[ Upstream commit d427da2323b093a65d8317783e76ab8fad2e2ef0 ]
There is no function to get all PCI devices in a system by matching
against the base class code only, ignoring the sub-class code and
the programming interface. Add pci_get_base_class() to suit the
need.
For example, if a driver wants to process all PCI display devices in
a system, it can do so like this:
pdev = NULL;
while ((pdev = pci_get_base_class(PCI_BASE_CLASS_DISPLAY, pdev))) {
do_something_for_pci_display_device(pdev);
}
Link: https://lore.kernel.org/r/20230825062714.6325-2-sui.jingfeng@linux.dev
Signed-off-by: Sui Jingfeng <suijingfeng@loongson.cn>
[bhelgaas: reword commit log]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Stable-dep-of: c2bc958b2b03 ("fbdev: vesafb: Detect VGA compatibility from screen info's VESA attributes")
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 11a1f4bc47362700fcbde717292158873fb847ed upstream.
Keith reports a use-after-free when a DPC event occurs concurrently to
hot-removal of the same portion of the hierarchy:
The dpc_handler() awaits readiness of the secondary bus below the
Downstream Port where the DPC event occurred. To do so, it polls the
config space of the first child device on the secondary bus. If that
child device is concurrently removed, accesses to its struct pci_dev
cause the kernel to oops.
That's because pci_bridge_wait_for_secondary_bus() neglects to hold a
reference on the child device. Before v6.3, the function was only
called on resume from system sleep or on runtime resume. Holding a
reference wasn't necessary back then because the pciehp IRQ thread
could never run concurrently. (On resume from system sleep, IRQs are
not enabled until after the resume_noirq phase. And runtime resume is
always awaited before a PCI device is removed.)
However starting with v6.3, pci_bridge_wait_for_secondary_bus() is also
called on a DPC event. Commit 53b54ad074 ("PCI/DPC: Await readiness
of secondary bus after reset"), which introduced that, failed to
appreciate that pci_bridge_wait_for_secondary_bus() now needs to hold a
reference on the child device because dpc_handler() and pciehp may
indeed run concurrently. The commit was backported to v5.10+ stable
kernels, so that's the oldest one affected.
Add the missing reference acquisition.
Abridged stack trace:
BUG: unable to handle page fault for address: 00000000091400c0
CPU: 15 PID: 2464 Comm: irq/53-pcie-dpc 6.9.0
RIP: pci_bus_read_config_dword+0x17/0x50
pci_dev_wait()
pci_bridge_wait_for_secondary_bus()
dpc_reset_link()
pcie_do_recovery()
dpc_handler()
Fixes: 53b54ad074 ("PCI/DPC: Await readiness of secondary bus after reset")
Closes: https://lore.kernel.org/r/20240612181625.3604512-3-kbusch@meta.com/
Link: https://lore.kernel.org/linux-pci/8e4bcd4116fd94f592f2bf2749f168099c480ddf.1718707743.git.lukas@wunner.de
Reported-by: Keith Busch <kbusch@kernel.org>
Tested-by: Keith Busch <kbusch@kernel.org>
Signed-off-by: Lukas Wunner <lukas@wunner.de>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: stable@vger.kernel.org # v5.10+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit a4bbcac11d3cea85822af8b40daed7e96bca5068 upstream.
The LS7A chipset can be used as part of a PCIe Root Complex with
Loongson-3C6000 and similar CPUs. In this case, DEV_LS7A_PCIE_PORT5 has a
PCI_CLASS_BRIDGE_HOST class code, and it is a Type 0 Function whose config
space provides access to Root Complex registers.
The DEV_LS7A_PCIE_PORT5 has an MSI Capability, and its MSI Enable bit must
be set before other devices below the Root Complex can use MSI. This is
not the standard PCI behavior of MSI Enable, so the normal PCI MSI code
does not set it.
Set the DEV_LS7A_PCIE_PORT5 MSI Enable bit via a quirk so other devices
below the Root Complex can use MSI.
[kwilczynski: exit early to reduce indentation; commit log]
Link: https://lore.kernel.org/linux-pci/20240612065315.2048110-1-chenhuacai@loongson.cn
Signed-off-by: Sheng Wu <wusheng@loongson.cn>
Signed-off-by: Huacai Chen <chenhuacai@loongson.cn>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
[bhelgaas: commit log]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 840b7a5edf88fe678c60dee88a135647c0ea4375 upstream.
Rockchip platforms use 'GPIO_ACTIVE_HIGH' flag in the devicetree definition
for ep_gpio. This means, whatever the logical value set by the driver for
the ep_gpio, physical line will output the same logic level.
For instance,
gpiod_set_value_cansleep(rockchip->ep_gpio, 0); --> Level low
gpiod_set_value_cansleep(rockchip->ep_gpio, 1); --> Level high
But while requesting the ep_gpio, GPIOD_OUT_HIGH flag is currently used.
Now, this also causes the physical line to output 'high' creating trouble
for endpoint devices during host reboot.
When host reboot happens, the ep_gpio will initially output 'low' due to
the GPIO getting reset to its POR value. Then during host controller probe,
it will output 'high' due to GPIOD_OUT_HIGH flag. Then during
rockchip_pcie_host_init_port(), it will first output 'low' and then 'high'
indicating the completion of controller initialization.
On the endpoint side, each output 'low' of ep_gpio is accounted for PERST#
assert and 'high' for PERST# deassert. With the above mentioned flow during
host reboot, endpoint will witness below state changes for PERST#:
(1) PERST# assert - GPIO POR state
(2) PERST# deassert - GPIOD_OUT_HIGH while requesting GPIO
(3) PERST# assert - rockchip_pcie_host_init_port()
(4) PERST# deassert - rockchip_pcie_host_init_port()
Now the time interval between (2) and (3) is very short as both happen
during the driver probe(), and this results in a race in the endpoint.
Because, before completing the PERST# deassertion in (2), endpoint got
another PERST# assert in (3).
A proper way to fix this issue is to change the GPIOD_OUT_HIGH flag in (2)
to GPIOD_OUT_LOW. Because the usual convention is to request the GPIO with
a state corresponding to its 'initial/default' value and let the driver
change the state of the GPIO when required.
As per that, the ep_gpio should be requested with GPIOD_OUT_LOW as it
corresponds to the POR value of '0' (PERST# assert in the endpoint). Then
the driver can change the state of the ep_gpio later in
rockchip_pcie_host_init_port() as per the initialization sequence.
This fixes the firmware crash issue in Qcom based modems connected to
Rockpro64 based board.
Fixes: e77f847df5 ("PCI: rockchip: Add Rockchip PCIe controller support")
Closes: https://lore.kernel.org/mhi/20240402045647.GG2933@thinkpad/
Link: https://lore.kernel.org/linux-pci/20240416-pci-rockchip-perst-fix-v1-1-4800b1d4d954@linaro.org
Reported-by: Slark Xiao <slark_xiao@163.com>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Niklas Cassel <cassel@kernel.org>
Cc: stable@vger.kernel.org # v4.9
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit 28b8d7793b8573563b3d45321376f36168d77b1e upstream.
PERST# is active low according to the PCIe specification.
However, the existing pcie-dw-rockchip.c driver does:
gpiod_set_value(..., 0); msleep(100); gpiod_set_value(..., 1);
when asserting + deasserting PERST#.
