On Wednesday 28 March 2018 05:20 PM, Niklas Cassel wrote:
> Since a 64-bit BAR consists of a BAR pair, and since there is no
> BAR after BAR_5, BAR_5 cannot be 64-bits wide.
> 
> This sanity check is done in pci_epc_clear_bar(), so that we don't need
> to do this sanity check in all epc->ops->clear_bar() implementations.
> 
> Signed-off-by: Niklas Cassel <[email protected]>

Acked-by: Kishon Vijay Abraham I <[email protected]>
> ---
> Kishon made a review comment that he wanted this:
> https://marc.info/?l=linux-kernel&m=152161168226203
> 
> Personally, I don't think that this check is needed,
> since pci_epc_set_bar() already does this check,
> and no one should modify the flags after pci_epc_set_bar()
> has been called.
> 
> If everyone agrees, then this patch could be dropped.
> 
>  drivers/pci/endpoint/pci-epc-core.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/pci/endpoint/pci-epc-core.c 
> b/drivers/pci/endpoint/pci-epc-core.c
> index eccc942043cb..b0ee42739c3c 100644
> --- a/drivers/pci/endpoint/pci-epc-core.c
> +++ b/drivers/pci/endpoint/pci-epc-core.c
> @@ -285,7 +285,9 @@ void pci_epc_clear_bar(struct pci_epc *epc, u8 func_no,
>  {
>       unsigned long flags;
>  
> -     if (IS_ERR_OR_NULL(epc) || func_no >= epc->max_functions)
> +     if (IS_ERR_OR_NULL(epc) || func_no >= epc->max_functions ||
> +         (epf_bar->barno == BAR_5 &&
> +          epf_bar->flags & PCI_BASE_ADDRESS_MEM_TYPE_64))
>               return;
>  
>       if (!epc->ops->clear_bar)
> 

Reply via email to