> On Feb 9, 2021, at 6:05 PM, Qiuxu Zhuo <qiuxu.z...@intel.com> wrote:
>
> On a Sapphire Rapids server, it failed to inject correctable errors
> to the RCiEP device e8:02.0 which was associated with the RCEC device
> e8:00.4. See the following error log before applying the patch:
>
> aer-inject -s e8:02.0 examples/correctable
> Error: Failed to write, No such device
>
> This was because rcec_assoc_rciep() mistakenly used "rciep->devfn" as
> device number to check whether the corresponding bit was set in
> the RCiEPBitmap of the RCEC. So that the RCiEP device e8:02.0 wasn't
> linked to the RCEC and resulted in the above error.
>
> Fix it by using PCI_SLOT() to convert rciep->devfn to device number.
> Ensure that the RCiEP devices associated with the RCEC are linked to
> the RCEC as the RCEC is enumerated. After applying the patch, correctable
> errors can be injected to the RCiEP successfully.
>
> Reported-and-tested-by: Wen Jin <wen....@intel.com>
> Signed-off-by: Qiuxu Zhuo <qiuxu.z...@intel.com>
Reviewed-by: Sean V Kelley <sean.v.kel...@intel.com>
Thanks,
Sean
> ---
> drivers/pci/pcie/rcec.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/pci/pcie/rcec.c b/drivers/pci/pcie/rcec.c
> index 2c5c552994e4..d0bcd141ac9c 100644
> --- a/drivers/pci/pcie/rcec.c
> +++ b/drivers/pci/pcie/rcec.c
> @@ -32,7 +32,7 @@ static bool rcec_assoc_rciep(struct pci_dev *rcec, struct
> pci_dev *rciep)
>
> /* Same bus, so check bitmap */
> for_each_set_bit(devn, &bitmap, 32)
> - if (devn == rciep->devfn)
> + if (devn == PCI_SLOT(rciep->devfn))
> return true;
>
> return false;
> --
> 2.17.1
>
