[ Upstream commit 0a967f5bfd ]
The VT-d spec requires (10.4.4 Global Command Register, TE
field) that:
Hardware implementations supporting DMA draining must drain
any in-flight DMA read/write requests queued within the
Root-Complex before completing the translation enable
command and reflecting the status of the command through
the TES field in the Global Status register.
Unfortunately, some integrated graphic devices fail to do
so after some kind of power state transition. As the
result, the system might stuck in iommu_disable_translati
on(), waiting for the completion of TE transition.
This adds RPLS to a quirk list for those devices and skips
TE disabling if the qurik hits.
Link: https://gitlab.freedesktop.org/drm/intel/-/issues/4898
Tested-by: Raviteja Goud Talla <ravitejax.goud.talla@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Acked-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Tejas Upadhyay <tejaskumarx.surendrakumar.upadhyay@intel.com>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20220302043256.191529-1-tejaskumarx.surendrakumar.upadhyay@intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit da8669ff41 upstream.
The page fault handling framework in the IOMMU core explicitly states
that it doesn't handle PCI PASID Stop Marker and the IOMMU drivers must
discard them before reporting faults. This handles Stop Marker messages
in prq_event_thread() before reporting events to the core.
The VT-d driver explicitly drains the pending page requests when a CPU
page table (represented by a mm struct) is unbound from a PASID according
to the procedures defined in the VT-d spec. The Stop Marker messages do
not need a response. Hence, it is safe to drop the Stop Marker messages
silently if any of them is found in the page request queue.
Fixes: d5b9e4bfe0 ("iommu/vt-d: Report prq to io-pgfault framework")
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20220421113558.3504874-1-baolu.lu@linux.intel.com
Link: https://lore.kernel.org/r/20220423082330.3897867-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 59bf3557cf upstream.
Calculate the appropriate mask for non-size-aligned page selective
invalidation. Since psi uses the mask value to mask out the lower order
bits of the target address, properly flushing the iotlb requires using a
mask value such that [pfn, pfn+pages) all lie within the flushed
size-aligned region. This is not normally an issue because iova.c
always allocates iovas that are aligned to their size. However, iovas
which come from other sources (e.g. userspace via VFIO) may not be
aligned.
To properly flush the IOTLB, both the start and end pfns need to be
equal after applying the mask. That means that the most efficient mask
to use is the index of the lowest bit that is equal where all higher
bits are also equal. For example, if pfn=0x17f and pages=3, then
end_pfn=0x181, so the smallest mask we can use is 8. Any differences
above the highest bit of pages are due to carrying, so by xnor'ing pfn
and end_pfn and then masking out the lower order bits based on pages, we
get 0xffffff00, where the first set bit is the mask we want to use.
Fixes: 6fe1010d6d ("vfio/type1: DMA unmap chunking")
Cc: stable@vger.kernel.org
Signed-off-by: David Stevens <stevensd@chromium.org>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20220401022430.1262215-1-stevensd@google.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20220410013533.3959168-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When enabling VMD and IOMMU scalable mode, the following kernel panic
call trace/kernel log is shown in Eagle Stream platform (Sapphire Rapids
CPU) during booting:
pci 0000:59:00.5: Adding to iommu group 42
...
vmd 0000:59:00.5: PCI host bridge to bus 10000:80
pci 10000:80:01.0: [8086:352a] type 01 class 0x060400
pci 10000:80:01.0: reg 0x10: [mem 0x00000000-0x0001ffff 64bit]
pci 10000:80:01.0: enabling Extended Tags
pci 10000:80:01.0: PME# supported from D0 D3hot D3cold
pci 10000:80:01.0: DMAR: Setup RID2PASID failed
pci 10000:80:01.0: Failed to add to iommu group 42: -16
pci 10000:80:03.0: [8086:352b] type 01 class 0x060400
pci 10000:80:03.0: reg 0x10: [mem 0x00000000-0x0001ffff 64bit]
pci 10000:80:03.0: enabling Extended Tags
pci 10000:80:03.0: PME# supported from D0 D3hot D3cold
------------[ cut here ]------------
kernel BUG at lib/list_debug.c:29!
