[ Upstream commit e5f72cb9ce ]
After ebe091ad81 ("PCI: Use pbus_select_window_for_type() during IO
window sizing") and ae88d0b9c5 ("PCI: Use pbus_select_window_for_type()
during mem window sizing"), many bridge windows can't get resources
assigned:
pci 0006:05:00.0: bridge window [??? 0x00001000-0x00001fff flags 0x20080000]: can't assign; no space
pci 0006:05:00.0: bridge window [??? 0x00001000-0x00001fff flags 0x20080000]: failed to assign
Those commits replace find_bus_resource_of_type() with
pbus_select_window_for_type(), and the latter lacks resource type
validation.
Add the resource type validation back to pbus_select_window_for_type() to
match the original behavior.
Fixes: 74afce3dfc ("PCI: Add bridge window selection functions")
Link: https://bugzilla.kernel.org/show_bug.cgi?id=221072
Signed-off-by: Kai-Heng Feng <kaihengf@nvidia.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://patch.msgid.link/20260210142058.82701-1-kaihengf@nvidia.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 5907a90551 ]
12d8:b404 is apparently another PCI ID for Pericom PI7C9X2G404 (as
identified by the chip silkscreen and lspci).
It is also affected by the PI7C9X2G errata (e.g. a network card attached
to it fails under load when P2P Redirect Request is enabled), so apply
the same quirk to this PCI ID too.
PCI bridge [0604]: Pericom Semiconductor PI7C9X2G404 EV/SV PCIe2 4-Port/4-Lane Packet Switch [12d8:b404] (rev 01)
Fixes: acd61ffb2f ("PCI: Add ACS quirk for Pericom PI7C9X2G switches")
Closes: https://lore.kernel.org/all/a1d926f0-4cb5-4877-a4df-617902648d80@green-communications.fr/
Signed-off-by: Nicolas Cavallari <nicolas.cavallari@green-communications.fr>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20260119160915.26456-1-nicolas.cavallari@green-communications.fr
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9abf79c8d7 ]
Previously program_hpx_type2() applied PCIe settings unconditionally,
which could incorrectly change bits like Extended Tag Field Enable and
Enable Relaxed Ordering.
When _HPX was added to ACPI r3.0, the intent of the PCIe Setting
Record (Type 2) in sec 6.2.7.3 was to configure AER registers when the
OS does not own the AER Capability:
The PCI Express setting record contains ... [the AER] Uncorrectable
Error Mask, Uncorrectable Error Severity, Correctable Error Mask
... to be used when configuring registers in the Advanced Error
Reporting Extended Capability Structure ...
OSPM [1] will only evaluate _HPX with Setting Record – Type 2 if
OSPM is not controlling the PCI Express Advanced Error Reporting
capability.
ACPI r3.0b, sec 6.2.7.3, added more AER registers, including registers
in the PCIe Capability with AER-related bits, and the restriction that
the OS use this only when it owns PCIe native hotplug:
... when configuring PCI Express registers in the Advanced Error
Reporting Extended Capability Structure *or PCI Express Capability
Structure* ...
An OS that has assumed ownership of native hot plug but does not
... have ownership of the AER register set must use ... the Type 2
record to program the AER registers ...
However, since the Type 2 record also includes register bits that
have functions other than AER, the OS must ignore values ... that
are not applicable.
Restrict program_hpx_type2() to only the intended purpose:
- Apply settings only when OS owns PCIe native hotplug but not AER,
- Only touch the AER-related bits (Error Reporting Enables) in Device
Control
- Don't touch Link Control at all, since nothing there seems AER-related,
but log _HPX settings for debugging purposes
Note that Read Completion Boundary is now configured elsewhere, since it is
unrelated to _HPX.
[1] Operating System-directed configuration and Power Management
Fixes: 40abb96c51 ("[PATCH] pciehp: Fix programming hotplug parameters")
Signed-off-by: Håkon Bugge <haakon.bugge@oracle.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20260129175237.727059-3-haakon.bugge@oracle.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1a6845aaa6 ]
Commit e42010d820 ("PCI: Set Read Completion Boundary to 128 iff Root
Port supports it (_HPX)") worked around a bogus _HPX type 2 record, which
caused program_hpx_type2() to set the RCB in an endpoint even though the
Root Port did not have the RCB bit set.
e42010d820 fixed that by setting the RCB in the endpoint only when it was
set in the Root Port.
In retrospect, program_hpx_type2() is intended for AER-related settings,
and the RCB should be configured elsewhere so it doesn't depend on the
presence or contents of an _HPX record.
Explicitly program the RCB from pci_configure_device() so it matches the
Root Port's RCB. The Root Port may not be visible to virtualized guests;
in that case, leave RCB alone.
Fixes: e42010d820 ("PCI: Set Read Completion Boundary to 128 iff Root Port supports it (_HPX)")
Signed-off-by: Håkon Bugge <haakon.bugge@oracle.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20260129175237.727059-2-haakon.bugge@oracle.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f7245901de ]
of_pci_bus_find_domain_nr() allows its parent parameter to be NULL but
of_pci_bus_release_domain_nr() (that undoes its effect) doesn't -- that
means it's going to blow up while calling of_get_pci_domain_nr() if the
parent parameter indeed happens to be NULL. Add the missing NULL check.
Found by Linux Verification Center (linuxtesting.org) with the Svace static
analysis tool.
Fixes: c14f7ccc9f ("PCI: Assign PCI domain IDs by ida_alloc()")
Signed-off-by: Sergey Shtylyov <s.shtylyov@auroraos.dev>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20260127203944.28588-1-s.shtylyov@auroraos.dev
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f909e3ee3e ]
calculate_memsize() applies lower bound to the resource size before
aligning the resource size making it impossible to shrink bridge window
resources. I've not found any justification for this lower bound and
nothing indicated it was to work around some HW issue.
Prior to the commit 3baeae3603 ("PCI: Use pci_release_resource() instead
of release_resource()"), releasing a bridge window during BAR resize
resulted in clearing start and end address of the resource. Clearing
addresses destroys the resource size as a side-effect, therefore nullifying
the effect of the old size lower bound.
After the commit 3baeae3603 ("PCI: Use pci_release_resource() instead of
release_resource()"), BAR resize uses the aligned old size, which results
in exceeding what fits into the parent window in some cases:
xe 0030:03:00.0: [drm] Attempting to resize bar from 256MiB -> 16384MiB
xe 0030:03:00.0: BAR 0 [mem 0x620c000000000-0x620c000ffffff 64bit]: releasing
xe 0030:03:00.0: BAR 2 [mem 0x6200000000000-0x620000fffffff 64bit pref]: releasing
pci 0030:02:01.0: bridge window [mem 0x6200000000000-0x620001fffffff 64bit pref]: releasing
pci 0030:01:00.0: bridge window [mem 0x6200000000000-0x6203fbff0ffff 64bit pref]: releasing
pci 0030:00:00.0: bridge window [mem 0x6200000000000-0x6203fbff0ffff 64bit pref]: was not released (still contains assigned resources)
pci 0030:00:00.0: Assigned bridge window [mem 0x6200000000000-0x6203fbff0ffff 64bit pref] to [bus 01-04] free space at [mem 0x6200400000000-0x62007ffffffff 64bit pref]
pci 0030:00:00.0: Assigned bridge window [mem 0x6200000000000-0x6203fbff0ffff 64bit pref] to [bus 01-04] cannot fit 0x4000000000 required for 0030:01:00.0 bridging to [bus 02-04]
The old size of 0x6200000000000-0x6203fbff0ffff resource was used as the
lower bound which results in 0x4000000000 size request due to alignment.
