registered */
> > +	list_for_each_entry(pallocator, &allocators_list, list) {
> > +		if (pallocator->ops == ops) {
> > +			pr_err("IOASID allocator already
> > registered\n");
> > +			ret = -EEXIST;
> > +			goto out_free;
> > +		} else if (use_same_ops(pallocator->ops, ops)) {
> > +			/*
> > +			 * If the new allocator shares the same
> > ops,
> > +			 * then they will share the same IOASID
> > space.
> > +			 * We should put them under the same
> > xarray.
> > +			 */
> > +			list_add_tail(&ops->list,
> > &pallocator->slist);
> > +			refcount_inc(&pallocator->users);
> > +			pr_info("New IOASID allocator ops shared
> > %u times\n",
> > +				refcount_read(&pallocator->users));
> > +			goto out_free;
> > +		}
> > +	}
> > +	list_add_tail(&ia_data->list, &allocators_list);
> > +
> > +	goto out_unlock;  
> I commented on this in Jean-Phillipe's set as well I think.
> Cleaner to just have a separate unlock copy here and return.
> 
fixed by Jean in his tree.
> > +
> > +out_free:
> > +	kfree(ia_data);
> > +out_unlock:
> > +	mutex_unlock(&ioasid_allocator_lock);
> > +	return ret;
> > +}
> > +EXPORT_SYMBOL_GPL(ioasid_register_allocator);
> > +
> > +static inline bool has_shared_ops(struct ioasid_allocator_data
> > *allocator) +{
> > +	return refcount_read(&allocator->users) > 1;
> > +}
> > +
> > +/**
> > + * ioasid_unregister_allocator - Remove a custom IOASID allocator
> > ops
> > + * @ops: the custom allocator to be removed
> > + *
> > + * Remove an allocator from the list, activate the next allocator
> > in
> > + * the order it was registered. Or revert to default allocator if
> > all
> > + * custom allocators are unregistered without outstanding IOASIDs.
> > + */
> > +void ioasid_unregister_allocator(struct ioasid_allocator_ops *ops)
> > +{
> > +	struct ioasid_allocator_data *pallocator;
> > +	struct ioasid_allocator_ops *sops;
> > +
> > +	mutex_lock(&ioasid_allocator_lock);
> > +	if (list_empty(&allocators_list)) {
> > +		pr_warn("No custom IOASID allocators active!\n");
> > +		goto exit_unlock;
> > +	}
> > +
> > +	list_for_each_entry(pallocator, &allocators_list, list) {
> > +		if (use_same_ops(pallocator->ops, ops)) {  
> Could reduce the depth with.
> 
> 		if (!use_same_ops(pallocator->ops, ops))
> 			continue;
> 
> Trade off between having the unusual flow that will resort and the
> code going off the side of your screen (if you use a very small
> screen ;)
> 
fixed by Jean in his tree.
> > +			if (refcount_read(&pallocator->users) ==
> > 1) {
> > +				/* No shared helper functions */
> > +				list_del(&pallocator->list);
> > +				/*
> > +				 * All IOASIDs should have been
> > freed before
> > +				 * the last allocator that shares
> > the same ops
> > +				 * is unregistered.
> > +				 */
> > +
> > WARN_ON(!xa_empty(&pallocator->xa));
> > +				kfree(pallocator);
> > +				if (list_empty(&allocators_list)) {
> > +					pr_info("No custom IOASID
> > allocators, switch to default.\n");
> > +					active_allocator =
> > &default_allocator;
> > +				} else if (pallocator ==
> > active_allocator) {
> > +					active_allocator =
> > list_entry(&allocators_list, struct ioasid_allocator_data, list);  
> 
> What is the intent here?  You seem to be calling the list entry
> dereference on the list head itself.  That will get you something
> fairly random on the stack I think.
> 
> list_first_entry maybe?
> 
The intent here is to pick a new active_allocator if the one being
unregistered happens to be the active allocator. It should be
list_first_entry. Thanks.
