Array initialization in ext memory

 Array initialization in ext memory

Hi! i have use p89c669 microcontroller and am29f040b [512 kb] flash.When i using the external code memory i can't initialize the array that have more than 4 values.if i initialize more than that the program will not be executed.if i comment the array the program was executed.the code was

const char red [256] = {0x00,0x00...0x00,0x00};

please help me!

regrads,
K.T.Venkatesan.

 RE: Array initialization in ext memory

I don't know what your problem is, but just want to inform you that for many compilers, it is better to not add any initialization for variables that are expected to be zero on startup. Unless you make special arrangements, the startup code will automatically zero-fill all variables that haven't got specific initialization values.

 [ouestion on rs232]

Hi ,
I have designed a circuit with µPSD3200 which communicates with PC , µPSD3200 data to pc and gets data from pc with rs232 .
i want to send and receive numbers from zero two FF , not ascii , just hex code from 00 to FF (in decimal : 0 to 255) .
but , i have read that the computer and micro sends ascii code , and ascii codes does not support the whole range of 00 to FF , so what shall i do ?

omid

 RE: [ouestion on rs232] - off-topic

That has nothing to do with the subject of this thread, does it?

Start a new thread for your new subject!

 RE: startup code

"the startup code will automatically zero-fill all variables that haven't got specific initialization values."

That is to say, all non-automatic variables...

 More detail needed

"the program will not be executed"

You need to provide more detail that that!

What debugging have you done to determine exactly how & where it is failing?
Does it work in the simulator?

Are you sure that the startup code is correctly doing anything and evereything necessary to use external memory?

 initialize or define?

i can't initialize the array that have more than 4 values
I ran into a problem once which is that if an array crosses a 64k boundary in memory it will fail.

Erik

 RE: initialize or define?

Hi! in that case can we use two sperate arrays that each have 60kb.

regards,
K.T.Venkatesan.

 Misunderstanding?

"in that case can we use two sperate arrays that each have 60kb"

I think Erik is saying that the issue is not the actual size of the array, but that it crosses a 64K address boundary...?

 it is

Erik is saying that the issue is not the actual size of the array, but that it crosses a 64K address boundary

 RE: it is

Erik is saying that the issue is not the actual size of the array, but that it crosses a 64K address boundary then, then only way is seperate the 512kb into 8 blocks, am i right?

 More detail still needed

"the program will not be executed"

You still need to provide more detail than that!

 RE: More detail still needed

I have experience of the product and I might be able to help more than some other posters here; but only if you give more details.

I guess you are using banking for code space? also for data space?

If so, then what are the banking constraints?

 RE: More detail still needed

Hi! i use no more banking.

 RE: More detail still needed

Hi! Actualy i use the p89c669 microcontroller and amd 512 kb flash.now i had write a program for led blinking.after that program was loaded in the flash [external code memory] and the EA pin in the microcontroller was grounded.when i power on the target pcb the led was blinking as the per code.when i add an array array [76800] ={0x00,..0xfd}; in code the compiler give the error error c249:'!':SEGMENT TOO LARGE .

Last time i use the array as array[255] = {0x00,...0xff}; and load it in target board the leds was not blinking.then i change the array format as extern unsigned char const far array [256] ={0x00,...0xff} the problem was solved.

Regards,
K.T.Venkatesan.

 RE: More detail still needed

I think a close inspection of the linker map file would be appropriate.

 Surely not?

extern unsigned char const far array [256] ={0x00,...0xff}

You can't have an initialiser on an extern declaration, can you?

 RE: Surely not?

Why use the 'far'? I thought that was used on cores that support 24 bit addressing.

Are you beyond 64K? Don't cross the boundary!

Check the map file.

 RE: Surely not?

Why use the 'far'? I thought that was used on cores that support 24 bit addressing.
which the (soon to be discontinued) '669 does

Erik

 RE: Surely not?

You can't have an initialiser on an extern declaration, can you?

Yes. Well, yes on a "definition" but not on a "declaration", to be pedantic, but the syntax is legal, and from the code snippets above you can't really tell whether it was meant to be a declaration. The initializer makes it a definition, and you'll get some sort of "multiply defined" diagnostic message if the syntax shows up in a .h file included in several .c files (for example).

From the ANSI C89 spec:

3.7.2 External object definitions

Semantics

   If the declaration of an identifier for an object has file scope
and an initializer, the declaration is an external definition for the
identifier.

   A declaration of an identifier for an object that has file scope
without an initializer, and without a storage-class specifier or with
the storage-class specifier static , constitutes a tentative
definition.  If a translation unit contains one or more tentative
definitions for an identifier, and the translation unit contains no
external definition for that identifier, then the behavior is exactly
as if the translation unit contains a file scope declaration of that
identifier, with the composite type as of the end of the translation
unit, with an initializer equal to 0.

