Re: [PATCH 2/5] arm/pci: Use official API to iterate over PCI buses

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Calling a sleepable lock from a hardware interrupt handler causes a 'scheduling while atomic' BUG.
--

commit f34de1e1be1288e13d00eee56b2fe60eb5f4ec85
Author: Gerd Bayer <gbayer@linux.ibm.com>

arm/pci: Use official API to iterate over PCI buses

This commit replaces iterating over pci_root_buses with pci_find_next_bus().
This change is intended to allow pci_root_buses to become private to the PCI
core.

> diff --git a/arch/arm/kernel/bios32.c b/arch/arm/kernel/bios32.c
> index ac0e890510daf..35642c9ba054c 100644
> --- a/arch/arm/kernel/bios32.c
> +++ b/arch/arm/kernel/bios32.c
> @@ -59,9 +59,9 @@ static void pcibios_bus_report_status(struct pci_bus *bus, u_int status_mask, in
>  
>  void pcibios_report_status(u_int status_mask, int warn)
>  {
> -	struct pci_bus *bus;
> +	struct pci_bus *bus = NULL;
>  
> -	list_for_each_entry(bus, &pci_root_buses, node)
> +	while ((bus = pci_find_next_bus(bus)) != NULL)

Can calling pci_find_next_bus() from an interrupt handler lead to a
scheduling while atomic bug?

In arch/arm/mach-footbridge/dc21285.c, pcibios_report_status() is called
directly from primary hardware interrupt handlers, such as dc21285_abort_irq()
and dc21285_serr_irq(). These handlers execute in an atomic hardirq context
where sleeping is not permitted.

Since pci_find_next_bus() unconditionally acquires the pci_bus_sem read-write
semaphore using down_read(), this introduces a blocking operation into that
atomic path:

dc21285_abort_irq() [hardirq context]
  pcibios_report_status()
    pci_find_next_bus()
      down_read(&pci_bus_sem) [sleeps]

Does this path need an alternative approach to safely iterate over the buses
without taking a sleeping lock?

>  		pcibios_bus_report_status(bus, status_mask, warn);
>  }

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260515-priv_root_buses-v1-0-f8e393c57390@linux.ibm.com?part=2