> 
> > +					pr_info("IOASID allocator
> > changed");
> > +				}
> > +				break;
> > +			}
> > +			/*
> > +			 * Find the matching shared ops to delete,
> > +			 * but keep outstanding IOASIDs
> > +			 */
> > +			list_for_each_entry(sops,
> > &pallocator->slist, list) {
> > +				if (sops == ops) {
> > +					list_del(&ops->list);
> > +					if
> > (refcount_dec_and_test(&pallocator->users))
> > +						pr_err("no shared
> > ops\n");
> > +					break;
> > +				}
> > +			}
> > +			break;
> > +		}
> > +	}
> > +
> > +exit_unlock:
> > +	mutex_unlock(&ioasid_allocator_lock);
> > +}
> > +EXPORT_SYMBOL_GPL(ioasid_unregister_allocator);
> > +
> > +/**
> > + * ioasid_set_data - Set private data for an allocated ioasid
> > + * @ioasid: the ID to set data
> > + * @data:   the private data
> > + *
> > + * For IOASID that is already allocated, private data can be set
> > + * via this API. Future lookup can be done via ioasid_find.
> > + */
> > +int ioasid_set_data(ioasid_t ioasid, void *data)
> > +{
> > +	struct ioasid_data *ioasid_data;
> > +	int ret = 0;
> > +
> > +	mutex_lock(&ioasid_allocator_lock);
> > +	ioasid_data = xa_load(&active_allocator->xa, ioasid);
> > +	if (ioasid_data)
> > +		ioasid_data->private = data;
> > +	else
> > +		ret = -ENOENT;
> > +	mutex_unlock(&ioasid_allocator_lock);
> > +
> > +	/* getter may use the private data */
> > +	synchronize_rcu();
> > +
> > +	return ret;
> > +}
> > +EXPORT_SYMBOL_GPL(ioasid_set_data);
> > +
> > +/**
> > + * ioasid_alloc - Allocate an IOASID
> > + * @set: the IOASID set
> > + * @min: the minimum ID (inclusive)
> > + * @max: the maximum ID (inclusive)
> > + * @private: data private to the caller
> > + *
> > + * Allocate an ID between @min and @max. Return the allocated ID
> > on success,
> > + * or INVALID_IOASID on failure. The @private pointer is stored
> > internally
> > + * and can be retrieved with ioasid_find().
> > + */
> > +ioasid_t ioasid_alloc(struct ioasid_set *set, ioasid_t min,
> > ioasid_t max,
> > +		      void *private)
> > +{
> > +	ioasid_t id = INVALID_IOASID;
> > +	struct ioasid_data *data;
> > +
> > +	data = kzalloc(sizeof(*data), GFP_KERNEL);
> > +	if (!data)
> > +		return INVALID_IOASID;
> > +
> > +	data->set = set;
> > +	data->private = private;
> > +
> > +	mutex_lock(&ioasid_allocator_lock);
> > +
> > +	id = active_allocator->ops->alloc(min, max, data);
> > +	if (id == INVALID_IOASID) {
> > +		pr_err("Failed ASID allocation %lu\n",
> > active_allocator->flags);
> > +			mutex_unlock(&ioasid_allocator_lock);
> > +			goto exit_free;
> > +	}
> > +	if (active_allocator->flags & IOASID_ALLOCATOR_CUSTOM) {
> > +		/* Custom allocator needs framework to store and
> > track allocation results */
> > +		min = id;
> > +		max = id + 1;  
> 
> Looking at xa_limit definition these are inclusive limits so why do
> we let it take either value?
> 
you are right, it was idr based which was exclusive. Jean already
fixed it.
> > +
> > +		if (xa_alloc(&active_allocator->xa, &id, data,
> > XA_LIMIT(min, max), GFP_KERNEL)) {
> > +			pr_err("Failed to alloc ioasid from %d to
> > %d\n", min, max);
> > +			active_allocator->ops->free(id, NULL);
> > +			goto exit_free;
> > +		}
> > +	}
> > +	data->id = id;
> > +
> > +	mutex_unlock(&ioasid_allocator_lock);
> > +
> > +exit_free:  
> As with the version of the patch without custom allocator, feels like
> we can simplify this section as we can't get here with id ==
> INVALID_IOASID unless we went to the exit_free.