invalid opcode: 0000 [#1] PREEMPT SMP NOPTI
CPU: 0 PID: 7 Comm: kworker/0:1 Not tainted 5.17.0-rc3+ #7
Hardware name: Lenovo ThinkSystem SR650V3/SB27A86647, BIOS ESE101Y-1.00 01/13/2022
Workqueue: events work_for_cpu_fn
RIP: 0010:__list_add_valid.cold+0x26/0x3f
Code: 9a 4a ab ff 4c 89 c1 48 c7 c7 40 0c d9 9e e8 b9 b1 fe ff 0f
0b 48 89 f2 4c 89 c1 48 89 fe 48 c7 c7 f0 0c d9 9e e8 a2 b1
fe ff <0f> 0b 48 89 d1 4c 89 c6 4c 89 ca 48 c7 c7 98 0c d9
9e e8 8b b1 fe
RSP: 0000:ff5ad434865b3a40 EFLAGS: 00010246
RAX: 0000000000000058 RBX: ff4d61160b74b880 RCX: ff4d61255e1fffa8
RDX: 0000000000000000 RSI: 00000000fffeffff RDI: ffffffff9fd34f20
RBP: ff4d611d8e245c00 R08: 0000000000000000 R09: ff5ad434865b3888
R10: ff5ad434865b3880 R11: ff4d61257fdc6fe8 R12: ff4d61160b74b8a0
R13: ff4d61160b74b8a0 R14: ff4d611d8e245c10 R15: ff4d611d8001ba70
FS: 0000000000000000(0000) GS:ff4d611d5ea00000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ff4d611fa1401000 CR3: 0000000aa0210001 CR4: 0000000000771ef0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe07f0 DR7: 0000000000000400
PKRU: 55555554
Call Trace:
<TASK>
intel_pasid_alloc_table+0x9c/0x1d0
dmar_insert_one_dev_info+0x423/0x540
? device_to_iommu+0x12d/0x2f0
intel_iommu_attach_device+0x116/0x290
__iommu_attach_device+0x1a/0x90
iommu_group_add_device+0x190/0x2c0
__iommu_probe_device+0x13e/0x250
iommu_probe_device+0x24/0x150
iommu_bus_notifier+0x69/0x90
blocking_notifier_call_chain+0x5a/0x80
device_add+0x3db/0x7b0
? arch_memremap_can_ram_remap+0x19/0x50
? memremap+0x75/0x140
pci_device_add+0x193/0x1d0
pci_scan_single_device+0xb9/0xf0
pci_scan_slot+0x4c/0x110
pci_scan_child_bus_extend+0x3a/0x290
vmd_enable_domain.constprop.0+0x63e/0x820
vmd_probe+0x163/0x190
local_pci_probe+0x42/0x80
work_for_cpu_fn+0x13/0x20
process_one_work+0x1e2/0x3b0
worker_thread+0x1c4/0x3a0
? rescuer_thread+0x370/0x370
kthread+0xc7/0xf0
? kthread_complete_and_exit+0x20/0x20
ret_from_fork+0x1f/0x30
</TASK>
Modules linked in:
---[ end trace 0000000000000000 ]---
...
Kernel panic - not syncing: Fatal exception
Kernel Offset: 0x1ca00000 from 0xffffffff81000000 (relocation range: 0xffffffff80000000-0xffffffffbfffffff)
---[ end Kernel panic - not syncing: Fatal exception ]---
The following 'lspci' output shows devices '10000:80:*' are subdevices of
the VMD device 0000:59:00.5:
$ lspci
...
0000:59:00.5 RAID bus controller: Intel Corporation Volume Management Device NVMe RAID Controller (rev 20)
...
10000:80:01.0 PCI bridge: Intel Corporation Device 352a (rev 03)
10000:80:03.0 PCI bridge: Intel Corporation Device 352b (rev 03)
10000:80:05.0 PCI bridge: Intel Corporation Device 352c (rev 03)
10000:80:07.0 PCI bridge: Intel Corporation Device 352d (rev 03)
10000:81:00.0 Non-Volatile memory controller: Intel Corporation NVMe Datacenter SSD [3DNAND, Beta Rock Controller]
10000:82:00.0 Non-Volatile memory controller: Intel Corporation NVMe Datacenter SSD [3DNAND, Beta Rock Controller]
The symptom 'list_add double add' is caused by the following failure
message:
pci 10000:80:01.0: DMAR: Setup RID2PASID failed
pci 10000:80:01.0: Failed to add to iommu group 42: -16
pci 10000:80:03.0: [8086:352b] type 01 class 0x060400
Device 10000:80:01.0 is the subdevice of the VMD device 0000:59:00.5,
so invoking intel_pasid_alloc_table() gets the pasid_table of the VMD
device 0000:59:00.5. Here is call path:
intel_pasid_alloc_table
pci_for_each_dma_alias
get_alias_pasid_table
search_pasid_table
pci_real_dma_dev() in pci_for_each_dma_alias() gets the real dma device
which is the VMD device 0000:59:00.5. However, pte of the VMD device
0000:59:00.5 has been configured during this message "pci 0000:59:00.5:
Adding to iommu group 42". So, the status -EBUSY is returned when
configuring pasid entry for device 10000:80:01.0.
