Re: [PATCH 4/9] libata: normalize port_info, port_operations and sht tables

[Tejun Heo <htejun@gmail.com>]

Tejun Heo wrote:
>>> * No reason to set ata_pci_default_filter() for PIO-only drivers.
>> and your patches add the calls for the CS5520 ?
>>
>>> diff --git a/drivers/ata/pata_cs5520.c b/drivers/ata/pata_cs5520.c
>>> index 972ed9f..5614e76 100644
>>> --- a/drivers/ata/pata_cs5520.c
>>> +++ b/drivers/ata/pata_cs5520.c
>>> @@ -160,6 +160,7 @@ static struct scsi_host_template cs5520_sht = {
>>>  static struct ata_port_operations cs5520_port_ops = {
>>>  	.set_piomode		= cs5520_set_piomode,
>>>  	.set_dmamode		= cs5520_set_dmamode,
>>> +	.mode_filter		= ata_pci_default_filter,
>> This case is wrong. CS5520 doesn't do DMA (just VDMA which we don't
>> support) and in addition doesn't use BAR4 so its not a generic PCI
>> controller and this is asking for trouble later.

Hmm... cs5520's vdma looks like BMDMA and behaves like one.  BMDMA
methods are used and, although it doesn't use BAR4, ioaddr->bmdma_addr
is initialized, so as long as libata SFF layer is concerned, it can be
considered a BMDMA controller.

Thanks.

-- 
tejun