[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Getting Bullseye/Bookworm kernel running on Loongson 3A boards



On Sun, Jul 23, 2023 at 06:16:39PM +0200, Aurelien Jarno wrote:
> Dear MIPS porters,
> 
> DSA would like to upgrade the remaining mips*el buildds to bullseye (and
> later to bookworm), however we have trouble running the Bullseye or
> Bookworm kernel on the LS3A-RS780-1w ones.
> 
> Both Bullseye and Bookworm kernels boot fine up to the userland, but the
> network card does not work. It *seems* to an interrupt issue as the
> corresponding counter in /proc/interrupts does not change, and in
> addition the kernel logs show the following warnings:
> 
> [    5.077846] pcieport 0000:00:02.0: of_irq_parse_pci: failed with rc=-22
> [    5.084744] pcieport 0000:00:03.0: of_irq_parse_pci: failed with rc=-22
> [    5.170772] ahci 0000:00:11.0: of_irq_parse_pci: failed with rc=-22
> [    7.376554] pci 0000:00:07.0: of_irq_parse_pci: failed with rc=-22
> [    7.444769] ata_generic 0000:00:14.1: of_irq_parse_pci: failed with rc=-22
> [    7.495993] ehci-pci 0000:00:12.2: of_irq_parse_pci: failed with rc=-22
> [    7.624868] ehci-pci 0000:00:13.2: of_irq_parse_pci: failed with rc=-22
> [    7.753230] ohci-pci 0000:00:12.0: of_irq_parse_pci: failed with rc=-22
> [    7.886362] ohci-pci 0000:00:12.1: of_irq_parse_pci: failed with rc=-22
> [    8.018045] ohci-pci 0000:00:13.0: of_irq_parse_pci: failed with rc=-22
> [    8.150168] ohci-pci 0000:00:13.1: of_irq_parse_pci: failed with rc=-22
> [    8.294391] ohci-pci 0000:00:14.5: of_irq_parse_pci: failed with rc=-22
> [   18.398222] snd_hda_intel 0000:00:14.2: of_irq_parse_pci: failed with rc=-22
> [   18.547932] pci 0000:00:01.0: of_irq_parse_pci: failed with rc=-22
> 
> Any idea how to solve this issue? I have attached a full 6.1 kernel boot
> log for reference.

asking on linux-mips@vger.kernel.org might get more feedback on the issue...
>From the quick look at kernel code and device tree of the board IMHO
the DTS is missing interrupt mapping entries for the PCI bus.

Could someone from Loongson have a look at the issue ?

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea.                                                [ RFC1925, 2.3 ]


Reply to: