Re: [PATCH v2 1/2] PCI: dwc: Fix a 64bit bug in dw_pcie_ep_raise_msix_irq()

[Ilpo Järvinen]

On Fri, 19 Jan 2024, Dan Carpenter wrote:

> The "msg_addr" variable is u64.  However, the "aligned_offset" is an
> unsigned int.  This means that when the code does:
> 
>         msg_addr &= ~aligned_offset;

Wouldn't it be more obvious to replace the entire line with this:
	msg_addr = ALIGN_DOWN(msg_addr, epc->mem->window.page_size);

+ add the #include for it. It should handle the casting to the same type 
internally.

-- 
 i.


> 
> it will unintentionally zero out the high 32 bits.  Declare
> "aligned_offset" as a u64 to address this bug.
> 
> Fixes: 2217fffcd63f ("PCI: dwc: endpoint: Fix dw_pcie_ep_raise_msix_irq() alignment support")
> Signed-off-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> ---
> v2: fix a typo in the commit message
> 
>  drivers/pci/controller/dwc/pcie-designware-ep.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/controller/dwc/pcie-designware-ep.c b/drivers/pci/controller/dwc/pcie-designware-ep.c
> index 5befed2dc02b..2b6607c23541 100644
> --- a/drivers/pci/controller/dwc/pcie-designware-ep.c
> +++ b/drivers/pci/controller/dwc/pcie-designware-ep.c
> @@ -525,7 +525,7 @@ int dw_pcie_ep_raise_msix_irq(struct dw_pcie_ep *ep, u8 func_no,
>  	struct dw_pcie_ep_func *ep_func;
>  	struct pci_epc *epc = ep->epc;
>  	u32 reg, msg_data, vec_ctrl;
> -	unsigned int aligned_offset;
> +	u64 aligned_offset;
>  	u32 tbl_offset;
>  	u64 msg_addr;
>  	int ret;
>