Re: [PATCH] Submitting PROMISE IDE Controllers Driver Patch

[Hank Yang]

> 1. Are there any praticular reasons for the fact that you
>     define the PCI_VENDOR_ID_PROMISE in ide-disk.c instead of
>     the kernel global ID header.

I find out it's already defined in pci_ids.h so shouldnt be needed here.
We will update this point later.
 
> 2. Why are you gurading all lba48 accesses by the promise
>     vendor id?

That's because we don't want to influence the on-board IDE and other IDE
controller's R/W routine. So it will just works on PROMISE controllers.
 
> 3. Are there any reasons for enabling the IDEDMA_TIMEOUT handling
>     unconditionally?

If we enable IDEDMA_TIMEOUT, kernel will look for
ide_dma_timeout_revovery() to retry again when drive's DMA timeout.

Thank you for yours efforts to look at this patch.

Best Regards
Hank Yang