Problem with doing intersegmental syscalls :( HELP!!!!!!

So, I have got two kernels,
One is the main one called from my own second stage bootloader known as NTLOAD.SYS, as APOS16.SYS (Kernel Filename),
Now that I need to extend my kernel API and keeping my kernel size limited, I've got to do intersegmental syscalls like :

Code:

; Assuming I'm in segment 0x2000
; I would do this
jmp 0x3000:0x0000
jmp 0x3000:0x0003
jmp 0x3000:0x0004
; and so on........

Basically I would load a file called BOOTDRV.SYS which contains these calls,
Now I have written some code but they don't seem to work.
My Kernel Caller,

Code:

; Load the BOOTDRV.SYS (Boot Drivers) in memory at 0x3000
        ; First Initialize the FDD
        call init_FDD
        ; Set up segments
        cli                             ; Clear interrupts
        mov ax, 0
        mov ss, ax                      
        mov sp, 0FFFFh
        sti                             
        cld                             
        mov ax,2000h                    
        mov ds, ax                      
        mov es, ax                      
        mov fs, ax                      
        mov ax,0x3000 ; ---> Reserved for BOOTDRV.SYS
        mov gs,ax
        call load_kernel
        call os_command_line ; Start the Interface
load_kernel:
        mov ax, .pos16_sys
        mov cx, 28768
        call os_load_file
        jc os_fatal_error
        push es
        push 0x3000
        pop es
        mov si, 28768
        mov di, 0
        mov cx, bx
        rep movsb
        mov ax, 0000h
        mov es, ax
        mov word [es:0014h], 0x3000
        mov word [es:0016h], .BOOT_BREAK
        mov word [es:006Ch], 0x3000 
        mov word [es:006Eh], .BOOT_BREAK
        pop es
        ret
        .pos16_sys db 'BOOTDRV.SYS',0
.BOOT_BREAK:
        cli
        pop ax
        pop ax
        push 2000h
        push os_command_line
        sti
        iret

And here's how my extended kernel looks like :

Code:

; POS16 Extended API
jmp os_lol ; 0x3000:0x0000
SEGMENT_RETURN:
        pushf
        pop word [gs:flags_tmp]
        cmp byte [gs:internal_call], 1
        jge .internal_return
        mov word [gs:return_ax_tmp], ax
        mov ax, fs
        mov ds, ax
        mov es, ax
        pop ax
        push 0x2000
        push ax
        mov ax, [gs:return_ax_tmp]
        push word [gs:flags_tmp]
        popf
        retf
        .internal_return:
                ret
flags_tmp                       dw 0
internal_call                   dw 0            ; cancels os_return
return_ax_tmp                   dw 0
os_lol:
        mov ah,0x0E
        jmp .loop
.loop:
        lodsb
        cmp al,0
        je .done
        int 10h
.done:
        jmp SEGMENT_RETURN

And I perform the call like this :

Code:

ext_print equ 3000h:0000h
mov si,lol2
call ext_print
jmp back_2_interface
lol2 db 'LOL! Printed from the Extended Kernel',0

When I execute this my OS halts and then the screen is almost unreadable?
Is there something wrong?[/code]