* [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-04 18:41 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-04 18:41 UTC (permalink / raw)
To: Shawn Lin
Cc: Lorenzo Pieralisi, Andrew Murray, Bjorn Helgaas, Heiko Stuebner,
Robin Murphy, linux-rockchip, linux-pci, linux-arm-kernel,
linux-kernel
I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
"Normal" PCIe cards work (mostly) just fine on this board. The PCIe
switches (I tried Pericom and ASMedia based switches) also work fine on
other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
also recognises the switch, but fails to initialize the buses behind the
bridge properly, see syslog from linux-5.6.0.
Any ideas what I do wrong, or any suggestions what I can test here?
Thanks,
Soeren
Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
regulator
Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
regulator
Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
f8000000.pcie: host bridge /pcie@f8000000 ranges:
Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
f8000000.pcie: Parsing ranges property...
Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
f8000000.pcie: PCI host bridge to bus 0000:00
Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
bus resource [bus 00-1f]
Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
bus resource [mem 0xfa000000-0xfbdfffff]
Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
scanning bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
[1d87:0100] type 01 class 0x060400
Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
supports D1
Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
supported from D0 D1 D3hot
Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
disabled
Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
for bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
scanning [bus 00-00] behind bridge, pass 0
Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
bridge configuration invalid ([bus 00-00]), reconfiguring
Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
scanning [bus 00-00] behind bridge, pass 1
Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
scanning bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
[1b21:1182] type 01 class 0x060400
Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
Payload Size set to 256 (was 128, max 256)
Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
enabling Extended Tags
Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
supported from D0 D3hot D3cold
Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
disabled
Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
for bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
scanning [bus 00-00] behind bridge, pass 0
Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
bridge configuration invalid ([bus 00-00]), reconfiguring
Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
scanning [bus 00-00] behind bridge, pass 1
Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
scanning bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
synchronous external abort: 96000210 [#1] PREEMPT SMP
Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
pcie_rockchip_host(+) brcmfmac brcmutil
Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
modprobe Not tainted 5.6.0 #1
Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
RockPro64 v2.1 (DT)
Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
daIf -PAN -UAO)
Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
x28: 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
x26: 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
x24: ffffffc011003644
Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
x22: ffffffc011003584
Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
x20: 0000000000000004
Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
x18: 00000000fffffff0
Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
x16: 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
x14: ffffffc010be2e28
Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
x12: ffffffc010be2000
Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
x10: ffffffc010be2470
Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
x8 : 0000000000000001
Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
x6 : 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
x4 : 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
x2 : 000000000080000b
Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
x0 : ffffffc012000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
pci_bus_read_config_dword+0x80/0xd0
Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
pci_bus_read_dev_vendor_id+0x48/0x68
Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
pci_scan_single_device+0x7c/0xd8
Apr 4 19:50:38 rockpro64 kernel: [ 74.691690] pci_scan_slot+0x34/0x118
Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
pci_scan_child_bus_extend+0x60/0x2cc
Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
pci_scan_bridge_extend+0x340/0x578
Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
pci_scan_child_bus_extend+0x20c/0x2cc
Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
pci_scan_bridge_extend+0x340/0x578
Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
pci_scan_child_bus_extend+0x20c/0x2cc
Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
pci_scan_child_bus+0x10/0x18
Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
pci_scan_root_bus_bridge+0x78/0xd0
Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
platform_drv_probe+0x50/0xa0
Apr 4 19:50:38 rockpro64 kernel: [ 74.775758] really_probe+0xd8/0x300
Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
driver_probe_device+0x54/0xe8
Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
device_driver_attach+0x6c/0x78
Apr 4 19:50:38 rockpro64 kernel: [ 74.798461] __driver_attach+0x54/0xd0
Apr 4 19:50:38 rockpro64 kernel: [ 74.805744] bus_for_each_dev+0x70/0xc0
Apr 4 19:50:38 rockpro64 kernel: [ 74.813119] driver_attach+0x20/0x28
Apr 4 19:50:38 rockpro64 kernel: [ 74.820101] bus_add_driver+0x178/0x1d8
Apr 4 19:50:38 rockpro64 kernel: [ 74.827249] driver_register+0x60/0x110
Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
__platform_driver_register+0x44/0x50
Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.852443] do_one_initcall+0x74/0x1a8
Apr 4 19:50:38 rockpro64 kernel: [ 74.859430] do_init_module+0x50/0x1f0
Apr 4 19:50:38 rockpro64 kernel: [ 74.866276] load_module+0x1c0c/0x2158
Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
__do_sys_finit_module+0xd0/0xe8
Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
__arm64_sys_finit_module+0x1c/0x28
Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
el0_svc_common.constprop.1+0x7c/0xe8
Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
el0_sync_handler+0x12c/0x1b0
Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
f94002a0 8b130013 (b9400273)
Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
181d7993f92f3f3d ]---
^ permalink raw reply [flat|nested] 15+ messages in thread
* [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-04 18:41 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-04 18:41 UTC (permalink / raw)
To: Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner, linux-pci, linux-kernel,
linux-rockchip, Bjorn Helgaas, Robin Murphy, linux-arm-kernel,
Andrew Murray
I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
"Normal" PCIe cards work (mostly) just fine on this board. The PCIe
switches (I tried Pericom and ASMedia based switches) also work fine on
other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
also recognises the switch, but fails to initialize the buses behind the
bridge properly, see syslog from linux-5.6.0.
Any ideas what I do wrong, or any suggestions what I can test here?
Thanks,
Soeren
Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
regulator
Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
regulator
Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
f8000000.pcie: host bridge /pcie@f8000000 ranges:
Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
f8000000.pcie: Parsing ranges property...
Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
f8000000.pcie: PCI host bridge to bus 0000:00
Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
bus resource [bus 00-1f]
Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
bus resource [mem 0xfa000000-0xfbdfffff]
Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
scanning bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
[1d87:0100] type 01 class 0x060400
Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
supports D1
Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
supported from D0 D1 D3hot
Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
disabled
Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
for bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
scanning [bus 00-00] behind bridge, pass 0
Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
bridge configuration invalid ([bus 00-00]), reconfiguring
Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
scanning [bus 00-00] behind bridge, pass 1
Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
scanning bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
[1b21:1182] type 01 class 0x060400
Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
Payload Size set to 256 (was 128, max 256)
Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
enabling Extended Tags
Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
supported from D0 D3hot D3cold
Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
disabled
Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
for bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
scanning [bus 00-00] behind bridge, pass 0
Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
bridge configuration invalid ([bus 00-00]), reconfiguring
Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
scanning [bus 00-00] behind bridge, pass 1
Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
scanning bus
Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
synchronous external abort: 96000210 [#1] PREEMPT SMP
Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
pcie_rockchip_host(+) brcmfmac brcmutil
Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
modprobe Not tainted 5.6.0 #1
Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
RockPro64 v2.1 (DT)
Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
daIf -PAN -UAO)
Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
x28: 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
x26: 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
x24: ffffffc011003644
Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
x22: ffffffc011003584
Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
x20: 0000000000000004
Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
x18: 00000000fffffff0
Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
x16: 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
x14: ffffffc010be2e28
Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
x12: ffffffc010be2000
Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
x10: ffffffc010be2470
Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
x8 : 0000000000000001
Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
x6 : 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
x4 : 0000000000000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
x2 : 000000000080000b
Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
x0 : ffffffc012000000
Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
pci_bus_read_config_dword+0x80/0xd0
Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
pci_bus_read_dev_vendor_id+0x48/0x68
Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
pci_scan_single_device+0x7c/0xd8
Apr 4 19:50:38 rockpro64 kernel: [ 74.691690] pci_scan_slot+0x34/0x118
Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
pci_scan_child_bus_extend+0x60/0x2cc
Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
pci_scan_bridge_extend+0x340/0x578
Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
pci_scan_child_bus_extend+0x20c/0x2cc
Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
pci_scan_bridge_extend+0x340/0x578
Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
pci_scan_child_bus_extend+0x20c/0x2cc
Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
pci_scan_child_bus+0x10/0x18
Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
pci_scan_root_bus_bridge+0x78/0xd0
Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
platform_drv_probe+0x50/0xa0
Apr 4 19:50:38 rockpro64 kernel: [ 74.775758] really_probe+0xd8/0x300
Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
driver_probe_device+0x54/0xe8
Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
device_driver_attach+0x6c/0x78
Apr 4 19:50:38 rockpro64 kernel: [ 74.798461] __driver_attach+0x54/0xd0
Apr 4 19:50:38 rockpro64 kernel: [ 74.805744] bus_for_each_dev+0x70/0xc0
Apr 4 19:50:38 rockpro64 kernel: [ 74.813119] driver_attach+0x20/0x28
Apr 4 19:50:38 rockpro64 kernel: [ 74.820101] bus_add_driver+0x178/0x1d8
Apr 4 19:50:38 rockpro64 kernel: [ 74.827249] driver_register+0x60/0x110
Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
__platform_driver_register+0x44/0x50
Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
Apr 4 19:50:38 rockpro64 kernel: [ 74.852443] do_one_initcall+0x74/0x1a8
Apr 4 19:50:38 rockpro64 kernel: [ 74.859430] do_init_module+0x50/0x1f0
Apr 4 19:50:38 rockpro64 kernel: [ 74.866276] load_module+0x1c0c/0x2158
Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
__do_sys_finit_module+0xd0/0xe8
Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
__arm64_sys_finit_module+0x1c/0x28
Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
el0_svc_common.constprop.1+0x7c/0xe8
Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
el0_sync_handler+0x12c/0x1b0
Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
f94002a0 8b130013 (b9400273)
Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
181d7993f92f3f3d ]---
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
2020-04-04 18:41 ` Soeren Moch
@ 2020-04-06 12:52 ` Robin Murphy
-1 siblings, 0 replies; 15+ messages in thread
From: Robin Murphy @ 2020-04-06 12:52 UTC (permalink / raw)
To: Soeren Moch, Shawn Lin
Cc: Lorenzo Pieralisi, Andrew Murray, Bjorn Helgaas, Heiko Stuebner,
linux-rockchip, linux-pci, linux-arm-kernel, linux-kernel
On 2020-04-04 7:41 pm, Soeren Moch wrote:
> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
> switches (I tried Pericom and ASMedia based switches) also work fine on
> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
> also recognises the switch, but fails to initialize the buses behind the
> bridge properly, see syslog from linux-5.6.0.