That exceeds what can fit into the parent window.
Since the lower bound never even was enforced fully because the resource
addresses were cleared when the bridge window is released, remove the
old_size lower bound entirely and trust the calculated bridge window size
is enough.
This same problem may occur on io window side but seems less likely to
cause issues due to general difference in alignment. Removing the lower
bound may have other unforeseen consequences in case of io window so it's
better to leave it as -next material if no problem is reported related to
io window sizing (BAR resize shouldn't touch io windows anyway).
Fixes: 3baeae3603 ("PCI: Use pci_release_resource() instead of release_resource()")
Reported-by: Simon Richter <Simon.Richter@hogyros.de>
Link: https://lore.kernel.org/r/f9a8c975-f5d3-4dd2-988e-4371a1433a60@hogyros.de/
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20251219174036.16738-6-ilpo.jarvinen@linux.intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3958bf16e2 ]
New way to calculate the bridge window head alignment produces tight-fit,
that is, it does not leave any gaps between the resources. Similarly,
relaxed tail alignment does not leave extra tail room.
Start to use bridge window calculation that does not over-estimate the size
of the required window.
pbus_upstream_space_available() can be removed.
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Malte Schröder <malte+lkml@tnxip.de>
Link: https://patch.msgid.link/20251219174036.16738-4-ilpo.jarvinen@linux.intel.com
Stable-dep-of: f909e3ee3e ("PCI: Remove old_size limit from bridge window sizing")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit bc75c8e507 ]
The calculation of bridge window head alignment is done by
calculate_mem_align() [*]. With the default bridge window alignment, it
is used for both head and tail alignment.
The selected head alignment does not always result in tight-fitting
resources (gap at d4f00000-d4ffffff):
d4800000-dbffffff : PCI Bus 0000:06
d4800000-d48fffff : PCI Bus 0000:07
d4800000-d4803fff : 0000:07:00.0
d4800000-d4803fff : nvme
d4900000-d49fffff : PCI Bus 0000:0a
d4900000-d490ffff : 0000:0a:00.0
d4900000-d490ffff : r8169
d4910000-d4913fff : 0000:0a:00.0
d4a00000-d4cfffff : PCI Bus 0000:0b
d4a00000-d4bfffff : 0000:0b:00.0
d4a00000-d4bfffff : 0000:0b:00.0
d4c00000-d4c07fff : 0000:0b:00.0
d4d00000-d4dfffff : PCI Bus 0000:15
d4d00000-d4d07fff : 0000:15:00.0
d4d00000-d4d07fff : xhci-hcd
d4e00000-d4efffff : PCI Bus 0000:16
d4e00000-d4e7ffff : 0000:16:00.0
d4e80000-d4e803ff : 0000:16:00.0
d4e80000-d4e803ff : ahci
d5000000-dbffffff : PCI Bus 0000:0c
This has not caused problems (for years) with the default bridge window
tail alignment that grossly over-estimates the required tail alignment
leaving more tail room than necessary. With the introduction of relaxed
tail alignment that leaves no extra tail room whatsoever, any gaps will
immediately turn into assignment failures.
Introduce head alignment calculation that ensures no gaps are left and
apply the new approach when using relaxed alignment. We may want to
consider using it for the normal alignment eventually, but as the first
step, solve only the problem with the relaxed tail alignment.
([*] I don't understand the algorithm in calculate_mem_align().)
Link: https://git.kernel.org/history/history/c/5d0a8965aea9 ("[PATCH] 2.5.14: New PCI allocation code (alpha, arm, parisc) [2/2]")
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=220775
Reported-by: Malte Schröder <malte+lkml@tnxip.de>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Malte Schröder <malte+lkml@tnxip.de>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20251219174036.16738-3-ilpo.jarvinen@linux.intel.com
Stable-dep-of: f909e3ee3e ("PCI: Remove old_size limit from bridge window sizing")
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 959ac08a2c ]
Per PCIe r7.0, sec 2.2.6.2.1 and 7.5.3.4, a Requester may not use 8-bit Tags
unless its Extended Tag Field Enable is set, but all Receivers/Completers
must handle 8-bit Tags correctly regardless of their Extended Tag Field
Enable.
Some devices do not handle 8-bit Tags as Completers, so add a quirk for
them. If we find such a device, we disable Extended Tags for the entire
hierarchy to make peer-to-peer DMA possible.
The 3ware 9650SA seems to have issues with handling 8-bit tags. Mark it as
broken.
This fixes PCI Parity Errors like :
3w-9xxx: scsi0: ERROR: (0x06:0x000C): PCI Parity Error: clearing.
3w-9xxx: scsi0: ERROR: (0x06:0x000D): PCI Abort: clearing.
3w-9xxx: scsi0: ERROR: (0x06:0x000E): Controller Queue Error: clearing.
3w-9xxx: scsi0: ERROR: (0x06:0x0010): Microcontroller Error: clearing.
Fixes: 60db3a4d8c ("PCI: Enable PCIe Extended Tags if supported")
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=202425
Signed-off-by: Jörg Wedekind <joerg@wedekind.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20260119143114.21948-1-joerg@wedekind.de
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 83014d82a1 ]
When mapping a p2pdma page the page reference count is initialised to 1
prior to calling vm_insert_page(). This is to avoid vm_insert_page()
warning if the page refcount is zero. Prior to setting the page count there
is a check to ensure the page is currently free (ie. has a zero reference
count).
However vm_insert_page() can fail. In this case the pages are freed back to
the genalloc pool, but that does not reset the page refcount. So a future
allocation of the same page will see the elevated page refcount from the
previous set_page_count() call triggering the VM_WARN_ON_ONCE_PAGE checking
that the page is free.
Fix this by resetting the page refcount to zero using set_page_count().
Note that put_page() is not used because that would result in freeing the
page twice due to implicitly calling p2pdma_folio_free().
Fixes: b7e2823787 ("mm/mm_init: move p2pdma page refcount initialisation to p2pdma")
Signed-off-by: Alistair Popple <apopple@nvidia.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Acked-by: Balbir Singh <balbirs@nvidia.com>
Link: https://patch.msgid.link/20260112005440.998543-1-apopple@nvidia.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 62171369cf ]
In pcie_ptm_create_debugfs(), if devm_kasprintf() fails after successfully
allocating ptm_debugfs with kzalloc(), the function returns without freeing
the allocated memory, resulting in a memory leak.