This is of course wrong, but because all the device trees for this
compatible string have also incorrectly marked this GPIO as ACTIVE_HIGH:
$ git grep -B 10 reset-gpios arch/arm64/boot/dts/rockchip/rk3568*
$ git grep -B 10 reset-gpios arch/arm64/boot/dts/rockchip/rk3588*
The actual toggling of PERST# is correct, and we cannot change it anyway,
since that would break device tree compatibility.
However, this driver does request the GPIO to be initialized as
GPIOD_OUT_HIGH, which does cause a silly sequence where PERST# gets
toggled back and forth for no good reason.
Fix this by requesting the GPIO to be initialized as GPIOD_OUT_LOW (which
for this driver means PERST# asserted).
This will avoid an unnecessary signal change where PERST# gets deasserted
(by devm_gpiod_get_optional()) and then gets asserted (by
rockchip_pcie_start_link()) just a few instructions later.
Before patch, debug prints on EP side, when booting RC:
[ 845.606810] pci: PERST# asserted by host!
[ 852.483985] pci: PERST# de-asserted by host!
[ 852.503041] pci: PERST# asserted by host!
[ 852.610318] pci: PERST# de-asserted by host!
After patch, debug prints on EP side, when booting RC:
[ 125.107921] pci: PERST# asserted by host!
[ 132.111429] pci: PERST# de-asserted by host!
This extra, very short, PERST# assertion + deassertion has been reported to
cause issues with certain WLAN controllers, e.g. RTL8822CE.
Fixes: 0e898eb8df ("PCI: rockchip-dwc: Add Rockchip RK356X host controller driver")
Link: https://lore.kernel.org/linux-pci/20240417164227.398901-1-cassel@kernel.org
Tested-by: Heiko Stuebner <heiko@sntech.de>
Tested-by: Jianfeng Liu <liujianfeng1994@gmail.com>
Signed-off-by: Niklas Cassel <cassel@kernel.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Heiko Stuebner <heiko@sntech.de>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Cc: stable@vger.kernel.org # v5.15+
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit fea93a3e5d5e6a09eb153866d2ce60ea3287a70d upstream.
The intent of the code snippet is to always return 0 for both
PCI_INTERRUPT_LINE and PCI_INTERRUPT_PIN.
The check misses PCI_INTERRUPT_PIN. This patch fixes that.
This is discovered by this call in VFIO:
pci_read_config_byte(vdev->pdev, PCI_INTERRUPT_PIN, &pin);
The old code does not set *val to 0 because it misses the check for
PCI_INTERRUPT_PIN. Garbage is returned in that case.
Fixes: 4daace0d8c ("PCI: hv: Add paravirtual PCI front-end for Microsoft Hyper-V VMs")
Link: https://lore.kernel.org/linux-pci/20240701202606.129606-1-wei.liu@kernel.org
Signed-off-by: Wei Liu <wei.liu@kernel.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Reviewed-by: Michael Kelley <mhklinux@outlook.com>
Cc: stable@kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit c2a57ee0f2f1ad8c970ff58b78a43e85abbdeb7f ]
When PERST# assert and deassert happens on the PERST# supported platforms,
both iATU0 and iATU6 will map inbound window to BAR0. DMA will access the
area that was previously allocated (iATU0) for BAR0, instead of the new
area (iATU6) for BAR0.
Right now, this isn't an issue because both iATU0 and iATU6 should
translate inbound accesses to BAR0 to the same allocated memory area.
However, having two separate inbound mappings for the same BAR is a
disaster waiting to happen.
The mappings between PCI BAR and iATU inbound window are maintained in the
dw_pcie_ep::bar_to_atu[] array. While allocating a new inbound iATU map for
a BAR, dw_pcie_ep_inbound_atu() API checks for the availability of the
existing mapping in the array and if it is not found (i.e., value in the
array indexed by the BAR is found to be 0), it allocates a new map value
using find_first_zero_bit().
The issue is the existing logic failed to consider the fact that the map
value '0' is a valid value for BAR0, so find_first_zero_bit() will return
'0' as the map value for BAR0 (note that it returns the first zero bit
position).
Due to this, when PERST# assert + deassert happens on the PERST# supported
platforms, the inbound window allocation restarts from BAR0 and the
existing logic to find the BAR mapping will return '6' for BAR0 instead of
'0' due to the fact that it considers '0' as an invalid map value.
Fix this issue by always incrementing the map value before assigning to
bar_to_atu[] array and then decrementing it while fetching. This will make
sure that the map value '0' always represents the invalid mapping."
Fixes: 4284c88fff ("PCI: designware-ep: Allow pci_epc_set_bar() update inbound map address")
Closes: https://lore.kernel.org/linux-pci/ZXsRp+Lzg3x%2Fnhk3@x1-carbon/
Link: https://lore.kernel.org/linux-pci/20240412160841.925927-1-Frank.Li@nxp.com
Reported-by: Niklas Cassel <Niklas.Cassel@wdc.com>
Tested-by: Niklas Cassel <niklas.cassel@wdc.com>
Signed-off-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Niklas Cassel <niklas.cassel@wdc.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 912315715d7b74f7abdb6f063ebace44ee288af9 ]
All EP specific resources are enabled during PERST# deassert. As a counter
operation, all resources should be disabled during PERST# assert. There is
no point in skipping that if the link was not enabled.
This will also result in enablement of the resources twice if PERST# got
deasserted again. So remove the check from qcom_pcie_perst_assert() and
disable all the resources unconditionally.
Fixes: f55fee56a6 ("PCI: qcom-ep: Add Qualcomm PCIe Endpoint controller driver")
Link: https://lore.kernel.org/linux-pci/20240430-pci-epf-rework-v4-1-22832d0d456f@linaro.org
Tested-by: Niklas Cassel <cassel@kernel.org>
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Niklas Cassel <cassel@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6bba3c0ac5dc54737998a0982b2e272242c87e0f ]
There are two issues related to epf_ntb_epc_cleanup():
1) It should call epf_ntb_config_sspad_bar_clear()
2) The epf_ntb_bind() function should call epf_ntb_epc_cleanup()
to cleanup.
I also changed the ordering a bit. Unwinding should be done in the
mirror order from how they are allocated.
Fixes: e35f56bb03 ("PCI: endpoint: Support NTB transfer between RC and EP")
Link: https://lore.kernel.org/linux-pci/aaffbe8d-7094-4083-8146-185f4a84e8a1@moroto.mountain
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 8e0f5a96c534f781e8c57ca30459448b3bfe5429 ]
Smatch complains about inconsistent NULL checking in vpci_scan_bus():
drivers/pci/endpoint/functions/pci-epf-vntb.c:1024 vpci_scan_bus() error: we previously assumed 'vpci_bus' could be null (see line 1021)
Instead of printing an error message and then crashing we should return
an error code and clean up.
Also the NULL check is reversed so it prints an error for success
instead of failure.
Fixes: e35f56bb03 ("PCI: endpoint: Support NTB transfer between RC and EP")
Link: https://lore.kernel.org/linux-pci/68e0f6a4-fd57-45d0-945b-0876f2c8cb86@moroto.mountain
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c93637e6a4c4e1d0e85ef7efac78d066bbb24d96 ]
Avoid large backtrace, it is sufficient to warn the user that there has
been a link problem. Either the link has failed and the system is in need
of maintenance, or the link continues to work and user has been informed.