>
> Any ideas what I do wrong, or any suggestions what I can test here?
See the thread here:
https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
The conclusion there seems to be that the RK3399 root complex just
doesn't handle certain types of response in a sensible manner, and
there's not much that can reasonably be done to change that.
Robin.
>
> Thanks,
> Soeren
>
>
> Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
> f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
> regulator
> Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
> f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
> regulator
> Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
> f8000000.pcie: host bridge /pcie@f8000000 ranges:
> Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
> f8000000.pcie: Parsing ranges property...
> Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
> f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
> f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
> f8000000.pcie: PCI host bridge to bus 0000:00
> Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
> bus resource [bus 00-1f]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
> bus resource [mem 0xfa000000-0xfbdfffff]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
> bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
> Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
> scanning bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
> [1d87:0100] type 01 class 0x060400
> Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
> supports D1
> Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
> supported from D0 D1 D3hot
> Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
> disabled
> Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
> for bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
> scanning [bus 00-00] behind bridge, pass 0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
> bridge configuration invalid ([bus 00-00]), reconfiguring
> Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
> scanning [bus 00-00] behind bridge, pass 1
> Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
> scanning bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
> [1b21:1182] type 01 class 0x060400
> Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
> Payload Size set to 256 (was 128, max 256)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
> enabling Extended Tags
> Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
> supported from D0 D3hot D3cold
> Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
> disabled
> Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
> Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
> 0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
> for bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
> scanning [bus 00-00] behind bridge, pass 0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
> bridge configuration invalid ([bus 00-00]), reconfiguring
> Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
> scanning [bus 00-00] behind bridge, pass 1
> Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
> scanning bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
> synchronous external abort: 96000210 [#1] PREEMPT SMP
> Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
> pcie_rockchip_host(+) brcmfmac brcmutil
> Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
> modprobe Not tainted 5.6.0 #1
> Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
> RockPro64 v2.1 (DT)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
> daIf -PAN -UAO)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
> rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
> Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
> x28: 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
> x26: 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
> x24: ffffffc011003644
> Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
> x22: ffffffc011003584
> Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
> x20: 0000000000000004
> Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
> x18: 00000000fffffff0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
> x16: 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
> x14: ffffffc010be2e28
> Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
> x12: ffffffc010be2000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
> x10: ffffffc010be2470
> Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
> x8 : 0000000000000001
> Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
> x6 : 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
> x4 : 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
> x2 : 000000000080000b
> Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
> x0 : ffffffc012000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
> Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
> pci_bus_read_config_dword+0x80/0xd0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
> pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
> pci_bus_read_dev_vendor_id+0x48/0x68
> Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
> pci_scan_single_device+0x7c/0xd8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.691690] pci_scan_slot+0x34/0x118
> Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
> pci_scan_child_bus_extend+0x60/0x2cc
> Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
> pci_scan_bridge_extend+0x340/0x578
> Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
> pci_scan_child_bus_extend+0x20c/0x2cc
> Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
> pci_scan_bridge_extend+0x340/0x578
> Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
> pci_scan_child_bus_extend+0x20c/0x2cc
> Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
> pci_scan_child_bus+0x10/0x18
> Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
> pci_scan_root_bus_bridge+0x78/0xd0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
> rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
> platform_drv_probe+0x50/0xa0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.775758] really_probe+0xd8/0x300
> Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
> driver_probe_device+0x54/0xe8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
> device_driver_attach+0x6c/0x78
> Apr 4 19:50:38 rockpro64 kernel: [ 74.798461] __driver_attach+0x54/0xd0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.805744] bus_for_each_dev+0x70/0xc0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.813119] driver_attach+0x20/0x28
> Apr 4 19:50:38 rockpro64 kernel: [ 74.820101] bus_add_driver+0x178/0x1d8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.827249] driver_register+0x60/0x110
> Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
> __platform_driver_register+0x44/0x50
> Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
> rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.852443] do_one_initcall+0x74/0x1a8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.859430] do_init_module+0x50/0x1f0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.866276] load_module+0x1c0c/0x2158
> Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
> __do_sys_finit_module+0xd0/0xe8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
> __arm64_sys_finit_module+0x1c/0x28
> Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
> el0_svc_common.constprop.1+0x7c/0xe8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
> Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
> el0_sync_handler+0x12c/0x1b0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
> Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
> f94002a0 8b130013 (b9400273)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
> 181d7993f92f3f3d ]---
>
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-06 12:52 ` Robin Murphy
0 siblings, 0 replies; 15+ messages in thread
From: Robin Murphy @ 2020-04-06 12:52 UTC (permalink / raw)
To: Soeren Moch, Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner, linux-pci, linux-kernel,
linux-rockchip, Bjorn Helgaas, linux-arm-kernel, Andrew Murray
On 2020-04-04 7:41 pm, Soeren Moch wrote:
> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
> switches (I tried Pericom and ASMedia based switches) also work fine on
> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
> also recognises the switch, but fails to initialize the buses behind the
> bridge properly, see syslog from linux-5.6.0.
>
> Any ideas what I do wrong, or any suggestions what I can test here?
See the thread here:
https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
The conclusion there seems to be that the RK3399 root complex just
doesn't handle certain types of response in a sensible manner, and
there's not much that can reasonably be done to change that.
Robin.
>
> Thanks,
> Soeren
>
>
> Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
> f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
> regulator
> Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
> f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
> regulator
> Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
> f8000000.pcie: host bridge /pcie@f8000000 ranges:
> Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
> f8000000.pcie: Parsing ranges property...
> Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
> f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
> f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
> f8000000.pcie: PCI host bridge to bus 0000:00
> Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
> bus resource [bus 00-1f]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
> bus resource [mem 0xfa000000-0xfbdfffff]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
> bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
> Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
> scanning bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
> [1d87:0100] type 01 class 0x060400
> Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
> supports D1
> Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
> supported from D0 D1 D3hot
> Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
> disabled
> Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
> for bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
> scanning [bus 00-00] behind bridge, pass 0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
> bridge configuration invalid ([bus 00-00]), reconfiguring
> Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
> scanning [bus 00-00] behind bridge, pass 1
> Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
> scanning bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
> [1b21:1182] type 01 class 0x060400
> Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
> Payload Size set to 256 (was 128, max 256)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
> enabling Extended Tags
> Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
> supported from D0 D3hot D3cold
> Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
> disabled
> Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
> Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
> 0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
> for bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
> scanning [bus 00-00] behind bridge, pass 0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
> bridge configuration invalid ([bus 00-00]), reconfiguring
> Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
> scanning [bus 00-00] behind bridge, pass 1
> Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
> scanning bus
> Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
> synchronous external abort: 96000210 [#1] PREEMPT SMP
> Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
> pcie_rockchip_host(+) brcmfmac brcmutil
> Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
> modprobe Not tainted 5.6.0 #1
> Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
> RockPro64 v2.1 (DT)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
> daIf -PAN -UAO)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
> rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
> Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
> x28: 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
> x26: 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
> x24: ffffffc011003644
> Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
> x22: ffffffc011003584
> Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
> x20: 0000000000000004
> Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
> x18: 00000000fffffff0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
> x16: 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
> x14: ffffffc010be2e28
> Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
> x12: ffffffc010be2000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
> x10: ffffffc010be2470
> Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
> x8 : 0000000000000001
> Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
> x6 : 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
> x4 : 0000000000000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
> x2 : 000000000080000b
> Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
> x0 : ffffffc012000000
> Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
> Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
> pci_bus_read_config_dword+0x80/0xd0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
> pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
> pci_bus_read_dev_vendor_id+0x48/0x68
> Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
> pci_scan_single_device+0x7c/0xd8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.691690] pci_scan_slot+0x34/0x118
> Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
> pci_scan_child_bus_extend+0x60/0x2cc
> Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
> pci_scan_bridge_extend+0x340/0x578
> Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
> pci_scan_child_bus_extend+0x20c/0x2cc
> Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
> pci_scan_bridge_extend+0x340/0x578
> Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
> pci_scan_child_bus_extend+0x20c/0x2cc
> Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
> pci_scan_child_bus+0x10/0x18
> Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
> pci_scan_root_bus_bridge+0x78/0xd0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
> rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
> platform_drv_probe+0x50/0xa0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.775758] really_probe+0xd8/0x300
> Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
> driver_probe_device+0x54/0xe8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
> device_driver_attach+0x6c/0x78
> Apr 4 19:50:38 rockpro64 kernel: [ 74.798461] __driver_attach+0x54/0xd0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.805744] bus_for_each_dev+0x70/0xc0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.813119] driver_attach+0x20/0x28
> Apr 4 19:50:38 rockpro64 kernel: [ 74.820101] bus_add_driver+0x178/0x1d8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.827249] driver_register+0x60/0x110
> Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
> __platform_driver_register+0x44/0x50
> Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
> rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
> Apr 4 19:50:38 rockpro64 kernel: [ 74.852443] do_one_initcall+0x74/0x1a8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.859430] do_init_module+0x50/0x1f0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.866276] load_module+0x1c0c/0x2158
> Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
> __do_sys_finit_module+0xd0/0xe8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
> __arm64_sys_finit_module+0x1c/0x28
> Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
> el0_svc_common.constprop.1+0x7c/0xe8
> Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
> Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
> el0_sync_handler+0x12c/0x1b0
> Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
> Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
> f94002a0 8b130013 (b9400273)
> Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
> 181d7993f92f3f3d ]---
>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-06 17:12 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-06 17:12 UTC (permalink / raw)
To: Robin Murphy, Shawn Lin
Cc: Lorenzo Pieralisi, Andrew Murray, Bjorn Helgaas, Heiko Stuebner,
linux-rockchip, linux-pci, linux-arm-kernel, linux-kernel
On 06.04.20 14:52, Robin Murphy wrote:
> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>> switches (I tried Pericom and ASMedia based switches) also work fine on
>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>> also recognises the switch, but fails to initialize the buses behind the
>> bridge properly, see syslog from linux-5.6.0.
>>
>> Any ideas what I do wrong, or any suggestions what I can test here?
>
> See the thread here:
>
> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>
Thanks Robin!
I also found out in the meantime that device enumeration fails in this
fatal way when probing non-existent devices. So if I hack my complete
bus topology into rockchip_pcie_valid_device, then all existing devices
come up properly. Of course this is not how PCIe should work.
>
> The conclusion there seems to be that the RK3399 root complex just
> doesn't handle certain types of response in a sensible manner, and
> there's not much that can reasonably be done to change that.
Hm, at least there is the promising suggestion to take over the SError
handler, maybe in ATF, as workaround.
I'm happy to test whatever becomes available.
Thanks,
Soeren
>
> Robin.
>
>>
>> Thanks,
>> Soeren
>>
>>
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
>> f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
>> regulator
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
>> f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
>> regulator
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
>> f8000000.pcie: host bridge /pcie@f8000000 ranges:
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
>> f8000000.pcie: Parsing ranges property...
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
>> f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
>> f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
>> f8000000.pcie: PCI host bridge to bus 0000:00
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
>> bus resource [bus 00-1f]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
>> bus resource [mem 0xfa000000-0xfbdfffff]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
>> bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
>> [1d87:0100] type 01 class 0x060400
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
>> supports D1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
>> supported from D0 D1 D3hot
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
>> disabled
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
>> for bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
>> scanning [bus 00-00] behind bridge, pass 0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
>> bridge configuration invalid ([bus 00-00]), reconfiguring
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
>> scanning [bus 00-00] behind bridge, pass 1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
>> [1b21:1182] type 01 class 0x060400
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
>> Payload Size set to 256 (was 128, max 256)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
>> enabling Extended Tags
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
>> supported from D0 D3hot D3cold
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
>> disabled
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
>> Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
>> 0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
>> for bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
>> scanning [bus 00-00] behind bridge, pass 0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
>> bridge configuration invalid ([bus 00-00]), reconfiguring
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
>> scanning [bus 00-00] behind bridge, pass 1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
>> synchronous external abort: 96000210 [#1] PREEMPT SMP
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
>> pcie_rockchip_host(+) brcmfmac brcmutil
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
>> modprobe Not tainted 5.6.0 #1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
>> RockPro64 v2.1 (DT)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
>> daIf -PAN -UAO)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
>> rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
>> x28: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
>> x26: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
>> x24: ffffffc011003644
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
>> x22: ffffffc011003584
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
>> x20: 0000000000000004
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
>> x18: 00000000fffffff0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
>> x16: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
>> x14: ffffffc010be2e28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
>> x12: ffffffc010be2000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
>> x10: ffffffc010be2470
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
>> x8 : 0000000000000001
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
>> x6 : 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
>> x4 : 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
>> x2 : 000000000080000b
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
>> x0 : ffffffc012000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
>> pci_bus_read_config_dword+0x80/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
>> pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
>> pci_bus_read_dev_vendor_id+0x48/0x68
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
>> pci_scan_single_device+0x7c/0xd8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.691690]
>> pci_scan_slot+0x34/0x118
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
>> pci_scan_child_bus_extend+0x60/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
>> pci_scan_bridge_extend+0x340/0x578
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
>> pci_scan_child_bus_extend+0x20c/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
>> pci_scan_bridge_extend+0x340/0x578
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
>> pci_scan_child_bus_extend+0x20c/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
>> pci_scan_child_bus+0x10/0x18
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
>> pci_scan_root_bus_bridge+0x78/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
>> rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
>> platform_drv_probe+0x50/0xa0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.775758]
>> really_probe+0xd8/0x300
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
>> driver_probe_device+0x54/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
>> device_driver_attach+0x6c/0x78
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.798461]
>> __driver_attach+0x54/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.805744]
>> bus_for_each_dev+0x70/0xc0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.813119]
>> driver_attach+0x20/0x28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.820101]
>> bus_add_driver+0x178/0x1d8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.827249]
>> driver_register+0x60/0x110
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
>> __platform_driver_register+0x44/0x50
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
>> rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.852443]
>> do_one_initcall+0x74/0x1a8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.859430]
>> do_init_module+0x50/0x1f0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.866276]
>> load_module+0x1c0c/0x2158
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
>> __do_sys_finit_module+0xd0/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
>> __arm64_sys_finit_module+0x1c/0x28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
>> el0_svc_common.constprop.1+0x7c/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
>> el0_sync_handler+0x12c/0x1b0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
>> f94002a0 8b130013 (b9400273)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
>> 181d7993f92f3f3d ]---
>>
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-06 17:12 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-06 17:12 UTC (permalink / raw)
To: Robin Murphy, Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner,
linux-pci-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-rockchip-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r, Bjorn Helgaas,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r, Andrew Murray
On 06.04.20 14:52, Robin Murphy wrote:
> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>> switches (I tried Pericom and ASMedia based switches) also work fine on
>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>> also recognises the switch, but fails to initialize the buses behind the
>> bridge properly, see syslog from linux-5.6.0.
>>
>> Any ideas what I do wrong, or any suggestions what I can test here?
>
> See the thread here:
>
> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>
Thanks Robin!
I also found out in the meantime that device enumeration fails in this
fatal way when probing non-existent devices. So if I hack my complete
bus topology into rockchip_pcie_valid_device, then all existing devices
come up properly. Of course this is not how PCIe should work.
>
> The conclusion there seems to be that the RK3399 root complex just
> doesn't handle certain types of response in a sensible manner, and
> there's not much that can reasonably be done to change that.
Hm, at least there is the promising suggestion to take over the SError
handler, maybe in ATF, as workaround.
I'm happy to test whatever becomes available.
Thanks,
Soeren
>
> Robin.
>
>>
>> Thanks,
>> Soeren
>>
>>
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
>> f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
>> regulator
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
>> f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
>> regulator
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
>> f8000000.pcie: host bridge /pcie@f8000000 ranges:
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
>> f8000000.pcie: Parsing ranges property...
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
>> f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
>> f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
>> f8000000.pcie: PCI host bridge to bus 0000:00
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
>> bus resource [bus 00-1f]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
>> bus resource [mem 0xfa000000-0xfbdfffff]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
>> bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
>> [1d87:0100] type 01 class 0x060400
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
>> supports D1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
>> supported from D0 D1 D3hot
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
>> disabled
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
>> for bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
>> scanning [bus 00-00] behind bridge, pass 0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
>> bridge configuration invalid ([bus 00-00]), reconfiguring
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
>> scanning [bus 00-00] behind bridge, pass 1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
>> [1b21:1182] type 01 class 0x060400
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
>> Payload Size set to 256 (was 128, max 256)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
>> enabling Extended Tags
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
>> supported from D0 D3hot D3cold
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
>> disabled
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
>> Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
>> 0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
>> for bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
>> scanning [bus 00-00] behind bridge, pass 0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
>> bridge configuration invalid ([bus 00-00]), reconfiguring
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
>> scanning [bus 00-00] behind bridge, pass 1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
>> synchronous external abort: 96000210 [#1] PREEMPT SMP
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
>> pcie_rockchip_host(+) brcmfmac brcmutil
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
>> modprobe Not tainted 5.6.0 #1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
>> RockPro64 v2.1 (DT)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
>> daIf -PAN -UAO)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
>> rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
>> x28: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
>> x26: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
>> x24: ffffffc011003644
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
>> x22: ffffffc011003584
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
>> x20: 0000000000000004
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
>> x18: 00000000fffffff0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
>> x16: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
>> x14: ffffffc010be2e28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
>> x12: ffffffc010be2000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
>> x10: ffffffc010be2470
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
>> x8 : 0000000000000001
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
>> x6 : 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
>> x4 : 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
>> x2 : 000000000080000b
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
>> x0 : ffffffc012000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
>> pci_bus_read_config_dword+0x80/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
>> pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
>> pci_bus_read_dev_vendor_id+0x48/0x68
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
>> pci_scan_single_device+0x7c/0xd8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.691690]
>> pci_scan_slot+0x34/0x118
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
>> pci_scan_child_bus_extend+0x60/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
>> pci_scan_bridge_extend+0x340/0x578
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
>> pci_scan_child_bus_extend+0x20c/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
>> pci_scan_bridge_extend+0x340/0x578
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
>> pci_scan_child_bus_extend+0x20c/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
>> pci_scan_child_bus+0x10/0x18
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
>> pci_scan_root_bus_bridge+0x78/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
>> rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
>> platform_drv_probe+0x50/0xa0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.775758]
>> really_probe+0xd8/0x300
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
>> driver_probe_device+0x54/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
>> device_driver_attach+0x6c/0x78
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.798461]
>> __driver_attach+0x54/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.805744]
>> bus_for_each_dev+0x70/0xc0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.813119]
>> driver_attach+0x20/0x28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.820101]
>> bus_add_driver+0x178/0x1d8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.827249]
>> driver_register+0x60/0x110
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
>> __platform_driver_register+0x44/0x50
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
>> rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.852443]
>> do_one_initcall+0x74/0x1a8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.859430]
>> do_init_module+0x50/0x1f0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.866276]
>> load_module+0x1c0c/0x2158
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
>> __do_sys_finit_module+0xd0/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
>> __arm64_sys_finit_module+0x1c/0x28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
>> el0_svc_common.constprop.1+0x7c/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
>> el0_sync_handler+0x12c/0x1b0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
>> f94002a0 8b130013 (b9400273)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
>> 181d7993f92f3f3d ]---
>>
_______________________________________________
Linux-rockchip mailing list
Linux-rockchip@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-rockchip
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-06 17:12 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-06 17:12 UTC (permalink / raw)
To: Robin Murphy, Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner, linux-pci, linux-kernel,
linux-rockchip, Bjorn Helgaas, linux-arm-kernel, Andrew Murray
On 06.04.20 14:52, Robin Murphy wrote:
> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>> switches (I tried Pericom and ASMedia based switches) also work fine on
>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>> also recognises the switch, but fails to initialize the buses behind the
>> bridge properly, see syslog from linux-5.6.0.
