Not all devices using this driver are backed by a PCI device, so it's not fair to assume 'to_pci_dev()' is valid. Fortunately, we only were using this to convert straight back to a bare 'device', which means we were only doing no-op pointer arithmetic, and not actually accessing potentially out-of-bounds memory. But this is still bad practice. Just use 'ar->dev' directly, since that's really all we're looking for. Fixes: 209b2a68de76 ("ath10k: add platform regulatory domain support") Signed-off-by: Brian Norris <briannorris@xxxxxxxxxxxx> --- drivers/net/wireless/ath/ath10k/mac.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/net/wireless/ath/ath10k/mac.c b/drivers/net/wireless/ath/ath10k/mac.c index a1c2801ded10..5c4ee66382d9 100644 --- a/drivers/net/wireless/ath/ath10k/mac.c +++ b/drivers/net/wireless/ath/ath10k/mac.c @@ -8313,7 +8313,6 @@ static u32 ath10k_mac_wrdd_get_mcc(struct ath10k *ar, union acpi_object *wrdd) static int ath10k_mac_get_wrdd_regulatory(struct ath10k *ar, u16 *rd) { - struct pci_dev __maybe_unused *pdev = to_pci_dev(ar->dev); acpi_handle root_handle; acpi_handle handle; struct acpi_buffer wrdd = {ACPI_ALLOCATE_BUFFER, NULL}; @@ -8321,7 +8320,7 @@ static int ath10k_mac_get_wrdd_regulatory(struct ath10k *ar, u16 *rd) u32 alpha2_code; char alpha2[3]; - root_handle = ACPI_HANDLE(&pdev->dev); + root_handle = ACPI_HANDLE(ar->dev); if (!root_handle) return -EOPNOTSUPP; -- 2.19.1.930.g4563a0d9d0-goog