The message from the warning can be looked up in the sources.
This makes an actual link issue less verbose.
First of all, this controller has a limitation in that the controller
driver has to assist the hardware with transition to L1 link state by
writing L1IATN to PMCTRL register, the L1 and L0 link state switching
is not fully automatic on this controller.
In case of an ASMedia ASM1062 PCIe SATA controller which does not support
ASPM, on entry to suspend or during platform pm_test, the SATA controller
enters D3hot state and the link enters L1 state. If the SATA controller
wakes up before rcar_pcie_wakeup() was called and returns to D0, the link
returns to L0 before the controller driver even started its transition to
L1 link state. At this point, the SATA controller did send an PM_ENTER_L1
DLLP to the PCIe controller and the PCIe controller received it, and the
PCIe controller did set PMSR PMEL1RX bit.
Once rcar_pcie_wakeup() is called, if the link is already back in L0 state
and PMEL1RX bit is set, the controller driver has no way to determine if
it should perform the link transition to L1 state, or treat the link as if
it is in L0 state. Currently the driver attempts to perform the transition
to L1 link state unconditionally, which in this specific case fails with a
PMSR L1FAEG poll timeout, however the link still works as it is already
back in L0 state.
Reduce this warning verbosity. In case the link is really broken, the
rcar_pcie_config_access() would fail, otherwise it will succeed and any
system with this controller and ASM1062 can suspend without generating
a backtrace.
Fixes: 84b5761462 ("PCI: rcar: Finish transition to L1 state in rcar_pcie_config_access()")
Link: https://lore.kernel.org/linux-pci/20240511235513.77301-1-marek.vasut+renesas@mailbox.org
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit a231707a91f323af1e5d9f1722055ec2fc1c7775 ]
If IORESOURCE_MEM is not provided in Device Tree due to
any error, resource_list_first_type() will return NULL and
pci_parse_request_of_pci_ranges() will just emit a warning.
This will cause a NULL pointer dereference. Fix this bug by adding NULL
return check.
Found by Linux Verification Center (linuxtesting.org) with SVACE.
Fixes: 0f71c60ffd ("PCI: dwc: Remove storing of PCI resources")
Link: https://lore.kernel.org/linux-pci/20240505061517.11527-1-amishin@t-argos.ru
Suggested-by: Bjorn Helgaas <helgaas@kernel.org>
Suggested-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Aleksandr Mishin <amishin@t-argos.ru>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9ffa0e70b2daf9b0271e4960b7c8a2350e2cda08 ]
After 6ab15b5e70 ("PCI: dwc: keystone: Convert .scan_bus() callback to
use add_bus"), ks_pcie_v3_65_add_bus() enabled BAR 0 for both v3.65a and
v4.90a devices. On the AM654x SoC, which uses v4.90a, enabling BAR 0
causes Completion Timeouts when setting up MSI-X. These timeouts delay
boot of the AM654x by about 45 seconds.
Move the BAR 0 initialization to ks_pcie_msi_host_init(), which is only
used for v3.65a devices, and remove ks_pcie_v3_65_add_bus().
[bhelgaas: commit log]
Fixes: 6ab15b5e70 ("PCI: dwc: keystone: Convert .scan_bus() callback to use add_bus")
Link: https://lore.kernel.org/linux-pci/20240328085041.2916899-3-s-vadapalli@ti.com
Suggested-by: Bjorn Helgaas <helgaas@kernel.org>
Suggested-by: Niklas Cassel <cassel@kernel.org>
Suggested-by: Serge Semin <fancer.lancer@gmail.com>
Signed-off-by: Siddharth Vadapalli <s-vadapalli@ti.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Niklas Cassel <cassel@kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 903534fa7d30214d8ba840ab1cd9e917e0c88e41 ]
pbus_size_mem() keeps the size of the optional resources in
children_add_size. When calculating the PCI bridge window size,
calculate_memsize() lower bounds size by old_size before adding
children_add_size and performing the window size alignment. This
results in double counting for the resources in children_add_size
because old_size may be based on the previous size of the bridge
window after it has already included children_add_size (that is,
size1 in pbus_size_mem() from an earlier invocation of that
function).
As a result, on repeated remove of the bus & rescan cycles the resource
size keeps increasing when children_add_size is non-zero as can be seen
from this extract:
iomem0: 23fffd00000-23fffdfffff : PCI Bus 0000:03 # 1MiB
iomem1: 20000000000-200001fffff : PCI Bus 0000:03 # 2MiB
iomem2: 20000000000-200002fffff : PCI Bus 0000:03 # 3MiB
iomem3: 20000000000-200003fffff : PCI Bus 0000:03 # 4MiB
iomem4: 20000000000-200004fffff : PCI Bus 0000:03 # 5MiB
Solve the double counting by moving old_size check later in
calculate_memsize() so that children_add_size is already accounted for.
After the patch, the bridge window retains its size as expected:
iomem0: 23fffd00000-23fffdfffff : PCI Bus 0000:03 # 1MiB
iomem1: 20000000000-200000fffff : PCI Bus 0000:03 # 1MiB
iomem2: 20000000000-200000fffff : PCI Bus 0000:03 # 1MiB
Fixes: a4ac9fea01 ("PCI : Calculate right add_size")
Link: https://lore.kernel.org/r/20240507102523.57320-2-ilpo.jarvinen@linux.intel.com
Tested-by: Lidong Wang <lidong.wang@intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 9eee5330656bf92f51cb1f09b2dc9f8cf975b3d1 upstream.
KFENCE reports the following UAF:
BUG: KFENCE: use-after-free read in __pci_enable_msi_range+0x2c0/0x488
Use-after-free read at 0x0000000024629571 (in kfence-#12):
__pci_enable_msi_range+0x2c0/0x488
pci_alloc_irq_vectors_affinity+0xec/0x14c
pci_alloc_irq_vectors+0x18/0x28
kfence-#12: 0x0000000008614900-0x00000000e06c228d, size=104, cache=kmalloc-128
allocated by task 81 on cpu 7 at 10.808142s:
__kmem_cache_alloc_node+0x1f0/0x2bc
kmalloc_trace+0x44/0x138
msi_alloc_desc+0x3c/0x9c
msi_domain_insert_msi_desc+0x30/0x78
msi_setup_msi_desc+0x13c/0x184
__pci_enable_msi_range+0x258/0x488
pci_alloc_irq_vectors_affinity+0xec/0x14c
pci_alloc_irq_vectors+0x18/0x28
freed by task 81 on cpu 7 at 10.811436s:
msi_domain_free_descs+0xd4/0x10c
msi_domain_free_locked.part.0+0xc0/0x1d8
msi_domain_alloc_irqs_all_locked+0xb4/0xbc
pci_msi_setup_msi_irqs+0x30/0x4c
__pci_enable_msi_range+0x2a8/0x488
pci_alloc_irq_vectors_affinity+0xec/0x14c
pci_alloc_irq_vectors+0x18/0x28
Descriptor allocation done in:
__pci_enable_msi_range
msi_capability_init
msi_setup_msi_desc
msi_insert_msi_desc
msi_domain_insert_msi_desc
msi_alloc_desc
...
