Add bind() and unbind() operations to the IOMMU API.
iommu_sva_bind_device() binds a device to an mm, and returns a handle to
the bond, which is released by calling iommu_sva_unbind_device().

Each mm bound to devices gets a PASID (by convention, a 20-bit system-wide
ID representing the address space), which can be retrieved with
iommu_sva_get_pasid(). When programming DMA addresses, device drivers
include this PASID in a device-specific manner, to let the device access
the given address space. Since the process memory may be paged out, device
and IOMMU must support I/O page faults (e.g. PCI PRI).

Using iommu_sva_set_ops(), device drivers provide an mm_exit() callback
that is called by the IOMMU driver if the process exits before the device
driver called unbind(). In mm_exit(), device driver should disable DMA
from the given context, so that the core IOMMU can reallocate the PASID.
Whether the process exited or nor, the device driver should always release
the handle with unbind().

To use these functions, device driver must first enable the
IOMMU_DEV_FEAT_SVA device feature with iommu_dev_enable_feature().
Signed-off-by: NJean-Philippe Brucker <jean-philippe.brucker@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Sharing a physical PCI device in a finer-granularity way
is becoming a consensus in the industry. IOMMU vendors
are also engaging efforts to support such sharing as well
as possible. Among the efforts, the capability of support
finer-granularity DMA isolation is a common requirement
due to the security consideration. With finer-granularity
DMA isolation, subsets of a PCI function can be isolated
from each others by the IOMMU. As a result, there is a
request in software to attach multiple domains to a physical
PCI device. One example of such use model is the Intel
Scalable IOV [1] [2]. The Intel vt-d 3.0 spec [3] introduces
the scalable mode which enables PASID granularity DMA
isolation.

This adds the APIs to support multiple domains per device.
In order to ease the discussions, we call it 'a domain in
auxiliary mode' or simply 'auxiliary domain' when multiple
domains are attached to a physical device.

The APIs include:

* iommu_dev_has_feature(dev, IOMMU_DEV_FEAT_AUX)
  - Detect both IOMMU and PCI endpoint devices supporting
    the feature (aux-domain here) without the host driver
    dependency.

* iommu_dev_feature_enabled(dev, IOMMU_DEV_FEAT_AUX)
  - Check the enabling status of the feature (aux-domain
    here). The aux-domain interfaces are available only
    if this returns true.

* iommu_dev_enable/disable_feature(dev, IOMMU_DEV_FEAT_AUX)
  - Enable/disable device specific aux-domain feature.

* iommu_aux_attach_device(domain, dev)
  - Attaches @domain to @dev in the auxiliary mode. Multiple
    domains could be attached to a single device in the
    auxiliary mode with each domain representing an isolated
    address space for an assignable subset of the device.

* iommu_aux_detach_device(domain, dev)
  - Detach @domain which has been attached to @dev in the
    auxiliary mode.

* iommu_aux_get_pasid(domain, dev)
  - Return ID used for finer-granularity DMA translation.
    For the Intel Scalable IOV usage model, this will be
    a PASID. The device which supports Scalable IOV needs
    to write this ID to the device register so that DMA
    requests could be tagged with a right PASID prefix.

This has been updated with the latest proposal from Joerg
posted here [5].

Many people involved in discussions of this design.

Kevin Tian <kevin.tian@intel.com>
Liu Yi L <yi.l.liu@intel.com>
Ashok Raj <ashok.raj@intel.com>
Sanjay Kumar <sanjay.k.kumar@intel.com>
Jacob Pan <jacob.jun.pan@linux.intel.com>
Alex Williamson <alex.williamson@redhat.com>
Jean-Philippe Brucker <jean-philippe.brucker@arm.com>
Joerg Roedel <joro@8bytes.org>

and some discussions can be found here [4] [5].

