Re: [colorforth] bug in bit32?

[albert@xxxxxxxxxxxxxxxxxx (Albert van der Horst)]

On Thu, May 05, 2005 at 02:13:09PM -0400, Mark Slicker wrote:
> On Thu, 5 May 2005, Albert van der Horst wrote:
>
> >On Wed, May 04, 2005 at 04:54:26PM -0400, Mark Slicker wrote:
> >>Here is bit16 and bit32 from ftp://ftp.ultratechnology.com/GEN.ASM:
> ><SNIP>
> >
> >>Notice above by the lablel 'b32', 'shl  EAX, 1'. I think this should be
> >>'shl  AX, 1' like in 'bit16'. I think bit32 apears to work because the
> >>previous row in the bitmap will be shifted into the upper 16 bits of EAX,
> >>and will be displayed on the next call to bit32.
> >
> >As far as I understand it, this keeping of 16 bits for the next call of
> >bit32 is intended. Why would you call it bit32 if it only handles
> >16 bits? It does handle 32 bits, in two calls.
>
> I did not name it, but the name does not necesarilly imply what should be
> the correct logic for this function. bit16 is used to write a row of the
> icon bitmap. A row is 16 bits, bit16 is called 24 times by emit. This
> displays the 16x24 icons we are familer with. bit32 is also called 24
> times but instead by emit2 (2emit). The difference between bit16 and bit32
> is that bit32 writes four pixels (2x2 square) for each bit in the bitmap
> row. The effect of bit32 being defined the way it is, is that magnified
> characters (2emit), when displayed, will be shifted down by one pixel and
> the last row of the bitmap will be cut off. Maybe this behavior is
> intended, but from my view it looks clearly like a bug.

I expect EAX to be three time a character value, such that it contains
all zero's in the upper 16 bits.

Indeed bit16 is called 24 times. But only the first time
something is deposited at [DI]. After that zero's have been
shifted into AX.
Similarly bit32 is called 24 times. Again only the first
time something is deposited at [DI]. The second time
EAX is already zero, in view of the values supplied.
After that again EAX is certainly zero.

So this would mean that it doesn't matter whether a 16 or 32 bits
register is shifted. Apart from that I don't claim to understand
much of what is going on.

Your conclusion could be checked by placing a character C on the
lowest line. It should look bad with one line missing.

> Mark
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
> For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
> Main web page - http://www.colorforth.com
>

--
Albert van der Horst,Oranjestr 8,3511 RA UTRECHT,THE NETHERLANDS
Economic growth -- like all pyramid schemes -- ultimately falters.
albert@xxxxxxxxxxxxxxxxxx http://home.hccnet.nl/a.w.m.van.der.horst

---------------------------------------------------------------------
To unsubscribe, e-mail: colorforth-unsubscribe@xxxxxxxxxxxxxxxxxx
For additional commands, e-mail: colorforth-help@xxxxxxxxxxxxxxxxxx
Main web page - http://www.colorforth.com