linux.git/drivers/iommu, branch v6.1.2

iommu/mediatek: Fix forever loop in error handling

2022-12-31T12:32:53+00:00

[ Upstream commit 462e768b55a2331324ff72e74706261134369826 ]

There is a typo so this loop does i++ where i-- was intended.  It will
result in looping until the kernel crashes.

Fixes: 26593928564c ("iommu/mediatek: Add error path for loop of mm_dts_parse")
Signed-off-by: Dan Carpenter 
Reviewed-by: Yong Wu 
Link: https://lore.kernel.org/r/Y5C3mTam2nkbaz6o@kili
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu/sun50i: Remove IOMMU_DOMAIN_IDENTITY

2022-12-31T12:32:51+00:00

[ Upstream commit ef5bb8e7a7127218f826b9ccdf7508e7a339f4c2 ]

This driver treats IOMMU_DOMAIN_IDENTITY the same as UNMANAGED, which
cannot possibly be correct.

UNMANAGED domains are required to start out blocking all DMAs. This seems
to be what this driver does as it allocates a first level 'dt' for the IO
page table that is 0 filled.

Thus UNMANAGED looks like a working IO page table, and so IDENTITY must be
a mistake. Remove it.

Fixes: 4100b8c229b3 ("iommu: Add Allwinner H6 IOMMU driver")
Signed-off-by: Jason Gunthorpe 
Reviewed-by: Robin Murphy 
Link: https://lore.kernel.org/r/0-v1-97f0adf27b5e+1f0-s50_identity_jgg@nvidia.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu/mediatek: Validate number of phandles associated with "mediatek,larbs"

2022-12-31T12:32:51+00:00

[ Upstream commit ef693a8440926884bfd9cc3d6d36f65719513350 ]

Fix the smatch warnings:
drivers/iommu/mtk_iommu.c:878 mtk_iommu_mm_dts_parse() error: uninitialized
symbol 'larbnode'.

If someone abuse the dtsi node(Don't follow the definition of dt-binding),
for example "mediatek,larbs" is provided as boolean property, "larb_nr"
will be zero and cause abnormal.

To fix this problem and improve the code safety, add some checking
for the invalid input from dtsi, e.g. checking the larb_nr/larbid valid
range, and avoid "mediatek,larb-id" property conflicts in the smi-larb
nodes.

Fixes: d2e9a1102cfc ("iommu/mediatek: Contain MM IOMMU flow with the MM TYPE")
Reported-by: kernel test robot 
Reported-by: Dan Carpenter 
Signed-off-by: Guenter Roeck 
Signed-off-by: Yong Wu 
Reviewed-by: AngeloGioacchino Del Regno 
Reviewed-by: Matthias Brugger 
Link: https://lore.kernel.org/r/20221018024258.19073-5-yong.wu@mediatek.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu/mediatek: Add error path for loop of mm_dts_parse

2022-12-31T12:32:51+00:00

[ Upstream commit 26593928564cf5b576ff05d3cbd958f57c9534bb ]

The mtk_iommu_mm_dts_parse will parse the smi larbs nodes. if the i+1
larb is parsed fail, we should put_device for the i..0 larbs.

There are two places need to comment:
1) The larbid may be not linear mapping, we should loop whole
   the array in the error path.
2) I move this line position: "data->larb_imu[id].dev = &plarbdev->dev;"
   before "if (!plarbdev->dev.driver)", That means set
   data->larb_imu[id].dev before the error path. then we don't need
   "platform_device_put(plarbdev)" again in probe_defer case. All depend
   on "put_device" of the error path in error cases.

Fixes: d2e9a1102cfc ("iommu/mediatek: Contain MM IOMMU flow with the MM TYPE")
Signed-off-by: Yong Wu 
Reviewed-by: AngeloGioacchino Del Regno 
Reviewed-by: Matthias Brugger 
Link: https://lore.kernel.org/r/20221018024258.19073-4-yong.wu@mediatek.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu/mediatek: Use component_match_add

2022-12-31T12:32:51+00:00

[ Upstream commit b5765a1b44bea9dfcae69c53ffeb4c689d0922a7 ]

In order to simplify the error patch(avoid call of_node_put), Use
component_match_add instead component_match_add_release since we are only
interested in the "device" here. Then we could always call of_node_put in
normal path.

Strictly this is not a fixes patch, but it is a prepare for adding the
error path, thus I add a Fixes tag too.

