Re: [PATCH v8 1/4] s390/pci: Hold fmb_lock when enabling or disabling PCI devices

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] pci/s390: Race between pcibios_disable_device() and debugfs write leaves FMB enabled
--

commit 6846c80d4f9b53691af0a39418f819b11bc5390b
Author: Omar Elghoul <oelghoul@linux.ibm.com>

s390/pci: Hold fmb_lock when enabling or disabling PCI devices

Ensure that fmb_lock is held by pcibios_enable_device() and
pcibios_disable_device() when calling zpci_fmb_enable_device() or
zpci_fmb_disable_device(), respectively. Additionally, assert that the
fmb_lock is held within the latter two functions.

> diff --git a/arch/s390/pci/pci.c b/arch/s390/pci/pci.c
> index d64c544b32b0c..c055a9ad09724 100644
> --- a/arch/s390/pci/pci.c
> +++ b/arch/s390/pci/pci.c

[ ... ]

> @@ -648,7 +654,9 @@ void pcibios_disable_device(struct pci_dev *pdev)
>  {
>  	struct zpci_dev *zdev = to_zpci(pdev);
>  
> +	mutex_lock(&zdev->fmb_lock);
>  	zpci_fmb_disable_device(zdev);
> +	mutex_unlock(&zdev->fmb_lock);
>  	zpci_debug_exit_device(zdev);
>  }

[Severity: Medium]
Does this create a race condition with debugfs writes?

Releasing fmb_lock before calling zpci_debug_exit_device() creates a window
where a blocked debugfs writer to the statistics file via pci_perf_seq_write()
can wake up, acquire the lock, and re-enable the Function Measurement Block
(FMB).

zpci_debug_exit_device() will then wait for the writer to finish and tear down
the debugfs files, leaving the device disabled but with FMB actively running
in hardware.

Should zpci_debug_exit_device() be called before acquiring fmb_lock to drain
active accesses and prevent new ones?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261005154557.57801-1-oelghoul@linux.ibm.com?part=1