Re: Stubs needed by some Windows Update Installs

[Dmitry Timoshkov]

"Aric Stewart"  wrote:
> @@ -297,7 +297,7 @@
>  @ stub SetupSetSourceListA
>  @ stub SetupSetSourceListW
>  @ stdcall SetupTermDefaultQueueCallback(ptr)
> -@ stub SetupTerminateFileLog
> +@ stdcall SetupTerminateFileLog(ptr)

The argument here should be 'long'.

Stubs needed by some Windows Update Installs

[Robert Shearman]

Dmitry Timoshkov wrote:
> "Aric Stewart"  wrote:
> > @@ -297,7 +297,7 @@
> >  @ stub SetupSetSourceListA
> >  @ stub SetupSetSourceListW
> >  @ stdcall SetupTermDefaultQueueCallback(ptr)
> > -@ stub SetupTerminateFileLog
> > +@ stdcall SetupTerminateFileLog(ptr)
>
> The argument here should be 'long'.

No, Aric is correct, although it doesn't make a difference on the 32-bit
Intel platform. On others sizeof(long) might not be equal to sizeof(void*)
and "void *" is the real type of handles.

Rob

Stubs needed by some Windows Update Installs

[Dmitry Timoshkov]

"Robert Shearman"  wrote:
> No, Aric is correct, although it doesn't make a difference on the 32-bit
> Intel platform. On others sizeof(long) might not be equal to sizeof(void*)
> and "void *" is the real type of handles.

Alexandre explained that already. That's a historical thing. Anyway, Alexandre
has fixed this already when he committed the patch.

Stubs needed by some Windows Update Installs

[Robert Shearman]

Dmitry Timoshkov wrote:
> "Robert Shearman"  wrote:
> > No, Aric is correct, although it doesn't make a difference on the 32-bit
> > Intel platform. On others sizeof(long) might not be equal to
>
> sizeof(void*)
> > and "void *" is the real type of handles.
>
> Alexandre explained that already. That's a historical thing.
> Anyway, Alexandre
> has fixed this already when he committed the patch.

Yes, but it should be a janitorial task to convert all handles that are
longs in the .spec file to ptr. Why make it harder on ourselves in the long
run?

Rob