Bug 212261

Summary:	pci 0000:00:00.2: can't derive routing for PCI INT A
Product:	Drivers	Reporter:	Erik Quaeghebeur (bugzilla.kernel.org)
Component:	PCI	Assignee:	drivers_pci (drivers_pci)
Status:	NEW ---
Severity:	normal	CC:	adm, bjorn, CoelacanthusHex, hxss, icegood1980, istankovic, josh.mcd31, kh.bugreport, konoha02, linux.alucard, marcel, mateusz.kaduk, rafael.ristovski, rcdtch, snafu109, whenov
Priority:	P1
Hardware:	All
OS:	Linux
Kernel Version:	5.14.14	Subsystem:
Regression:	No	Bisected commit-id:
Attachments:	output of lspci -vvv (as superuser) dmesg output acpidump output from Lenovo P14s AMD Gen 1 acpidump from Asus TUF B450M-Pro Gaming acpidump from ASRock B450M Pro4-F R2.0 acpidump thinkpad e495

Description Erik Quaeghebeur 2021-03-13 13:19:39 UTC

I'm on a Lenovo T14 AMD, with a Ryzen 4750U. In the logs, I find

pci 0000:00:00.2: AMD-Vi: IOMMU performance counters supported
pci 0000:00:00.2: can't derive routing for PCI INT A
pci 0000:00:00.2: PCI INT A: not connected

where the last two lines are warnings. Various bugreports related to this can be found elsewhere.

I guess there is either something wrong with the hardware/firmware or with the kernel. I'd hope this warning can be fixed. Feel free to ask for more information.

Comment 1 Erik Quaeghebeur 2021-03-13 22:17:31 UTC

Other reports about this issue found on the web:
* https://bugs.launchpad.net/dell-sputnik/+bug/1881121
* https://www.spinics.net/lists/linux-pci/msg95035.html
* https://bugzilla.kernel.org/show_bug.cgi?id=201753 (mentioned as side issue)

Comment 2 Erik Quaeghebeur 2021-03-13 22:24:23 UTC

Created attachment 295841 [details]
output of lspci -vvv (as superuser)

The following was stderr output

pcilib: sysfs_read_vpd: read failed: Input/output error
pcilib: sysfs_read_vpd: read failed: Input/output error
pcilib: sysfs_read_vpd: read failed: Input/output error
pcilib: sysfs_read_vpd: read failed: Input/output error
pcilib: sysfs_read_vpd: read failed: Input/output error

Comment 3 Erik Quaeghebeur 2021-03-13 22:26:44 UTC

I do not know whether this is a regression or not. As far as I know, the warning has no important practical consequences. But there are plenty of issues still with AMD Ryzen hardware, so who knows whether this plays a role or is a symptom.

Comment 4 Joshua McDonagh 2021-03-14 07:05:41 UTC

Is MCELOG output available?

Comment 5 Erik Quaeghebeur 2021-03-14 16:24:45 UTC

Created attachment 295849 [details]
dmesg output

Comment 6 Erik Quaeghebeur 2021-03-14 16:34:21 UTC

(In reply to Joshua McDonagh from comment #4)
> Is MCELOG output available?

Not currently. Do you want me to install mcelog and see if a log is generated (or mcelog messages are sent to the journal)? Is there any specific configuration for mcelog I should be aware of? (I read that mcelog does not work on AMD systems, but perhaps this is outdated.)

Comment 7 Joshua McDonagh 2021-03-14 20:27:02 UTC

MCELOG is functional with AMD processors- MCELOG will help give a more detailed view of I/O errors for PCI but usually only triggers when cpu root port is cause of the issue. 


seems to be a common issue in relation to Zen processors -  looking into this more, will update if any resolution found for your issue

Comment 8 Qwerty 2021-08-08 12:02:46 UTC

This issue also happen on 4750GE CPU & kernel 5.13.5-1-MANJARO
> kernel: pci 0000:00:00.2: can't derive routing for PCI INT A
> kernel: pci 0000:00:00.2: PCI INT A: not connected

lspci -vvv
>00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Renoir IOMMU
>       Subsystem: Advanced Micro Devices, Inc. [AMD] Renoir IOMMU
>       Control: I/O- Mem- BusMaster- SpecCycle- MemWINV- VGASnoop- ParErr-
>Stepping- SERR- FastB2B- DisINTx-
>       Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort-
><MAbort- >SERR- <PERR- INTx-
>       Interrupt: pin A routed to IRQ -2147483648
>       Capabilities: <access denied>

Let me know if i should try some command/s.

Comment 9 Bjorn Helgaas 2021-08-10 22:15:47 UTC

Check to be sure your BIOS is up to date.