It then invokes dmar_remove_one_dev_info() to release
'struct device_domain_info *' from iommu_devinfo_cache. But, the pasid
table is not released because of the following statement in
__dmar_remove_one_dev_info():
if (info->dev && !dev_is_real_dma_subdevice(info->dev)) {
...
intel_pasid_free_table(info->dev);
}
The subsequent dmar_insert_one_dev_info() operation of device
10000:80:03.0 allocates 'struct device_domain_info *' from
iommu_devinfo_cache. The allocated address is the same address that
is released previously for device 10000:80:01.0. Finally, invoking
device_attach_pasid_table() causes the issue.
`git bisect` points to the offending commit 474dd1c650 ("iommu/vt-d:
Fix clearing real DMA device's scalable-mode context entries"), which
releases the pasid table if the device is not the subdevice by
checking the returned status of dev_is_real_dma_subdevice().
Reverting the offending commit can work around the issue.
The solution is to prevent from allocating pasid table if those
devices are subdevices of the VMD device.
Fixes: 474dd1c650 ("iommu/vt-d: Fix clearing real DMA device's scalable-mode context entries")
Cc: stable@vger.kernel.org # v5.14+
Signed-off-by: Adrian Huang <ahuang12@lenovo.com>
Link: https://lore.kernel.org/r/20220216091307.703-1-adrianhuang0701@gmail.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20220221053348.262724-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
After commit e3beca48a4 ("irqdomain/treewide: Keep firmware node
unconditionally allocated"). For tear down scenario, fn is only freed
after fail to allocate ir_domain, though it also should be freed in case
dmar_enable_qi returns error.
Besides free fn, irq_domain and ir_msi_domain need to be removed as well
if intel_setup_irq_remapping fails to enable queued invalidation.
Improve the rewinding path by add out_free_ir_domain and out_free_fwnode
lables per Baolu's suggestion.
Fixes: e3beca48a4 ("irqdomain/treewide: Keep firmware node unconditionally allocated")
Suggested-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Guoqing Jiang <guoqing.jiang@linux.dev>
Link: https://lore.kernel.org/r/20220119063640.16864-1-guoqing.jiang@linux.dev
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20220128031002.2219155-3-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
page->freelist is for the use of slab. We already have the ability
to free a list of pages in the core mm, but it requires the use of a
list_head and for the pages to be chained together through page->lru.
Switch the Intel IOMMU and IOVA code over to using free_pages_list().
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
[rm: split from original patch, cosmetic tweaks, fix fq entries]
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/2115b560d9a0ce7cd4b948bd51a2b7bde8fdfd59.1639753638.git.robin.murphy@arm.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Remove dma_to_buf_pfn function, which is not used in the codebase.
This was pointed by clang with the following warning:
'dma_to_mm_pfn' [-Wunused-function]
static inline unsigned long dma_to_mm_pfn(unsigned long dma_pfn)
^
https://lore.kernel.org/r/YYhY7GqlrcTZlzuA@fedora
drivers/iommu/intel/iommu.c:136:29: warning: unused function
Signed-off-by: Maíra Canal <maira.canal@usp.br>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20211217083817.1745419-4-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The find.h APIs are designed to be used only on unsigned long arguments.
This can technically result in a over-read, but it is harmless in this
case. Regardless, fix it to avoid the warning seen under -Warray-bounds,
which we'd like to enable globally:
In file included from ./include/linux/bitmap.h:9,
from drivers/iommu/intel/iommu.c:17:
drivers/iommu/intel/iommu.c: In function 'domain_context_mapping_one':
./include/linux/find.h:119:37: warning: array subscript 'long unsigned int[0]' is partly outside array bounds of 'int[1]' [-Warray-bounds]
119 | unsigned long val = *addr & GENMASK(size - 1, 0);
| ^~~~~
drivers/iommu/intel/iommu.c:2115:18: note: while referencing 'max_pde'
2115 | int pds, max_pde;
| ^~~~~~~
Signed-off-by: Kees Cook <keescook@chromium.org>
Acked-by: Yury Norov <yury.norov@gmail.com>
Link: https://lore.kernel.org/r/20211215232432.2069605-1-keescook@chromium.org
Signed-off-by: Joerg Roedel <jroedel@suse.de>
When supporting only the .map and .unmap callbacks of iommu_ops,
the IOMMU driver can make assumptions about the size and alignment
used for mappings based on the driver provided pgsize_bitmap. VT-d
previously used essentially PAGE_MASK for this bitmap as any power
of two mapping was acceptably filled by native page sizes.