Free ptm_debugfs before returning in the devm_kasprintf() error path and in
pcie_ptm_destroy_debugfs().
Fixes: 132833405e ("PCI: Add debugfs support for exposing PTM context")
Signed-off-by: Aadityarangan Shridhar Iyengar <adiyenga@cisco.com>
[bhelgaas: squash additional fix from Mani:
https://lore.kernel.org/r/pdp4xc4d5ee3e547mmdro5riui3mclduqdl7j6iclfbozo2a4c@7m3qdm6yrhuv]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Manivannan Sadhasivam <mani@kernel.org>
Link: https://patch.msgid.link/20260111163650.33168-1-adiyenga@cisco.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 01464a3fdf ]
pcie_port_probe_service() unconditionally calls get_device() (unless it
fails). So drop that reference also unconditionally as it's fine for a
PCIe driver to not have a remove callback.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@baylibre.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Reviewed-by: Jonathan Cameron <jonathan.cameron@huawei.com>
Link: https://patch.msgid.link/e1c68c3b3f1af8427e98ca5e2c79f8bf0ebe2ce4.1764688034.git.u.kleine-koenig@baylibre.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 613f3255a3 ]
Sophgo 2044 Root Ports advertise L0 and L1 capabilities without supporting
them. Since commit f3ac2ff148 ("PCI/ASPM: Enable all ClockPM and ASPM
states for devicetree platforms") force enabled ASPM on all device tree
platforms, the issue became evident and the SG2044 Root Port started
breaking.
Hence, disable the L0s and L1 capabilities in the LINKCAP register for the
SG2044 Root Ports, so that these states won't get enabled.
Fixes: 467d9c0348 ("PCI: dwc: Add Sophgo SG2044 PCIe controller driver in Root Complex mode")
Signed-off-by: Inochi Amaoto <inochiama@gmail.com>
[mani: reworded description and corrected fixes tag]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Tested-by: Han Gao <gaohan@iscas.ac.cn>
Link: https://patch.msgid.link/20260109040756.731169-1-inochiama@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 73711730a1 ]
The bit for enabling extended tags is Reserved and Preserved (RsvdP) for
VFs, according to PCIe r7.0 section 7.5.3.4 table 7.21. Hence, bail out
early from pci_configure_extended_tags() if the device is a VF.
Otherwise, we may see incorrect log messages such as:
kernel: pci 0000:af:00.2: enabling Extended Tags
(af:00.2 is a VF)
Fixes: 60db3a4d8c ("PCI: Enable PCIe Extended Tags if supported")
Signed-off-by: Håkon Bugge <haakon.bugge@oracle.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Link: https://patch.msgid.link/20251112095442.1913258-1-haakon.bugge@oracle.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit cb500023a7 ]
Commit b7e2823787 has already changed the initial page refcount of
p2pdma page from one to zero, however, in p2pmem_alloc_mmap() it uses
"VM_WARN_ON_ONCE_PAGE(!page_ref_count(page))" to assert the initial page
refcount should not be zero and the following will be reported when
CONFIG_DEBUG_VM is enabled:
page: refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x380400000
flags: 0x20000000002000(reserved|node=0|zone=4)
raw: 0020000000002000 ff1100015e3ab440 0000000000000000 0000000000000000
raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
page dumped because: VM_WARN_ON_ONCE_PAGE(!page_ref_count(page))
------------[ cut here ]------------
WARNING: CPU: 5 PID: 449 at drivers/pci/p2pdma.c:240 p2pmem_alloc_mmap+0x83a/0xa60
Fix by using "page_ref_count(page)" as the assertion condition.
Fixes: b7e2823787 ("mm/mm_init: move p2pdma page refcount initialisation to p2pdma")
Signed-off-by: Hou Tao <houtao1@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Link: https://patch.msgid.link/20251220040446.274991-3-houtao@huaweicloud.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6220694c52 ]
When vm_insert_page() fails in p2pmem_alloc_mmap(), p2pmem_alloc_mmap()
doesn't invoke percpu_ref_put() to free the per-CPU ref of pgmap acquired
after gen_pool_alloc_owner(), and memunmap_pages() will hang forever when
trying to remove the PCI device.
Fix it by adding the missed percpu_ref_put().
Fixes: 7e9c7ef83d ("PCI/P2PDMA: Allow userspace VMA allocations through sysfs")
Signed-off-by: Hou Tao <houtao1@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Alistair Popple <apopple@nvidia.com>
Link: https://patch.msgid.link/20251220040446.274991-2-houtao@huaweicloud.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 4d98208450 ]
When transitioning to D3cold, __pci_set_power_state() first transitions to
D3hot. If the device was already in D3hot, this adds excess work:
(a) read/modify/write PMCSR; and
(b) excess delay (pci_dev_d3_sleep()).
For (b), we already performed the necessary delay on the previous D3hot
entry; this was extra noticeable when evaluating runtime PM transition
latency.
Check whether we're already in the target state before continuing.
Note that __pci_set_power_state() already does this same check for other
state transitions, but D3cold is special because __pci_set_power_state()
converts it to D3hot for the purposes of PMCSR.
This seems to be an oversight in commit 0aacdc9574 ("PCI/PM: Clean up
pci_set_low_power_state()").
Fixes: 0aacdc9574 ("PCI/PM: Clean up pci_set_low_power_state()")
Signed-off-by: Brian Norris <briannorris@google.com>
Signed-off-by: Brian Norris <briannorris@chromium.org>
[bhelgaas: reverse test to match other "dev->current_state == state" cases]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20251003154008.1.I7a21c240b30062c66471329567a96dceb6274358@changeid
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f42b3c053b ]
In xilinx_pcie_init_irq_domain(), if xilinx_allocate_msi_domains() fails
after pcie->leg_domain has been successfully created via
irq_domain_create_linear(), the function returns directly without cleaning
up the allocated IRQ domain, resulting in a resource leak. In
xilinx_free_msi_domains(), pcie->leg_domain is also neglected.
Add irq_domain_remove() call in the error path to properly release the
IRQ domain before returning the error. Also rename
xilinx_free_msi_domains() to xilinx_free_irq_domains() and add the release
of pcie->leg_domain to it.
Fixes: 313b64c3ae ("PCI: xilinx: Convert to MSI domains")
Suggested-by: Manivannan Sadhasivam <mani@kernel.org>
Signed-off-by: Haotian Zhang <vulab@iscas.ac.cn>
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Link: https://patch.msgid.link/20251219021615.965-1-vulab@iscas.ac.cn
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 7f0cdcddf8 ]
In mtk_pcie_init_irq_domain(), if mtk_pcie_allocate_msi_domains()
fails after port->irq_domain has been successfully created via
irq_domain_create_linear(), the function returns directly without
cleaning up the allocated IRQ domain, resulting in a resource leak.
Add irq_domain_remove() call in the error path to properly release the
INTx IRQ domain before returning the error.