[1] https://software.intel.com/en-us/download/intel-scalable-io-virtualization-technical-specification
[2] https://schd.ws/hosted_files/lc32018/00/LC3-SIOV-final.pdf
[3] https://software.intel.com/en-us/download/intel-virtualization-technology-for-directed-io-architecture-specification
[4] https://lkml.org/lkml/2018/7/26/4
[5] https://www.spinics.net/lists/iommu/msg31874.html

Cc: Ashok Raj <ashok.raj@intel.com>
Cc: Jacob Pan <jacob.jun.pan@linux.intel.com>
Cc: Kevin Tian <kevin.tian@intel.com>
Cc: Liu Yi L <yi.l.liu@intel.com>
Suggested-by: NKevin Tian <kevin.tian@intel.com>
Suggested-by: NJean-Philippe Brucker <jean-philippe.brucker@arm.com>
Suggested-by: NJoerg Roedel <jroedel@suse.de>
Signed-off-by: NLu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: NJean-Philippe Brucker <jean-philippe.brucker@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Print the warning about the fall-back to IOMMU_DOMAIN_DMA in
iommu_group_get_for_dev() only when such a domain was
actually allocated.

Otherwise the user will get misleading warnings in the
kernel log when the iommu driver used doesn't support
IOMMU_DOMAIN_DMA and IOMMU_DOMAIN_IDENTITY.

Fixes: fccb4e3b ('iommu: Allow default domain type to be set on the kernel command line')
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Use dev_printk() when possible so the IOMMU messages are more consistent
with other messages related to the device.

E.g., I think these messages related to surprise hotplug:

  pciehp 0000:80:10.0:pcie004: Slot(36): Link Down
  iommu: Removing device 0000:87:00.0 from group 12
  pciehp 0000:80:10.0:pcie004: Slot(36): Card present
  pcieport 0000:80:10.0: Data Link Layer Link Active not set in 1000 msec

would be easier to read as these (also requires some PCI changes not
included here):

  pci 0000:80:10.0: Slot(36): Link Down
  pci 0000:87:00.0: Removing from iommu group 12
  pci 0000:80:10.0: Slot(36): Card present
  pci 0000:80:10.0: Data Link Layer Link Active not set in 1000 msec
Signed-off-by: NBjorn Helgaas <bhelgaas@google.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Introduce iotlb_sync_map() callback that is invoked in the end of
iommu_map(). This new callback allows IOMMU drivers to avoid syncing
after mapping of each contiguous chunk and sync only when the whole
mapping is completed, optimizing performance of the mapping operation.
Signed-off-by: NDmitry Osipenko <digetx@gmail.com>
Reviewed-by: NRobin Murphy <robin.murphy@arm.com>
Reviewed-by: NThierry Reding <treding@nvidia.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

This check needs to be there and got lost at some point
during development. Add it again.

Fixes: 641fb0ef ('iommu/of: Don't call iommu_ops->add_device directly')
Reported-by: NMarek Szyprowski <m.szyprowski@samsung.com>
Reported-by: Nkernelci.org bot <bot@kernelci.org>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Put them into separate functions and call those where the
plain ops have been called before.
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

These wrappers will be used to easily change the location of
the field later when all users are converted.
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Historically a lot of these existed because we did not have
a distinction between what was modular code and what was providing
support to modules via EXPORT_SYMBOL and friends.  That changed
when we forked out support for the latter into the export.h file.
This means we should be able to reduce the usage of module.h
in code that is obj-y Makefile or bool Kconfig.

The advantage in removing such instances is that module.h itself
sources about 15 other headers; adding significantly to what we feed
cpp, and it can obscure what headers we are effectively using.

Since module.h might have been the implicit source for init.h
(for __init) and for export.h (for EXPORT_SYMBOL) we consider each
instance for the presence of either and replace as needed.

Cc: Joerg Roedel <joro@8bytes.org>
Cc: iommu@lists.linux-foundation.org
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

The original motivation for iommu_map_sg() was to give IOMMU drivers the
chance to map an IOVA-contiguous scatterlist as efficiently as they
could. It turns out that there isn't really much driver-specific
business involved there, so now that the default implementation is
mandatory let's just improve that - the main thing we're after is to use
larger pages wherever possible, and as long as domain->pgsize_bitmap
reflects reality, iommu_map() can already do that in a generic way. All
we need to do is detect physically-contiguous segments and batch them
into a single map operation, since whatever we do here is transparent to
our caller and not bound by any segment-length restrictions on the list
itself.