However, with the .map_pages and .unmap_pages interface we're now
getting page-size and count arguments. If we simply combine these
as (page-size * count) and make use of the previous map/unmap
functions internally, any size and alignment assumptions are very
different.
As an example, a given vfio device assignment VM will often create
a 4MB mapping at IOVA pfn [0x3fe00 - 0x401ff]. On a system that
does not support IOMMU super pages, the unmap_pages interface will
ask to unmap 1024 4KB pages at the base IOVA. dma_pte_clear_level()
will recurse down to level 2 of the page table where the first half
of the pfn range exactly matches the entire pte level. We clear the
pte, increment the pfn by the level size, but (oops) the next pte is
on a new page, so we exit the loop an pop back up a level. When we
then update the pfn based on that higher level, we seem to assume
that the previous pfn value was at the start of the level. In this
case the level size is 256K pfns, which we add to the base pfn and
get a results of 0x7fe00, which is clearly greater than 0x401ff,
so we're done. Meanwhile we never cleared the ptes for the remainder
of the range. When the VM remaps this range, we're overwriting valid
ptes and the VT-d driver complains loudly, as reported by the user
report linked below.
The fix for this seems relatively simple, if each iteration of the
loop in dma_pte_clear_level() is assumed to clear to the end of the
level pte page, then our next pfn should be calculated from level_pfn
rather than our working pfn.
Fixes: 3f34f12597 ("iommu/vt-d: Implement map/unmap_pages() iommu_ops callback")
Reported-by: Ajay Garg <ajaygargnsit@gmail.com>
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Tested-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com>
Link: https://lore.kernel.org/all/20211002124012.18186-1-ajaygargnsit@gmail.com/
Link: https://lore.kernel.org/r/163659074748.1617923.12716161410774184024.stgit@omen
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20211126135556.397932-3-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The __domain_mapping() always removes the pages in the range from
'iov_pfn' to 'end_pfn', but the 'end_pfn' is always the last pfn
of the range that the caller wants to map.
This would introduce too many duplicated removing and leads the
map operation take too long, for example:
Map iova=0x100000,nr_pages=0x7d61800
iov_pfn: 0x100000, end_pfn: 0x7e617ff
iov_pfn: 0x140000, end_pfn: 0x7e617ff
iov_pfn: 0x180000, end_pfn: 0x7e617ff
iov_pfn: 0x1c0000, end_pfn: 0x7e617ff
iov_pfn: 0x200000, end_pfn: 0x7e617ff
...
it takes about 50ms in total.
We can reduce the cost by recalculate the 'end_pfn' and limit it
to the boundary of the end of this pte page.
Map iova=0x100000,nr_pages=0x7d61800
iov_pfn: 0x100000, end_pfn: 0x13ffff
iov_pfn: 0x140000, end_pfn: 0x17ffff
iov_pfn: 0x180000, end_pfn: 0x1bffff
iov_pfn: 0x1c0000, end_pfn: 0x1fffff
iov_pfn: 0x200000, end_pfn: 0x23ffff
...
it only need 9ms now.
This also removes a meaningless BUG_ON() in __domain_mapping().
Signed-off-by: Longpeng(Mike) <longpeng2@huawei.com>
Tested-by: Liujunjie <liujunjie23@huawei.com>
Link: https://lore.kernel.org/r/20211008000433.1115-1-longpeng2@huawei.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20211014053839.727419-10-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
update_pasid() and its call chain are currently unused in the tree because
Thomas disabled the ENQCMD feature. The feature will be re-enabled shortly
using a different approach and update_pasid() and its call chain will not
be used in the new approach.
Remove the useless functions.
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Link: https://lore.kernel.org/r/20210920192349.2602141-1-fenghua.yu@intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20211014053839.727419-8-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The IOMMU VT-d implementation uses the first level for GPA->HPA translation
by default. Although both the first level and the second level could handle
the DMA translation, they're different in some way. For example, the second
level translation has separate controls for the Access/Dirty page tracking.
With the first level translation, there's no such control. On the other
hand, the second level translation has the page-level control for forcing
snoop, but the first level only has global control with pasid granularity.
This uses the second level for GPA->HPA translation so that we can provide
a consistent hardware interface for use cases like dirty page tracking for
live migration.
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20210926114535.923263-1-baolu.lu@linux.intel.com
Link: https://lore.kernel.org/r/20211014053839.727419-6-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
An iommu domain could be allocated and mapped before it's attached to any
device. This requires that in scalable mode, when the domain is allocated,
the format (FL or SL) of the page table must be determined. In order to
achieve this, the platform should support consistent SL or FL capabilities
on all IOMMU's. This adds a check for this and aborts IOMMU probing if it
doesn't meet this requirement.