Fixes: 43e6409db6 ("PCI: mediatek: Add MSI support for MT2712 and MT7622")
Signed-off-by: Haotian Zhang <vulab@iscas.ac.cn>
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Link: https://patch.msgid.link/20251119023308.476-1-vulab@iscas.ac.cn
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 7c5c7d06bd upstream.
The asynchronous creation of sub-groups by a delayed work could lead to a
NULL pointer dereference when the driver directory is removed before the
work completes.
The crash can be easily reproduced with the following commands:
# cd /sys/kernel/config/pci_ep/functions/pci_epf_test
# for i in {1..20}; do mkdir test && rmdir test; done
BUG: kernel NULL pointer dereference, address: 0000000000000088
...
Call Trace:
configfs_register_group+0x3d/0x190
pci_epf_cfs_work+0x41/0x110
process_one_work+0x18f/0x350
worker_thread+0x25a/0x3a0
Fix this issue by using configfs_add_default_group() API which does not
have the deadlock problem as configfs_register_group() and does not require
the delayed work handler.
Fixes: e85a2d7837 ("PCI: endpoint: Add support in configfs to associate two EPCs with EPF")
Signed-off-by: Liu Song <liu.song13@zte.com.cn>
[mani: slightly reworded the description and added stable list]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@kernel.org
Link: https://patch.msgid.link/20250710143845409gLM6JdlwPhlHG9iX3F6jK@zte.com.cn
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 0cc13256b6 ]
Though I couldn't confirm ASPM L0s support with the Qcom hardware team, a
bug report from Dmitry suggests that L0s is broken on this legacy SoC.
Hence, remove L0s support from the Root Port Link Capabilities in this SoC.
Since qcom_pcie_clear_aspm_l0s() is now used by more than one SoC config,
call it from qcom_pcie_host_init() instead.
Reported-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Closes: https://lore.kernel.org/linux-pci/4cp5pzmlkkht2ni7us6p3edidnk25l45xrp6w3fxguqcvhq2id@wjqqrdpkypkf
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com>
Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com>
Link: https://patch.msgid.link/20251126081718.8239-1-mani@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 383d89699c upstream.
In 2009, commit c82f63e411 ("PCI: check saved state before restore")
changed the behavior of pci_restore_state() such that it became necessary
to call pci_save_state() afterwards, lest recovery from subsequent PCI
errors fails.
The commit has just been reverted and so all the pci_save_state() after
pci_restore_state() calls that have accumulated in the tree are now
superfluous. Drop them.
Two drivers chose a different approach to achieve the same result:
drivers/scsi/ipr.c and drivers/net/ethernet/intel/e1000e/netdev.c set the
pci_dev's "state_saved" flag to true before calling pci_restore_state().
Drop this as well.
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Dave Jiang <dave.jiang@intel.com>
Acked-by: Giovanni Cabiddu <giovanni.cabiddu@intel.com> # qat
Link: https://patch.msgid.link/c2b28cc4defa1b743cf1dedee23c455be98b397a.1760274044.git.lukas@wunner.de
Cc: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a2f1e22390 upstream.
When the PCI core gained power management support in 2002, it introduced
pci_save_state() and pci_restore_state() helpers to restore Config Space
after a D3hot or D3cold transition, which implies a Soft or Fundamental
Reset (PCIe r7.0 sec 5.8):
https://git.kernel.org/tglx/history/c/a5287abe398b
In 2006, EEH and AER were introduced to recover from errors by performing
a reset. Because errors can occur at any time, drivers began calling
pci_save_state() on probe to ensure recoverability.
In 2009, recoverability was foiled by commit c82f63e411 ("PCI: check
saved state before restore"): It amended pci_restore_state() to bail out
if the "state_saved" flag has been cleared. The flag is cleared by
pci_restore_state() itself, hence a saved state is now allowed to be
restored only once and is then invalidated. That doesn't seem to make
sense because the saved state should be good enough to be reused.
Soon after, drivers began to work around this behavior by calling
pci_save_state() immediately after pci_restore_state(), see e.g. commit
b94f2d775a ("igb: call pci_save_state after pci_restore_state").
Hilariously, two drivers even set the "saved_state" flag to true before
invoking pci_restore_state(), see ipr_reset_restore_cfg_space() and
e1000_io_slot_reset().
Despite these workarounds, recoverability at all times is not guaranteed:
E.g. when a PCIe port goes through a runtime suspend and resume cycle,
the "saved_state" flag is cleared by:
pci_pm_runtime_resume()
pci_pm_default_resume_early()
pci_restore_state()
... and hence on a subsequent AER event, the port's Config Space cannot be
restored. Riana reports a recovery failure of a GPU-integrated PCIe
switch and has root-caused it to the behavior of pci_restore_state().
Another workaround would be necessary, namely calling pci_save_state() in
pcie_port_device_runtime_resume().
The motivation of commit c82f63e411 was to prevent restoring state if
pci_save_state() hasn't been called before. But that can be achieved by
saving state already on device addition, after Config Space has been
initialized. A desirable side effect is that devices become recoverable
even if no driver gets bound. This renders the commit unnecessary, so
revert it.
Reported-by: Riana Tauro <riana.tauro@intel.com> # off-list
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Riana Tauro <riana.tauro@intel.com>
Reviewed-by: Rafael J. Wysocki (Intel) <rafael@kernel.org>
Link: https://patch.msgid.link/9e34ce61c5404e99ffdd29205122c6fb334b38aa.1763483367.git.lukas@wunner.de
Cc: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 269031b15c upstream.
Commit 7ffb791423 ("x86/kaslr: Reduce KASLR entropy on most x86 systems")
is too narrow. The effect being mitigated in that commit is caused by
ZONE_DEVICE which PCI_P2PDMA has a dependency. ZONE_DEVICE, in general,
lets any physical address be added to the direct-map. I.e. not only ACPI
hotplug ranges, CXL Memory Windows, or EFI Specific Purpose Memory, but
also any PCI MMIO range for the DEVICE_PRIVATE and PCI_P2PDMA cases. Update
the mitigation, limit KASLR entropy, to apply in all ZONE_DEVICE=y cases.
Distro kernels typically have PCI_P2PDMA=y, so the practical exposure of
this problem is limited to the PCI_P2PDMA=n case.
A potential path to recover entropy would be to walk ACPI and determine the
limits for hotplug and PCI MMIO before kernel_randomize_memory(). On
smaller systems that could yield some KASLR address bits. This needs
additional investigation to determine if some limited ACPI table scanning
can happen this early without an open coded solution like
arch/x86/boot/compressed/acpi.c needs to deploy.
