www.elektronik.si

[.]

Brisana vsebina odstranjenega uporabnika.

[Sokrat]

Funkcija CRC16 zahteva kot prvi parameter kazalec na tip char (pointer na 1 byte), ti pa podas kazalec na strukturo, deklarirano zgoraj, ki ni tipa char ? Konec koncev se vsak podatek lahko obravnava kot byte (ali karkoli drugega), tako da nekaj najbrz res deluje, vprasanje pa je, ali je to res to kar bi ti rad (odvisno od tega, kako so v tvoji arhitekturi razporejeni posamezni bajti znotraj longa in koliko je long dolg).

[NeoTO]

Če veš kaj si mislil s tem narediti, ti svetujem naslednje:
Ustvari nov pointer tipa char in mu priredi naslov tvojega pointerja Data.

[Sokrat]

So polja tipa long sploh nujna v tem primeru ? Domnevam namrec, da PICi nimajo ukazov za delo z 32-bitnimi integerji

[.]

Brisana vsebina odstranjenega uporabnika.

[Sokrat]

Mare, struktura v zgornjem primeru je velika (dolga) 32 bajtov. Tvoj primer je dolg samo 8 bajtov, zato bi bil izracun CRC s tisto vrednostjo dolzine (32) nepricakovan (odvisno od tega, kaj je v RAMu za tisto strukturo), pa se ostala koda bi morda delala narobe, ce se zanasa na dolzino 32 bajtov.

Torej ali

[.]

Brisana vsebina odstranjenega uporabnika.

[Sokrat]

No, jaz bi na tvojem mestu modificiral funkcijo CRC16. Spremeni prvi parameter v njeni deklaraciji v pointer na tisto tvojo strukturo, v telesu funkcije pa dodaj se eno spremenljivko, ki bio pointer na uchar (unsigned char * kazalec) in takoj po klicu funkcije spremenljivki kazalec pripisi vrednost prvega parametra, torej pointerja na strukturo.

Druga (najbrz slabsa) moznost je, da strukturo deklariras kot polje uchar-ov in sam premetavas podatke (ko imas carry pac stejes naprej v naslednji spremenljivki - vsi mikroprocesorji/MCUji imajo carry flag ravno zato), kar pa je bolj glomazno - zdaj to pocne prevajalnik namesto tebe in ti skrije "grdo" delo ... bolj ali manj optimalno.

Se eno vprasanje, ker ne vemo prakticno nicesar o tvoji aplikaciji: morda bi bilo bolj ucinkovito, da bi namesto CRC kode in retransmita uporabljal Hammingovo kodiranje ("ECC"). Tako se ti napake prvega reda odpravljajo same, ob napakah drugega reda pa sele zahtevas retransmit. Pri CRC bos retransmit zahteval ob napaki prvega reda, ce ti jo bo funkcija seveda odkrila (obstaja moznost, da je ne bo, kar sicer velja za vsako tako preverjanje). Koliko napak pa pricakujes pri tvojih prenosih ?

[NeoTO]

Kaj pa če ustvariš union za tvojo strukturo?

Iz PICmicro MCU C®
An introduction to programming
The Microchip PIC in CCS C
By Nigel Gardner

8.4 Introduction to Unions
A union is defined as a single memory location that is shared by two or more
variables. The variables that share the memory location may be of different
data types. However, you may only use one variable at a time. A union looks
very much like a structure. The general format of the union is:
union tag-name
{
type element1;
type element2;
.
.
type elementn;
} variable-list;
Again, the tag-name is the name of the union and the variable-list are the
variables that have a union type tag-name. The difference between unions
and structures is that each member of the union shares the same data space.
For example, the following union contains three members: an integer, a
character array, and a double.
union u_type
{
int i;
char c[3];
double d;
} temp;
The way that a union appears in memory is shown below. We will use the
previous example to illustrate a union. The integer uses two bytes, the
character array uses three bytes and the double uses four bytes.
<--------------------------------------------------double--------------------------------------------------->
<---------c[2]----------> <---------c[1]----------> <---------c[0]---------->
<---------------------integer---------------------->
element0 element1 element2 element3
Accessing the members of the union is the same as with structures, you use a
period. The statement temp.i will access the two byte integer member i of
the union temp and temp.d will access the four byte double d. If you are
accessing the union through a pointer, you would use the arrow operator just
like structures.
It is important to note that the size of the union is fixed at complier time to
accommodate the largest member of the union. Assuming that doubles are
four bytes long, the union temp will have a length of four bytes.
103
A good example of using a union is when an 8-bit microcontroller has an
external 12-bit A/D converter connected to a serial port. The microcontroller
reads the A/D in two bytes. So we might set up a union that has two unsigned
chars and a signed short as the members.
union sample
{
unsigned char bytes[2];
signed short word;
}
When you want to read the A/D, you would read two bytes of data from the
A/D and store them in the bytes array. Then, whenever you want to use the
12-bit sample you would use word to access the 12-bit number.[/i]

[.]

Brisana vsebina odstranjenega uporabnika.

[Sokrat]

[Proteus]

[NeoTO]

Prilagam ti svoj programček za paketno prenašanje. Poglej če si lahko kaj pomagaš!

[.]

Brisana vsebina odstranjenega uporabnika.

[Sokrat]

Na kratko: dodaj & 0xff k TMR1L = timer_setupin, torej