>Start AT91Bootstrap...
Copy application ...
Run ...

.section start
.text

#include "include/part.h"

/*----------------------------------------------------------------------------
Area Definition
----------------
Must be defined as function to put first in the code as it must be mapped
at offset 0 of the flash EBI_CSR0, ie. at address 0 before remap.
_---------------------------------------------------------------------------*/

/* Application startup entry point */

.globl reset
.align 4
reset:

/* Exception vectors (should be a branch to be detected as a valid code by the rom */
_exception_vectors:
b reset_vector /* reset */
b undef_vector /* Undefined Instruction */
b swi_vector /* Software Interrupt */
b pabt_vector /* Prefetch Abort */
b dabt_vector /* Data Abort */
b rsvd_vector /* reserved */
b irq_vector /* IRQ : read the AIC */
b fiq_vector /* FIQ */

undef_vector:
b undef_vector
swi_vector:
b swi_vector
pabt_vector:
b pabt_vector
dabt_vector:
b dabt_vector
rsvd_vector:
b rsvd_vector
irq_vector:
b irq_vector
fiq_vector:
b fiq_vector
reset_vector:

/* Init the stack */
_init_stack:
ldr sp,=TOP_OF_MEM

/* Test if main oscillator is enabled */
ldr r0,=AT91C_PMC_SR
ldr r1, [r0]
ldr r2,=AT91C_PMC_MOSCS
ands r1, r1, r2
bne _switch_to_mosc


/* Enable the main oscillator */
_enable_mosc:
ldr r0,=AT91C_PMC_MOR
mov r1, #(0x40 << 8)
ldr r2,=AT91C_CKGR_MOSCEN
orr r1, r1, r2
str r1, [r0]
ldr r0,=AT91C_PMC_SR
1:
ldr r1, [r0]
ldr r2,=AT91C_PMC_MOSCS
ands r1, r1, r2
beq 1b

/* Test if MCK == SLOW CLOCK */
_switch_to_mosc:
ldr r0,=AT91C_PMC_MCKR
ldr r1,=AT91C_PMC_CSS
ldr r2, [r0]
and r2, r2, r1
mov r1, #0
cmp r1, r2
/* No => Do nothing */
bne _init_bss
/* Yes => Switch to the main oscillator */
ldr r1,=AT91C_PMC_CSS_MAIN_CLK
ldr r2,=AT91C_PMC_PRES_CLK
orr r1, r1, r2
str r1, [r0]
ldr r0,=AT91C_PMC_SR
1:
ldr r1, [r0]
ldr r2,=AT91C_PMC_MCKRDY
ands r1, r1, r2
beq 1b

/* Copy the data section in RAM at .data link address */
_init_data:
ldr r2, =_lp_data
ldmia r2, {r1, r3, r4}
1:
cmp r3, r4
ldrcc r2, [r1], #4
strcc r2, [r3], #4
bcc 1b

/* Initialize the bss segment */
_init_bss:
adr r2, _lp_bss
ldmia r2, {r3, r4}
mov r2, #0
1:
cmp r3, r4
strcc r2, [r3], #4
bcc 1b

/* Branch on C code Main function (with interworking) */
_branch_main:
ldr r4, = main
mov lr, pc
bx r4

/* Branch to the application at the end of the bootstrap init */
//_go:
// ldr r1, =MACH_TYPE
// mov lr, pc
// bx r0

.align
_lp_data:
.word _etext
.word _sdata
.word _edata

_lp_bss:
.word _sbss
.word _ebss

#include "include/part.h"
#include "include/main.h"
#include "include/debug.h"

int main () {
    hw_init();

    dbg_print("Hello World!...\n\r");
    return 0;
}

# Link Address and Top_of_Memory
#LINK_ADDR=0x200000
#TOP_OF_MEMORY=0x301000
LINK_ADDR=0x21D00000
TOP_OF_MEMORY=0x301000

# Link Address and Top_of_Memory
LINK_ADDR=0x21D00000
TOP_OF_MEMORY=0x30100000

>Start AT91Bootstrap...
Copy application ...
Run ...

void dbgu_send_byte(char byte)
{
    while(!(AT91C_BASE_DBGU->DBGU_CSR & AT91C_US_TXRDY));
    AT91C_BASE_DBGU->DBGU_THR = byte;
}

int main(void)
{
    dbgu_send_byte('A');
    dbgu_send_byte('B');
    dbgu_send_byte('C');
    dbgu_send_byte('\r');
    dbgu_send_byte('\n');
    ...
}

#include "include/part.h"
#include "include/main.h"
#include "include/debug.h"

int main () {

while (1) {
dbg_print("Hello World!\n\r");
}

return 0;
}[/code]

и соответственно получил вывод в цикле.

Только единственное я кое что не понял, у меня все заработало только после того как я убрал инициализацию стека и Text, Data сегментов из cstartup.s

Cstartup у меня сейчас выглядит следующим образом:
[code].section start
.text

#include "include/part.h"

/*----------------------------------------------------------------------------
Area Definition
----------------
Must be defined as function to put first in the code as it must be mapped
at offset 0 of the flash EBI_CSR0, ie. at address 0 before remap.
_---------------------------------------------------------------------------*/

/* Application startup entry point */

.globl reset
.align 4
reset:

/* Exception vectors (should be a branch to be detected as a valid code by the rom */
_exception_vectors:
b reset_vector /* reset */
b undef_vector /* Undefined Instruction */
b swi_vector /* Software Interrupt */
b pabt_vector /* Prefetch Abort */
b dabt_vector /* Data Abort */
b rsvd_vector /* reserved */
b irq_vector /* IRQ : read the AIC */
b fiq_vector /* FIQ */

undef_vector:
b undef_vector
swi_vector:
b swi_vector
pabt_vector:
b pabt_vector
dabt_vector:
b dabt_vector
rsvd_vector:
b rsvd_vector
irq_vector:
b irq_vector
fiq_vector:
b fiq_vector
reset_vector:

/* Branch on C code Main function (with interworking) */
_branch_main:
ldr r0, = main
mov lr, pc
bx r0

/* Branch to the application at the end of the bootstrap init */
//_go:
// ldr r1, =MACH_TYPE
// mov lr, pc
// bx r0

.align
_lp_data:
.word _etext
.word _sdata
.word _edata

_lp_bss:
.word _sbss
.word _ebss