Re comment #4, what is the MCELOG connection?  The "can't derive routing for PCI INT A" issue should be a firmware or ACPI thing.  If somebody collects acpidump information, we might be able to trace through the _PRT and figure out what's going on.

The "pcilib: sysfs_read_vpd: read failed: Input/output error" messages while running lspci are unrelated.  My guess is they're related to the Realtek device at 02:00.0.  There might be corresponding notes in dmesg, too.

Comment 10 Erik Quaeghebeur 2021-08-21 21:06:14 UTC

(In reply to Bjorn Helgaas from comment #9)
> Check to be sure your BIOS is up to date.

I have the latest BIOS (as of today). The issue is still there.
 
> […] If somebody collects
> acpidump information, we might be able to trace through the _PRT and figure
> out what's going on.

I can call acpidump. Please tell me how to invoke it and how to save the output to attach to this report and I'll do it gladly.

Comment 11 Erik Quaeghebeur 2021-10-25 08:31:16 UTC

(In reply to Erik Quaeghebeur from comment #10)
> (In reply to Bjorn Helgaas from comment #9)
>  
> > […] If somebody collects
> > acpidump information, we might be able to trace through the _PRT and figure
> > out what's going on.
> 
> I can call acpidump. Please tell me how to invoke it and how to save the
> output to attach to this report and I'll do it gladly.

Pinging this bug to get the info necessary to do a useful acpidump.

Comment 12 Qwerty 2021-10-25 08:46:44 UTC

i have mentioned here earlier ( https://bugzilla.kernel.org/show_bug.cgi?id=212261#c8 ) that i have seen:
"pci 0000:00:00.2: can't derive routing for PCI INT A"

but now when i do: sudo dmesg|grep -Ei "0000:00:00.2|pci"
(exctracting that 0000 string or string "pci" from dmesg)

i no longer see any issue or warning. Only see three regular lines regarding keyboard and mouse.

Previously i was on: 5.13.5-1-MANJARO
Now i am on: 5.13.19-2-MANJARO

So i guess i can now un-follow this bug? If no, let me know which commands to run to get targeted and anonymized debug info.

Comment 13 Marcel Ziswiler 2021-11-16 23:38:04 UTC

I am on Fedora Silverblue 35 running 5.14.17-301.fc35.x86_64 and latest LENOVO 20UES1Y600/20UES1Y600, BIOS R1BET66W(1.35 ) 07/30/2021 still getting the same:

Nov 16 15:16:14 fedora kernel: pci 0000:00:00.2: can't derive routing for PCI INT A
Nov 16 15:16:14 fedora kernel: pci 0000:00:00.2: PCI INT A: not connected

Let me know how I may help provide more information. Thanks!

Comment 14 Sergey S 2021-12-04 15:26:01 UTC

```
2021-12-04T18:13:05+0300 hp kernel:   #2  #3  #4  #5  #6  #7  #8  #9 #10 #11 #12 #13 #14 #15
2021-12-04T18:13:05+0300 hp kernel: pci 0000:00:00.2: can't derive routing for PCI INT A
2021-12-04T18:13:05+0300 hp kernel: pci 0000:00:00.2: PCI INT A: not connected
```

 OS: Arch Linux 
 Kernel: x86_64 Linux 5.15.6-arch1-1
 CPU: AMD Ryzen 7 5800U with Radeon Graphics @ 16x 1.9GHz

Comment 15 Sergey S 2021-12-04 16:10:58 UTC

# lspci -Dvs 0000:00:00.2
0000:00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Renoir IOMMU
	Subsystem: Advanced Micro Devices, Inc. [AMD] Renoir IOMMU
	Flags: bus master, fast devsel, latency 0, IRQ -2147483648
	Capabilities: [40] Secure device <?>
	Capabilities: [64] MSI: Enable- Count=1/4 Maskable- 64bit+
	Capabilities: [74] HyperTransport: MSI Mapping Enable+ Fixed+

Comment 16 Qwerty 2021-12-05 19:32:41 UTC

(In reply to Qwerty from comment #12)
> i no longer see any issue or warning
> Previously i was on: 5.13.5-1-MANJARO
> Now i am on: 5.13.19-2-MANJARO

update: no under kernel 5.15.2-2-MANJARO i see it again:
kernel: PCI: CLS 64 bytes, default 64
kernel: pci 0000:00:00.2: AMD-Vi: IOMMU performance counters supported
kernel: pci 0000:00:00.2: can't derive routing for PCI INT A
kernel: pci 0000:00:00.2: PCI INT A: not connected

$ sudo lspci -Dvs 0000:00:00.2
0000:00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Renoir IOMMU
	Subsystem: Advanced Micro Devices, Inc. [AMD] Renoir IOMMU
	Flags: bus master, fast devsel, latency 0, IRQ -2147483648
	Capabilities: [40] Secure device <?>
	Capabilities: [64] MSI: Enable- Count=1/4 Maskable- 64bit+
	Capabilities: [74] HyperTransport: MSI Mapping Enable+ Fixed+

Comment 17 Matt H 2022-03-27 12:16:18 UTC

On a Lenovo P14s AMD Gen 1:

[    0.408536] pci 0000:00:00.2: AMD-Vi: IOMMU performance counters supported
[    0.408603] pci 0000:00:00.2: can't derive routing for PCI INT A
[    0.408604] pci 0000:00:00.2: PCI INT A: not connected

$ sudo lspci -Dvs 0000:00:00.2
0000:00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Renoir/Cezanne IOMMU
	Subsystem: Lenovo Renoir/Cezanne IOMMU
	Flags: bus master, fast devsel, latency 0, IRQ -2147483648
	Capabilities: [40] Secure device <?>
	Capabilities: [64] MSI: Enable- Count=1/4 Maskable- 64bit+
	Capabilities: [74] HyperTransport: MSI Mapping Enable+ Fixed+

I'm attaching output of `acpidump`.

Comment 18 Matt H 2022-03-27 12:18:04 UTC

Created attachment 300624 [details]
acpidump output from Lenovo P14s AMD Gen 1

Comment 19 Karri Hämäläinen 2022-07-07 12:51:05 UTC

Created attachment 301360 [details]
acpidump from Asus TUF B450M-Pro Gaming

Same with Ryzen 2200G.

Systems:

AMD Ryzen 3 2200G
AsRock B450M Pro4-F R2.0
Bios P3.10 (latest)
AGESA 1.2.0.6

AMD Ryzen 3 2200G
Asus TUF B450M-Pro Gaming
Bios 3802 (latest)
AGESA 1.2.0.7

OS tested:
Debian Bullseye 11.3: Kernel 5.10.120
Debian Sid: Kernel 5.18.5
Debian Sid: Experimental Kernel 5.19.0-rc4

No change.


From Asus MB with Kernel 5.18.5:

[    0.282567] pci 0000:00:00.2: AMD-Vi: IOMMU performance counters supported
[    0.282605] pci 0000:00:00.2: can't derive routing for PCI INT A
[    0.282606] pci 0000:00:00.2: PCI INT A: not connected


0000:00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Raven/Raven2 IOMMU
        Subsystem: ASUSTeK Computer Inc. PRIME B450M-A Motherboard
        Flags: bus master, fast devsel, latency 0, IRQ 25
        Capabilities: [40] Secure device <?>
        Capabilities: [64] MSI: Enable+ Count=1/4 Maskable- 64bit+
        Capabilities: [74] HyperTransport: MSI Mapping Enable+ Fixed+

Attached acpidump from same.

Comment 20 Karri Hämäläinen 2022-07-15 13:56:59 UTC

Created attachment 301438 [details]
acpidump from ASRock B450M Pro4-F R2.0

Same with Ryzen Pro 4650G.

Systems:

AMD Ryzen Pro 5 4650G
ASRock B450M Pro4-F R2.0
Bios P3.10 (latest)
AGESA 1.2.0.6


OS tested:
Debian Sid: Kernel 5.18.5


[    0.273902] pci 0000:00:00.2: AMD-Vi: IOMMU performance counters supported
[    0.273932] pci 0000:00:00.2: can't derive routing for PCI INT A
[    0.273933] pci 0000:00:00.2: PCI INT A: not connected


00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Renoir/Cezanne IOMMU
        Subsystem: Advanced Micro Devices, Inc. [AMD] Renoir/Cezanne IOMMU
        Control: I/O- Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0
        Interrupt: pin A routed to IRQ -2147483648
        Capabilities: [40] Secure device <?>
        Capabilities: [64] MSI: Enable- Count=1/4 Maskable- 64bit+
                Address: 0000000000000000  Data: 0000
        Capabilities: [74] HyperTransport: MSI Mapping Enable+ Fixed+


Attached acpidump from same.


Didn't do cross-motherboard or cross-kernel tests for now.
Can still do if asked, but changing processors contains some risk of damage.

