On 12/03/2019 15:59, Marc Gonzalez wrote:
On 27/02/2019 16:31, Marc Gonzalez wrote:
FTR, this is just a work-in-progress status, not a formal submission.
I still can't get the ethernet card to play ball. As soon as I set the link up,
the system hangs a few seconds, then reboots.
Here are the relevant DT nodes:
(There's probably an error or 12 lurking in there)
anoc1_smmu: arm,smmu@1680000 {
compatible = "qcom,msm8996-smmu-v2", "qcom,smmu-v2";
reg = <0x01680000 0x10000>;
#iommu-cells = <0>;
That's almost certainly not what you want - note that the binding is
only defined for 1 or 2 cells. The current Linux driver does actually
happen to handle 0 cells, in which case it will assume every device uses
Stream ID 0, but if the funky mappings in the SDM845 patch are anything
to go by I doubt that's correct.
#global-interrupts = <0>;
Ouch - not having a global fault interrupt will make this even more
painful to debug. At least with that you can figure out what Stream IDs
are in play by forcing stuff to fault.
Robin.
interrupts =
<GIC_SPI 364 IRQ_TYPE_EDGE_RISING>,
<GIC_SPI 365 IRQ_TYPE_EDGE_RISING>,
<GIC_SPI 366 IRQ_TYPE_EDGE_RISING>,
<GIC_SPI 367 IRQ_TYPE_EDGE_RISING>,
<GIC_SPI 368 IRQ_TYPE_EDGE_RISING>,
<GIC_SPI 369 IRQ_TYPE_EDGE_RISING>;
};
pcie0: pci@1c00000 {
compatible = "qcom,pcie-msm8996";
reg-names =
"parf",
"dbi",
"elbi",
"config";
reg =
<0x01c00000 0x2000>,
<0x1b000000 0xf1d>,
<0x1b000f20 0xa8>,
<0x1b100000 0x100000>;
device_type = "pci";
linux,pci-domain = <0>;
bus-range = <0x00 0xff>;
#address-cells = <3>;
#size-cells = <2>;
power-domains = <&gcc PCIE_0_GDSC>;
num-lanes = <1>;
phy-names = "pciephy";
phys = <&pciephy>;
ranges =
/*** downstream I/O ***/
<0x01000000 0x0 0x1b200000 0x1b200000 0x0 0x100000>,
/*** non-prefetchable memory ***/
<0x02000000 0x0 0x1b300000 0x1b300000 0x0 0xd00000>;
#interrupt-cells = <1>;
interrupt-names = "msi";
interrupts = <GIC_SPI 405 IRQ_TYPE_LEVEL_HIGH>;
interrupt-map-mask = <0 0 0 0x7>;
interrupt-map =
<0 0 0 1 &intc 0 135 IRQ_TYPE_LEVEL_HIGH>, /* int_a */
<0 0 0 2 &intc 0 136 IRQ_TYPE_LEVEL_HIGH>, /* int_b */
<0 0 0 3 &intc 0 138 IRQ_TYPE_LEVEL_HIGH>, /* int_c */
<0 0 0 4 &intc 0 139 IRQ_TYPE_LEVEL_HIGH>; /* int_d */
clock-names =
"pipe",
"bus_master",
"bus_slave",
"cfg",
"aux";
clocks =
<&gcc GCC_PCIE_0_PIPE_CLK>,
<&gcc GCC_PCIE_0_MSTR_AXI_CLK>,
<&gcc GCC_PCIE_0_SLV_AXI_CLK>,
<&gcc GCC_PCIE_0_CFG_AHB_CLK>,
<&gcc GCC_PCIE_0_AUX_CLK>;
iommu-map = <0 &anoc1_smmu 0 1>;
iommu-map-mask = <0>;
/* PCIe Fundamental Reset */
perst-gpios = <&tlmm 35 GPIO_ACTIVE_LOW>;
};
phy@1c06000 {
compatible = "qcom,msm8998-qmp-pcie-phy";
reg = <0x01c06000 0x18c>;
#clock-cells = <1>;
#address-cells = <1>;
#size-cells = <1>;
ranges;
clock-names =
"aux",
"cfg_ahb",
"ref";
clocks =
<&gcc GCC_PCIE_PHY_AUX_CLK>,
<&gcc GCC_PCIE_0_CFG_AHB_CLK>,
<&gcc GCC_PCIE_CLKREF_CLK>;
vdda-phy-supply = <&vreg_l1a_0p875>;
vdda-pll-supply = <&vreg_l2a_1p2>;
reset-names =
"phy",
"common",
"cfg";
resets =
<&gcc GCC_PCIE_PHY_BCR>,
<&gcc GCC_PCIE_PHY_COM_BCR>,
<&gcc GCC_PCIE_PHY_NOCSR_COM_PHY_BCR>;
pciephy: lane@1c06800 {
reg = <0x01c06200 0x128>, <0x01c06400 0x1fc>, <0x01c06800 0x20c>;
#phy-cells = <0>;
clock-names = "pipe0";
clocks = <&gcc GCC_PCIE_0_PIPE_CLK>;
clock-output-names = "pcie_0_pipe_clk_src";
};
};
And here's the boot log, augmented with some custom stack dumps:
https://pastebin.ubuntu.com/p/2PYTtnBfPV/
What stands out:
pci 0000:01:00.0: [Firmware Bug]: disabling VPD access (can't determine size of non-standard VPD format)
OF: /soc/pci@1c00000: iommu-map, using mask 00000000, rid-base: 00000000, out-base: 00000000, length: 00000001, rid: 00000000 -> 00000000
OF: /soc/pci@1c00000: iommu-map, using mask 00000000, rid-base: 00000000, out-base: 00000000, length: 00000001, rid: 00000100 -> 00000000
OF: /soc/pci@1c00000: iommu-map, using mask 00000000, rid-base: 00000000, out-base: 00000000, length: 00000001, rid: 00000000 -> 00000000
(no idea what "out-base" is supposed to be set to)
atl1c 0000:01:00.0: Adding to iommu group 0
atl1c 0000:01:00.0: assign IRQ: got 0
iommu_map | iommu_map_sg | iommu_dma_alloc | __iommu_alloc_attrs | dma_alloc_attrs | atl1c_open | __dev_open | __dev_change_flags | dev_change_flags | devinet_ioctl | inet_ioctl | sock_ioctl | do_vfs_ioctl | ksys_ioctl | __arm64_sys_ioctl |
iommu_map | __iommu_dma_map | iommu_dma_map_page | __iommu_map_page | atl1c_alloc_rx_buffer | atl1c_configure | atl1c_up | atl1c_open | __dev_open | __dev_change_flags | dev_change_flags | devinet_ioctl | inet_ioctl | sock_ioctl | do_vfs_ioctl |
iommu_map | __iommu_dma_map | iommu_dma_map_page | __iommu_map_page | atl1c_alloc_rx_buffer | atl1c_configure | atl1c_up | atl1c_common_task | process_one_work | worker_thread | kthread | ret_from_fork |
In case anyone spots any glaring mistake...
Regards.