Freed in case of failure in __msi_domain_alloc_locked()
__pci_enable_msi_range
msi_capability_init
pci_msi_setup_msi_irqs
msi_domain_alloc_irqs_all_locked
msi_domain_alloc_locked
__msi_domain_alloc_locked => fails
msi_domain_free_locked
...
That failure propagates back to pci_msi_setup_msi_irqs() in
msi_capability_init() which accesses the descriptor for unmasking in the
error exit path.
Cure it by copying the descriptor and using the copy for the error exit path
unmask operation.
[ tglx: Massaged change log ]
Fixes: bf6e054e0e ("genirq/msi: Provide msi_device_populate/destroy_sysfs()")
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Mostafa Saleh <smostafa@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Bjorn Heelgas <bhelgaas@google.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20240624203729.1094506-1-smostafa@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 6613443ffc49d03e27f0404978f685c4eac43fba ]
On runtime resume, pci_dev_wait() is called:
pci_pm_runtime_resume()
pci_pm_bridge_power_up_actions()
pci_bridge_wait_for_secondary_bus()
pci_dev_wait()
While a device is runtime suspended along with its PCI hierarchy, the
device could get disconnected. In such case, the link will not come up no
matter how long pci_dev_wait() waits for it.
Besides the above mentioned case, there could be other ways to get the
device disconnected while pci_dev_wait() is waiting for the link to come
up.
Make pci_dev_wait() exit if the device is already disconnected to avoid
unnecessary delay.
The use cases of pci_dev_wait() boil down to two:
1. Waiting for the device after reset
2. pci_bridge_wait_for_secondary_bus()
The callers in both cases seem to benefit from propagating the
disconnection as error even if device disconnection would be more
analoguous to the case where there is no device in the first place which
return 0 from pci_dev_wait(). In the case 2, it results in unnecessary
marking of the devices disconnected again but that is just harmless extra
work.
Also make sure compiler does not become too clever with dev->error_state
and use READ_ONCE() to force a fetch for the up-to-date value.
Link: https://lore.kernel.org/r/20240208132322.4811-1-ilpo.jarvinen@linux.intel.com
Reported-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Tested-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 256df20c590bf0e4d63ac69330cf23faddac3e08 ]
Hewlett-Packard HP Pavilion 17 Notebook PC/1972 is an Intel Ivy Bridge
system with a muxless AMD Radeon dGPU. Attempting to use the dGPU fails
with the following sequence:
ACPI Error: Aborting method \AMD3._ON due to previous error (AE_AML_LOOP_TIMEOUT) (20230628/psparse-529)
radeon 0000:01:00.0: not ready 1023ms after resume; waiting
radeon 0000:01:00.0: not ready 2047ms after resume; waiting
radeon 0000:01:00.0: not ready 4095ms after resume; waiting
radeon 0000:01:00.0: not ready 8191ms after resume; waiting
radeon 0000:01:00.0: not ready 16383ms after resume; waiting
radeon 0000:01:00.0: not ready 32767ms after resume; waiting
radeon 0000:01:00.0: not ready 65535ms after resume; giving up
radeon 0000:01:00.0: Unable to change power state from D3cold to D0, device inaccessible
The issue is that the Root Port the dGPU is connected to can't handle the
transition from D3cold to D0 so the dGPU can't properly exit runtime PM.
The existing logic in pci_bridge_d3_possible() checks for systems that are
newer than 2015 to decide that D3 is safe. This would nominally work for
an Ivy Bridge system (which was discontinued in 2015), but this system
appears to have continued to receive BIOS updates until 2017 and so this
existing logic doesn't appropriately capture it.
Add the system to bridge_d3_blacklist to prevent D3cold from being used.
Link: https://lore.kernel.org/r/20240307163709.323-1-mario.limonciello@amd.com
Reported-by: Eric Heintzmann <heintzmann.eric@free.fr>
Closes: https://gitlab.freedesktop.org/drm/amd/-/issues/3229
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Eric Heintzmann <heintzmann.eric@free.fr>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 2dba285caba53f309d6060fca911b43d63f41697 upstream.
Remove wrong mask on subsys_vendor_id. Both the Vendor ID and Subsystem
Vendor ID are u16 variables and are written to a u32 register of the
controller. The Subsystem Vendor ID was always 0 because the u16 value
was masked incorrectly with GENMASK(31,16) resulting in all lower 16
bits being set to 0 prior to the shift.
Remove both masks as they are unnecessary and set the register correctly
i.e., the lower 16-bits are the Vendor ID and the upper 16-bits are the
Subsystem Vendor ID.
This is documented in the RK3399 TRM section 17.6.7.1.17
[kwilczynski: removed unnecesary newline]
Fixes: cf590b0783 ("PCI: rockchip: Add EP driver for Rockchip PCIe controller")
Link: https://lore.kernel.org/linux-pci/20240403144508.489835-1-rick.wertenbroek@gmail.com
Signed-off-by: Rick Wertenbroek <rick.wertenbroek@gmail.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Damien Le Moal <dlemoal@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit e2e78a294a8a863898b781dbcf90e087eda3155d ]
The "Downstream Port Containment related Enhancements" ECN of Jan 28, 2019
(document 12888 below), defined the EDR_PORT_LOCATE_DSM function with
Revision ID 5 with a return value encoding (Bits 2:0 = Function, Bits 7:3 =
Device, Bits 15:8 = Bus). When the ECN was integrated into PCI Firmware
r3.3, sec 4.6.13, Bit 31 was added to indicate success or failure.
Check Bit 31 for failure in acpi_dpc_port_get().
Link: https://lore.kernel.org/r/20240501022543.1626025-1-sathyanarayanan.kuppuswamy@linux.intel.com
Link: https://members.pcisig.com/wg/PCI-SIG/document/12888
Fixes: ac1c8e35a3 ("PCI/DPC: Add Error Disconnect Recover (EDR) support")
Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
[bhelgaas: split into two patches, update commit log]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Satish Thatchanamurthy <Satish.Thatchanamurt@Dell.com> # one platform
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit f24ba846133d0edec785ac6430d4daf6e9c93a09 ]
The "Downstream Port Containment related Enhancements" ECN of Jan 28, 2019
(document 12888 below), defined the EDR_PORT_DPC_ENABLE_DSM function with
Revision ID 5 with Arg3 being an integer. But when the ECN was integrated
into PCI Firmware r3.3, sec 4.6.12, it was defined as Revision ID 6 with
Arg3 being a package containing an integer.
The implementation in acpi_enable_dpc() supplies a package as Arg3 (arg4 in
the code), but it previously specified Revision ID 5. Align this with PCI
Firmware r3.3 by using Revision ID 6.
If firmware implemented per the ECN, its Revision 5 function would receive
a package as Arg3 when it expects an integer, so acpi_enable_dpc() would
likely fail. If such firmware exists and lacks a Revision 6 function that
expects a package, we may have to add support for Revision 5.
Link: https://lore.kernel.org/r/20240501022543.1626025-1-sathyanarayanan.kuppuswamy@linux.intel.com
Link: https://members.pcisig.com/wg/PCI-SIG/document/12888
Fixes: ac1c8e35a3 ("PCI/DPC: Add Error Disconnect Recover (EDR) support")
Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
[bhelgaas: split into two patches, update commit log]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Satish Thatchanamurthy <Satish.Thatchanamurt@Dell.com> # one platform
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 73cb3a35f94db723c0211ad099bce55b2155e3f0 ]
Two changes were made in link retraining logic independent of each other.
