We are running a survey to help us improve the experience for all of our members. If you see the survey appear, please take the time to tell us about your experience if you can.
Hi, I've encountered a problem trying to implement autovectoring with an EZ-USB Development Board (AN2131-DK001) using the Keil evaluation DK51. I've scoured the net to see if anyone else has encountered the same problem, but to no avail. I have the following code in an .a51 file.
CSEG AT 43H equ $ + 2 DB 02H DB HIGH 4100h DS 1
#pragma intvector (0x40FD) //start at 4100h #pragma interval (4) void Sudav_ISR (void) interrupt 0 { } . . .
1) It is true that byte@45 "is automatically modified by the EZ-USB hardware and is therefore not writeable". 2 ) The DS instruction is the correct choice. I was just pointing out that there was no basis for Byte@45 == 0. 3 ) I have not tried your method. I have used:
#pragma NOIV // Do not generate interrupt vectors void ISR_Sof( void ) interrupt 0 { EZUSB_IRQ_CLEAR(); //do 1st USBIRQ = bmSOF; //Clear SOF IRQ ++mUsbFrame; }
NAME USBJmpTbl extrn code (ISR_Sudav, ISR_Sof, ISR_Sutok, ISR_Susp, ISR_Ures, ISR_IBN, ISR_Ep0in, ISR_Ep0out, ISR_Ep1in, ISR_Ep1out, ISR_Ep2in, ISR_Ep2out, ISR_Ep3in, ISR_Ep3out, ISR_Ep4in, ISR_Ep4out, ISR_Ep5in, ISR_Ep5out, ISR_Ep6in, ISR_Ep6out, ISR_Ep7in, ISR_Ep7out) public USB_AutoVector, USB_Jump_Table ;------------------------------------------------------------------------------ ; Interrupt Vectors ;------------------------------------------------------------------------------ CSEG AT 43H USB_AutoVector equ $ + 2 ljmp USB_Jump_Table ; Autovector will replace byte 45 ;------------------------------------------------------------------------------ ; USB Jump Table ;------------------------------------------------------------------------------ ?PR?USB_JUMP_TABLE?USBJT segment code page ; Place jump table on a page boundary RSEG ?PR?USB_JUMP_TABLE?USBJT ; autovector jump table USB_Jump_Table: ljmp ISR_Sudav ; