[PATCH 3/5] iommu/amd: Fix group refcounting

[Robin Murphy]

If acpihid_device_group() finds an existing group for the relevant
devid, it should be taking an additional reference on that group.

Signed-off-by: Robin Murphy <***@arm.com>
---
drivers/iommu/amd_iommu.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index 754595ee11b6..4ef05691ceb1 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -373,6 +373,8 @@ static struct iommu_group *acpihid_device_group(struct device *dev)

if (!entry->group)
entry->group = generic_device_group(dev);
+ else
+ __iommu_group_get(entry->group);

return entry->group;
}

[Robin Murphy]

When arm_smmu_device_group() finds an existing group due to Stream ID
aliasing, it should be taking an additional reference on that group.

Reported-by: Sricharan R <***@codeaurora.org>
Signed-off-by: Robin Murphy <***@arm.com>
---
drivers/iommu/arm-smmu.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c
index 8f7281444551..9f760cfba4fa 100644
--- a/drivers/iommu/arm-smmu.c
+++ b/drivers/iommu/arm-smmu.c
@@ -1478,7 +1478,7 @@ static struct iommu_group *arm_smmu_device_group(struct device *dev)
}

if (group)
- return group;
+ return __iommu_group_get(group);

if (dev_is_pci(dev))
group = pci_device_group(dev);

[Robin Murphy]

For each subsequent device assigned to the m4u_group after its initial
allocation, we need to take an additional reference.

Signed-off-by: Robin Murphy <***@arm.com>
---
drivers/iommu/mtk_iommu.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/drivers/iommu/mtk_iommu.c b/drivers/iommu/mtk_iommu.c
index b12c12d74c33..f5591f6aef49 100644
--- a/drivers/iommu/mtk_iommu.c
+++ b/drivers/iommu/mtk_iommu.c
@@ -410,6 +410,8 @@ static struct iommu_group *mtk_iommu_device_group(struct device *dev)
data->m4u_group = iommu_group_alloc();
if (IS_ERR(data->m4u_group))
dev_err(dev, "Failed to allocate M4U IOMMU group\n");
+ } else {
+ __iommu_group_get(data->m4u_group);
}
return data->m4u_group;
}

[Robin Murphy]

For each subsequent device assigned to the m4u_group after its initial
allocation, we need to take an additional reference.

Signed-off-by: Robin Murphy <***@arm.com>
---
drivers/iommu/mtk_iommu_v1.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/drivers/iommu/mtk_iommu_v1.c b/drivers/iommu/mtk_iommu_v1.c
index b8aeb0768483..1b736990373f 100644
--- a/drivers/iommu/mtk_iommu_v1.c
+++ b/drivers/iommu/mtk_iommu_v1.c
@@ -502,6 +502,8 @@ static struct iommu_group *mtk_iommu_device_group(struct device *dev)
data->m4u_group = iommu_group_alloc();
if (IS_ERR(data->m4u_group))
dev_err(dev, "Failed to allocate M4U IOMMU group\n");
+ } else {
+ __iommu_group_get(data->m4u_group);
}
return data->m4u_group;
}

[Sricharan]

Hi,

-----Original Message-----
Sent: Wednesday, November 09, 2016 6:17 PM
Subject: [PATCH 1/5] iommu: Allow taking a reference on a group directly
iommu_group_get_for_dev() expects that the IOMMU driver's device_group
callback return a group with a reference held for the given device.
Whilst allocating a new group is fine, and pci_device_group() correctly
handles reusing an existing group, there is no general means for IOMMU
drivers doing their own group lookup to take additional references on an
existing group pointer without having to also store device pointers or
resort to elaborate trickery.
Add an IOMMU-driver-specific function to fill the hole.
---
drivers/iommu/iommu.c | 14 ++++++++++++++
include/linux/iommu.h | 1 +
2 files changed, 15 insertions(+)
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 9a2f1960873b..b0b052bc6bb5 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -552,6 +552,20 @@ struct iommu_group *iommu_group_get(struct device *dev)
EXPORT_SYMBOL_GPL(iommu_group_get);
/**
+ * __iommu_group_get - Increment reference on a group
+ *
+ * This function may be called by internal iommu driver group management
+ * when the context of a struct device pointer is not available. It is
+ * not for general use. Returns the given group for convenience.
+ */
+struct iommu_group *__iommu_group_get(struct iommu_group *group)
+{
+ kobject_get(group->devices_kobj);
+ return group;
+}
+
+/**