Comment 21 Nelson G 2022-08-10 19:52:39 UTC

Created attachment 301545 [details]
acpidump thinkpad e495

Hello, just wanted to report I got this as well, Thinkpad E495,  ryzen 3500u

$ sudo lspci -Dvs 0000:00:00.2
0000:00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Raven/Raven2 IOMMU
	Subsystem: Lenovo Raven/Raven2 IOMMU
	Flags: fast devsel, IRQ 25
	Capabilities: [40] Secure device <?>
	Capabilities: [64] MSI: Enable+ Count=1/4 Maskable- 64bit+
	Capabilities: [74] HyperTransport: MSI Mapping Enable+ Fixed+

Comment 22 lebaan 2023-10-06 16:24:26 UTC

I, as well.

% lspci -Dvs 0000:00:14.0
0000:00:14.0 USB controller: Intel Corporation 7 Series/C210 Series Chipset Family USB xHCI Host Controller (rev 04) (prog-if 30 [XHCI])
        Subsystem: Hewlett-Packard Company Elitebook 8470p
        Flags: bus master, medium devsel, latency 0, IRQ 31
        Memory at d4420000 (64-bit, non-prefetchable) [size=64K]
        Capabilities: <access denied>
        Kernel driver in use: xhci_hcd

Comment 23 Archttila 2023-12-30 11:19:56 UTC

Same with Ryzen 5700G

Dec 30 11:28:48 desktop kernel: pci 0000:00:00.2: can't derive routing for PCI INT A
Dec 30 11:28:48 desktop kernel: pci 0000:00:00.2: PCI INT A: not connected
Dec 30 11:28:48 desktop kernel: pci 0000:00:00.2: AMD-Vi: Found IOMMU cap 0x40

Arch
6.6.8-AMD-znver3

Asrock DeskMini X300
/X300M-STX, BIOS P1.80F 05/09/2023

Comment 24 Mateusz Kaduk 2024-02-15 19:11:26 UTC

I have this as well with Lenovo Legion 9 16IRX8

[    1.401876] intel-lpss 0000:00:15.0: enabling device (0004 -> 0006)
[    1.401968] intel-lpss 0000:00:15.0: can't derive routing for PCI INT A
[    1.401970] intel-lpss 0000:00:15.0: PCI INT A: not connected
[    1.402000] intel-lpss: probe of 0000:00:15.0 failed with error -22
[    1.414068] intel-lpss 0000:00:15.2: enabling device (0004 -> 0006)
[    1.414149] intel-lpss 0000:00:15.2: can't derive routing for PCI INT C
[    1.414149] intel-lpss 0000:00:15.2: PCI INT C: not connected
[    1.414166] intel-lpss: probe of 0000:00:15.2 failed with error -22
[    1.426224] intel-lpss 0000:00:19.0: enabling device (0004 -> 0006)
[    1.426306] intel-lpss 0000:00:19.0: can't derive routing for PCI INT A
[    1.426307] intel-lpss 0000:00:19.0: PCI INT A: not connected
[    1.426327] intel-lpss: probe of 0000:00:19.0 failed with error -22
[    1.438199] intel-lpss 0000:00:19.1: enabling device (0004 -> 0006)
[    1.438270] intel-lpss 0000:00:19.1: can't derive routing for PCI INT B
[    1.438271] intel-lpss 0000:00:19.1: PCI INT B: not connected
[    1.438295] intel-lpss: probe of 0000:00:19.1 failed with error -22
[    1.450213] intel-lpss 0000:00:1e.0: enabling device (0004 -> 0006)
[    1.450311] intel-lpss 0000:00:1e.0: can't derive routing for PCI INT A
[    1.450311] intel-lpss 0000:00:1e.0: PCI INT A: not connected
[    1.450327] intel-lpss: probe of 0000:00:1e.0 failed with error -22
[    1.462348] intel-lpss 0000:00:1e.3: enabling device (0004 -> 0006)
[    1.462409] intel-lpss 0000:00:1e.3: can't derive routing for PCI INT D
[    1.462410] intel-lpss 0000:00:1e.3: PCI INT D: not connected
[    1.462425] intel-lpss: probe of 0000:00:1e.3 failed with error -22

Comment 25 Ivan Stanković 2024-03-17 16:55:32 UTC

Same with HP EliteBook 845 G9, with Ryzen 7 PRO 6850U:

kernel: pci 0000:00:00.2: can't derive routing for PCI INT A 
kernel: pci 0000:00:00.2: PCI INT A: not connected           


0000:00:00.2 IOMMU: Advanced Micro Devices, Inc. [AMD] Family 17h-19h IOMMU
        Subsystem: Advanced Micro Devices, Inc. [AMD] Family 17h-19h IOMMU 
        Flags: fast devsel, IRQ 25                                         
        Capabilities: [40] Secure device <?>                               
        Capabilities: [64] MSI: Enable+ Count=1/4 Maskable- 64bit+         
        Capabilities: [74] HyperTransport: MSI Mapping Enable+ Fixed+