Re: [sdcc-devel] [ sdcc-Bugs-434347 ] Global constant array initialisation.

[Johan K. <joh...@id...>]

Because of the overwhelming response I received on my "printing on windoze"
request (none ;), it took a while before I had time to tackle this one (that
one did feed my bank account so it had to go first). Now it should be fixed.

Johan

P.S. If someone wants to know how to print a window or a canvas or a device
context or whatever in windoze: I know how. But I don't think I will tell
you :)

----- Original Message -----
From: Johan Knol <joh...@id...>
To: <sdc...@li...>
Sent: Tuesday, June 19, 2001 9:04 PM
Subject: Re: [sdcc-devel] [ sdcc-Bugs-434347 ] Global constant array
initialisation.


> Const used to be a storage class by it self that in the process (allways!)
> was changed to code, not anymore. Therefore now the "if (!tree->initMode)"
> in decorateType() comes in handy, however the tree->initMode didn't get
set
> while emitting maps. Now it does. This reveils however two more small
> problems. The first one being that global const arrays aren't by default
in
> cseg, and the trailing 0 in the array isn't casted to a code pointer
(which
> I don't think if ever did). Still looking....
>
> Johan
>     The best way to use a computer is to NOT use it AT ALL!
>
> P.S. I added this to the test case:
>
> void test (char index) {
>   static void (*const _tests[])(void) = {
>     testIncrement,
>     0
>   };
>   (*(_tests[index]))();
> }
>
> > static void
> > testIncrement(void)
> > {
> >     volatile  int i;
> >     i = 0;
> >     i--;
> > }
> >
> > static void (*const _tests[])(void) = {
> > testIncrement,
> > 0
> > };
> >
> > error   *** Attempt to assign value to a constant
> > variable
>
>
>
>
> _______________________________________________
> sdcc-devel mailing list
> sdc...@li...
> http://lists.sourceforge.net/lists/listinfo/sdcc-devel
>