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20210926114535.923263-1-baolu.lu@linux.intel.com
Link: https://lore.kernel.org/r/20211014053839.727419-5-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
When the dmar translation fault happens, the kernel prints a single line
fault reason with corresponding hexadecimal code defined in the Intel VT-d
specification.
Currently, when user wants to debug the translation fault in detail,
debugfs is used for dumping the dmar_translation_struct, which is not
available when the kernel failed to boot.
Dump the DMAR translation structure, pagewalk the IO page table and print
the page table entry when the fault happens.
This takes effect only when CONFIG_DMAR_DEBUG is enabled.
Signed-off-by: Kyung Min Park <kyung.min.park@intel.com>
Link: https://lore.kernel.org/r/20210815203845.31287-1-kyung.min.park@intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20211014053839.727419-3-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Handling of intel_iommu kernel command line option should return "true" to
indicate option is valid and so avoid logging it as unknown by the core
parsing code.
Also log unknown sub-options at the notice level to let user know of
potential typos or similar.
Reported-by: Eero Tamminen <eero.t.tamminen@intel.com>
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://lore.kernel.org/r/20210831112947.310080-1-tvrtko.ursulin@linux.intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20211014053839.727419-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
719a193356 ("iommu/vt-d: Tweak the description of a DMA fault") changed
the DMA fault reason from hex to decimal. It also added "0x" prefixes to
the PCI bus/device, e.g.,
- DMAR: [INTR-REMAP] Request device [00:00.5]
+ DMAR: [INTR-REMAP] Request device [0x00:0x00.5]
These no longer match dev_printk() and other similar messages in
dmar_match_pci_path() and dmar_acpi_insert_dev_scope().
Drop the "0x" prefixes from the bus and device addresses.
Fixes: 719a193356 ("iommu/vt-d: Tweak the description of a DMA fault")
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://lore.kernel.org/r/20210903193711.483999-1-helgaas@kernel.org
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210922054726.499110-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
pasid_mutex and dev->iommu->param->lock are held while unbinding mm is
flushing IO page fault workqueue and waiting for all page fault works to
finish. But an in-flight page fault work also need to hold the two locks
while unbinding mm are holding them and waiting for the work to finish.
This may cause an ABBA deadlock issue as shown below:
idxd 0000:00:0a.0: unbind PASID 2
======================================================
WARNING: possible circular locking dependency detected
5.14.0-rc7+ #549 Not tainted [ 186.615245] ----------
dsa_test/898 is trying to acquire lock:
ffff888100d854e8 (¶m->lock){+.+.}-{3:3}, at:
iopf_queue_flush_dev+0x29/0x60
but task is already holding lock:
ffffffff82b2f7c8 (pasid_mutex){+.+.}-{3:3}, at:
intel_svm_unbind+0x34/0x1e0
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #2 (pasid_mutex){+.+.}-{3:3}:
__mutex_lock+0x75/0x730
mutex_lock_nested+0x1b/0x20
intel_svm_page_response+0x8e/0x260
iommu_page_response+0x122/0x200
iopf_handle_group+0x1c2/0x240
process_one_work+0x2a5/0x5a0
worker_thread+0x55/0x400
kthread+0x13b/0x160
ret_from_fork+0x22/0x30
-> #1 (¶m->fault_param->lock){+.+.}-{3:3}:
__mutex_lock+0x75/0x730
mutex_lock_nested+0x1b/0x20
iommu_report_device_fault+0xc2/0x170
prq_event_thread+0x28a/0x580
irq_thread_fn+0x28/0x60
irq_thread+0xcf/0x180
kthread+0x13b/0x160
ret_from_fork+0x22/0x30
-> #0 (¶m->lock){+.+.}-{3:3}:
__lock_acquire+0x1134/0x1d60
lock_acquire+0xc6/0x2e0
__mutex_lock+0x75/0x730
mutex_lock_nested+0x1b/0x20
iopf_queue_flush_dev+0x29/0x60
intel_svm_drain_prq+0x127/0x210
intel_svm_unbind+0xc5/0x1e0
iommu_sva_unbind_device+0x62/0x80
idxd_cdev_release+0x15a/0x200 [idxd]
__fput+0x9c/0x250
____fput+0xe/0x10
task_work_run+0x64/0xa0
exit_to_user_mode_prepare+0x227/0x230
syscall_exit_to_user_mode+0x2c/0x60
do_syscall_64+0x48/0x90
entry_SYSCALL_64_after_hwframe+0x44/0xae
other info that might help us debug this:
Chain exists of:
¶m->lock --> ¶m->fault_param->lock --> pasid_mutex
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(pasid_mutex);
lock(¶m->fault_param->lock);
lock(pasid_mutex);
lock(¶m->lock);
*** DEADLOCK ***
2 locks held by dsa_test/898:
#0: ffff888100cc1cc0 (&group->mutex){+.+.}-{3:3}, at:
iommu_sva_unbind_device+0x53/0x80
#1: ffffffff82b2f7c8 (pasid_mutex){+.+.}-{3:3}, at:
intel_svm_unbind+0x34/0x1e0
stack backtrace:
CPU: 2 PID: 898 Comm: dsa_test Not tainted 5.14.0-rc7+ #549
Hardware name: Intel Corporation Kabylake Client platform/KBL S
DDR4 UD IMM CRB, BIOS KBLSE2R1.R00.X050.P01.1608011715 08/01/2016
Call Trace:
dump_stack_lvl+0x5b/0x74
dump_stack+0x10/0x12
print_circular_bug.cold+0x13d/0x142
check_noncircular+0xf1/0x110
__lock_acquire+0x1134/0x1d60
lock_acquire+0xc6/0x2e0
? iopf_queue_flush_dev+0x29/0x60
? pci_mmcfg_read+0xde/0x240
__mutex_lock+0x75/0x730
? iopf_queue_flush_dev+0x29/0x60
? pci_mmcfg_read+0xfd/0x240
? iopf_queue_flush_dev+0x29/0x60
mutex_lock_nested+0x1b/0x20
iopf_queue_flush_dev+0x29/0x60
intel_svm_drain_prq+0x127/0x210
? intel_pasid_tear_down_entry+0x22e/0x240
intel_svm_unbind+0xc5/0x1e0
iommu_sva_unbind_device+0x62/0x80
idxd_cdev_release+0x15a/0x200
pasid_mutex protects pasid and svm data mapping data. It's unnecessary
to hold pasid_mutex while flushing the workqueue. To fix the deadlock
issue, unlock pasid_pasid during flushing the workqueue to allow the works
to be handled.
Fixes: d5b9e4bfe0 ("iommu/vt-d: Report prq to io-pgfault framework")
Reported-and-tested-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: https://lore.kernel.org/r/20210826215918.4073446-1-fenghua.yu@intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210828070622.2437559-3-baolu.lu@linux.intel.com
[joro: Removed timing information from kernel log messages]
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The mm->pasid will be used in intel_svm_free_pasid() after load_pasid()
during unbinding mm. Clearing it in load_pasid() will cause PASID cannot
be freed in intel_svm_free_pasid().
Additionally mm->pasid was updated already before load_pasid() during pasid
allocation. No need to update it again in load_pasid() during binding mm.
Don't update mm->pasid to avoid the issues in both binding mm and unbinding
mm.
Fixes: 4048377414 ("iommu/vt-d: Use iommu_sva_alloc(free)_pasid() helpers")
Reported-and-tested-by: Dave Jiang <dave.jiang@intel.com>
Co-developed-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: https://lore.kernel.org/r/20210826215918.4073446-1-fenghua.yu@intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210828070622.2437559-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Kernel doc validator is unhappy with the following
.../perf.c:16: warning: Function parameter or member 'latency_lock' not described in 'DEFINE_SPINLOCK'
.../perf.c:16: warning: expecting prototype for perf.c(). Prototype was for DEFINE_SPINLOCK() instead
Drop kernel doc annotation since the top comment is not in the required format.
Signed-off-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/r/20210729163538.40101-1-andriy.shevchenko@linux.intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210818134852.1847070-8-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The minimum per-IOMMU PRQ queue size is one 4K page, this is more entries
than the hardcoded limit of 32 in the current VT-d code. Some devices can
support up to 512 outstanding PRQs but underutilized by this limit of 32.
Although, 32 gives some rough fairness when multiple devices share the same
IOMMU PRQ queue, but far from optimal for customized use case. This extends
the per-IOMMU PRQ queue size to four 4K pages and let the devices have as
many outstanding page requests as they can.
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210720013856.4143880-1-baolu.lu@linux.intel.com
Link: https://lore.kernel.org/r/20210818134852.1847070-7-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
We preset the access and dirty bits for IOVA over first level usage only
for the kernel DMA (i.e., when domain type is IOMMU_DOMAIN_DMA). We should
also preset the FL A/D for user space DMA usage. The idea is that even the
user space A/D bit memory write is unnecessary. We should avoid it to
minimize the overhead.