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Kees Cook <kees@kernel.org>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Logan Gunthorpe <logang@deltatee.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Lorenzo Stoakes <lorenzo.stoakes@oracle.com>
Cc: "Liam R. Howlett" <Liam.Howlett@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Mike Rapoport <rppt@kernel.org>
Cc: Suren Baghdasaryan <surenb@google.com>
Cc: Michal Hocko <mhocko@suse.com>
Fixes: 7ffb791423 ("x86/kaslr: Reduce KASLR entropy on most x86 systems")
Cc: <stable@vger.kernel.org>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Reviewed-by: Balbir Singh <balbirs@nvidia.com>
Tested-by: Yasunori Goto <y-goto@fujitsu.com>
Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
Link: http://patch.msgid.link/692e08b2516d4_261c1100a3@dwillia2-mobl4.notmuch
Signed-off-by: Dave Jiang <dave.jiang@intel.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit fd390ff144 ]
Some systems ship with multiple display class devices but not all
of them are VGA devices. If the "only" VGA device on the system is not
used for displaying the image on the screen marking it as `boot_vga`
because nothing was found is totally wrong.
This behavior actually leads to mistakes of the wrong device being
advertised to userspace and then userspace can make incorrect decisions.
As there is an accurate `boot_display` sysfs file stop lying about
`boot_vga` by assuming if nothing is found it's the right device.
Reported-by: Aaron Erhardt <aer@tuxedocomputers.com>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=220712
Tested-by: Aaron Erhardt <aer@tuxedocomputers.com>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Fixes: ad90860bd1 ("fbcon: Use screen info to find primary device")
Tested-by: Luke D. Jones <luke@ljones.dev>
Signed-off-by: Mario Limonciello (AMD) <superm1@kernel.org>
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patch.msgid.link/20260106044638.52906-1-superm1@kernel.org
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit df27c03b9e upstream.
Previously meson_pcie_link_up() only returned true if the link was in the
L0 state. This was incorrect because hardware autonomously manages
transitions between L0, L0s, and L1 while both components on the link stay
in D0. Those states should all be treated as "link is active".
Returning false when the device was in L0s or L1 broke config accesses
because dw_pcie_other_conf_map_bus() fails if the link is down, which
caused errors like this:
meson-pcie fc000000.pcie: error: wait linkup timeout
pci 0000:01:00.0: BAR 0: error updating (0xfc700004 != 0xffffffff)
Remove the LTSSM state check, timeout, speed check, and error message from
meson_pcie_link_up(), the dw_pcie_ops.link_up() method, so it is a simple
boolean check of whether the link is active. Timeouts and error messages
are handled at a higher level, e.g., dw_pcie_wait_for_link().
Fixes: 9c0ef6d34f ("PCI: amlogic: Add the Amlogic Meson PCIe controller driver")
Reported-by: Linnaea Lavia <linnaea-von-lavia@live.com>
Closes: https://lore.kernel.org/r/DM4PR05MB102707B8CDF84D776C39F22F2C7F0A@DM4PR05MB10270.namprd05.prod.outlook.com
[bhelgaas: squash removal of unused WAIT_LINKUP_TIMEOUT by
Martin Blumenstingl <martin.blumenstingl@googlemail.com>:
https://patch.msgid.link/20260105125625.239497-1-martin.blumenstingl@googlemail.com]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Linnaea Lavia <linnaea-von-lavia@live.com>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on BananaPi M2S
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20251103221930.1831376-1-helgaas@kernel.org
Link: https://patch.msgid.link/20260105125625.239497-1-martin.blumenstingl@googlemail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit eff0306b10 upstream.
First of all, the driver was parsing the 'dbi' register region as 'elbi'.
This was due to DT mistakenly passing 'dbi' as 'elbi'. Since the DT is
now fixed to supply 'dbi' region, this driver can rely on the DWC core
driver to parse and map it.
However, to support the old DTs, if the 'elbi' region is found in DT, parse
and map the region as both 'dw_pcie::elbi_base' as 'dw_pcie::dbi_base'.
This will allow the driver to work with both broken and fixed DTs.
Also, skip parsing the 'elbi' region in DWC core if 'pci->elbi_base' was
already populated.
Fixes: 9c0ef6d34f ("PCI: amlogic: Add the Amlogic Meson PCIe controller driver")
Fixes: c96992a24b ("PCI: dwc: Add support for ELBI resource mapping")
Reported-by: Linnaea Lavia <linnaea-von-lavia@live.com>
Closes: https://lore.kernel.org/linux-pci/DM4PR05MB102707B8CDF84D776C39F22F2C7F0A@DM4PR05MB10270.namprd05.prod.outlook.com/
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@oss.qualcomm.com>
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Tested-by: Neil Armstrong <neil.armstrong@linaro.org> # on Bananapi-M2S
Reviewed-by: Neil Armstrong <neil.armstrong@linaro.org>
Cc: stable@vger.kernel.org # 6.2
Link: https://patch.msgid.link/20251101-pci-meson-fix-v1-3-c50dcc56ed6a@oss.qualcomm.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 9583f9d229 upstream.
caab002d50 ("PCI: brcmstb: Disable L0s component of ASPM if requested")
set PCI_EXP_LNKCAP_ASPM_L1 and (optionally) PCI_EXP_LNKCAP_ASPM_L0S in
PCI_EXP_LNKCAP (aka PCIE_RC_CFG_PRIV1_LINK_CAPABILITY in brcmstb).
But instead of using PCI_EXP_LNKCAP_ASPM_L1 and PCI_EXP_LNKCAP_ASPM_L0S
directly, it used PCIE_LINK_STATE_L1 and PCIE_LINK_STATE_L0S, which are
Linux-created values that only coincidentally matched the PCIe spec.
b478e162f2 ("PCI/ASPM: Consolidate link state defines") later changed
them so they no longer matched the PCIe spec, so the bits ended up in the
wrong place in PCI_EXP_LNKCAP.
Use PCI_EXP_LNKCAP_ASPM_L0S to clear L0s support when there's an
'aspm-no-l0s' property. Rely on brcmstb hardware to advertise L0s and/or
L1 support otherwise.
Fixes: caab002d50 ("PCI: brcmstb: Disable L0s component of ASPM if requested")
Reported-by: Bjorn Helgaas <bhelgaas@google.com>
Closes: https://lore.kernel.org/linux-pci/20250925194424.GA2197200@bhelgaas
Signed-off-by: Jim Quinlan <james.quinlan@broadcom.com>
[mani: reworded subject and description, added closes tag and CCed stable]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
[bhelgaas: commit log]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20251003170436.1446030-1-james.quinlan@broadcom.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 894f475f88 upstream.
When a PCI device is suspended, it is normally the PCI core's job to save
Config Space and put the device into a low power state. However drivers
are allowed to assume these responsibilities. When they do, the PCI core
can tell by looking at the state_saved flag in struct pci_dev: The flag
is cleared before commencing the suspend sequence and it is set when
pci_save_state() is called. If the PCI core finds the flag set late in
the suspend sequence, it refrains from calling pci_save_state() itself.
But there are two corner cases where the PCI core neglects to clear the
flag before commencing the suspend sequence:
* If a driver has legacy PCI PM callbacks, pci_legacy_suspend() neglects
to clear the flag. The (stale) flag is subsequently queried by
pci_legacy_suspend() itself and pci_legacy_suspend_late().