   If the declaration of an identifier for an object is a tentative
definition and has internal linkage, the declared type shall not be an
incomplete type.

Examples

         int i1 = 1;          /*  definition, external linkage */
         static int i2 = 2;   /*  definition, internal linkage */
         extern int i3 = 3;   /*  definition, external linkage */
         int i4;              /*  tentative definition, external linkage */
         static int i5;       /*  tentative definition, internal linkage */

         int i1;   /*  valid tentative definition, refers to previous */
         int i2;   /*  $3.1.2.2 renders undefined, linkage disagreement */
         int i3;   /*  valid tentative definition, refers to previous */
         int i4;   /*  valid tentative definition, refers to previous */
         int i5;   /*  $3.1.2.2 renders undefined, linkage disagreement */



         extern int i1; /* refers to previous, whose linkage is external */
         extern int i2; /* refers to previous, whose linkage is internal */
         extern int i3; /* refers to previous, whose linkage is external */
         extern int i4; /* refers to previous, whose linkage is external */
         extern int i5; /* refers to previous, whose linkage is internal */

Since I'm not a fan of programming languages doing much (if anything) implicitly, I prefer always to repeat the extern/static on the definition as well as the declarations. This at least earns me a diagnostic message when an object changes from external to internal, and serves as a useful reminder that "this object is accessible from outside", and keeps the syntax parallel with the "static" case.

 RE: Surely not?

Yes!i can initialize the array with extern
i see this in keil web site http://www.keil.com/support/docs/3128.htm

K.T.V

 You mis-read it

The whole point of that article is that you do the initialisation in assembler - not in 'C'!

 Have you done as suggested?

Have you examined the link map file ?

 RE: Have you done as suggested?

yes! here it is!

LX51 LINKER/LOCATER V4.02                                                               03/11/2008  11:52:22  PAGE 1


LX51 LINKER/LOCATER V4.02, INVOKED BY:
C:\KEIL\C51\BIN\LX51.EXE COMMON {START_MX.obj}, COMMON {main.obj}, BANK1 {code bank 1.obj}, BANK1 {.\L51_BANK.obj} TO LC
>>  color BANKAREA (0X008000, 0X00FFFF) CLASSES (EDATA (0X7F0000-0X7F04FF))


CPU MODE:     PHILIPS 80C51MX (ROM HUGE)
MEMORY MODEL: LARGE


INPUT MODULES INCLUDED:
  START_MX.obj (?C_STARTUP)
         COMMENT TYPE 0: AX51 V3.01
  main.obj (MAIN)
         COMMENT TYPE 0: CX51 V8.01
  code bank 1.obj (CODE_BANK_1)
         COMMENT TYPE 0: CX51 V8.01
  .\L51_BANK.obj (?BANK?SWITCHING)
         COMMENT TYPE 0: AX51 V3.01
  C:\KEIL\C51\LIB\CH51L.LIB (MALLOC)
         COMMENT TYPE 0: CX51 V7.20f
  C:\KEIL\C51\LIB\CH51L.LIB (INIT_MEM)
         COMMENT TYPE 0: CX51 V7.20f


ACTIVE MEMORY CLASSES OF MODULE:  LC color (?C_STARTUP)

BASE        START       END         USED      MEMORY CLASS
==========================================================
7F0000H     7F0000H     7F04FFH               EDATA
800000H     800000H     80FFFFH     0001CCH   CODE
7F0000H     7F0000H     7F00FFH     000001H   IDATA
000000H     800000H     FFFFFFH     0002FAH   HCONST
000000H     800000H     FFFFFFH     000453H   ECODE
000000H     000000H     00FFFFH     000020H   XDATA
000000H     000000H     FFFFFFH     000042H   HDATA
7F0000H     7F0000H     7F007FH     000008H   DATA

 RE: Have you done as suggested?

Hmmm,

You said before that you didn't use banking??!!

 RE: Have you done as suggested?

yes! today only i try the banking!

 why 'trying'

'trying' is the bane of many projects.
the result of a 'try' may come out right even if what you 'try' is the wrong approach.

the dead easy way to handle this is to do two things
a) make an .a51 module that contain ALL xdata and make them accessible from C by extern, then you have full control of the locations
b) access the biggie in an assembler routine

you will be home free

Erik

 by the way

I have done the exact same as above

Erik

 RE: by the way

Hi! the array problem was solved, now i using the banking method.but the bank size was limitted to 32kb.i can't enter more than 32kb in the keil/options for target1/bank area .

my heartly thanks to all!