Acked-by: ***@codeaurora.org

Regards,
Sricharan

[Will Deacon]

iommu_group_get_for_dev() expects that the IOMMU driver's device_group
callback return a group with a reference held for the given device.
Whilst allocating a new group is fine, and pci_device_group() correctly
handles reusing an existing group, there is no general means for IOMMU
drivers doing their own group lookup to take additional references on an
existing group pointer without having to also store device pointers or
resort to elaborate trickery.
Add an IOMMU-driver-specific function to fill the hole.
---
drivers/iommu/iommu.c | 14 ++++++++++++++
include/linux/iommu.h | 1 +
2 files changed, 15 insertions(+)
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 9a2f1960873b..b0b052bc6bb5 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -552,6 +552,20 @@ struct iommu_group *iommu_group_get(struct device *dev)
EXPORT_SYMBOL_GPL(iommu_group_get);
/**
+ * __iommu_group_get - Increment reference on a group
+ *
+ * This function may be called by internal iommu driver group management
+ * when the context of a struct device pointer is not available. It is
+ * not for general use. Returns the given group for convenience.
+ */
+struct iommu_group *__iommu_group_get(struct iommu_group *group)
+{
+ kobject_get(group->devices_kobj);
+ return group;
+}

This probably either wants sticking in a header or exporting to modules.
That said, why do we need the underscores and the comment about internal
group management? That's pretty much already the case for iommu_group_get.

Of course, removing the underscores gives you a naming conflict, but we
could just call it something like "iommu_group_get_ref".

Will

[Robin Murphy]

Post by Will Deacon

iommu_group_get_for_dev() expects that the IOMMU driver's device_group
callback return a group with a reference held for the given device.
Whilst allocating a new group is fine, and pci_device_group() correctly
handles reusing an existing group, there is no general means for IOMMU
drivers doing their own group lookup to take additional references on an
existing group pointer without having to also store device pointers or
resort to elaborate trickery.
Add an IOMMU-driver-specific function to fill the hole.
---
drivers/iommu/iommu.c | 14 ++++++++++++++
include/linux/iommu.h | 1 +
2 files changed, 15 insertions(+)
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 9a2f1960873b..b0b052bc6bb5 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -552,6 +552,20 @@ struct iommu_group *iommu_group_get(struct device *dev)
EXPORT_SYMBOL_GPL(iommu_group_get);
/**
+ * __iommu_group_get - Increment reference on a group
+ *
+ * This function may be called by internal iommu driver group management
+ * when the context of a struct device pointer is not available. It is
+ * not for general use. Returns the given group for convenience.
+ */
+struct iommu_group *__iommu_group_get(struct iommu_group *group)
+{
+ kobject_get(group->devices_kobj);
+ return group;
+}

This probably either wants sticking in a header or exporting to modules.
That said, why do we need the underscores and the comment about internal
group management? That's pretty much already the case for iommu_group_get.

The definition of struct iommu_group is private to iommu.c, so any
touching of the members has to be in here. The comment is to contrast
with iommu_group_get()'s "This function is called by iommu drivers and
users". This one is explicitly not for users of the API (DMA mapping,
VFIO, etc.), as they really have no business messing with refcounts
directly, and should always be operating in the context of a device;
it's only for the benefit of anyone *implementing* the API. And since
IOMMU drivers aren't modular (yet... ;)) there's no cause for an export.

Post by Will Deacon
Of course, removing the underscores gives you a naming conflict, but we
could just call it something like "iommu_group_get_ref".