> 
> If we did take the exit_free it is always INVALID_IOASID ( I think,
> but haven't checked...)
> 
> So just have a return from the good path and always run
> this block i the bad one.
>  
sounds good, fixed in Jean's tree.
> > +	if (id == INVALID_IOASID) {
> > +		kfree(data);
> > +		return INVALID_IOASID;
> > +	}
> > +	return id;
> > +}
> > +EXPORT_SYMBOL_GPL(ioasid_alloc);
> > +
> > +/**
> > + * ioasid_free - Free an IOASID
> > + * @ioasid: the ID to remove
> > + */
> > +void ioasid_free(ioasid_t ioasid)
> > +{
> > +	struct ioasid_data *ioasid_data;
> > +
> > +	mutex_lock(&ioasid_allocator_lock);
> > +
> > +	ioasid_data = xa_load(&active_allocator->xa, ioasid);
> > +	if (!ioasid_data) {
> > +		pr_err("Trying to free unknown IOASID %u\n",
> > ioasid);
> > +		goto exit_unlock;
> > +	}
> > +
> > +	active_allocator->ops->free(ioasid,
> > active_allocator->ops->pdata);
> > +	/* Custom allocator needs additional steps to free the xa
> > */  
> xa element perhaps?
> 
> > +	if (active_allocator->flags & IOASID_ALLOCATOR_CUSTOM) {
> > +		ioasid_data = xa_erase(&active_allocator->xa,
> > ioasid);
> > +		kfree_rcu(ioasid_data, rcu);
> > +	}
> > +
> > +exit_unlock:
> > +	mutex_unlock(&ioasid_allocator_lock);
> > +}
> > +EXPORT_SYMBOL_GPL(ioasid_free);
> > +
> > +/**
> > + * ioasid_find - Find IOASID data
> > + * @set: the IOASID set
> > + * @ioasid: the IOASID to find
> > + * @getter: function to call on the found object
> > + *
> > + * The optional getter function allows to take a reference to the
> > found object
> > + * under the rcu lock. The function can also check if the object
> > is still valid:
> > + * if @getter returns false, then the object is invalid and NULL
> > is returned.
> > + *
> > + * If the IOASID has been allocated for this set, return the
> > private pointer
> > + * passed to ioasid_alloc. Private data can be NULL if not set.
> > Return an error
> > + * if the IOASID is not found or does not belong to the set.
> > + */
> > +void *ioasid_find(struct ioasid_set *set, ioasid_t ioasid,
> > +		  bool (*getter)(void *))
> > +{
> > +	void *priv = NULL;  
> 
> Always set so no need to initialize.
> 
right
> > +	struct ioasid_data *ioasid_data;
> > +
> > +	rcu_read_lock();
> > +	ioasid_data = xa_load(&active_allocator->xa, ioasid);
> > +	if (!ioasid_data) {
> > +		priv = ERR_PTR(-ENOENT);
> > +		goto unlock;
> > +	}
> > +	if (set && ioasid_data->set != set) {
> > +		/* data found but does not belong to the set */
> > +		priv = ERR_PTR(-EACCES);
> > +		goto unlock;
> > +	}
> > +	/* Now IOASID and its set is verified, we can return the
> > private data */
> > +	priv = ioasid_data->private;
> > +	if (getter && !getter(priv))
> > +		priv = NULL;
> > +unlock:
> > +	rcu_read_unlock();
> > +
> > +	return priv;
> > +}
> > +EXPORT_SYMBOL_GPL(ioasid_find);
> > +
> > +MODULE_AUTHOR("Jean-Philippe Brucker
> > <jean-philippe.brucker@xxxxxxx>"); +MODULE_AUTHOR("Jacob Pan
> > <jacob.jun.pan@xxxxxxxxxxxxxxx>"); +MODULE_DESCRIPTION("IO Address
> > Space ID (IOASID) allocator"); +MODULE_LICENSE("GPL");
> > diff --git a/include/linux/ioasid.h b/include/linux/ioasid.h
> > new file mode 100644
> > index 0000000..8c8d1c5
> > --- /dev/null