Speaking of efficiency, there's really very little point in duplicating
the checks that iommu_map() is going to do anyway, so those get cleared
up in the process.
Signed-off-by: NRobin Murphy <robin.murphy@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Add a generic command line option to enable lazy unmapping via IOVA
flush queues, which will initally be suuported by iommu-dma. This echoes
the semantics of "intel_iommu=strict" (albeit with the opposite default
value), but in the driver-agnostic fashion of "iommu.passthrough".
Signed-off-by: NZhen Lei <thunder.leizhen@huawei.com>
[rm: move handling out of SMMUv3 driver, clean up documentation]
Signed-off-by: NRobin Murphy <robin.murphy@arm.com>
[will: dropped broken printk when parsing command-line option]
Signed-off-by: NWill Deacon <will.deacon@arm.com>

This patch fixes a typo in iommu.c.
Signed-off-by: NRami Rosen <ramirose@gmail.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

The external interface to get/set window attributes is already
abstracted behind iommu_domain_{get,set}_attr(), so there's no real
reason for the internal interface to be different. Since we only have
one window-based driver anyway, clean up the core code by just moving
the DOMAIN_ATTR_WINDOWS handling directly into the PAMU driver.
Signed-off-by: NRobin Murphy <robin.murphy@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

While iommu_get_domain_for_dev() is the robust way for arbitrary IOMMU
API callers to retrieve the domain pointer, for DMA ops domains it
doesn't scale well for large systems and multi-queue devices, since the
momentary refcount adjustment will lead to exclusive cacheline contention
when multiple CPUs are operating in parallel on different mappings for
the same device.

In the case of DMA ops domains, however, this refcounting is actually
unnecessary, since they already imply that the group exists and is
managed by platform code and IOMMU internals (by virtue of
iommu_group_get_for_dev()) such that a reference will already be held
for the lifetime of the device. Thus we can avoid the bottleneck by
providing a fast lookup specifically for the DMA code to retrieve the
default domain it already knows it has set up - a simple read-only
dereference plays much nicer with cache-coherency protocols.
Signed-off-by: NRobin Murphy <robin.murphy@arm.com>
Tested-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Implement bus specific support for the fsl-mc bus including
registering arm_smmu_ops and bus specific device add operations.
Signed-off-by: NNipun Gupta <nipun.gupta@nxp.com>
Reviewed-by: NRobin Murphy <robin.murphy@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

All iommu drivers use the default_iommu_map_sg implementation, and there
is no good reason to ever override it.  Just expose it as iommu_map_sg
directly and remove the indirection, specially in our post-spectre world
where indirect calls are horribly expensive.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

This allows the default behavior to be controlled by a kernel config
option instead of changing the commandline for the kernel to include
"iommu.passthrough=on" or "iommu=pt" on machines where this is desired.

Likewise, for machines where this config option is enabled, it can be
disabled at boot time with "iommu.passthrough=off" or "iommu=nopt".

Also corrected iommu=pt documentation for IA-64, since it has no code that
parses iommu= at all.
Signed-off-by: NOlof Johansson <olof@lixom.net>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

While we could print it at setup time, this is an easier way to match
each device to their default IOMMU allocation type.
Signed-off-by: NOlof Johansson <olof@lixom.net>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Provide base enablement for using debugfs to expose internal data of an
IOMMU driver. When called, create the /sys/kernel/debug/iommu directory.

Emit a strong warning at boot time to indicate that this feature is
enabled.

This function is called from iommu_init, and creates the initial DebugFS
directory. Drivers may then call iommu_debugfs_new_driver_dir() to
instantiate a device-specific directory to expose internal data.
It will return a pointer to the new dentry structure created in
/sys/kernel/debug/iommu, or NULL in the event of a failure.

Since the IOMMU driver can not be removed from the running system, there
is no need for an "off" function.
Signed-off-by: NGary R Hook <gary.hook@amd.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

@name parameter has been removed.
Signed-off-by: NLu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

strtobool() does check for NULL parameter already. No need to repeat.

While here, switch to kstrtobool() and unshadow actual error code
(which is still -EINVAL).

No functional change intended.
Signed-off-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Currently, iommu_unmap, iommu_unmap_fast and iommu_map_sg return
size_t.  However, some of the return values are error codes (< 0),
which can be misinterpreted as large size. Therefore, returning size 0
instead to signify failure to map/unmap.

Cc: Joerg Roedel <joro@8bytes.org>
Cc: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: NSuravee Suthikulpanit <suravee.suthikulpanit@amd.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

The result of iommu_group_get() was being blindly used in both
attach and detach which results in a dereference when trying
to work with an unknown device.
Signed-off-by: NJordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: NAlex Williamson <alex.williamson@redhat.com>

With the current IOMMU-API the hardware TLBs have to be
flushed in every iommu_ops->unmap() call-back.

For unmapping large amounts of address space, like it
happens when a KVM domain with assigned devices is
destroyed, this causes thousands of unnecessary TLB flushes
in the IOMMU hardware because the unmap call-back runs for
every unmapped physical page.

With the TLB Flush Interface and the new iommu_unmap_fast()
function introduced here the need to clean the hardware TLBs
is removed from the unmapping code-path. Users of
iommu_unmap_fast() have to explicitly call the TLB-Flush
functions to sync the page-table changes to the hardware.

Three functions for TLB-Flushes are introduced:

	* iommu_flush_tlb_all() - Flushes all TLB entries
	                          associated with that
				  domain. TLBs entries are
				  flushed when this function
				  returns.

	* iommu_tlb_range_add() - This will add a given
				  range to the flush queue
				  for this domain.

	* iommu_tlb_sync() - Flushes all queued ranges from
			     the hardware TLBs. Returns when
			     the flush is finished.

The semantic of this interface is intentionally similar to
the iommu_gather_ops from the io-pgtable code.

Cc: Alex Williamson <alex.williamson@redhat.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Robin Murphy <robin.murphy@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

The recently-removed FIXME in iommu_get_domain_for_dev() turns out to
have been a little misleading, since that check is still worthwhile even
when groups *are* universal. We have a few IOMMU-aware drivers which
only care whether their device is already attached to an existing domain
or not, for which the previous behaviour of iommu_get_domain_for_dev()
was ideal, and who now crash if their device does not have an IOMMU.

With IOMMU groups now serving as a reliable indicator of whether a
device has an IOMMU or not (barring false-positives from VFIO no-IOMMU
mode), drivers could arguably do this:

	group = iommu_group_get(dev);
	if (group) {
		domain = iommu_get_domain_for_dev(dev);
		iommu_group_put(group);
	}

However, rather than duplicate that code across multiple callsites,
particularly when it's still only the domain they care about, let's skip
straight to the next step and factor out the check into the common place
it applies - in iommu_get_domain_for_dev() itself. Sure, it ends up
looking rather familiar, but now it's backed by the reasoning of having
a robust API able to do the expected thing for all devices regardless.

Fixes: 05f80300 ("iommu: Finish making iommu_group support mandatory")
Reported-by: NShawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: NRobin Murphy <robin.murphy@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

This new call-back will be used to check if the domain attach need be
deferred for now. If yes, the domain attach/detach will return directly.
Signed-off-by: NBaoquan He <bhe@redhat.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Now that all the drivers properly implementing the IOMMU API support
groups (I'm ignoring the etnaviv GPU MMUs which seemingly only do just
enough to convince the ARM DMA mapping ops), we can remove the FIXME
workarounds from the core code. In the process, it also seems logical to
make the .device_group callback non-optional for drivers calling
iommu_group_get_for_dev() - the current callers all implement it anyway,
and it doesn't make sense for any future callers not to either.
Signed-off-by: NRobin Murphy <robin.murphy@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

This callback should never return NULL. Print a warning if
that happens so that we notice and can fix it.
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

The generic device_group call-backs in iommu.c return NULL
in case of error. Since they are getting ERR_PTR values from
iommu_group_alloc(), just pass them up instead.
Reported-by: NGerald Schaefer <gerald.schaefer@de.ibm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

The function is in no fast-path, there is no need for it to
be static inline in a header file. This also removes the
need to include iommu trace-points in iommu.h.
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

In iommu_bus_notifier(), when action is
BUS_NOTIFY_ADD_DEVICE, it will return 'ops->add_device(dev)'
directly. But ops->add_device will return ERR_VAL, such as
-ENODEV. These value will make notifier_call_chain() not to
traverse the remain nodes in struct notifier_block list.