>>
>> Any ideas what I do wrong, or any suggestions what I can test here?
>
> See the thread here:
>
> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>
Thanks Robin!
I also found out in the meantime that device enumeration fails in this
fatal way when probing non-existent devices. So if I hack my complete
bus topology into rockchip_pcie_valid_device, then all existing devices
come up properly. Of course this is not how PCIe should work.
>
> The conclusion there seems to be that the RK3399 root complex just
> doesn't handle certain types of response in a sensible manner, and
> there's not much that can reasonably be done to change that.
Hm, at least there is the promising suggestion to take over the SError
handler, maybe in ATF, as workaround.
I'm happy to test whatever becomes available.
Thanks,
Soeren
>
> Robin.
>
>>
>> Thanks,
>> Soeren
>>
>>
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
>> f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
>> regulator
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
>> f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
>> regulator
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
>> f8000000.pcie: host bridge /pcie@f8000000 ranges:
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
>> f8000000.pcie: Parsing ranges property...
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
>> f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
>> f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
>> f8000000.pcie: PCI host bridge to bus 0000:00
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
>> bus resource [bus 00-1f]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
>> bus resource [mem 0xfa000000-0xfbdfffff]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
>> bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
>> [1d87:0100] type 01 class 0x060400
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
>> supports D1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
>> supported from D0 D1 D3hot
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
>> disabled
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
>> for bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
>> scanning [bus 00-00] behind bridge, pass 0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
>> bridge configuration invalid ([bus 00-00]), reconfiguring
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
>> scanning [bus 00-00] behind bridge, pass 1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
>> [1b21:1182] type 01 class 0x060400
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
>> Payload Size set to 256 (was 128, max 256)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
>> enabling Extended Tags
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
>> supported from D0 D3hot D3cold
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
>> disabled
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
>> Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
>> 0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
>> for bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
>> scanning [bus 00-00] behind bridge, pass 0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
>> bridge configuration invalid ([bus 00-00]), reconfiguring
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
>> scanning [bus 00-00] behind bridge, pass 1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
>> scanning bus
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
>> synchronous external abort: 96000210 [#1] PREEMPT SMP
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
>> pcie_rockchip_host(+) brcmfmac brcmutil
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
>> modprobe Not tainted 5.6.0 #1
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
>> RockPro64 v2.1 (DT)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
>> daIf -PAN -UAO)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
>> rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
>> x28: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
>> x26: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
>> x24: ffffffc011003644
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
>> x22: ffffffc011003584
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
>> x20: 0000000000000004
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
>> x18: 00000000fffffff0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
>> x16: 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
>> x14: ffffffc010be2e28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
>> x12: ffffffc010be2000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
>> x10: ffffffc010be2470
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
>> x8 : 0000000000000001
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
>> x6 : 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
>> x4 : 0000000000000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
>> x2 : 000000000080000b
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
>> x0 : ffffffc012000000
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
>> pci_bus_read_config_dword+0x80/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
>> pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
>> pci_bus_read_dev_vendor_id+0x48/0x68
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
>> pci_scan_single_device+0x7c/0xd8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.691690]
>> pci_scan_slot+0x34/0x118
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
>> pci_scan_child_bus_extend+0x60/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
>> pci_scan_bridge_extend+0x340/0x578
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
>> pci_scan_child_bus_extend+0x20c/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
>> pci_scan_bridge_extend+0x340/0x578
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
>> pci_scan_child_bus_extend+0x20c/0x2cc
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
>> pci_scan_child_bus+0x10/0x18
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
>> pci_scan_root_bus_bridge+0x78/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
>> rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
>> platform_drv_probe+0x50/0xa0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.775758]
>> really_probe+0xd8/0x300
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
>> driver_probe_device+0x54/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
>> device_driver_attach+0x6c/0x78
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.798461]
>> __driver_attach+0x54/0xd0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.805744]
>> bus_for_each_dev+0x70/0xc0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.813119]
>> driver_attach+0x20/0x28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.820101]
>> bus_add_driver+0x178/0x1d8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.827249]
>> driver_register+0x60/0x110
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
>> __platform_driver_register+0x44/0x50
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
>> rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.852443]
>> do_one_initcall+0x74/0x1a8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.859430]
>> do_init_module+0x50/0x1f0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.866276]
>> load_module+0x1c0c/0x2158
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
>> __do_sys_finit_module+0xd0/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
>> __arm64_sys_finit_module+0x1c/0x28
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
>> el0_svc_common.constprop.1+0x7c/0xe8
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
>> el0_sync_handler+0x12c/0x1b0
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
>> f94002a0 8b130013 (b9400273)
>> Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
>> 181d7993f92f3f3d ]---
>>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-14 11:35 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-14 11:35 UTC (permalink / raw)
To: Robin Murphy, Shawn Lin
Cc: Lorenzo Pieralisi, Andrew Murray, Bjorn Helgaas, Heiko Stuebner,
linux-rockchip, linux-pci, linux-arm-kernel, linux-kernel
On 06.04.20 19:12, Soeren Moch wrote:
> On 06.04.20 14:52, Robin Murphy wrote:
>> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>>> switches (I tried Pericom and ASMedia based switches) also work fine on
>>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>>> also recognises the switch, but fails to initialize the buses behind the
>>> bridge properly, see syslog from linux-5.6.0.
>>>
>>> Any ideas what I do wrong, or any suggestions what I can test here?
>> See the thread here:
>>
>> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>>
> Thanks Robin!
>
> I also found out in the meantime that device enumeration fails in this
> fatal way when probing non-existent devices. So if I hack my complete
> bus topology into rockchip_pcie_valid_device, then all existing devices
> come up properly. Of course this is not how PCIe should work.
>> The conclusion there seems to be that the RK3399 root complex just
>> doesn't handle certain types of response in a sensible manner, and
>> there's not much that can reasonably be done to change that.
> Hm, at least there is the promising suggestion to take over the SError
> handler, maybe in ATF, as workaround.
Unfortunately it seems to be not that easy. Only when PCIe device
probing runs on one of the Cortex-A72 cores of rk3399 we see the SError.
When probing runs on one of the A53 cores, we get a synchronous external
abort instead.
Is this expected to see different error types on big.LITTLE systems? Or
is this another special property of the rk3399 pcie controller?
For the SError handling there was an example in the above mentioned
thread. Is a similar example available for SEA handling?
Thanks,
Soeren
> I'm happy to test whatever becomes available.
>
> Thanks,
> Soeren
>> Robin.
>>
>>> Thanks,
>>> Soeren
>>>
>>>
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
>>> f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
>>> regulator
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
>>> f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
>>> regulator
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
>>> f8000000.pcie: host bridge /pcie@f8000000 ranges:
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
>>> f8000000.pcie: Parsing ranges property...
