Re: more about serial ports: do they even work?

[Stefano Stabellini <stefano.stabellini@eu.citrix.com>]

David S. Ahern wrote:

> 
> Stefano Stabellini wrote:
>> Michael Tokarev wrote:
>>
>>> Other than that, an.. excellent idea, I wanted to propose
>>> just that when I first saw all this stuff, but was somewhat
>>> afraid.  And I *think* there's at least *some* sense.  Qemu
>>> is a CPU emulator and may work on another arch where those
>>> bits are defined differently.  Maybe that was the reason for
>>> all this converting - to be safe than sorry, so to say.  No?
>>>
>> Yes, this is exactly the reason why they were introduced in the first place.
>> Let's suppose that the guest defines those constants differently: we
>> need to parse them and covert them appropriately to the host format.
>> CHR_IOCTL_SERIAL_SET_TIOCM and CHR_TIOCM_DTR correspond to the guest
>> version of TIOCMSET and TIOCM_DTR and can be defined differently
>> depending on the particular guest arch.
> 
> The following works for me.  It fixes the existing checks in place for
> the GET and replicates that for the SET. The ioctl initialization is
> needed in the SET is needed.

It looks pretty good to me.

Acked-by: stefano.stabellini@eu.citrix.com