Ideally, this would be the iommu_group_get() to precisely match
iommu_group_put(), and the existing function would renamed something
like iommu_dev_group_get() (or perhaps even all external uses converted
over to iommu_group_get_for_dev()), but that would be an awful lot of
churn for little obvious benefit. Similarly, I nearly added the below
hunk, but it didn't seem worth the bother.

Robin.

----->8-----
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 9a2f1960873b..89d509c59019 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -545,7 +545,7 @@ struct iommu_group *iommu_group_get(struct device *dev)
struct iommu_group *group = dev->iommu_group;

if (group)
- kobject_get(group->devices_kobj);
+ __iommu_group_get(group);

return group;
}

[Will Deacon]

Post by Robin Murphy

Post by Will Deacon

iommu_group_get_for_dev() expects that the IOMMU driver's device_group
callback return a group with a reference held for the given device.
Whilst allocating a new group is fine, and pci_device_group() correctly
handles reusing an existing group, there is no general means for IOMMU
drivers doing their own group lookup to take additional references on an
existing group pointer without having to also store device pointers or
resort to elaborate trickery.
Add an IOMMU-driver-specific function to fill the hole.
---
drivers/iommu/iommu.c | 14 ++++++++++++++
include/linux/iommu.h | 1 +
2 files changed, 15 insertions(+)
diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index 9a2f1960873b..b0b052bc6bb5 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -552,6 +552,20 @@ struct iommu_group *iommu_group_get(struct device *dev)
EXPORT_SYMBOL_GPL(iommu_group_get);
/**
+ * __iommu_group_get - Increment reference on a group
+ *
+ * This function may be called by internal iommu driver group management
+ * when the context of a struct device pointer is not available. It is
+ * not for general use. Returns the given group for convenience.
+ */
+struct iommu_group *__iommu_group_get(struct iommu_group *group)
+{
+ kobject_get(group->devices_kobj);
+ return group;
+}

This probably either wants sticking in a header or exporting to modules.
That said, why do we need the underscores and the comment about internal
group management? That's pretty much already the case for iommu_group_get.

The definition of struct iommu_group is private to iommu.c, so any
touching of the members has to be in here. The comment is to contrast
with iommu_group_get()'s "This function is called by iommu drivers and
users". This one is explicitly not for users of the API (DMA mapping,
VFIO, etc.), as they really have no business messing with refcounts
directly, and should always be operating in the context of a device;

But they can already do it if they want to, using the horrible group id
hack that's been doing the rounds. The IOMMU API is already low-level
enough, so I don't think trying to split it up like this is helpful.
Hell, people can even just dip in and bump the kobject directly, or grab a
handle to a device in the group already and call iommu_group_get.

That said, it doesn't look like iommu_group_get_by_id actually has any
callers in tree, so maybe we could kill it.

Post by Robin Murphy
it's only for the benefit of anyone *implementing* the API. And since
IOMMU drivers aren't modular (yet... ;)) there's no cause for an export.

Post by Will Deacon
Of course, removing the underscores gives you a naming conflict, but we
could just call it something like "iommu_group_get_ref".

Ideally, this would be the iommu_group_get() to precisely match
iommu_group_put(), and the existing function would renamed something
like iommu_dev_group_get() (or perhaps even all external uses converted
over to iommu_group_get_for_dev()), but that would be an awful lot of
churn for little obvious benefit. Similarly, I nearly added the below
hunk, but it didn't seem worth the bother.

I'd still rather the new function was renamed. We already have the group,
so calling a weird underscore version of iommu_group_get is really
counter-intuitive.

Joerg -- do you have a preference?

Will

[Joerg Roedel]

Post by Will Deacon
I'd still rather the new function was renamed. We already have the group,
so calling a weird underscore version of iommu_group_get is really
counter-intuitive.
Joerg -- do you have a preference?

iommu_group_ref_get sound best to me. Unfortunatly C has no function
overloading ;)


Joerg