The commit e7e3975636 ("PCI/ASPM: Avoid link retraining race") added a
check to pcie_retrain_link() to ensure no Link Training is currently active
to address the Implementation Note in PCIe r6.1 sec 7.5.3.7. At that time
pcie_wait_for_retrain() only checked for the Link Training (LT) bit being
cleared.
The commit 680e9c47a2 ("PCI: Add support for polling DLLLA to
pcie_retrain_link()") generalized pcie_wait_for_retrain() into
pcie_wait_for_link_status() which can wait either for LT or the Data Link
Layer Link Active (DLLLA) bit with 'use_lt' argument and supporting waiting
for either cleared or set using 'active' argument.
In the merge commit 1abb473903 ("Merge branch 'pci/enumeration'"), those
two divergent branches converged. The merge changed LT bit checking added
in the commit e7e3975636 ("PCI/ASPM: Avoid link retraining race") to now
wait for completion of any ongoing Link Training using DLLLA bit being set
if 'use_lt' is false.
When 'use_lt' is false, the pseudo-code steps of what occurs in
pcie_retrain_link():
1. Wait for DLLLA==1
2. Trigger link to retrain
3. Wait for DLLLA==1
Step 3 waits for the link to come up from the retraining triggered by Step
2. As Step 1 is supposed to wait for any ongoing retraining to end, using
DLLLA also for it does not make sense because link training being active is
still indicated using LT bit, not with DLLLA.
Correct the pcie_wait_for_link_status() parameters in Step 1 to only wait
for LT==0 to ensure there is no ongoing Link Training.
This only impacts the Target Speed quirk, which is the only case where
waiting for DLLLA bit is used. It currently works in the problematic case
by means of link training getting initiated by hardware repeatedly and
respecting the new link parameters set by the caller, which then make
training succeed and bring the link up, setting DLLLA and causing
pcie_wait_for_link_status() to return success. We are not supposed to rely
on luck and need to make sure that LT transitioned through the inactive
state though before we initiate link training by hand via RL (Retrain Link)
bit.
Fixes: 1abb473903 ("Merge branch 'pci/enumeration'")
Link: https://lore.kernel.org/r/20240423130820.43824-1-ilpo.jarvinen@linux.intel.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 19326006a21da26532d982254677c892dae8f29b ]
Tegra194 PCIe probe path is taking failure path in success case for
Endpoint mode. Return success from the switch case instead of going
into the failure path.
Fixes: c57247f940 ("PCI: tegra: Add support for PCIe endpoint mode in Tegra194")
Link: https://lore.kernel.org/linux-pci/20240408093053.3948634-1-vidyas@nvidia.com
Signed-off-by: Vidya Sagar <vidyas@nvidia.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Reviewed-by: Jon Hunter <jonathanh@nvidia.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit 1e560864159d002b453da42bd2c13a1805515a20 upstream.
A last minute revert in 6.7-final introduced a potential deadlock when
enabling ASPM during probe of Qualcomm PCIe controllers as reported by
lockdep:
============================================
WARNING: possible recursive locking detected
6.7.0 #40 Not tainted
--------------------------------------------
kworker/u16:5/90 is trying to acquire lock:
ffffacfa78ced000 (pci_bus_sem){++++}-{3:3}, at: pcie_aspm_pm_state_change+0x58/0xdc
but task is already holding lock:
ffffacfa78ced000 (pci_bus_sem){++++}-{3:3}, at: pci_walk_bus+0x34/0xbc
other info that might help us debug this:
Possible unsafe locking scenario:
CPU0
----
lock(pci_bus_sem);
lock(pci_bus_sem);
*** DEADLOCK ***
Call trace:
print_deadlock_bug+0x25c/0x348
__lock_acquire+0x10a4/0x2064
lock_acquire+0x1e8/0x318
down_read+0x60/0x184
pcie_aspm_pm_state_change+0x58/0xdc
pci_set_full_power_state+0xa8/0x114
pci_set_power_state+0xc4/0x120
qcom_pcie_enable_aspm+0x1c/0x3c [pcie_qcom]
pci_walk_bus+0x64/0xbc
qcom_pcie_host_post_init_2_7_0+0x28/0x34 [pcie_qcom]
The deadlock can easily be reproduced on machines like the Lenovo ThinkPad
X13s by adding a delay to increase the race window during asynchronous
probe where another thread can take a write lock.
Add a new pci_set_power_state_locked() and associated helper functions that
can be called with the PCI bus semaphore held to avoid taking the read lock
twice.
Link: https://lore.kernel.org/r/ZZu0qx2cmn7IwTyQ@hovoldconsulting.com
Link: https://lore.kernel.org/r/20240130100243.11011-1-johan+linaro@kernel.org
Fixes: f93e71aea6c6 ("Revert "PCI/ASPM: Remove pcie_aspm_pm_state_change()"")
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: <stable@vger.kernel.org> # 6.7
[bhelgaas: backported to v6.6.y, which contains 8cc22ba3f7 ("Revert
"PCI/ASPM: Remove pcie_aspm_pm_state_change()""), a backport of
f93e71aea6c6. This omits the drivers/pci/controller/dwc/pcie-qcom.c hunk
that updates qcom_pcie_enable_aspm(), which was added by 9f4f3dfad8cf
("PCI: qcom: Enable ASPM for platforms supporting 1.9.0 ops"), which is not
present in v6.6.28.]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 9a9eec4765737b9b2a8d6ae03de6480a5f12dd5c ]
Use FIELD_GET() to remove dependencies on the field position, i.e., the
shift value. No functional change intended.
Link: https://lore.kernel.org/r/20231018113254.17616-5-ilpo.jarvinen@linux.intel.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit b5ff74c1ef50fe08e384026875fec660fadfaedd ]
For a physical PCI device that is passed through to a Hyper-V guest VM,
current code specifies the VMBus ring buffer size as 4 pages. But this
is an inappropriate dependency, since the amount of ring buffer space
needed is unrelated to PAGE_SIZE. For example, on x86 the ring buffer
size ends up as 16 Kbytes, while on ARM64 with 64 Kbyte pages, the ring
size bloats to 256 Kbytes. The ring buffer for PCI pass-thru devices
is used for only a few messages during device setup and removal, so any
space above a few Kbytes is wasted.
Fix this by declaring the ring buffer size to be a fixed 16 Kbytes.
Furthermore, use the VMBUS_RING_SIZE() macro so that the ring buffer
header is properly accounted for, and so the size is rounded up to a
page boundary, using the page size for which the kernel is built. While
w/64 Kbyte pages this results in a 64 Kbyte ring buffer header plus a
64 Kbyte ring buffer, that's the smallest possible with that page size.
It's still 128 Kbytes better than the current code.
Link: https://lore.kernel.org/linux-pci/20240216202240.251818-1-mhklinux@outlook.com
Signed-off-by: Michael Kelley <mhklinux@outlook.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Reviewed-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Reviewed-by: Ilpo Jarvinen <ilpo.jarvinen@linux.intel.com>
Reviewed-by: Long Li <longli@microsoft.com>
Cc: <stable@vger.kernel.org> # 5.15.x
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 72e34b8593e08a0ee759b7a038e0b178418ea6f8 ]
The commit message in commit fc9a77040b ("PCI: designware-ep: Configure
Resizable BAR cap to advertise the smallest size") claims that it modifies
the Resizable BAR capability to only advertise support for 1 MB size BARs.
