r15720 - wtf?

[Aleksey Bragin]

Hi,
James - please, could you tell me the purpose of 15720 commit? "Adding Cromwell to Usb" means that you are doing only non-coding updates most of all, but turns out commit touches more than just xml build system integration.

But, *why* you changed variable name, in struct even(!!!!), interface to pinterface? You understand, and I already told you, it's Linux's source code, and it will just add more conflicts when doing 2.5->2.6 upgrade.

Please, in future, submit this kind of patches to Linux.
Though I thank you for the changes you have done to other files, xml files, etc.

WBR,
Aleksey Bragin.

[James Tabor]

Hi,
Aleksey Bragin wrote:
> Hi,
>
> James - please, could you tell me the purpose of 15720 commit? "Adding 
> Cromwell to Usb" means that you are doing only non-coding updates most 
> of all, but turns out commit touches more than just xml build system 
> integration.

Cromwell works! Why not have it in the system! There are no USBPDO-x or USBFDO-x,
but at lest it loads. I fixed the problem with the sys directory. It builds
clean now.
> But, *why* you changed variable name, in struct even(!!!!), interface to 
> pinterface? You understand, and I already told you, it's Linux's source 
> code, and it will just add more conflicts when doing 2.5->2.6 upgrade.

Blame the new build system!

I had no choice, interface was creating conflicts, I could not locate the
reason why ATM. I think it is a __USE_W32API thing, Interface must be used
some where else. Oh, Relax man, I'll fix it! 8^D

I had to comment out IoGetDeviceProperty, there is no pnp support yet so why make
this call? For the rest of the code, I will clean up the mess I made. Some things
should not have been committed.

I use 2.6.11ac7, I keep up with Linux all the time, importing for me is not
a problem.
> Please, in future, submit this kind of patches to Linux.
> Though I thank you for the changes you have done to other files, xml 
> files, etc.
>
> WBR,
> Aleksey Bragin.

Research Research and Research!

Here is the problem,,

Well, put simply,
Usbd and Usbport are Microsoft closed proto-calls. So, using linux as a base
we can create our own drivers. BTW Uhci.sys loads and recognizes the usb ports.
With my hardware it does. 8^)

Expect Cromwell to change into real WDM like files, look at serial and serenum.
Both BTW work with M$.

I'm doing more work tonight.

Sorry for the mess and Thanks!
James

[Hyperion]

James Tabor ha scritto:
> I had no choice, interface was creating conflicts, I could not locate the
> reason why ATM. I think it is a __USE_W32API thing, Interface must be 
> used
> some where else. Oh, Relax man, I'll fix it! 8^D

it's a COM thing, to declare interface pointers (COM is used in kernel 
mode for a few driver models, so it kind of makes sense). #undefine 
interface after including DDK headers should do the trick

[Ge van Geldorp]

> From: James Tabor
> > But, *why* you changed variable name, in struct even(!!!!), 
> > interface to pinterface?
>
> Blame the new build system!
>
> I had no choice, interface was creating conflicts, I could 
> not locate the reason why ATM. I think it is a __USE_W32API 
> thing, Interface must be used some where else.

My first guess would be that "#define interface struct" in
w32api/include/basetyps.h is messing things up for you.

Ge van Geldorp.