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
>>> f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
>>> f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
>>> f8000000.pcie: PCI host bridge to bus 0000:00
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
>>> bus resource [bus 00-1f]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
>>> bus resource [mem 0xfa000000-0xfbdfffff]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
>>> bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
>>> [1d87:0100] type 01 class 0x060400
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
>>> supports D1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
>>> supported from D0 D1 D3hot
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
>>> disabled
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
>>> for bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
>>> scanning [bus 00-00] behind bridge, pass 0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
>>> bridge configuration invalid ([bus 00-00]), reconfiguring
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
>>> scanning [bus 00-00] behind bridge, pass 1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
>>> [1b21:1182] type 01 class 0x060400
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
>>> Payload Size set to 256 (was 128, max 256)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
>>> enabling Extended Tags
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
>>> supported from D0 D3hot D3cold
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
>>> disabled
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
>>> Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
>>> 0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
>>> for bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
>>> scanning [bus 00-00] behind bridge, pass 0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
>>> bridge configuration invalid ([bus 00-00]), reconfiguring
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
>>> scanning [bus 00-00] behind bridge, pass 1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
>>> synchronous external abort: 96000210 [#1] PREEMPT SMP
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
>>> pcie_rockchip_host(+) brcmfmac brcmutil
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
>>> modprobe Not tainted 5.6.0 #1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
>>> RockPro64 v2.1 (DT)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
>>> daIf -PAN -UAO)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
>>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
>>> rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
>>> x28: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
>>> x26: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
>>> x24: ffffffc011003644
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
>>> x22: ffffffc011003584
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
>>> x20: 0000000000000004
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
>>> x18: 00000000fffffff0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
>>> x16: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
>>> x14: ffffffc010be2e28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
>>> x12: ffffffc010be2000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
>>> x10: ffffffc010be2470
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
>>> x8 : 0000000000000001
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
>>> x6 : 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
>>> x4 : 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
>>> x2 : 000000000080000b
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
>>> x0 : ffffffc012000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
>>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
>>> pci_bus_read_config_dword+0x80/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
>>> pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
>>> pci_bus_read_dev_vendor_id+0x48/0x68
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
>>> pci_scan_single_device+0x7c/0xd8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.691690]
>>> pci_scan_slot+0x34/0x118
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
>>> pci_scan_child_bus_extend+0x60/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
>>> pci_scan_bridge_extend+0x340/0x578
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
>>> pci_scan_child_bus_extend+0x20c/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
>>> pci_scan_bridge_extend+0x340/0x578
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
>>> pci_scan_child_bus_extend+0x20c/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
>>> pci_scan_child_bus+0x10/0x18
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
>>> pci_scan_root_bus_bridge+0x78/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
>>> rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
>>> platform_drv_probe+0x50/0xa0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.775758]
>>> really_probe+0xd8/0x300
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
>>> driver_probe_device+0x54/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
>>> device_driver_attach+0x6c/0x78
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.798461]
>>> __driver_attach+0x54/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.805744]
>>> bus_for_each_dev+0x70/0xc0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.813119]
>>> driver_attach+0x20/0x28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.820101]
>>> bus_add_driver+0x178/0x1d8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.827249]
>>> driver_register+0x60/0x110
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
>>> __platform_driver_register+0x44/0x50
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
>>> rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.852443]
>>> do_one_initcall+0x74/0x1a8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.859430]
>>> do_init_module+0x50/0x1f0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.866276]
>>> load_module+0x1c0c/0x2158
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
>>> __do_sys_finit_module+0xd0/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
>>> __arm64_sys_finit_module+0x1c/0x28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
>>> el0_svc_common.constprop.1+0x7c/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
>>> el0_sync_handler+0x12c/0x1b0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
>>> f94002a0 8b130013 (b9400273)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
>>> 181d7993f92f3f3d ]---
>>>
>
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-14 11:35 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-14 11:35 UTC (permalink / raw)
To: Robin Murphy, Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner,
linux-pci-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-rockchip-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r, Bjorn Helgaas,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r, Andrew Murray
On 06.04.20 19:12, Soeren Moch wrote:
> On 06.04.20 14:52, Robin Murphy wrote:
>> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>>> switches (I tried Pericom and ASMedia based switches) also work fine on
>>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>>> also recognises the switch, but fails to initialize the buses behind the
>>> bridge properly, see syslog from linux-5.6.0.
>>>
>>> Any ideas what I do wrong, or any suggestions what I can test here?
>> See the thread here:
>>
>> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>>
> Thanks Robin!
>
> I also found out in the meantime that device enumeration fails in this
> fatal way when probing non-existent devices. So if I hack my complete
> bus topology into rockchip_pcie_valid_device, then all existing devices
> come up properly. Of course this is not how PCIe should work.
>> The conclusion there seems to be that the RK3399 root complex just
>> doesn't handle certain types of response in a sensible manner, and
>> there's not much that can reasonably be done to change that.
> Hm, at least there is the promising suggestion to take over the SError
> handler, maybe in ATF, as workaround.
Unfortunately it seems to be not that easy. Only when PCIe device
probing runs on one of the Cortex-A72 cores of rk3399 we see the SError.
When probing runs on one of the A53 cores, we get a synchronous external
abort instead.
Is this expected to see different error types on big.LITTLE systems? Or
is this another special property of the rk3399 pcie controller?
For the SError handling there was an example in the above mentioned
thread. Is a similar example available for SEA handling?
Thanks,
Soeren
> I'm happy to test whatever becomes available.
>
> Thanks,
> Soeren
>> Robin.
>>
>>> Thanks,
>>> Soeren
>>>
>>>
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
>>> f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
>>> regulator
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
>>> f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
>>> regulator
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
>>> f8000000.pcie: host bridge /pcie@f8000000 ranges:
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
>>> f8000000.pcie: Parsing ranges property...
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
>>> f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
>>> f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
>>> f8000000.pcie: PCI host bridge to bus 0000:00
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
>>> bus resource [bus 00-1f]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
>>> bus resource [mem 0xfa000000-0xfbdfffff]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
>>> bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
>>> [1d87:0100] type 01 class 0x060400
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
>>> supports D1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
>>> supported from D0 D1 D3hot
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
>>> disabled
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
>>> for bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
>>> scanning [bus 00-00] behind bridge, pass 0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
>>> bridge configuration invalid ([bus 00-00]), reconfiguring
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
>>> scanning [bus 00-00] behind bridge, pass 1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
>>> [1b21:1182] type 01 class 0x060400
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
>>> Payload Size set to 256 (was 128, max 256)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
>>> enabling Extended Tags
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
>>> supported from D0 D3hot D3cold
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
>>> disabled
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
>>> Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
>>> 0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
>>> for bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
>>> scanning [bus 00-00] behind bridge, pass 0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
>>> bridge configuration invalid ([bus 00-00]), reconfiguring
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
>>> scanning [bus 00-00] behind bridge, pass 1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
>>> synchronous external abort: 96000210 [#1] PREEMPT SMP
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
>>> pcie_rockchip_host(+) brcmfmac brcmutil
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
>>> modprobe Not tainted 5.6.0 #1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
>>> RockPro64 v2.1 (DT)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
>>> daIf -PAN -UAO)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
>>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
>>> rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
>>> x28: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
>>> x26: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
>>> x24: ffffffc011003644
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
>>> x22: ffffffc011003584
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
>>> x20: 0000000000000004
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
>>> x18: 00000000fffffff0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
>>> x16: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
>>> x14: ffffffc010be2e28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
>>> x12: ffffffc010be2000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
>>> x10: ffffffc010be2470
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
>>> x8 : 0000000000000001
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
>>> x6 : 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
>>> x4 : 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
>>> x2 : 000000000080000b
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
>>> x0 : ffffffc012000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
>>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
>>> pci_bus_read_config_dword+0x80/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
>>> pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
>>> pci_bus_read_dev_vendor_id+0x48/0x68
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
>>> pci_scan_single_device+0x7c/0xd8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.691690]
>>> pci_scan_slot+0x34/0x118
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
>>> pci_scan_child_bus_extend+0x60/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
>>> pci_scan_bridge_extend+0x340/0x578
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
>>> pci_scan_child_bus_extend+0x20c/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
>>> pci_scan_bridge_extend+0x340/0x578
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
>>> pci_scan_child_bus_extend+0x20c/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
>>> pci_scan_child_bus+0x10/0x18
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
>>> pci_scan_root_bus_bridge+0x78/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
>>> rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
>>> platform_drv_probe+0x50/0xa0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.775758]
>>> really_probe+0xd8/0x300
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
>>> driver_probe_device+0x54/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
>>> device_driver_attach+0x6c/0x78
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.798461]
>>> __driver_attach+0x54/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.805744]
>>> bus_for_each_dev+0x70/0xc0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.813119]
>>> driver_attach+0x20/0x28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.820101]
>>> bus_add_driver+0x178/0x1d8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.827249]
>>> driver_register+0x60/0x110
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
>>> __platform_driver_register+0x44/0x50
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
>>> rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.852443]
>>> do_one_initcall+0x74/0x1a8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.859430]
>>> do_init_module+0x50/0x1f0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.866276]
>>> load_module+0x1c0c/0x2158
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
>>> __do_sys_finit_module+0xd0/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
>>> __arm64_sys_finit_module+0x1c/0x28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
>>> el0_svc_common.constprop.1+0x7c/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
>>> el0_sync_handler+0x12c/0x1b0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
>>> f94002a0 8b130013 (b9400273)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
>>> 181d7993f92f3f3d ]---
>>>
>
_______________________________________________
Linux-rockchip mailing list
Linux-rockchip@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-rockchip
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-14 11:35 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-14 11:35 UTC (permalink / raw)
To: Robin Murphy, Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner, linux-pci, linux-kernel,
linux-rockchip, Bjorn Helgaas, linux-arm-kernel, Andrew Murray
On 06.04.20 19:12, Soeren Moch wrote:
> On 06.04.20 14:52, Robin Murphy wrote:
>> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>>> switches (I tried Pericom and ASMedia based switches) also work fine on
>>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>>> also recognises the switch, but fails to initialize the buses behind the
>>> bridge properly, see syslog from linux-5.6.0.
>>>
>>> Any ideas what I do wrong, or any suggestions what I can test here?
>> See the thread here:
>>
>> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>>
> Thanks Robin!
>
> I also found out in the meantime that device enumeration fails in this
> fatal way when probing non-existent devices. So if I hack my complete
> bus topology into rockchip_pcie_valid_device, then all existing devices
> come up properly. Of course this is not how PCIe should work.
>> The conclusion there seems to be that the RK3399 root complex just
>> doesn't handle certain types of response in a sensible manner, and
>> there's not much that can reasonably be done to change that.
> Hm, at least there is the promising suggestion to take over the SError
> handler, maybe in ATF, as workaround.
