Serial port conformance tests

[Saulius Krasuckas]

* On Mon, 23 Mar 2009, Chris Teague wrote:
> 
> very useful tool would be a serial port loopback device.  Rather than 
> require conformance testers to attach a hardware loopback device (NULL 
> modem) to a physical port, could we create some virtual ports in wine 
> and connect them together? Maybe com98 and com99 for example?  Anything 
> written to one of the ports would be sent to the other one.

Chris, none of my very business, but do you need to test only Tx/Rx lines, 
o some more? (like DTR, RTS and the rest)


* On Tue, 24 Mar 2009, Alexandre Julliard wrote:
> 
> Serial data and ioctls are forwarded straight to the kernel, so you'd 
> need to put the loopback driver in the kernel, not in Wine.

Basic tests could probably cover Tx/Rx signals only, so IMHO none kernel 
module would be necessary -- some user space app can suffice (remserial, 
socat, nullmodem).  But that would tie such test to a specific 
configuration of a particular linux box...

Alexandre, for the core operation your proposal sounds very reasonable, 
but in the case of winetest, how it would work on Windows?

For tests, I suppose Wine really needs some stubby serial driver (eg one 
comming from the mentioned com0com), which would then communicate with 
ntdll/wineserver, where the missing functionality could probably be 
redirected to the linux kernel, when it's implemented there.  No?


With regard to Chris: personally I would start hacking at making com0com 
to work on Wine.