Re: make test fails (Kernel 2.6.12.3)

In article <[email protected]> you wrote:
> Peter T. Breuer wrote:
>  > Them please explain your compiler report. As you can see, args 3 and 4
>  > are long long.

It looks as though your compiler computes long long plus long as giving
a long.

Is that right? Morally? Factually?

> As some of the mistakes I have made have been as stupid, I did a fresh 
> untar of the source, and did a make LINUXDIR=/usr/src/linux-2.6.12.3 and 
> I get the same error for lock.c:

That's hardly a test! A test would be to look hard at the compile line
concerned, and repeat it by hand.

> so my guess is the problem is with GCC somewhere.

I agree - but there is no need to guess.

> I misunderstood the question. From man ioctl
> 
>    int ioctl(int d, int request, ...);

Oh, of course,  it would be the NEXT arg that was the interesting one.
I suppose it will be a char * or a long, and it makes no difference
which, so forget it - no problem.

You can amuse yourself by fixing the minor complaints, and then see how
things go. I'll add the corrections this end too.

Peter