* If a device has no driver or its driver has no PCI PM callbacks,
pci_pm_freeze() neglects to clear the flag. The (stale) flag is
subsequently queried by pci_pm_freeze_noirq().
The flag may be set prior to suspend if the device went through error
recovery: Drivers commonly invoke pci_restore_state() + pci_save_state()
to restore Config Space after reset.
The flag may also be set if drivers call pci_save_state() on probe to
allow for recovery from subsequent errors.
The result is that pci_legacy_suspend_late() and pci_pm_freeze_noirq()
don't call pci_save_state() and so the state that will be restored on
resume is the one recorded on last error recovery or on probe, not the one
that the device had on suspend. If the two states happen to be identical,
there's no problem.
Reinstate clearing the flag in pci_legacy_suspend() and pci_pm_freeze().
The two functions used to do that until commit 4b77b0a2ba ("PCI: Clear
saved_state after the state has been restored") deemed it unnecessary
because it assumed that it's sufficient to clear the flag on resume in
pci_restore_state(). The commit seemingly did not take into account that
pci_save_state() and pci_restore_state() are not only used by power
management code, but also for error recovery.
Devices without driver or whose driver has no PCI PM callbacks may be in
runtime suspend when pci_pm_freeze() is called. Their state has already
been saved, so don't clear the flag to skip a pointless pci_save_state()
in pci_pm_freeze_noirq().
None of the drivers with legacy PCI PM callbacks seem to use runtime PM,
so clear the flag unconditionally in their case.
Fixes: 4b77b0a2ba ("PCI: Clear saved_state after the state has been restored")
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Rafael J. Wysocki (Intel) <rafael@kernel.org>
Cc: stable@vger.kernel.org # v2.6.32+
Link: https://patch.msgid.link/094f2aad64418710daf0940112abe5a0afdc6bce.1763483367.git.lukas@wunner.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit ff529a9307 ]
pci_epc_mem_alloc_addr() allocates a CPU address from the ATU window phys
base and a page number. Set the ep->page_size so the resulting CPU address
is correctly aligned with the ATU required alignment.
Fixes: 151f3d29ba ("PCI: stm32-ep: Add PCIe Endpoint support for STM32MP25")
Signed-off-by: Christian Bruel <christian.bruel@foss.st.com>
[mani: added fixes tag]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20251114-atu_align_ep-v1-1-88da5366fa04@foss.st.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit fa81d60990 ]
If the host has deasserted PERST# and started link training before the link
is started on EP side, enabling LTSSM before the endpoint registers are
initialized in the perst_irq handler results in probing incorrect values.
Thus, wait for the PERST# level-triggered interrupt to start link training
at the end of initialization and cleanup the stm32_pcie_[start stop]_link
functions.
Fixes: 151f3d29ba ("PCI: stm32-ep: Add PCIe Endpoint support for STM32MP25")
Signed-off-by: Christian Bruel <christian.bruel@foss.st.com>
[mani: added fixes tag]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
[bhelgaas: wrap line]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20251114-perst_ep-v1-1-e7976317a890@foss.st.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 91c4c89db4 ]
pbus_reassign_bridge_resources() saves bridge windows into the saved
list before attempting to adjust resource assignments to perform a BAR
resize operation. If resource adjustments cannot be completed fully,
rollback is attempted by restoring the resource from the saved list.
The rollback, however, does not check whether the resources it restores were
assigned by the partial resize attempt. If restore changes addresses of the
resource, it can result in corrupting the resource tree.
An example of a corrupted resource tree with overlapping addresses:
6200000000000-6203fbfffffff : pciex@620c3c0000000
6200000000000-6203fbff0ffff : PCI Bus 0030:01
6200020000000-62000207fffff : 0030:01:00.0
6200000000000-6203fbff0ffff : PCI Bus 0030:02
A resource that are assigned into the resource tree must remain
unchanged. Thus, release such a resource before attempting to restore
and claim it back.
For simplicity, always do the release and claim back for the resource
even in the cases where it is restored to the same address range.
Note: this fix may "break" some cases where devices "worked" because
the resource tree corruption allowed address space double counting to
fit more resource than what can now be assigned without double
counting. The upcoming changes to BAR resizing should address those
scenarios (to the extent possible).
Fixes: 8bb705e3e7 ("PCI: Add pci_resize_resource() for resizing BARs")
Reported-by: Simon Richter <Simon.Richter@hogyros.de>
Link: https://lore.kernel.org/linux-pci/67840a16-99b4-4d8c-9b5c-4721ab0970a2@hogyros.de/
Reported-by: Alex Bennée <alex.bennee@linaro.org>
Link: https://lore.kernel.org/linux-pci/874irqop6b.fsf@draig.linaro.org/
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Alex Bennée <alex.bennee@linaro.org> # AVA, AMD GPU
Link: https://patch.msgid.link/20251113162628.5946-2-ilpo.jarvinen@linux.intel.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 95d9c3f0e4 ]
Commit under Fixes introduced support for PCIe EP mode on AM654x platforms.
When the mode happens to be either "DW_PCIE_RC_TYPE" or "DW_PCIE_EP_TYPE",
the PCIe Controller is configured accordingly. However, when the mode is
neither of them, an error message is displayed, but the driver probe
succeeds. Since this "invalid" mode is not associated with a functional
PCIe Controller, the probe should fail.
Fix the behavior by exiting "ks_pcie_probe()" with the return value of
"-EINVAL" in addition to displaying the existing error message when the
mode is invalid.
Fixes: 23284ad677 ("PCI: keystone: Add support for PCIe EP in AM654x Platforms")
Signed-off-by: Siddharth Vadapalli <s-vadapalli@ti.com>
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20251029080547.1253757-4-s-vadapalli@ti.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 25423cda14 ]
When Root Complex (RC) triggers a Doorbell interrupt to Endpoint (EP), it
triggers the below warning in the EP:
BUG: sleeping function called from invalid context at kernel/locking/mutex.c:271
Call trace:
__might_resched+0x130/0x158
__might_sleep+0x70/0x88
mutex_lock+0x2c/0x80
pci_epc_get_msi+0x78/0xd8
pci_epf_test_raise_irq.isra.0+0x74/0x138
pci_epf_test_doorbell_handler+0x34/0x50
The BUG arises because the EP's pci_epf_test_doorbell_handler() which is
running in the hard IRQ context is making an indirect call to
pci_epc_get_msi(), which uses mutex inside.
To fix the issue, convert the hard IRQ handler to a threaded IRQ handler to
allow it to call functions that can sleep during bottom half execution.
Also, register the threaded IRQ handler with IRQF_ONESHOT to keep the
interrupt line disabled until the threaded IRQ handler completes execution.
