Re: [Xen-devel] [PATCH V4] ns16550: Add support for UART present in Broadcom TruManage capable NetXtreme chips

["Jan Beulich" <JBeulich@xxxxxxxx>]

>>> On 21.11.13 at 23:50, Aravind Gopalakrishnan 
>>> <Aravind.Gopalakrishnan@xxxxxxx> wrote:
> @@ -432,9 +474,16 @@ static void __init ns16550_endboot(struct serial_port 
> *port)
>  {
>  #ifdef HAS_IOPORTS
>      struct ns16550 *uart = port->uart;
> -
> +    unsigned long sfn, efn;
> +    

Stray blanks.

>      if ( uart->remapped_io_base )
> +    {
> +        sfn = paddr_to_pfn((unsigned long) uart->io_base + PAGE_SIZE);
> +        efn = paddr_to_pfn((unsigned long) uart->io_base + uart->io_size - 
> 1);

These casts aren't correct for 32-bit hosts (even if we don't have
any that define HAS_IOPORTS). And anyway, using PFN_UP()
and PFN_DOWN() here respectively would be much easier to read.

> +        if ( iomem_deny_access(dom0, sfn, efn) != 0 )

You need to handle the case of sfn > efn, I think.

> +                    /* Handle 64 bit BAR if found */ 
> +                    if ( bar & PCI_BASE_ADDRESS_MEM_TYPE_64 )
> +                    {

Do you really need that (and the respective changes elsewhere)?
I'd be fine with simply bailing in that case for the time being.

> +                    }
> +                    else
> +                    {
> +                        len &= PCI_BASE_ADDRESS_MEM_MASK;
> +                        uart->io_size = len & ~(len - 1);

Once again: "len & -len" is the simpler alternative.

> +                /* IO based */
> +                else
> +                {
> +                    pci_conf_write32(0, b, d, f, 
> +                                     PCI_BASE_ADDRESS_0 + bar_idx*4, ~0u);
> +                    len = pci_conf_read32(0, b, d, f, PCI_BASE_ADDRESS_0);
> +                    pci_conf_write32(0, b, d, f, 
> +                                     PCI_BASE_ADDRESS_0 + bar_idx*4, bar);
> +                    len &= PCI_BASE_ADDRESS_IO_MASK;
> +                    len &= ~(len - 1);

len &= -len (PCI_BASE_ADDRESS_IO_MASK being ~0x03UL I don't
see why this would be incorrect, as you indicated in a reply to v3).

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel