kernel/git/stable/linux.git/drivers/pci/controller, branch linux-6.7.y

PCI: hv: Fix ring buffer size calculation

2024-04-03T13:11:22+00:00

[ 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 Signed-off-by: Krzysztof Wilczyński Reviewed-by: Kuppuswamy Sathyanarayanan Reviewed-by: Ilpo Jarvinen Reviewed-by: Long Li Cc: # 5.15.x Signed-off-by: Sasha Levin

PCI: dwc: endpoint: Fix advertised resizable BAR size

2024-04-03T13:11:22+00:00

[ Upstream commit 72e34b8593e08a0ee759b7a038e0b178418ea6f8 ] The commit message in commit fc9a77040b04 ("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: fc9a77040b04 ("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 Signed-off-by: Krzysztof Wilczyński Reviewed-by: Manivannan Sadhasivam Cc: # 5.2 Signed-off-by: Sasha Levin

PCI: qcom: Enable BDF to SID translation properly

2024-04-03T13:11:22+00:00

[ 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: 4c9398822106 ("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 Signed-off-by: Krzysztof Wilczyński Cc: stable@vger.kernel.org # 5.11 Signed-off-by: Sasha Levin

PCI: qcom: Disable ASPM L0s for sc8280xp, sa8540p and sa8295p

2024-04-03T13:11:21+00:00

[ Upstream commit d1997c98781459f7b6d0bf1858f538f48454a97b ] Commit 9f4f3dfad8cf ("PCI: qcom: Enable ASPM for platforms supporting 1.9.0 ops") started enabling ASPM unconditionally when the hardware claims to support it. This triggers Correctable Errors for some PCIe devices on machines like the Lenovo ThinkPad X13s when L0s is enabled, which could indicate an incomplete driver ASPM implementation or that the hardware does in fact not support L0s. This has now been confirmed by Qualcomm to be the case for sc8280xp and its derivate platforms (e.g. sa8540p and sa8295p). Specifically, the PHY configuration used on these platforms is not correctly tuned for L0s and there is currently no updated configuration available. Add a new flag to the driver configuration data and use it to disable ASPM L0s on sc8280xp, sa8540p and sa8295p for now. Note that only the 1.9.0 ops enable ASPM currently. Link: https://lore.kernel.org/r/20240306095651.4551-4-johan+linaro@kernel.org Fixes: 9f4f3dfad8cf ("PCI: qcom: Enable ASPM for platforms supporting 1.9.0 ops") Signed-off-by: Johan Hovold Signed-off-by: Lorenzo Pieralisi Reviewed-by: Manivannan Sadhasivam Cc: stable@vger.kernel.org # 6.7 Signed-off-by: Sasha Levin

PCI: brcmstb: Fix broken brcm_pcie_mdio_write() polling

2024-03-26T22:18:43+00:00

[ 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: ca5dcc76314d ("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 Signed-off-by: Stefan Wahren Signed-off-by: Krzysztof Wilczyński Acked-by: Florian Fainelli Signed-off-by: Sasha Levin

PCI: dwc: Fix a 64bit bug in dw_pcie_ep_raise_msix_irq()

2024-02-23T08:51:52+00:00

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 Signed-off-by: Bjorn Helgaas Reviewed-by: Niklas Cassel Reviewed-by: Ilpo Järvinen Reviewed-by: Manivannan Sadhasivam Cc: Signed-off-by: Greg Kroah-Hartman

PCI/ASPM: Fix deadlock when enabling ASPM

2024-02-16T18:14:30+00:00

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 Signed-off-by: Bjorn Helgaas Cc: # 6.7 Signed-off-by: Greg Kroah-Hartman

PCI: mediatek-gen3: Fix translation window size calculation

2024-01-25T23:45:26+00:00

[ 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: d3bf75b579b9 ("PCI: mediatek-gen3: Add MediaTek Gen3 driver for MT8192") Signed-off-by: Jianjun Wang Signed-off-by: Krzysztof Wilczyński Signed-off-by: Bjorn Helgaas Reviewed-by: AngeloGioacchino Del Regno Signed-off-by: Sasha Levin

PCI: keystone: Fix race condition when initializing PHYs

2024-01-25T23:45:25+00:00

[ 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: 49229238ab47 ("PCI: keystone: Cleanup PHY handling") Signed-off-by: Siddharth Vadapalli Signed-off-by: Krzysztof Wilczyński Acked-by: Ravi Gunasekaran Signed-off-by: Sasha Levin

PCI: xilinx-xdma: Fix error code in xilinx_pl_dma_pcie_init_irq_domain()

2024-01-25T23:45:25+00:00

[ Upstream commit 2324be17b5e05ac682e7c81fcbfc7b36a9b1becb ] Currently, if the function irq_domain_add_linear() fails to allocate a new IRQ domain and returns NULL, we would then still return a success from the xilinx_pl_dma_pcie_init_irq_domain() function regardless, as the PTR_ERR(NULL) would return a value of zero. This is not a desirable outcome. Thus, fix the incorrect error code and return the -ENOMEM error code if the irq_domain_add_linear() fails to allocate a new IRQ domain. [kwilczynski: commit log] Link: https://lore.kernel.org/linux-pci/20231030072757.3236546-1-harshit.m.mogalapalli@oracle.com Fixes: 8d786149d78c ("PCI: xilinx-xdma: Add Xilinx XDMA Root Port driver") Signed-off-by: Harshit Mogalapalli Signed-off-by: Krzysztof Wilczyński Signed-off-by: Sasha Levin