However, the commit writes all zeroes to PCI_REBAR_CAP (the register which
contains the possible BAR sizes that a BAR be resized to).
According to the spec, it is illegal to not have a bit set in
PCI_REBAR_CAP, and 1 MB is the smallest size allowed.
Set bit 4 in PCI_REBAR_CAP, so that we actually advertise support for a
1 MB BAR size.
Before:
Capabilities: [2e8 v1] Physical Resizable BAR
BAR 0: current size: 1MB
BAR 1: current size: 1MB
BAR 2: current size: 1MB
BAR 3: current size: 1MB
BAR 4: current size: 1MB
BAR 5: current size: 1MB
After:
Capabilities: [2e8 v1] Physical Resizable BAR
BAR 0: current size: 1MB, supported: 1MB
BAR 1: current size: 1MB, supported: 1MB
BAR 2: current size: 1MB, supported: 1MB
BAR 3: current size: 1MB, supported: 1MB
BAR 4: current size: 1MB, supported: 1MB
BAR 5: current size: 1MB, supported: 1MB
Fixes: fc9a77040b ("PCI: designware-ep: Configure Resizable BAR cap to advertise the smallest size")
Link: https://lore.kernel.org/linux-pci/20240307111520.3303774-1-cassel@kernel.org
Signed-off-by: Niklas Cassel <cassel@kernel.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Cc: <stable@vger.kernel.org> # 5.2
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit bf79e33cdd89db498e00a6131e937259de5f2705 ]
Qcom SoCs making use of ARM SMMU require BDF to SID translation table in
the driver to properly map the SID for the PCIe devices based on their BDF
identifier. This is currently achieved with the help of
qcom_pcie_config_sid_1_9_0() function for SoCs supporting the 1_9_0 config.
But With newer Qcom SoCs starting from SM8450, BDF to SID translation is
set to bypass mode by default in hardware. Due to this, the translation
table that is set in the qcom_pcie_config_sid_1_9_0() is essentially
unused and the default SID is used for all endpoints in SoCs starting from
SM8450.
This is a security concern and also warrants swapping the DeviceID in DT
while using the GIC ITS to handle MSIs from endpoints. The swapping is
currently done like below in DT when using GIC ITS:
/*
* MSIs for BDF (1:0.0) only works with Device ID 0x5980.
* Hence, the IDs are swapped.
*/
msi-map = <0x0 &gic_its 0x5981 0x1>,
<0x100 &gic_its 0x5980 0x1>;
Here, swapping of the DeviceIDs ensure that the endpoint with BDF (1:0.0)
gets the DeviceID 0x5980 which is associated with the default SID as per
the iommu mapping in DT. So MSIs were delivered with IDs swapped so far.
But this also means the Root Port (0:0.0) won't receive any MSIs (for PME,
AER etc...)
So let's fix these issues by clearing the BDF to SID bypass mode for all
SoCs making use of the 1_9_0 config. This allows the PCIe devices to use
the correct SID, thus avoiding the DeviceID swapping hack in DT and also
achieving the isolation between devices.
Fixes: 4c93988221 ("PCI: qcom: Add support for configuring BDF to SID mapping for SM8250")
Link: https://lore.kernel.org/linux-pci/20240307-pci-bdf-sid-fix-v1-1-9423a7e2d63c@linaro.org
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Cc: stable@vger.kernel.org # 5.11
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 002bf2fbc00e5c4b95fb167287e2ae7d1973281e ]
PM runtime can be done simultaneously with AER error handling. Avoid that
by using pm_runtime_get_sync() before and pm_runtime_put() after reset in
pcie_do_recovery() for all recovering devices.
pm_runtime_get_sync() will increase dev->power.usage_count counter to
prevent any possible future request to runtime suspend a device. It will
also resume a device, if it was previously in D3hot state.
I tested with igc device by doing simultaneous aer_inject and rpm
suspend/resume via /sys/bus/pci/devices/PCI_ID/power/control and can
reproduce:
igc 0000:02:00.0: not ready 65535ms after bus reset; giving up
pcieport 0000:00:1c.2: AER: Root Port link has been reset (-25)
pcieport 0000:00:1c.2: AER: subordinate device reset failed
pcieport 0000:00:1c.2: AER: device recovery failed
igc 0000:02:00.0: Unable to change power state from D3hot to D0, device inaccessible
The problem disappears when this patch is applied.
Link: https://lore.kernel.org/r/20240212120135.146068-1-stanislaw.gruszka@linux.intel.com
Signed-off-by: Stanislaw Gruszka <stanislaw.gruszka@linux.intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Acked-by: Rafael J. Wysocki <rafael@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 627c6db20703b5d18d928464f411d0d4ec327508 ]
Commit 5459c0b704 ("PCI/DPC: Quirk PIO log size for certain Intel Root
Ports") and commit 3b8803494a ("PCI/DPC: Quirk PIO log size for Intel Ice
Lake Root Ports") add quirks for Ice, Tiger and Alder Lake Root Ports.
System firmware for Raptor Lake still has the bug, so Linux logs the
warning below on several Raptor Lake systems like Dell Precision 3581 with
Intel Raptor Lake processor (0W18NX) system firmware/BIOS version 1.10.1.
pci 0000:00:07.0: [8086:a76e] type 01 class 0x060400
pci 0000:00:07.0: DPC: RP PIO log size 0 is invalid
pci 0000:00:07.1: [8086:a73f] type 01 class 0x060400
pci 0000:00:07.1: DPC: RP PIO log size 0 is invalid
Apply the quirk for Raptor Lake Root Ports as well.
This also enables the DPC driver to dump the RP PIO Log registers when DPC
is triggered.
Link: https://lore.kernel.org/r/20240305113057.56468-1-pmenzel@molgen.mpg.de
Reported-by: Niels van Aert <nvaert1986@hotmail.com>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218560
Signed-off-by: Paul Menzel <pmenzel@molgen.mpg.de>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: <stable@vger.kernel.org>
Cc: Mika Westerberg <mika.westerberg@linux.intel.com>
Cc: Niels van Aert <nvaert1986@hotmail.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9d5286d4e7f68beab450deddbb6a32edd5ecf4bf ]
A race condition between the .runtime_idle() callback and the .remove()
callback in the rtsx_pcr PCI driver leads to a kernel crash due to an
unhandled page fault [1].
The problem is that rtsx_pci_runtime_idle() is not expected to be running
after pm_runtime_get_sync() has been called, but the latter doesn't really
guarantee that. It only guarantees that the suspend and resume callbacks
will not be running when it returns.
However, if a .runtime_idle() callback is already running when
pm_runtime_get_sync() is called, the latter will notice that the runtime PM
status of the device is RPM_ACTIVE and it will return right away without
waiting for the former to complete. In fact, it cannot wait for
.runtime_idle() to complete because it may be called from that callback (it
arguably does not make much sense to do that, but it is not strictly
prohibited).
Thus in general, whoever is providing a .runtime_idle() callback needs
to protect it from running in parallel with whatever code runs after
pm_runtime_get_sync(). [Note that .runtime_idle() will not start after
pm_runtime_get_sync() has returned, but it may continue running then if it
has started earlier.]