Fixes: eff0c286aa ("PCI: endpoint: pci-epf-test: Add doorbell test support")
Signed-off-by: Bhanu Seshu Kumar Valluri <bhanuseshukumar@gmail.com>
[mani: reworded description a bit]
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Reviewed-by: Niklas Cassel <cassel@kernel.org>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Link: https://patch.msgid.link/20251014024109.42287-1-bhanuseshukumar@gmail.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit d312742f68 ]
Since the reliance on ARM-specific struct pci_sys_data was removed, this
driver can be compile-tested on other architectures.
While at it, make the help text a bit more generic, as some members of
the R-Car Gen2 family have a different number of internal PCI
controllers.
Fixes: 4a957563fe ("PCI: rcar-gen2: Convert to use modern host bridge probe functions")
Suggested-by: Ilpo Jarvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
[bhelgaas: add rcar-gen2 to subject]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Link: https://patch.msgid.link/00f75d6732eacce93f04ffaeedc415d2db714cd6.1759480426.git.geert+renesas@glider.be
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 932ec9dff6 ]
devm_pm_runtime_enable() is used in the probe, so pm_runtime_disable()
should not be called explicitly in the remove function.
Fixes: 1c72774df0 ("PCI: sg2042: Add Sophgo SG2042 PCIe driver")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Manivannan Sadhasivam <mani@kernel.org>
Tested-by: Chen Wang <unicorn_wang@outlook.com> # on Pioneerbox.
Acked-by: Chen Wang <unicorn_wang@outlook.com>
Link: https://patch.msgid.link/242eca0ff6601de7966a53706e9950fbcb10aac8.1759169586.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Sasha Levin <sashal@kernel.org>
-----BEGIN PGP SIGNATURE-----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=2eqF
-----END PGP SIGNATURE-----
Merge tag 'pci-v6.18-fixes-5' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci
Pull pci fixes from Bjorn Helgaas:
- Cache the ASPM L0s/L1 Supported bits early so quirks can override
them if necessary (Bjorn Helgaas)
- Add quirks for PA Semi and Freescale Root Ports and a HiSilicon Wi-Fi
device that are reported to have broken L0s and L1 (Shawn Lin, Bjorn
Helgaas)
* tag 'pci-v6.18-fixes-5' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci:
PCI/ASPM: Avoid L0s and L1 on Hi1105 [19e5:1105] Wi-Fi
PCI/ASPM: Avoid L0s and L1 on PA Semi [1959:a002] Root Ports
PCI/ASPM: Avoid L0s and L1 on Freescale [1957:0451] Root Ports
PCI/ASPM: Convert quirks to override advertised link states
PCI/ASPM: Add pcie_aspm_remove_cap() to override advertised link states
PCI/ASPM: Cache L0s/L1 Supported so advertised link states can be overridden
This Wi-Fi advertises the L0s and L1 capabilities but actually it doesn't
support them. This is confirmed by HiSilicon team in actual productization.
Signed-off-by: Shawn Lin <shawn.lin@rock-chips.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/1762916319-139532-1-git-send-email-shawn.lin@rock-chips.com
Christian reported that f3ac2ff148 ("PCI/ASPM: Enable all ClockPM and
ASPM states for devicetree platforms") broke booting on the A-EON AmigaOne
X1000.
Override the L0s and L1 Support advertised in Link Capabilities by the
X1000 Root Ports ([1959:a002]) so we don't try to enable those states.
Fixes: f3ac2ff148 ("PCI/ASPM: Enable all ClockPM and ASPM states for devicetree platforms")
Fixes: df5192d9bb ("PCI/ASPM: Enable only L0s and L1 for devicetree platforms")
Reported-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Link: https://lore.kernel.org/r/a41d2ca1-fcd9-c416-b111-a958e92e94bf@xenosoft.de
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Christian reported that f3ac2ff148 ("PCI/ASPM: Enable all ClockPM and
ASPM states for devicetree platforms") broke booting on the A-EON X5000.
Override the L0s and L1 Support advertised in Link Capabilities by the
X5000 Root Ports ([1957:0451]) so we don't try to enable those states.
Fixes: f3ac2ff148 ("PCI/ASPM: Enable all ClockPM and ASPM states for devicetree platforms")
Fixes: df5192d9bb ("PCI/ASPM: Enable only L0s and L1 for devicetree platforms")
Reported-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Link: https://lore.kernel.org/r/db5c95a1-cf3e-46f9-8045-a1b04908051a@xenosoft.de
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Shawn Lin <shawn.lin@rock-chips.com>
Reviewed-by: Lukas Wunner <lukas@wunner.de>
Link: https://patch.msgid.link/20251110222929.2140564-5-helgaas@kernel.org
Existing quirks to disable ASPM L0s and L1 use pci_disable_link_state(),
which disables ASPM states and prevents their use in the future. But since
they are FINAL quirks, they happen after ASPM has already been enabled.
Here's a typical call path:
pci_host_probe
pci_scan_root_bus_bridge
pci_scan_child_bus
pci_scan_slot
pci_scan_single_device
pci_device_add
pci_fixup_device(pci_fixup_header) # HEADER quirks
pcie_aspm_init_link_state
pcie_config_aspm_path
pcie_config_aspm_link
pcie_config_aspm_dev # ASPM may be enabled
pci_bus_add_devices
pci_bus_add_devices
pci_fixup_device(pci_fixup_final) # FINAL quirks
quirk_disable_aspm_l0s
pci_disable_link_state(dev, PCIE_LINK_STATE_L0S)
Sometimes enabling ASPM can make the link non-functional, so if we know
ASPM is broken on a device, we shouldn't enable it at all, even
temporarily.
Convert the existing quirks to use pcie_aspm_remove_cap() instead, which
overrides the ASPM Support advertised in PCIe Link Capabilities, and make
them HEADER quirks so they run before pcie_aspm_init_link_state() has a
chance to enable ASPM.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Shawn Lin <shawn.lin@rock-chips.com>
Reviewed-by: Lukas Wunner <lukas@wunner.de>
Link: https://patch.msgid.link/20251110222929.2140564-4-helgaas@kernel.org
Add pcie_aspm_remove_cap(). A quirk can use this to prevent use of ASPM
L0s or L1 link states, even if the device advertised support for them.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Shawn Lin <shawn.lin@rock-chips.com>
Reviewed-by: Lukas Wunner <lukas@wunner.de>
Link: https://patch.msgid.link/20251110222929.2140564-3-helgaas@kernel.org
Defective devices sometimes advertise support for ASPM L0s or L1 states
even if they don't work correctly.
Cache the L0s Supported and L1 Supported bits early in enumeration so
HEADER quirks can override the ASPM states advertised in Link Capabilities
before pcie_aspm_cap_init() enables ASPM.
