maping an unsigned char ofer a structure

My problem is the following. I have an unsigned char array packet[] and the following structures:

Code:

---

typedef struct set_serv_prop_s
{

        byte serviceType;
        union
        {
                heartbeatCommand_t        *heartbeatCommand;
                announcement_t                *announcement;
        } set_serv_propType;

} set_serv_prop_t;

typedef struct announcement_s
{

        union
        {
                announAllocCommand_t        *announAllocCommand;
                announAllocResponse_t        *announAllocResponse;
                announDownloadCommand_t        *announDownloadCommand;
                announRemCommand_t                *announRemCommand;
                announRemResponse_t                *announRemResponse;
                announMarkRemCommand_t        *announMarkRemCommand;
                announDownAbortCommand_t *announDownAbortCommand;
        } announcementType;

} announcement_t;

typedef struct announAllocCommand_s
{

        byte        encodingLaw;
        uint16        reserved;
        uint32        announcementSize;

} announAllocCommand_t;

---

the service type field tells what king of package it is.

i try to map it:

Code:

---

void read_set_serv_prop_from_pkt(set_serv_prop_t **head,
                                                                unsigned char* packet)
{
        int i;

        *head = (set_serv_prop_t*) packet.payload;
       
        for(i=0;i<packet.size;i++)
                printf("&#37;x ",packet.payload[i]);
        printf("\n");
       
        if((*head)->serviceType==0x05)
        {
                announAllocCommand_t *allocCommand = (*head)->set_serv_propType.announcement->announcementType.announAllocCommand;
                       
                        fprintf(OUTPUT, "\t%d\n",allocCommand->encodingLaw);
                        fprintf(OUTPUT, "\t%d\n",allocCommand->reserved);
                        fprintf(OUTPUT, "\t%d\n\n", allocCommand->announcementSize);
        }
}

---

and i get the output:

Code:

---

5 0 0 0 0 0 fa 0
        49
        12846
        -2147446592

---

5 - Service Type
0 - Encoding law
0 0 - Reserved
0 0 fa 0 -Announcement size

the output is totally wrong. What is the problem here? and how can I solve it ?

Please help... I'm very pressed by time

Code:

---

typedef struct announAllocCommand_s
{

        byte        encodingLaw;
        uint16        reserved;
        uint32        announcementSize;

} announAllocCommand_t;

---

This struct is probaboy ending up 8 bytes long. Is that what you expect?
The reason is:

Code:

---

byte encodingLaw;  // :@0, 1 byte long.
uint16 reserved; //: @2, 2 bytes long - 1 byte alignment inserted before this.
uint32 announcementSize;  //: @4 - 4 bytes long.

---

--
Mats

but i have the service type byte in the first structure. i want to map that char array on the structure

You could probably do something like this:

Code:

---

typedef struct announAllocCommand_s
{
        byte    serviceType;
        byte        encodingLaw;
        uint16        reserved;
        uint32        announcementSize;

} announAllocCommand_t;

typedef struct set_serv_prop_s
{

        union
        {
                byte serviceType;
                heartbeatCommand_t        heartbeatCommand;
                announcement_t                announcement;
        } set_serv_propType;

} set_serv_prop_t;

---

Note that I've removed the star from the "announceMent" and "heartbeadCommand", as I believe this is what you want - I expect that you are NOT sending pointers across the comms channel.

--
Mats

this last post makes sense to me. thank you very much. now i understand padding (or at least I think so :))

Quote:

---

Originally Posted by spank

this last post makes sense to me. thank you very much. now i understand padding (or at least I think so :))

---

Most compilers have a way that you can turn padding off, but unfortunately, every compiler manufacturer have their own idea of HOW you do that.

Padding is there to improve performance [and in some processors, bigger data items than bytes HAS to be aligned to certain boundaries, or the processor will throw an exception and abort the execution]. So if you have a single byte followed by a bigger type, e.g. a 32-bit integer, the 32-bit integer should be aligned to an even 4 byte (32-bit) address to make it efficent [and/or correct where relevant].

--
Mats

I am working on CodeWarrior and I build the target for a SC100 CCS Simulator. I'm forced to do this because I want to avoid endianness problems (the pc i'm working on is a intel P4 and the package is in network order). so in order to work only on big endian I use this.

Again thank you for helping me!