This patch revises iommu_bus_notifier() to return
NOTIFY_DONE when some errors happened in ops->add_device().
Signed-off-by: Nzhichang.yuan <yuanzhichang@hisilicon.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

The IOMMU core currently initialises the default domain for each group
to IOMMU_DOMAIN_DMA, under the assumption that devices will use
IOMMU-backed DMA ops by default. However, in some cases it is desirable
for the DMA ops to bypass the IOMMU for performance reasons, reserving
use of translation for subsystems such as VFIO that require it for
enforcing device isolation.

Rather than modify each IOMMU driver to provide different semantics for
DMA domains, instead we introduce a command line parameter that can be
used to change the type of the default domain. Passthrough can then be
specified using "iommu.passthrough=1" on the kernel command line.
Signed-off-by: NWill Deacon <will.deacon@arm.com>

The introduction of reserved regions has left a couple of rough edges
which we could do with sorting out sooner rather than later. Since we
are not yet addressing the potential dynamic aspect of software-managed
reservations and presenting them at arbitrary fixed addresses, it is
incongruous that we end up displaying hardware vs. software-managed MSI
regions to userspace differently, especially since ARM-based systems may
actually require one or the other, or even potentially both at once,
(which iommu-dma currently has no hope of dealing with at all). Let's
resolve the former user-visible inconsistency ASAP before the ABI has
been baked into a kernel release, in a way that also lays the groundwork
for the latter shortcoming to be addressed by follow-up patches.

For clarity, rename the software-managed type to IOMMU_RESV_SW_MSI, use
IOMMU_RESV_MSI to describe the hardware type, and document everything a
little bit. Since the x86 MSI remapping hardware falls squarely under
this meaning of IOMMU_RESV_MSI, apply that type to their regions as well,
so that we tell the same story to userspace across all platforms.

Secondly, as the various region types require quite different handling,
and it really makes little sense to ever try combining them, convert the
bitfield-esque #defines to a plain enum in the process before anyone
gets the wrong impression.

Fixes: d30ddcaa ("iommu: Add a new type field in iommu_resv_region")
Reviewed-by: NEric Auger <eric.auger@redhat.com>
CC: Alex Williamson <alex.williamson@redhat.com>
CC: David Woodhouse <dwmw2@infradead.org>
CC: kvm@vger.kernel.org
Signed-off-by: NRobin Murphy <robin.murphy@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

And also move its remaining functionality to
iommu_device_register() and 'struct iommu_device'.

Cc: Rob Herring <robh+dt@kernel.org>
Cc: Frank Rowand <frowand.list@gmail.com>
Cc: Matthias Brugger <matthias.bgg@gmail.com>
Cc: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: devicetree@vger.kernel.org
Cc: linux-arm-kernel@lists.infradead.org
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

This struct represents one hardware iommu in the iommu core
code. For now it only has the iommu-ops associated with it,
but that will be extended soon.

The register/unregister interface is also added, as well as
making use of it in the Intel and AMD IOMMU drivers.
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

The struct is used to link devices to iommu-groups, so
'struct group_device' is a better name. Further this makes
the name iommu_device available for a struct representing
hardware iommus.
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Rename the function to iommu_ops_from_fwnode(), because that
is what the function actually does. The new name is much
more descriptive about what the function does.
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

In case the device reserved region list is void, the returned value
of iommu_insert_device_resv_regions is uninitialized. Let's return 0
in that case.

This fixes commit 6c65fb31 ("iommu: iommu_get_group_resv_regions").
Reported-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NEric Auger <eric.auger@redhat.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

Move the assignment statement into if branch above, where it only
needs to be.
Signed-off-by: NZhen Lei <thunder.leizhen@huawei.com>
Reviewed-by: NRobin Murphy <robin.murphy@arm.com>
Signed-off-by: NJoerg Roedel <jroedel@suse.de>

A new iommu-group sysfs attribute file is introduced. It contains
the list of reserved regions for the iommu-group. Each reserved
region is described on a separate line:
- first field is the start IOVA address,
- second is the end IOVA address,
- third is the type.
Signed-off-by: NEric Auger <eric.auger@redhat.com>
Tested-by: NTomasz Nowicki <tomasz.nowicki@caviumnetworks.com>
Tested-by: NBharat Bhushan <bharat.bhushan@nxp.com>
Signed-off-by: NWill Deacon <will.deacon@arm.com>