One way to address that race condition is to call pm_runtime_barrier()
after pm_runtime_get_sync() (not before it, because a nonzero value of the
runtime PM usage counter is necessary to prevent runtime PM callbacks from
being invoked) to wait for the .runtime_idle() callback to complete should
it be running at that point. A suitable place for doing that is in
pci_device_remove() which calls pm_runtime_get_sync() before removing the
driver, so it may as well call pm_runtime_barrier() subsequently, which
will prevent the race in question from occurring, not just in the rtsx_pcr
driver, but in any PCI drivers providing .runtime_idle() callbacks.
Link: https://lore.kernel.org/lkml/20240229062201.49500-1-kai.heng.feng@canonical.com/ # [1]
Link: https://lore.kernel.org/r/5761426.DvuYhMxLoT@kreacher
Reported-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Ricky Wu <ricky_wu@realtek.com>
Acked-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 039741a8d7c9a01c1bc84a5ac5aa770a5e138a30 ]
The MDIO_WT_DONE() macro tests bit 31, which is always 0 (== done) as
readw_poll_timeout_atomic() does a 16-bit read. Replace with the readl
variant.
[kwilczynski: commit log]
Fixes: ca5dcc7631 ("PCI: brcmstb: Replace status loops with read_poll_timeout_atomic()")
Link: https://lore.kernel.org/linux-pci/20240217133722.14391-1-wahrenst@gmx.net
Signed-off-by: Jonathan Bell <jonathan@raspberrypi.com>
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Acked-by: Florian Fainelli <florian.fainelli@broadcom.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit baf67aefbe7d7deafa59ca49612d163f8889934c ]
Per PCIe r6.1, sec 2.2.6.2 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 9650SE 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.
Link: https://lore.kernel.org/r/20240219132811.8351-1-joerg@wedekind.de
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>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit aebfdfe39b9327a3077d0df8db3beb3160c9bdd0 ]
If device_register() fails in ntb_register_device(), the device name
allocated by dev_set_name() should be freed. As per the comment in
device_register(), callers should use put_device() to give up the
reference in the error path. So fix this by calling put_device() in the
error path so that the name can be freed in kobject_cleanup().
As a result of this, put_device() in the error path of
ntb_register_device() is removed and the actual error is returned.
Fixes: a1bd3baeb2 ("NTB: Add NTB hardware abstraction layer")
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Link: https://lore.kernel.org/r/20231201033057.1399131-1-yangyingliang@huaweicloud.com
[mani: reworded commit message]
Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit dec529b0b0572b32f9eb91c882dd1f08ca657efb ]
The commit in Fixes changed the logic on how resources are released and
introduced a new switchtec_exit_pci() that need to be called explicitly in
order to undo a corresponding switchtec_init_pci().
This was done in the remove function, but not in the probe.
Fix the probe now.
Fixes: df25461119d9 ("PCI: switchtec: Fix stdev_release() crash after surprise hot remove")
Link: https://lore.kernel.org/r/01446d2ccb91a578239915812f2b7dfbeb2882af.1703428183.git.christophe.jaillet@wanadoo.fr
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1e5c66afd4a40bb7be17cb33cbb1a1085f727730 ]
It is not allowed to sleep within a RCU read section, so use GFP_ATOMIC
instead of GFP_KERNEL here.
Link: https://lore.kernel.org/r/02d9ec4a10235def0e764ff1f5be881ba12e16e8.1704397858.git.christophe.jaillet@wanadoo.fr
Fixes: ae21f835a5 ("PCI/P2PDMA: Finish RCU conversion of pdev->p2pdma")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 6568d82512b0a64809acff3d7a747362fa4288c8 ]
The TLP Prefix Log Register consists of multiple DWORDs (PCIe r6.1 sec
7.9.14.13) but the loop in dpc_process_rp_pio_error() keeps reading from
the first DWORD, so we print only the first PIO TLP Prefix (duplicated
several times), and we never print the second, third, etc., Prefixes.
Add the iteration count based offset calculation into the config read.
Fixes: f20c4ea49e ("PCI/DPC: Add eDPC support")
Link: https://lore.kernel.org/r/20240118110815.3867-1-ilpo.jarvinen@linux.intel.com
Signed-off-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
[bhelgaas: add user-visible details to commit log]
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 39714fd73c6b60a8d27bcc5b431afb0828bf4434 ]
Make pci_dev_is_disconnected() public so that it can be called from
Intel VT-d driver to quickly fix/workaround the surprise removal
unplug hang issue for those ATS capable devices on PCIe switch downstream
hotplug capable ports.
Beside pci_device_is_present() function, this one has no config space
space access, so is light enough to optimize the normal pure surprise
removal and safe removal flow.
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Dan Carpenter <dan.carpenter@linaro.org>
Tested-by: Haorong Ye <yehaorong@bytedance.com>
Signed-off-by: Ethan Zhao <haifeng.zhao@linux.intel.com>
Link: https://lore.kernel.org/r/20240301080727.3529832-2-haifeng.zhao@linux.intel.com
Signed-off-by: Lu Baolu <baolu.lu@linux.intel.com>
Signed-off-by: Joerg Roedel <jroedel@suse.de>
Stable-dep-of: 4fc82cd907ac ("iommu/vt-d: Don't issue ATS Invalidation request when device is disconnected")
Signed-off-by: Sasha Levin <sashal@kernel.org>
commit db744ddd59be798c2627efbfc71f707f5a935a40 upstream.
While calculating the hardware interrupt number for a MSI interrupt, the
higher bits (i.e. from bit-5 onwards a.k.a domain_nr >= 32) of the PCI
domain number gets truncated because of the shifted value casting to return
type of pci_domain_nr() which is 'int'. This for example is resulting in
same hardware interrupt number for devices 0019:00:00.0 and 0039:00:00.0.
To address this cast the PCI domain number to 'irq_hw_number_t' before left
shifting it to calculate the hardware interrupt number.
Please note that this fixes the issue only on 64-bit systems and doesn't
change the behavior for 32-bit systems i.e. the 32-bit systems continue to
have the issue. Since the issue surfaces only if there are too many PCIe
controllers in the system which usually is the case in modern server
systems and they don't tend to run 32-bit kernels.
Fixes: 3878eaefb8 ("PCI/MSI: Enhance core to support hierarchy irqdomain")
Signed-off-by: Vidya Sagar <vidyas@nvidia.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Shanker Donthineni <sdonthineni@nvidia.com>
Cc: stable@vger.kernel.org
Link: https://lore.kernel.org/r/20240115135649.708536-1-vidyas@nvidia.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit b5d1b4b46f856da1473c7ba9a5cdfcb55c9b2478 upstream.
The "msg_addr" variable is u64. However, the "aligned_offset" is an
unsigned int. This means that when the code does:
msg_addr &= ~aligned_offset;
it will unintentionally zero out the high 32 bits. Use ALIGN_DOWN() to do
the alignment instead.
