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

[Aravind Gopalakrishnan <aravind.gopalakrishnan@amd.com>]

On 11/22/2013 6:29 AM, Jan Beulich wrote:
>>>> On 21.11.13 at 23:50, Aravind Gopalakrishnan <Aravind.Gopalakrishnan@amd.com> 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.

Will fix this..

>>       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.

Okay, will correct them..

>> +        if ( iomem_deny_access(dom0, sfn, efn) != 0 )
> You need to handle the case of sfn > efn, I think.

Will do..

>> +                    /* 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.

Yes, as the device has a 64 bit BAR.

>> +                    }
>> +                    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
>
>

Ah. I see it now.. (Originally I took it to mean just len = -(len).. 
which worked for mmio case...)
Anyway, Will change it..

Thanks,
-Aravind.