Suggested-by: Sanjay Kumar <sanjay.k.kumar@intel.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210720013856.4143880-1-baolu.lu@linux.intel.com
Link: https://lore.kernel.org/r/20210818134852.1847070-6-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The commit 8950dcd83a ("iommu/vt-d: Leave scalable mode default off")
leaves the scalable mode default off and end users could turn it on with
"intel_iommu=sm_on". Using the Intel IOMMU scalable mode for kernel DMA,
user-level device access and Shared Virtual Address have been enabled.
This enables the scalable mode by default if the hardware advertises the
support and adds kernel options of "intel_iommu=sm_on/sm_off" for end
users to configure it through the kernel parameters.
Suggested-by: Ashok Raj <ashok.raj@intel.com>
Suggested-by: Sanjay Kumar <sanjay.k.kumar@intel.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Cc: Kevin Tian <kevin.tian@intel.com>
Link: https://lore.kernel.org/r/20210720013856.4143880-1-baolu.lu@linux.intel.com
Link: https://lore.kernel.org/r/20210818134852.1847070-5-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
In preparation for the strict vs. non-strict decision for DMA domains
to be expressed in the domain type, make sure we expose our flush queue
awareness by accepting the new domain type, and test the specific
feature flag where we want to identify DMA domains in general. The DMA
ops reset/setup can simply be made unconditional, since iommu-dma
already knows only to touch DMA domains.
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Robin Murphy <robin.murphy@arm.com>
Link: https://lore.kernel.org/r/31a8ef868d593a2f3826a6a120edee81815375a7.1628682049.git.robin.murphy@arm.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
This fixes improper iotlb invalidation in intel_pasid_tear_down_entry().
When a PASID was used as nested mode, released and reused, the following
error message will appear:
[ 180.187556] Unexpected page request in Privilege Mode
[ 180.187565] Unexpected page request in Privilege Mode
[ 180.279933] Unexpected page request in Privilege Mode
[ 180.279937] Unexpected page request in Privilege Mode
Per chapter 6.5.3.3 of VT-d spec 3.3, when tear down a pasid entry, the
software should use Domain selective IOTLB flush if the PGTT of the pasid
entry is SL only or Nested, while for the pasid entries whose PGTT is FL
only or PT using PASID-based IOTLB flush is enough.
Fixes: 2cd1311a26 ("iommu/vt-d: Add set domain DOMAIN_ATTR_NESTING attr")
Signed-off-by: Kumar Sanjay K <sanjay.k.kumar@intel.com>
Signed-off-by: Liu Yi L <yi.l.liu@intel.com>
Tested-by: Yi Sun <yi.y.sun@intel.com>
Link: https://lore.kernel.org/r/20210817042425.1784279-1-yi.l.liu@intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210817124321.1517985-3-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
A PASID reference is increased whenever a device is bound to an mm (and
its PASID) successfully (i.e. the device's sdev user count is increased).
But the reference is not dropped every time the device is unbound
successfully from the mm (i.e. the device's sdev user count is decreased).
The reference is dropped only once by calling intel_svm_free_pasid() when
there isn't any device bound to the mm. intel_svm_free_pasid() drops the
reference and only frees the PASID on zero reference.
Fix the issue by dropping the PASID reference and freeing the PASID when
no reference on successful unbinding the device by calling
intel_svm_free_pasid() .
Fixes: 4048377414 ("iommu/vt-d: Use iommu_sva_alloc(free)_pasid() helpers")
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Link: https://lore.kernel.org/r/20210813181345.1870742-1-fenghua.yu@intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210817124321.1517985-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
As the Intel VT-d driver has switched to use the iommu_ops.map_pages()
callback, multiple pages of the same size will be mapped in a call.
There's no need to put the clflush'es in iotlb_sync_map() callback.
Move them back into __domain_mapping() to simplify the code.
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210720020615.4144323-4-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Implement the map_pages() and unmap_pages() callback for the Intel IOMMU
driver to allow calls from iommu core to map and unmap multiple pages of
the same size in one call. With map/unmap_pages() implemented, the prior
map/unmap callbacks are deprecated.
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210720020615.4144323-3-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The pgsize bitmap is used to advertise the page sizes our hardware supports
to the IOMMU core, which will then use this information to split physically
contiguous memory regions it is mapping into page sizes that we support.
Traditionally the IOMMU core just handed us the mappings directly, after
making sure the size is an order of a 4KiB page and that the mapping has
natural alignment. To retain this behavior, we currently advertise that we
support all page sizes that are an order of 4KiB.
We are about to utilize the new IOMMU map/unmap_pages APIs. We could change
this to advertise the real page sizes we support.
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210720020615.4144323-2-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
We only ever now set strict mode enabled in iommu_set_dma_strict(), so
just remove the argument.
Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/1626088340-5838-7-git-send-email-john.garry@huawei.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Make IOMMU_DEFAULT_LAZY default for when INTEL_IOMMU config is set,
as is current behaviour.
Also delete global flag intel_iommu_strict:
- In intel_iommu_setup(), call iommu_set_dma_strict(true) directly. Also
remove the print, as iommu_subsys_init() prints the mode and we have
already marked this param as deprecated.
- For cap_caching_mode() check in intel_iommu_setup(), call
iommu_set_dma_strict(true) directly; also reword the accompanying print
with a level downgrade and also add the missing '\n'.
- For Ironlake GPU, again call iommu_set_dma_strict(true) directly and
keep the accompanying print.
[jpg: Remove intel_iommu_strict]
Signed-off-by: Zhen Lei <thunder.leizhen@huawei.com>
Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/1626088340-5838-5-git-send-email-john.garry@huawei.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Now that the x86 drivers support iommu.strict, deprecate the custom
methods.
Signed-off-by: John Garry <john.garry@huawei.com>
Acked-by: Robin Murphy <robin.murphy@arm.com>
Reviewed-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/1626088340-5838-2-git-send-email-john.garry@huawei.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The commit 2b0140c696 ("iommu/vt-d: Use pci_real_dma_dev() for mapping")
fixes an issue of "sub-device is removed where the context entry is cleared
for all aliases". But this commit didn't consider the PASID entry and PASID
table in VT-d scalable mode. This fix increases the coverage of scalable
mode.
Suggested-by: Sanjay Kumar <sanjay.k.kumar@intel.com>
Fixes: 8038bdb855 ("iommu/vt-d: Only clear real DMA device's context entries")
Fixes: 2b0140c696 ("iommu/vt-d: Use pci_real_dma_dev() for mapping")
Cc: stable@vger.kernel.org # v5.6+
Cc: Jon Derrick <jonathan.derrick@intel.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210712071712.3416949-1-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
This fixes a bug in context cache clear operation. The code was not
following the correct invalidation flow. A global device TLB invalidation
should be added after the IOTLB invalidation. At the same time, it
uses the domain ID from the context entry. But in scalable mode, the
domain ID is in PASID table entry, not context entry.
Fixes: 7373a8cc38 ("iommu/vt-d: Setup context and enable RID2PASID support")
Cc: stable@vger.kernel.org # v5.0+
Signed-off-by: Sanjay Kumar <sanjay.k.kumar@intel.com>
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Link: https://lore.kernel.org/r/20210712071315.3416543-1-baolu.lu@linux.intel.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Passing a 64-bit address width to iommu_setup_dma_ops() is valid on
virtual platforms, but isn't currently possible. The overflow check in
iommu_dma_init_domain() prevents this even when @dma_base isn't 0. Pass
a limit address instead of a size, so callers don't have to fake a size
to work around the check.
The base and limit parameters are being phased out, because:
* they are redundant for x86 callers. dma-iommu already reserves the
first page, and the upper limit is already in domain->geometry.
* they can now be obtained from dev->dma_range_map on Arm.
But removing them on Arm isn't completely straightforward so is left for
future work. As an intermediate step, simplify the x86 callers by
passing dummy limits.
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Reviewed-by: Robin Murphy <robin.murphy@arm.com>
Link: https://lore.kernel.org/r/20210618152059.1194210-5-jean-philippe@linaro.org
Signed-off-by: Joerg Roedel <jroedel@suse.de>
The assignment of iommu from info->iommu occurs before info is null checked
hence leading to a potential null pointer dereference issue. Fix this by
assigning iommu and checking if iommu is null after null checking info.
Addresses-Coverity: ("Dereference before null check")
Fixes: 4c82b88696 ("iommu/vt-d: Allocate/register iopf queue for sva devices")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Link: https://lore.kernel.org/r/20210611135024.32781-1-colin.king@canonical.com
Signed-off-by: Joerg Roedel <jroedel@suse.de>
A recent commit broke the build on 32-bit x86. The linker throws these
messages:
ld: drivers/iommu/intel/perf.o: in function `dmar_latency_snapshot':
perf.c:(.text+0x40c): undefined reference to `__udivdi3'
ld: perf.c:(.text+0x458): undefined reference to `__udivdi3'
The reason are the 64-bit divides in dmar_latency_snapshot(). Use the
div_u64() helper function for those.
Fixes: 55ee5e67a5 ("iommu/vt-d: Add common code for dmar latency performance monitors")
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Link: https://lore.kernel.org/r/20210610083120.29224-1-joro@8bytes.org