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Shawn Lin <shawn.lin@rock-chips.com>
Reviewed-by: Lukas Wunner <lukas@wunner.de>
Link: https://patch.msgid.link/20251110222929.2140564-2-helgaas@kernel.org
-----BEGIN PGP SIGNATURE-----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=fYnS
-----END PGP SIGNATURE-----
Merge tag 'pci-v6.18-fixes-4' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci
Pull pci fixes from Bjorn Helgaas:
- Restore custom qcom ASPM enablement code so L1 PM Substates are
enabled as they were in v6.17 even though the PCI core now enables
just L0s and L1 by default (Bjorn Helgaas)
- Size prefetchable bridge windows only when they actually exist, to
avoid a WARN_ON() regression (Ilpo Järvinen)
* tag 'pci-v6.18-fixes-4' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci:
PCI: Do not size non-existing prefetchable window
Revert "PCI: qcom: Remove custom ASPM enablement code"
pbus_size_mem() should only be called for bridge windows that exist but
__pci_bus_size_bridges() may point 'pref' to a resource that does not exist
(has zero flags) in case of non-root buses.
When prefetchable bridge window does not exist, the same non-prefetchable
bridge window is sized more than once which may result in duplicating
entries into the realloc_head list. Duplicated entries are shown in this
log and trigger a WARN_ON() because realloc_head had residual entries after
the resource assignment algorithm:
pci 0000:00:03.0: [11ab:6820] type 01 class 0x060400 PCIe Root Port
pci 0000:00:03.0: PCI bridge to [bus 00]
pci 0000:00:03.0: bridge window [io 0x0000-0x0fff]
pci 0000:00:03.0: bridge window [mem 0x00000000-0x000fffff]
pci 0000:00:03.0: bridge window [mem 0x00200000-0x003fffff] to [bus 02] add_size 200000 add_align 200000
pci 0000:00:03.0: bridge window [mem 0x00200000-0x003fffff] to [bus 02] add_size 200000 add_align 200000
pci 0000:00:03.0: bridge window [mem 0xe0000000-0xe03fffff]: assigned
pci 0000:00:03.0: PCI bridge to [bus 02]
pci 0000:00:03.0: bridge window [mem 0xe0000000-0xe03fffff]
------------[ cut here ]------------
WARNING: CPU: 0 PID: 1 at drivers/pci/setup-bus.c:2373 pci_assign_unassigned_root_bus_resources+0x1bc/0x234
Check resource flags of 'pref' and only size the prefetchable window if the
resource has the IORESOURCE_PREFETCH flag.
Fixes: ae88d0b9c5 ("PCI: Use pbus_select_window_for_type() during mem window sizing")
Reported-by: Klaus Kudielka <klaus.kudielka@gmail.com>
Closes: https://lore.kernel.org/r/51e8cf1c62b8318882257d6b5a9de7fdaaecc343.camel@gmail.com/
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Klaus Kudielka <klaus.kudielka@gmail.com>
Link: https://patch.msgid.link/20251027132423.8841-1-ilpo.jarvinen@linux.intel.com
This reverts commit a729c16646.
Prior to a729c16646 ("PCI: qcom: Remove custom ASPM enablement code"),
the qcom controller driver enabled ASPM, including L0s, L1, and L1 PM
Substates, for all devices powered on at the time the controller driver
enumerates them.
ASPM was *not* enabled for devices powered on later by pwrctrl (unless the
kernel was built with PCIEASPM_POWERSAVE or PCIEASPM_POWER_SUPERSAVE, or
the user enabled ASPM via module parameter or sysfs).
After f3ac2ff148 ("PCI/ASPM: Enable all ClockPM and ASPM states for
devicetree platforms"), the PCI core enabled all ASPM states for all
devices whether powered on initially or by pwrctrl, so a729c16646 was
unnecessary and reverted.
But f3ac2ff148 was too aggressive and broke platforms that didn't support
CLKREQ# or required device-specific configuration for L1 Substates, so
df5192d9bb ("PCI/ASPM: Enable only L0s and L1 for devicetree platforms")
enabled only L0s and L1.
On Qualcomm platforms, this left L1 Substates disabled, which was a
regression. Revert a729c16646 so L1 Substates will be enabled on devices
that are initially powered on. Devices powered on by pwrctrl will be
addressed later.
Fixes: df5192d9bb ("PCI/ASPM: Enable only L0s and L1 for devicetree platforms")
Reported-by: Johan Hovold <johan@kernel.org>
Closes: https://lore.kernel.org/lkml/aPuXZlaawFmmsLmX@hovoldconsulting.com/
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Johan Hovold <johan@kernel.org>
Reviewed-by: Manivannan Sadhasivam <mani@kernel.org>
Link: https://patch.msgid.link/20251024210514.1365996-1-helgaas@kernel.org
-----BEGIN PGP SIGNATURE-----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=UcX7
-----END PGP SIGNATURE-----
Merge tag 'pci-v6.18-fixes-3' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci
Pull pci fixes from Bjorn Helgaas:
- Add DWC custom pci_ops for the root bus instead of overwriting the
DBI base address, which broke drivers that rely on the DBI address
for iATU programming; fixes an FU740 probe regression (Krishna
Chaitanya Chundru)
- Revert qcom ECAM enablement, which is rendered unnecessary by the DWC
custom pci_ops (Krishna Chaitanya Chundru)
- Fix longstanding MIPS Malta resource registration issues to avoid
exposing them when the next commit fixes the boot failure (Maciej W.
Rozycki)
- Use pcibios_align_resource() on MIPS Malta to fix boot failure caused
by using the generic pci_enable_resources() (Ilpo Järvinen)
- Enable only ASPM L0s and L1, not L1 PM Substates, for devicetree
platforms because we lack information required to configure L1
Substates; fixes regressions on powerpc and rockchip. A qcom
regression (L1 Substates no longer enabled) remains and will be
addressed next (Bjorn Helgaas)
* tag 'pci-v6.18-fixes-3' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci:
PCI/ASPM: Enable only L0s and L1 for devicetree platforms
MIPS: Malta: Use pcibios_align_resource() to block io range
MIPS: Malta: Fix PCI southbridge legacy resource reservations
MIPS: Malta: Fix keyboard resource preventing i8042 driver from registering
Revert "PCI: qcom: Prepare for the DWC ECAM enablement"
PCI: dwc: Use custom pci_ops for root bus DBI vs ECAM config access
This reverts commit 4660e50cf8.
Commit f6fd357f7a ("PCI: dwc: Prepare the driver for enabling ECAM
mechanism using iATU 'CFG Shift Feature'") enabled ECAM access by using
the config space start as DBI address.
However, this approach breaks vendor drivers that rely on the DBI address
for internal accesses, especially when the vendor config space is 256MB
aligned.
To resolve this, avoid using the DBI as the start of config space and
instead introduce a custom ECAM PCI ops implementation.
Revert the qcom specific ECAM preparation logic in 4660e50cf8 ("PCI:
qcom: Prepare for the DWC ECAM enablement") since it's no longer necessary.
Signed-off-by: Krishna Chaitanya Chundru <krishna.chundru@oss.qualcomm.com>
[bhelgaas: commit log]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Link: https://patch.msgid.link/20251017-ecam_fix-v1-2-f6faa3d0edf3@oss.qualcomm.com
