Hello list. :)

I=E2=80=99m working on a simple data acquisition project based on the=20
16F9877A.
  I=E2=80=99m using HI-TECH PICC Lite v8.02pl1 as my compiler.  No simulato=
r=20
or
ICD, all code is run on device.

I=E2=80=99ve gotten things to work more or less, but for reasons I don=E2=
=80=99t
understand.  I was wondering if some folks might be able to provide me
with the "why".

First problem is with strings- they don=E2=80=99t work unless I use
-NO_STRING_PACK. I thought that the 877A could read its own program
memory and string packing should work.  The USART demo from
microchipc.com also fails when string packing is enabled, so it isn=E2=80=
=99t=20
my
code.

The second problem puzzles me more.  I have a function that writes data
out to an 8-bit bi-directional bus.  However, the parameter isn=E2=80=99t=
=20
passed
properly.  No matter what I pass in to the function, the parameter=20
comes
in a 0x80.  Here=E2=80=99s some pseudo code.

void outtobus(unsigned char data, unsigned char flag, unsigned char=20
flag2) {
        //...some code... does not modify =E2=80=98data=E2=80=99... sets port as o=
utput
        putchhex(data); // always prints 0x80
        //...more code... puts data to bus...
}

void initbus() {
        putchhex(0b10101010); //prints 0xAA as expected.
        outtobus(0b10101010); //prints 0x80, and puts 0x80 out to bus!?
}

void main() {
        initbus();
        //other functions that call outtobus also always send 0x80?
}

My solution to this one was to switch the compiler optimization level
from none to '9'.  Again, not sure why it fails with no optimization.

I=E2=80=99m just not so conformable with the "it works now" way of thinking=
,=20
I=E2=80=99d
rather know way it didn=E2=80=99t work before so I don't repeat the same=20
problems!

Thanks for any insight.
Jake

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.