[openib-general] [PATCH 1/3 v2] osm: Changes for windows compatability

[Michael S. Tsirkin]

> Hi Hal.
> 
> Fixing windows compilation problems
> [V2 - Previous patch had an error]

I don't think "fixing windows compilation" is a real log description.
What kind of errors? Isn't there a better fix?

> Signed-off-by: Yevgeny Kliteynik <kliteyn at dev.mellanox.co.il>
> ---
>  osm/include/iba/ib_types.h |    2 ++
>  1 files changed, 2 insertions(+), 0 deletions(-)
> 
> diff --git a/osm/include/iba/ib_types.h b/osm/include/iba/ib_types.h
> index 723e8b9..ec65b64 100644
> --- a/osm/include/iba/ib_types.h
> +++ b/osm/include/iba/ib_types.h
> @@ -59,9 +59,10 @@ BEGIN_C_DECLS
>           #define OSM_EXPORT	__declspec(dllimport)
>      #endif
>      #define OSM_API __stdcall
> +    #define OSM_CDECL __cdecl
>  #else
>      #define OSM_EXPORT	extern
>      #define OSM_API
> +    #define OSM_CDECL
>      #define __ptr64
>  #endif
 
Why is this necessary at all?
http://msdn2.microsoft.com/en-us/library/zkwh89ks.aspx
	Microsoft Specific
	This is the default calling convention for C and C++ programs.

In other words it's the default, you don't have to declare it.

	Place the __cdecl modifier before a variable or a function name. Because the C
	naming and calling conventions are the default, the only time you need to use
	__cdecl is when you have specified the /Gz (stdcall) or /Gr (fastcall) compiler
	option. The /Gd compiler option forces the __cdecl calling convention.

So why are you compiling with /Gz, after the code is already littered with
OSM_API? And why is OSM_API necessary?

It seems to me the right thing might be to remove all of OSM_API/OSM_CDECL
from code, and just build everything on windows with consistent compiler flags.


-- 
MST