Fixes: 2217fffcd63f ("PCI: dwc: endpoint: Fix dw_pcie_ep_raise_msix_irq() alignment support")
Link: https://lore.kernel.org/r/af59c7ad-ab93-40f7-ad4a-7ac0b14d37f5@moroto.mountain
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Niklas Cassel <cassel@kernel.org>
Reviewed-by: Ilpo Järvinen <ilpo.jarvinen@linux.intel.com>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Niklas Cassel <cassel@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
[ Upstream commit 41044d5360685e78a869d40a168491a70cdb7e73 ]
The commit noted in fixes added a bogus requirement that runtime PM managed
devices need to be in the RPM_ACTIVE state for PME polling. In fact, only
devices in low power states should be polled.
However there's still a requirement that the device config space must be
accessible, which has implications for both the current state of the polled
device and the parent bridge, when present. It's not sufficient to assume
the bridge remains in D0 and cases have been observed where the bridge
passes the D0 test, but the PM state indicates RPM_SUSPENDING and config
space of the polled device becomes inaccessible during pci_pme_wakeup().
Therefore, since the bridge is already effectively required to be in the
RPM_ACTIVE state, formalize this in the code and elevate the PM usage count
to maintain the state while polling the subordinate device.
This resolves a regression reported in the bugzilla below where a
Thunderbolt/USB4 hierarchy fails to scan for an attached NVMe endpoint
downstream of a bridge in a D3hot power state.
Link: https://lore.kernel.org/r/20240123185548.1040096-1-alex.williamson@redhat.com
Fixes: d3fcd73603 ("PCI: Fix runtime PM race with PME polling")
Reported-by: Sanath S <sanath.s@amd.com>
Closes: https://bugzilla.kernel.org/show_bug.cgi?id=218360
Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Tested-by: Sanath S <sanath.s@amd.com>
Reviewed-by: Rafael J. Wysocki <rafael@kernel.org>
Cc: Lukas Wunner <lukas@wunner.de>
Cc: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 1291b716bbf969e101d517bfb8ba18d958f758b8 ]
When a device with AER detects an error, it logs error information in its
own AER Error Status registers. It may send an Error Message to the Root
Port (RCEC in the case of an RCiEP), which logs the fact that an Error
Message was received (Root Error Status) and the Requester ID of the
message source (Error Source Identification).
aer_print_port_info() prints the Requester ID from the Root Port Error
Source in the usual Linux "bb:dd.f" format, but when find_source_device()
finds no error details in the hierarchy below the Root Port, it printed the
raw Requester ID without decoding it.
Decode the Requester ID in the usual Linux format so it matches other
messages.
Sample message changes:
- pcieport 0000:00:1c.5: AER: Correctable error received: 0000:00:1c.5
- pcieport 0000:00:1c.5: AER: can't find device of ID00e5
+ pcieport 0000:00:1c.5: AER: Correctable error message received from 0000:00:1c.5
+ pcieport 0000:00:1c.5: AER: found no error details for 0000:00:1c.5
Link: https://lore.kernel.org/r/20231206224231.732765-3-helgaas@kernel.org
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Jonathan Cameron <Jonathan.Cameron@huawei.com>
Reviewed-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit df25461119d987b8c81d232cfe4411e91dcabe66 ]
A PCI device hot removal may occur while stdev->cdev is held open. The call
to stdev_release() then happens during close or exit, at a point way past
switchtec_pci_remove(). Otherwise the last ref would vanish with the
trailing put_device(), just before return.
At that later point in time, the devm cleanup has already removed the
stdev->mmio_mrpc mapping. Also, the stdev->pdev reference was not a counted
one. Therefore, in DMA mode, the iowrite32() in stdev_release() will cause
a fatal page fault, and the subsequent dma_free_coherent(), if reached,
would pass a stale &stdev->pdev->dev pointer.
Fix by moving MRPC DMA shutdown into switchtec_pci_remove(), after
stdev_kill(). Counting the stdev->pdev ref is now optional, but may prevent
future accidents.
Reproducible via the script at
https://lore.kernel.org/r/20231113212150.96410-1-dns@arista.com
Link: https://lore.kernel.org/r/20231122042316.91208-2-dns@arista.com
Signed-off-by: Daniel Stodden <dns@arista.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Dmitry Safonov <dima@arista.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit e585a37e5061f6d5060517aed1ca4ccb2e56a34c ]
By running a Van Gogh device (Steam Deck), the following message
was noticed in the kernel log:
pci 0000:04:00.3: PCI class overridden (0x0c03fe -> 0x0c03fe) so dwc3 driver can claim this instead of xhci
Effectively this means the quirk executed but changed nothing, since the
class of this device was already the proper one (likely adjusted by newer
firmware versions).
Check and perform the override only if necessary.
Link: https://lore.kernel.org/r/20231120160531.361552-1-gpiccoli@igalia.com
Signed-off-by: Guilherme G. Piccoli <gpiccoli@igalia.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: Huang Rui <ray.huang@amd.com>
Cc: Vicki Pfau <vi@endrift.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 3ed48c80b28d8dcd584d6ddaf00c75b7673e1a05 ]
Spectrum-{1,2,3,4} devices report that a D3hot->D0 transition causes a
reset (i.e., they advertise NoSoftRst-). However, this transition does
not have any effect on the device: It continues to be operational and
network ports remain up. Advertising this support makes it seem as if a
PM reset is viable for these devices. Mark it as unavailable to skip it
when testing reset methods.
Before:
# cat /sys/bus/pci/devices/0000\:03\:00.0/reset_method
pm bus
After:
# cat /sys/bus/pci/devices/0000\:03\:00.0/reset_method
bus
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit 9ccc1318cf4bd90601f221268e42c3374703d681 ]
When using the fls() helper, the translation table should be a power of
two; otherwise, the resulting value will not be correct.
For example, given fls(0x3e00000) - 1 = 25, the PCIe translation window
size will be set to 0x2000000 instead of the expected size 0x3e00000.
Fix the translation window by splitting the MMIO space into multiple tables
if its size is not a power of two.
[kwilczynski: commit log]
Link: https://lore.kernel.org/linux-pci/20231023081423.18559-1-jianjun.wang@mediatek.com
Fixes: d3bf75b579 ("PCI: mediatek-gen3: Add MediaTek Gen3 driver for MT8192")
Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
[ Upstream commit c12ca110c613a81cb0f0099019c839d078cd0f38 ]
The PCI driver invokes the PHY APIs using the ks_pcie_enable_phy()
function. The PHY in this case is the Serdes. It is possible that the
PCI instance is configured for two lane operation across two different
Serdes instances, using one lane of each Serdes.
In such a configuration, if the reference clock for one Serdes is
provided by the other Serdes, it results in a race condition. After the
Serdes providing the reference clock is initialized by the PCI driver by
invoking its PHY APIs, it is not guaranteed that this Serdes remains
powered on long enough for the PHY APIs based initialization of the
dependent Serdes. In such cases, the PLL of the dependent Serdes fails
to lock due to the absence of the reference clock from the former Serdes
which has been powered off by the PM Core.
Fix this by obtaining reference to the PHYs before invoking the PHY
initialization APIs and releasing reference after the initialization is
complete.
Link: https://lore.kernel.org/linux-pci/20230927041845.1222080-1-s-vadapalli@ti.com
Fixes: 49229238ab ("PCI: keystone: Cleanup PHY handling")
Signed-off-by: Siddharth Vadapalli <s-vadapalli@ti.com>
Signed-off-by: Krzysztof Wilczyński <kwilczynski@kernel.org>
Acked-by: Ravi Gunasekaran <r-gunasekaran@ti.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
Sui Jingfeng