Re: [PATCH 2.6.22-rc4]: libata: PATA-mode fixes for sis_sata

[yulin <yulin@starsoftcomm.com>]

Alan Cox 写道:
>> So one leader of our IT, who is a Windows expert, suggested we can use 
>> int 13h and I/O in Linux. But int 13h just is used in real mode.
>>
>> Can you give me some advice to persuade him? If it is feasible, can you 
>> introduce me?
>>     
>
> int13 is real mode only, runs polled and 16bit with one command at a
> time. As far as I know even Windows no longer uses it for anything but
> booting.
>
>   
>> Even can you give some good solutions to solve the compatibility problem?
>>     
>
> For hardware where we have info the right answer is to report bugs and
> get them fixed (or indeed send fixes). For the more general case the
> ata_generic driver will drive anything that is SFF8038i compliant and you
> can just add PCI identifiers to it for anything you hit.
>
> I am working on a generic ACPI driver to go with it for such cases.
>
> Alan
>
>
>   

Thank you for your help and explanations.

I will do some research work on ata_generic driver. Can you give me some 
advice or related information to do deeper research?

I alse found the problem of ACPI in our product, so we turn it off, but 
always caused some other problems. So i think a generic ACPI driver will 
be welcome.

And your mail is the best message received today before going home.

:)

 From YuLin