Unfortunately it seems to be not that easy. Only when PCIe device
probing runs on one of the Cortex-A72 cores of rk3399 we see the SError.
When probing runs on one of the A53 cores, we get a synchronous external
abort instead.
Is this expected to see different error types on big.LITTLE systems? Or
is this another special property of the rk3399 pcie controller?
For the SError handling there was an example in the above mentioned
thread. Is a similar example available for SEA handling?
Thanks,
Soeren
> I'm happy to test whatever becomes available.
>
> Thanks,
> Soeren
>> Robin.
>>
>>> Thanks,
>>> Soeren
>>>
>>>
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.501951] rockchip-pcie
>>> f8000000.pcie: f8000000.pcie supply vpcie1v8 not found, using dummy
>>> regulator
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.502906] rockchip-pcie
>>> f8000000.pcie: f8000000.pcie supply vpcie0v9 not found, using dummy
>>> regulator
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.572050] rockchip-pcie
>>> f8000000.pcie: host bridge /pcie@f8000000 ranges:
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573018] rockchip-pcie
>>> f8000000.pcie: Parsing ranges property...
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.573040] rockchip-pcie
>>> f8000000.pcie: MEM 0x00fa000000..0x00fbdfffff -> 0x00fa000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.574080] rockchip-pcie
>>> f8000000.pcie: IO 0x00fbe00000..0x00fbefffff -> 0x00fbe00000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.575420] rockchip-pcie
>>> f8000000.pcie: PCI host bridge to bus 0000:00
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576247] pci_bus 0000:00: root
>>> bus resource [bus 00-1f]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.576930] pci_bus 0000:00: root
>>> bus resource [mem 0xfa000000-0xfbdfffff]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.577739] pci_bus 0000:00: root
>>> bus resource [io 0x0000-0xfffff] (bus address [0xfbe00000-0xfbefffff])
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578876] pci_bus 0000:00:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.578918] pci 0000:00:00.0:
>>> [1d87:0100] type 01 class 0x060400
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.579734] pci 0000:00:00.0:
>>> supports D1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580252] pci 0000:00:00.0: PME#
>>> supported from D0 D1 D3hot
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.580952] pci 0000:00:00.0: PME#
>>> disabled
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585475] pci_bus 0000:00: fixups
>>> for bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585491] pci 0000:00:00.0:
>>> scanning [bus 00-00] behind bridge, pass 0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.585497] pci 0000:00:00.0:
>>> bridge configuration invalid ([bus 00-00]), reconfiguring
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586562] pci 0000:00:00.0:
>>> scanning [bus 00-00] behind bridge, pass 1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586725] pci_bus 0000:01:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.586792] pci 0000:01:00.0:
>>> [1b21:1182] type 01 class 0x060400
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.587785] pci 0000:01:00.0: Max
>>> Payload Size set to 256 (was 128, max 256)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.588625] pci 0000:01:00.0:
>>> enabling Extended Tags
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.589487] pci 0000:01:00.0: PME#
>>> supported from D0 D3hot D3cold
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590199] pci 0000:01:00.0: PME#
>>> disabled
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.590344] pci 0000:01:00.0: 2.000
>>> Gb/s available PCIe bandwidth, limited by 2.5 GT/s x1 link at
>>> 0000:00:00.0 (capable of 4.000 Gb/s with 5 GT/s x1 link)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598206] pci_bus 0000:01: fixups
>>> for bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598226] pci 0000:01:00.0:
>>> scanning [bus 00-00] behind bridge, pass 0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.598231] pci 0000:01:00.0:
>>> bridge configuration invalid ([bus 00-00]), reconfiguring
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599163] pci 0000:01:00.0:
>>> scanning [bus 00-00] behind bridge, pass 1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599443] pci_bus 0000:02:
>>> scanning bus
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.599460] Internal error:
>>> synchronous external abort: 96000210 [#1] PREEMPT SMP
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600271] Modules linked in:
>>> pcie_rockchip_host(+) brcmfmac brcmutil
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.600978] CPU: 3 PID: 565 Comm:
>>> modprobe Not tainted 5.6.0 #1
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.601607] Hardware name: Pine64
>>> RockPro64 v2.1 (DT)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602147] pstate: 60000085 (nZCv
>>> daIf -PAN -UAO)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.602666] pc :
>>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.603373] lr :
>>> rockchip_pcie_rd_conf+0x94/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604064] sp : ffffffc011003500
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604419] x29: ffffffc011003500
>>> x28: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.604986] x27: 0000000000000001
>>> x26: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.605552] x25: 0000000000000000
>>> x24: ffffffc011003644
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606117] x23: ffffff80f1792000
>>> x22: ffffffc011003584
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.606683] x21: ffffff80e98313c0
>>> x20: 0000000000000004
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607249] x19: ffffffc012200000
>>> x18: 00000000fffffff0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.607815] x17: 0000000000000000
>>> x16: 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608381] x15: ffffffc010b77c00
>>> x14: ffffffc010be2e28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.608947] x13: 0000000000000000
>>> x12: ffffffc010be2000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.609512] x11: ffffffc010b77000
>>> x10: ffffffc010be2470
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.610079] x9 : 0000000011821b21
>>> x8 : 0000000000000001
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.615455] x7 : 0000000000000000
>>> x6 : 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.621487] x5 : 0000000000200000
>>> x4 : 0000000000000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.627519] x3 : 0000000000c00008
>>> x2 : 000000000080000b
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.633551] x1 : ffffffc015c00008
>>> x0 : ffffffc012000000
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.639583] Call trace:
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.645785]
>>> rockchip_pcie_rd_conf+0x120/0x228 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.656354]
>>> pci_bus_read_config_dword+0x80/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.665083]
>>> pci_bus_generic_read_dev_vendor_id+0x30/0x1a8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.674722]
>>> pci_bus_read_dev_vendor_id+0x48/0x68
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.683382]
>>> pci_scan_single_device+0x7c/0xd8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.691690]
>>> pci_scan_slot+0x34/0x118
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.699155]
>>> pci_scan_child_bus_extend+0x60/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.707774]
>>> pci_scan_bridge_extend+0x340/0x578
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.716224]
>>> pci_scan_child_bus_extend+0x20c/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.724943]
>>> pci_scan_bridge_extend+0x340/0x578
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.733320]
>>> pci_scan_child_bus_extend+0x20c/0x2cc
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.741998]
>>> pci_scan_child_bus+0x10/0x18
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.749739]
>>> pci_scan_root_bus_bridge+0x78/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.757988]
>>> rockchip_pcie_probe+0x830/0xb90 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.768042]
>>> platform_drv_probe+0x50/0xa0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.775758]
>>> really_probe+0xd8/0x300
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.782939]
>>> driver_probe_device+0x54/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.790661]
>>> device_driver_attach+0x6c/0x78
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.798461]
>>> __driver_attach+0x54/0xd0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.805744]
>>> bus_for_each_dev+0x70/0xc0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.813119]
>>> driver_attach+0x20/0x28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.820101]
>>> bus_add_driver+0x178/0x1d8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.827249]
>>> driver_register+0x60/0x110
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.834308]
>>> __platform_driver_register+0x44/0x50
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.842299]
>>> rockchip_pcie_driver_init+0x20/0x1000 [pcie_rockchip_host]
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.852443]
>>> do_one_initcall+0x74/0x1a8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.859430]
>>> do_init_module+0x50/0x1f0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.866276]
>>> load_module+0x1c0c/0x2158
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.873100]
>>> __do_sys_finit_module+0xd0/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.880480]
>>> __arm64_sys_finit_module+0x1c/0x28
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.888157]
>>> el0_svc_common.constprop.1+0x7c/0xe8
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.896000] do_el0_svc+0x18/0x20
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.902285]
>>> el0_sync_handler+0x12c/0x1b0
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.909380] el0_sync+0x114/0x140
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.915692] Code: a8c37bfd d65f03c0
>>> f94002a0 8b130013 (b9400273)
>>> Apr 4 19:50:38 rockpro64 kernel: [ 74.925210] ---[ end trace
>>> 181d7993f92f3f3d ]---
>>>
>
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
2020-04-14 11:35 ` Soeren Moch
(?)
@ 2020-04-14 12:28 ` Robin Murphy
-1 siblings, 0 replies; 15+ messages in thread
From: Robin Murphy @ 2020-04-14 12:28 UTC (permalink / raw)
To: Soeren Moch, Shawn Lin
Cc: Lorenzo Pieralisi, Andrew Murray, Bjorn Helgaas, Heiko Stuebner,
linux-rockchip, linux-pci, linux-arm-kernel, linux-kernel
On 2020-04-14 12:35 pm, Soeren Moch wrote:
> On 06.04.20 19:12, Soeren Moch wrote:
>> On 06.04.20 14:52, Robin Murphy wrote:
>>> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>>>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>>>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>>>> switches (I tried Pericom and ASMedia based switches) also work fine on
>>>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>>>> also recognises the switch, but fails to initialize the buses behind the
>>>> bridge properly, see syslog from linux-5.6.0.
>>>>
>>>> Any ideas what I do wrong, or any suggestions what I can test here?
>>> See the thread here:
>>>
>>> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>>>
>> Thanks Robin!
>>
>> I also found out in the meantime that device enumeration fails in this
>> fatal way when probing non-existent devices. So if I hack my complete
>> bus topology into rockchip_pcie_valid_device, then all existing devices
>> come up properly. Of course this is not how PCIe should work.