Fixes: d2e9a1102cfc ("iommu/mediatek: Contain MM IOMMU flow with the MM TYPE")
Suggested-by: Robin Murphy 
Signed-off-by: Yong Wu 
Reviewed-by: AngeloGioacchino Del Regno 
Reviewed-by: Matthias Brugger 
Link: https://lore.kernel.org/r/20221018024258.19073-3-yong.wu@mediatek.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu/mediatek: Add platform_device_put for recovering the device refcnt

2022-12-31T12:32:51+00:00

[ Upstream commit dcb40e9fcce9bd251eaff19f3724131db522846c ]

Add platform_device_put to match with of_find_device_by_node.

Meanwhile, I add a new variable "pcommdev" which is for smi common device.
Otherwise, "platform_device_put(plarbdev)" for smi-common dev may be not
readable. And add a checking for whether pcommdev is NULL.

Fixes: d2e9a1102cfc ("iommu/mediatek: Contain MM IOMMU flow with the MM TYPE")
Signed-off-by: Yong Wu 
Reviewed-by: AngeloGioacchino Del Regno 
Reviewed-by: Matthias Brugger 
Link: https://lore.kernel.org/r/20221018024258.19073-2-yong.wu@mediatek.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu/fsl_pamu: Fix resource leak in fsl_pamu_probe()

2022-12-31T12:32:50+00:00

[ Upstream commit 73f5fc5f884ad0c5f7d57f66303af64f9f002526 ]

The fsl_pamu_probe() returns directly when create_csd() failed, leaving
irq and memories unreleased.
Fix by jumping to error if create_csd() returns error.

Fixes: 695093e38c3e ("iommu/fsl: Freescale PAMU driver and iommu implementation.")
Signed-off-by: Yuan Can 
Link: https://lore.kernel.org/r/20221121082022.19091-1-yuancan@huawei.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu/amd: Fix pci device refcount leak in ppr_notifier()

2022-12-31T12:32:50+00:00

[ Upstream commit 6cf0981c2233f97d56938d9d61845383d6eb227c ]

As comment of pci_get_domain_bus_and_slot() says, it returns
a pci device with refcount increment, when finish using it,
the caller must decrement the reference count by calling
pci_dev_put(). So call it before returning from ppr_notifier()
to avoid refcount leak.

Fixes: daae2d25a477 ("iommu/amd: Don't copy GCR3 table root pointer")
Signed-off-by: Yang Yingliang 
Link: https://lore.kernel.org/r/20221118093604.216371-1-yangyingliang@huawei.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu: Avoid races around device probe

2022-12-31T12:32:49+00:00

[ Upstream commit 01657bc14a3990c665375f77978631fee77b1fce ]

We currently have 3 different ways that __iommu_probe_device() may be
called, but no real guarantee that multiple callers can't tread on each
other, especially once asynchronous driver probe gets involved. It would
likely have taken a fair bit of luck to hit this previously, but commit
57365a04c921 ("iommu: Move bus setup to IOMMU device registration") ups
the odds since now it's not just omap-iommu that may trigger multiple
bus_iommu_probe() calls in parallel if probing asynchronously.

Add a lock to ensure we can't try to double-probe a device, and also
close some possible race windows to make sure we're truly robust against
trying to double-initialise a group via two different member devices.

Reported-by: Brian Norris 
Signed-off-by: Robin Murphy 
Tested-by: Brian Norris 
Fixes: 57365a04c921 ("iommu: Move bus setup to IOMMU device registration")
Link: https://lore.kernel.org/r/1946ef9f774851732eed78760a78ec40dbc6d178.1667591503.git.robin.murphy@arm.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin

iommu/mediatek: Check return value after calling platform_get_resource()

2022-12-31T12:32:49+00:00

[ Upstream commit 73b6924cdebc899de9b719e1319aa86c6bed4acf ]

platform_get_resource() may return NULL pointer, we need check its
return value to avoid null-ptr-deref in resource_size().

Fixes: 42d57fc58aeb ("iommu/mediatek: Initialise/Remove for multi bank dev")
Signed-off-by: Yang Yingliang 
Reviewed-by: Matthias Brugger 
Link: https://lore.kernel.org/r/20221029103550.3774365-1-yangyingliang@huawei.com
Signed-off-by: Joerg Roedel 
Signed-off-by: Sasha Levin