> > +++ b/include/linux/ioasid.h
> > @@ -0,0 +1,74 @@
> > +/* SPDX-License-Identifier: GPL-2.0 */
> > +#ifndef __LINUX_IOASID_H
> > +#define __LINUX_IOASID_H
> > +
> > +#define INVALID_IOASID ((ioasid_t)-1)
> > +typedef unsigned int ioasid_t;
> > +typedef ioasid_t (*ioasid_alloc_fn_t)(ioasid_t min, ioasid_t max,
> > void *data); +typedef void (*ioasid_free_fn_t)(ioasid_t ioasid,
> > void *data); +
> > +struct ioasid_set {
> > +	int dummy;
> > +};
> > +
> > +/**
> > + * struct ioasid_allocator_ops - IOASID allocator helper functions
> > and data
> > + *
> > + * @alloc:	helper function to allocate IOASID
> > + * @free:	helper function to free IOASID
> > + * @list:	for tracking ops that share helper functions but
> > not data
> > + * @pdata:	data belong to the allocator, provided when
> > calling alloc()  
> and free()
> 
> Seems odd to call out one and not the other.
> I'm not actually sure it's true though as I can't see pdata being
> provided to the alloc call.
> 
You are right, allocator data should be provided to both free and
alloc. There is a bug in my code, should have called with allocator
private data instead of IOASID private data.
I will fix on top of Jean's version.

Thank you so much,

> > + */
> > +struct ioasid_allocator_ops {
> > +	ioasid_alloc_fn_t alloc;
> > +	ioasid_free_fn_t free;
> > +	struct list_head list;
> > +	void *pdata;
> > +};
> > +
> > +#define DECLARE_IOASID_SET(name) struct ioasid_set name = { 0 }  
> 
> I was a little curious on how this would actually be used.
> Even more curious is that it isn't used that I can find.
> 
It can be used for data types or ownership permission.
For VT-d. We use this as the mm_struct of the PASID allocator to ensure
that only IOASID allocated by the same process can operate on it,
e.g. free or bind.
> > +
> > +#if IS_ENABLED(CONFIG_IOASID)
> > +ioasid_t ioasid_alloc(struct ioasid_set *set, ioasid_t min,
> > ioasid_t max,
> > +		      void *private);
> > +void ioasid_free(ioasid_t ioasid);
> > +
> > +void *ioasid_find(struct ioasid_set *set, ioasid_t ioasid,
> > +			bool (*getter)(void *));
> > +int ioasid_register_allocator(struct ioasid_allocator_ops
> > *allocator); +void ioasid_unregister_allocator(struct
> > ioasid_allocator_ops *allocator); +int ioasid_set_data(ioasid_t
> > ioasid, void *data); +
> > +#else /* !CONFIG_IOASID */
> > +static inline ioasid_t ioasid_alloc(struct ioasid_set *set,
> > ioasid_t min,
> > +				    ioasid_t max, void *private)
> > +{
> > +	return INVALID_IOASID;
> > +}
> > +
> > +static inline void ioasid_free(ioasid_t ioasid)
> > +{
> > +}
> > +
> > +static inline void *ioasid_find(struct ioasid_set *set, ioasid_t
> > ioasid,
> > +				bool (*getter)(void *))
> > +{
> > +	return NULL;
> > +}
> > +
> > +static inline int ioasid_register_allocator(struct
> > ioasid_allocator_ops *allocator) +{
> > +	return -ENODEV;
> > +}
> > +
> > +static inline void ioasid_unregister_allocator(struct
> > ioasid_allocator_ops *allocator) +{
> > +}
> > +
> > +static inline int ioasid_set_data(ioasid_t ioasid, void *data)
> > +{
> > +	return -ENODEV;
> > +}
> > +
> > +#endif /* CONFIG_IOASID */
> > +#endif /* __LINUX_IOASID_H */  
> 
> 

[Jacob Pan]