>>> The conclusion there seems to be that the RK3399 root complex just
>>> doesn't handle certain types of response in a sensible manner, and
>>> there's not much that can reasonably be done to change that.
>> Hm, at least there is the promising suggestion to take over the SError
>> handler, maybe in ATF, as workaround.
> Unfortunately it seems to be not that easy. Only when PCIe device
> probing runs on one of the Cortex-A72 cores of rk3399 we see the SError.
> When probing runs on one of the A53 cores, we get a synchronous external
> abort instead.
>
> Is this expected to see different error types on big.LITTLE systems? Or
> is this another special property of the rk3399 pcie controller?
As far as I'm aware, the CPU microarchitecture is indeed one of the
factors in whether it takes a given external abort synchronously or
asynchronously, so yes, I'd say that probably is expected. I wouldn't
necessarily even rely on a single microarchitecture only behaving one
way, since in principle it's possible that surrounding instructions
might affect whether the core still has enough context left to take the
exception synchronously or not at the point the abort does come back.
In general external aborts are a "should never happen" kind of thing, so
they're not necessarily expected to be recoverable (I think the RAS
extensions might add a more robustness in terms of reporting, but aren't
relevant here either way).
At this point I'm starting to wonder whether it might be possible to do
something similar to the Arm N1SDP workaround using the Cortex-M0,
albeit with the complication that probing would realistically have to be
explicitly invoked from the Linux driver due to clocks and external
regulators... :/
Robin.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-14 12:28 ` Robin Murphy
0 siblings, 0 replies; 15+ messages in thread
From: Robin Murphy @ 2020-04-14 12:28 UTC (permalink / raw)
To: Soeren Moch, Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner, linux-pci, linux-kernel,
linux-rockchip, Bjorn Helgaas, linux-arm-kernel, Andrew Murray
On 2020-04-14 12:35 pm, Soeren Moch wrote:
> On 06.04.20 19:12, Soeren Moch wrote:
>> On 06.04.20 14:52, Robin Murphy wrote:
>>> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>>>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>>>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>>>> switches (I tried Pericom and ASMedia based switches) also work fine on
>>>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>>>> also recognises the switch, but fails to initialize the buses behind the
>>>> bridge properly, see syslog from linux-5.6.0.
>>>>
>>>> Any ideas what I do wrong, or any suggestions what I can test here?
>>> See the thread here:
>>>
>>> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>>>
>> Thanks Robin!
>>
>> I also found out in the meantime that device enumeration fails in this
>> fatal way when probing non-existent devices. So if I hack my complete
>> bus topology into rockchip_pcie_valid_device, then all existing devices
>> come up properly. Of course this is not how PCIe should work.
>>> The conclusion there seems to be that the RK3399 root complex just
>>> doesn't handle certain types of response in a sensible manner, and
>>> there's not much that can reasonably be done to change that.
>> Hm, at least there is the promising suggestion to take over the SError
>> handler, maybe in ATF, as workaround.
> Unfortunately it seems to be not that easy. Only when PCIe device
> probing runs on one of the Cortex-A72 cores of rk3399 we see the SError.
> When probing runs on one of the A53 cores, we get a synchronous external
> abort instead.
>
> Is this expected to see different error types on big.LITTLE systems? Or
> is this another special property of the rk3399 pcie controller?
As far as I'm aware, the CPU microarchitecture is indeed one of the
factors in whether it takes a given external abort synchronously or
asynchronously, so yes, I'd say that probably is expected. I wouldn't
necessarily even rely on a single microarchitecture only behaving one
way, since in principle it's possible that surrounding instructions
might affect whether the core still has enough context left to take the
exception synchronously or not at the point the abort does come back.
In general external aborts are a "should never happen" kind of thing, so
they're not necessarily expected to be recoverable (I think the RAS
extensions might add a more robustness in terms of reporting, but aren't
relevant here either way).
At this point I'm starting to wonder whether it might be possible to do
something similar to the Arm N1SDP workaround using the Cortex-M0,
albeit with the complication that probing would realistically have to be
explicitly invoked from the Linux driver due to clocks and external
regulators... :/
Robin.
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-14 12:28 ` Robin Murphy
0 siblings, 0 replies; 15+ messages in thread
From: Robin Murphy @ 2020-04-14 12:28 UTC (permalink / raw)
To: Soeren Moch, Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner, linux-pci, linux-kernel,
linux-rockchip, Bjorn Helgaas, linux-arm-kernel, Andrew Murray
On 2020-04-14 12:35 pm, Soeren Moch wrote:
> On 06.04.20 19:12, Soeren Moch wrote:
>> On 06.04.20 14:52, Robin Murphy wrote:
>>> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>>>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>>>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>>>> switches (I tried Pericom and ASMedia based switches) also work fine on
>>>> other boards. The RK3399 PCIe controller with pcie_rockchip_host driver
>>>> also recognises the switch, but fails to initialize the buses behind the
>>>> bridge properly, see syslog from linux-5.6.0.
>>>>
>>>> Any ideas what I do wrong, or any suggestions what I can test here?
>>> See the thread here:
>>>
>>> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>>>
>> Thanks Robin!
>>
>> I also found out in the meantime that device enumeration fails in this
>> fatal way when probing non-existent devices. So if I hack my complete
>> bus topology into rockchip_pcie_valid_device, then all existing devices
>> come up properly. Of course this is not how PCIe should work.
>>> The conclusion there seems to be that the RK3399 root complex just
>>> doesn't handle certain types of response in a sensible manner, and
>>> there's not much that can reasonably be done to change that.
>> Hm, at least there is the promising suggestion to take over the SError
>> handler, maybe in ATF, as workaround.
> Unfortunately it seems to be not that easy. Only when PCIe device
> probing runs on one of the Cortex-A72 cores of rk3399 we see the SError.
> When probing runs on one of the A53 cores, we get a synchronous external
> abort instead.
>
> Is this expected to see different error types on big.LITTLE systems? Or
> is this another special property of the rk3399 pcie controller?
As far as I'm aware, the CPU microarchitecture is indeed one of the
factors in whether it takes a given external abort synchronously or
asynchronously, so yes, I'd say that probably is expected. I wouldn't
necessarily even rely on a single microarchitecture only behaving one
way, since in principle it's possible that surrounding instructions
might affect whether the core still has enough context left to take the
exception synchronously or not at the point the abort does come back.
In general external aborts are a "should never happen" kind of thing, so
they're not necessarily expected to be recoverable (I think the RAS
extensions might add a more robustness in terms of reporting, but aren't
relevant here either way).
At this point I'm starting to wonder whether it might be possible to do
something similar to the Arm N1SDP workaround using the Cortex-M0,
albeit with the complication that probing would realistically have to be
explicitly invoked from the Linux driver due to clocks and external
regulators... :/
Robin.
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
2020-04-14 12:28 ` Robin Murphy
@ 2020-04-27 19:32 ` Soeren Moch
-1 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-27 19:32 UTC (permalink / raw)
To: Robin Murphy, Shawn Lin
Cc: Lorenzo Pieralisi, Andrew Murray, Bjorn Helgaas, Heiko Stuebner,
linux-rockchip, linux-pci, linux-arm-kernel, linux-kernel
On 14.04.20 14:28, Robin Murphy wrote:
> On 2020-04-14 12:35 pm, Soeren Moch wrote:
>> On 06.04.20 19:12, Soeren Moch wrote:
>>> On 06.04.20 14:52, Robin Murphy wrote:
>>>> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>>>>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>>>>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>>>>> switches (I tried Pericom and ASMedia based switches) also work
>>>>> fine on
>>>>> other boards. The RK3399 PCIe controller with pcie_rockchip_host
>>>>> driver
>>>>> also recognises the switch, but fails to initialize the buses
>>>>> behind the
>>>>> bridge properly, see syslog from linux-5.6.0.
>>>>>
>>>>> Any ideas what I do wrong, or any suggestions what I can test here?
>>>> See the thread here:
>>>>
>>>> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>>>>
>>>>
>>> Thanks Robin!
>>>
>>> I also found out in the meantime that device enumeration fails in this
>>> fatal way when probing non-existent devices. So if I hack my complete
>>> bus topology into rockchip_pcie_valid_device, then all existing devices
>>> come up properly. Of course this is not how PCIe should work.
>>>> The conclusion there seems to be that the RK3399 root complex just
>>>> doesn't handle certain types of response in a sensible manner, and
>>>> there's not much that can reasonably be done to change that.
>>> Hm, at least there is the promising suggestion to take over the SError
>>> handler, maybe in ATF, as workaround.
>> Unfortunately it seems to be not that easy. Only when PCIe device
>> probing runs on one of the Cortex-A72 cores of rk3399 we see the SError.
>> When probing runs on one of the A53 cores, we get a synchronous external
>> abort instead.
>>
>> Is this expected to see different error types on big.LITTLE systems? Or
>> is this another special property of the rk3399 pcie controller?
>
> As far as I'm aware, the CPU microarchitecture is indeed one of the
> factors in whether it takes a given external abort synchronously or
> asynchronously, so yes, I'd say that probably is expected. I wouldn't
> necessarily even rely on a single microarchitecture only behaving one
> way, since in principle it's possible that surrounding instructions
> might affect whether the core still has enough context left to take
> the exception synchronously or not at the point the abort does come back.
>
> In general external aborts are a "should never happen" kind of thing,
> so they're not necessarily expected to be recoverable (I think the RAS
> extensions might add a more robustness in terms of reporting, but
> aren't relevant here either way).
>
Okay. In an ideal world we would not need software workarounds for
hardware bugs.
@Shawn: Can you point me to the rk3399 errata you mentioned in commit
712fa1777207c2f2703a6eb618a9699099cbe37b ?
Thanks.
> At this point I'm starting to wonder whether it might be possible to
> do something similar to the Arm N1SDP workaround using the Cortex-M0,
> albeit with the complication that probing would realistically have to
> be explicitly invoked from the Linux driver due to clocks and external
> regulators... :/
>
Sounds complicated.
For me I use the patch below. Of course this hack is not intended for
merging, just as reference to conclude this discussion.
If someone comes up with a better solution, I'm happy to test this.
Thanks,
Soeren
------------------------8<------------------------------------
From 9f2e26186bbf867f1baada057bcbd843c465c381 Mon Sep 17 00:00:00 2001
From: Soeren Moch <smoch@web.de>
Date: Fri, 17 Apr 2020 12:14:04 +0200
Subject: [PATCH] PCI: rockchip: rk3399: pcie switch support
Due to a hardware bug the rk3399 PCIe controller signals error conditions
to the cpu when scanning for PCIe devices, which are not available. So
PCIe bridges are not supported.
The rk3399 Cortex-A72 cores generate SError interrupts for these false
PCIe errors, Cortex-A53 cores generate Synchronuos External Aborts.
This hack enables PCIe device probing on buses behind bridges by
ignoring the generated SError. Device probing needs to be done on
Cortex-A72 cores, e.g. use
taskset -c 4 modprobe pcie_rockchip_host
Signed-off-by: Soeren Moch <smoch@web.de>
---
arch/arm64/kernel/traps.c | 10 +++++++++-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/arch/arm64/kernel/traps.c b/arch/arm64/kernel/traps.c
index cf402be5c573..da2b64d2613f 100644
--- a/arch/arm64/kernel/traps.c
+++ b/arch/arm64/kernel/traps.c
@@ -906,8 +906,16 @@ bool arm64_is_fatal_ras_serror(struct pt_regs
*regs, unsigned int esr)
asmlinkage void do_serror(struct pt_regs *regs, unsigned int esr)
{
- const bool was_in_nmi = in_nmi();
+ bool was_in_nmi;
+ /* ignore SError to enable rk3399 PCIe bus enumeration */
+ if (esr >> ESR_ELx_EC_SHIFT == ESR_ELx_EC_SERROR) {
+ pr_debug("ignoring SError Interrupt on CPU%d\n",
+ smp_processor_id());
+ return;
+ }
+
+ was_in_nmi = in_nmi();
if (!was_in_nmi)
nmi_enter();
--
2.17.1
^ permalink raw reply related [flat|nested] 15+ messages in thread
* Re: [BUG] PCI: rockchip: rk3399: pcie switch support
@ 2020-04-27 19:32 ` Soeren Moch
0 siblings, 0 replies; 15+ messages in thread
From: Soeren Moch @ 2020-04-27 19:32 UTC (permalink / raw)
To: Robin Murphy, Shawn Lin
Cc: Lorenzo Pieralisi, Heiko Stuebner, linux-pci, linux-kernel,
linux-rockchip, Bjorn Helgaas, linux-arm-kernel, Andrew Murray
On 14.04.20 14:28, Robin Murphy wrote:
> On 2020-04-14 12:35 pm, Soeren Moch wrote:
>> On 06.04.20 19:12, Soeren Moch wrote:
>>> On 06.04.20 14:52, Robin Murphy wrote:
>>>> On 2020-04-04 7:41 pm, Soeren Moch wrote:
>>>>> I want to use a PCIe switch on a RK3399 based RockPro64 V2.1 board.
>>>>> "Normal" PCIe cards work (mostly) just fine on this board. The PCIe
>>>>> switches (I tried Pericom and ASMedia based switches) also work
>>>>> fine on
>>>>> other boards. The RK3399 PCIe controller with pcie_rockchip_host
>>>>> driver
>>>>> also recognises the switch, but fails to initialize the buses
>>>>> behind the
>>>>> bridge properly, see syslog from linux-5.6.0.
>>>>>
>>>>> Any ideas what I do wrong, or any suggestions what I can test here?
>>>> See the thread here:
>>>>
>>>> https://lore.kernel.org/linux-pci/CAMdYzYoTwjKz4EN8PtD5pZfu3+SX+68JL+dfvmCrSnLL=K6Few@mail.gmail.com/
>>>>
>>>>
>>> Thanks Robin!
>>>
>>> I also found out in the meantime that device enumeration fails in this
>>> fatal way when probing non-existent devices. So if I hack my complete
>>> bus topology into rockchip_pcie_valid_device, then all existing devices
>>> come up properly. Of course this is not how PCIe should work.
>>>> The conclusion there seems to be that the RK3399 root complex just
>>>> doesn't handle certain types of response in a sensible manner, and
>>>> there's not much that can reasonably be done to change that.
>>> Hm, at least there is the promising suggestion to take over the SError
>>> handler, maybe in ATF, as workaround.
>> Unfortunately it seems to be not that easy. Only when PCIe device
>> probing runs on one of the Cortex-A72 cores of rk3399 we see the SError.
>> When probing runs on one of the A53 cores, we get a synchronous external
>> abort instead.
>>
>> Is this expected to see different error types on big.LITTLE systems? Or
>> is this another special property of the rk3399 pcie controller?
>
> As far as I'm aware, the CPU microarchitecture is indeed one of the
> factors in whether it takes a given external abort synchronously or
> asynchronously, so yes, I'd say that probably is expected. I wouldn't
> necessarily even rely on a single microarchitecture only behaving one
> way, since in principle it's possible that surrounding instructions
> might affect whether the core still has enough context left to take
> the exception synchronously or not at the point the abort does come back.
>
> In general external aborts are a "should never happen" kind of thing,
> so they're not necessarily expected to be recoverable (I think the RAS
> extensions might add a more robustness in terms of reporting, but
> aren't relevant here either way).
>
Okay. In an ideal world we would not need software workarounds for
hardware bugs.
@Shawn: Can you point me to the rk3399 errata you mentioned in commit
712fa1777207c2f2703a6eb618a9699099cbe37b ?
Thanks.
> At this point I'm starting to wonder whether it might be possible to
> do something similar to the Arm N1SDP workaround using the Cortex-M0,
> albeit with the complication that probing would realistically have to
> be explicitly invoked from the Linux driver due to clocks and external
> regulators... :/
>
Sounds complicated.
For me I use the patch below. Of course this hack is not intended for
merging, just as reference to conclude this discussion.
If someone comes up with a better solution, I'm happy to test this.
Thanks,
Soeren
------------------------8<------------------------------------
From 9f2e26186bbf867f1baada057bcbd843c465c381 Mon Sep 17 00:00:00 2001
From: Soeren Moch <smoch@web.de>
Date: Fri, 17 Apr 2020 12:14:04 +0200
Subject: [PATCH] PCI: rockchip: rk3399: pcie switch support
Due to a hardware bug the rk3399 PCIe controller signals error conditions
to the cpu when scanning for PCIe devices, which are not available. So
PCIe bridges are not supported.
The rk3399 Cortex-A72 cores generate SError interrupts for these false
PCIe errors, Cortex-A53 cores generate Synchronuos External Aborts.
This hack enables PCIe device probing on buses behind bridges by
ignoring the generated SError. Device probing needs to be done on
Cortex-A72 cores, e.g. use
taskset -c 4 modprobe pcie_rockchip_host
Signed-off-by: Soeren Moch <smoch@web.de>
---
arch/arm64/kernel/traps.c | 10 +++++++++-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/arch/arm64/kernel/traps.c b/arch/arm64/kernel/traps.c
index cf402be5c573..da2b64d2613f 100644
--- a/arch/arm64/kernel/traps.c
+++ b/arch/arm64/kernel/traps.c
@@ -906,8 +906,16 @@ bool arm64_is_fatal_ras_serror(struct pt_regs
*regs, unsigned int esr)
asmlinkage void do_serror(struct pt_regs *regs, unsigned int esr)
{
- const bool was_in_nmi = in_nmi();
+ bool was_in_nmi;
+ /* ignore SError to enable rk3399 PCIe bus enumeration */
+ if (esr >> ESR_ELx_EC_SHIFT == ESR_ELx_EC_SERROR) {
+ pr_debug("ignoring SError Interrupt on CPU%d\n",
+ smp_processor_id());
+ return;
+ }
+
+ was_in_nmi = in_nmi();
if (!was_in_nmi)
nmi_enter();
--
2.17.1
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply related [flat|nested] 15+ messages in thread
end of thread, other threads:[~2020-04-27 19:32 UTC | newest]
Thread overview: 15+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-04-04 18:41 [BUG] PCI: rockchip: rk3399: pcie switch support Soeren Moch
2020-04-04 18:41 ` Soeren Moch
2020-04-06 12:52 ` Robin Murphy
2020-04-06 12:52 ` Robin Murphy
2020-04-06 17:12 ` Soeren Moch
2020-04-06 17:12 ` Soeren Moch
2020-04-06 17:12 ` Soeren Moch
2020-04-14 11:35 ` Soeren Moch
2020-04-14 11:35 ` Soeren Moch
2020-04-14 11:35 ` Soeren Moch
2020-04-14 12:28 ` Robin Murphy
2020-04-14 12:28 ` Robin Murphy
2020-04-14 12:28 ` Robin Murphy
2020-04-27 19:32 ` Soeren Moch
2020-04-27 19:32 ` Soeren Moch
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.