iommu/vt-d: Fix NULL domain on device release
authorLu Baolu <baolu.lu@linux.intel.com>
Tue, 5 Mar 2024 12:21:18 +0000 (20:21 +0800)
committerJoerg Roedel <jroedel@suse.de>
Wed, 6 Mar 2024 16:35:57 +0000 (17:35 +0100)
commit81e921fd321614c2ad8ac333b041aae1da7a1c6d
tree8b7ab0d396698b7a9fbbf0aa74f3f1301086e5c8
parent0061ffe289e19caabeea8103e69cb0f1896e34d8
iommu/vt-d: Fix NULL domain on device release

In the kdump kernel, the IOMMU operates in deferred_attach mode. In this
mode, info->domain may not yet be assigned by the time the release_device
function is called. It leads to the following crash in the crash kernel:

    BUG: kernel NULL pointer dereference, address: 000000000000003c
    ...
    RIP: 0010:do_raw_spin_lock+0xa/0xa0
    ...
    _raw_spin_lock_irqsave+0x1b/0x30
    intel_iommu_release_device+0x96/0x170
    iommu_deinit_device+0x39/0xf0
    __iommu_group_remove_device+0xa0/0xd0
    iommu_bus_notifier+0x55/0xb0
    notifier_call_chain+0x5a/0xd0
    blocking_notifier_call_chain+0x41/0x60
    bus_notify+0x34/0x50
    device_del+0x269/0x3d0
    pci_remove_bus_device+0x77/0x100
    p2sb_bar+0xae/0x1d0
    ...
    i801_probe+0x423/0x740

Use the release_domain mechanism to fix it. The scalable mode context
entry which is not part of release domain should be cleared in
release_device().

Fixes: 586081d3f6b1 ("iommu/vt-d: Remove DEFER_DEVICE_DOMAIN_INFO")
Reported-by: Eric Badger <ebadger@purestorage.com>
Closes: https://lore.kernel.org/r/20240113181713.1817855-1-ebadger@purestorage.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20240305013305.204605-3-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
drivers/iommu/intel/iommu.c
drivers/iommu/intel/pasid.c
drivers/iommu/intel/pasid.h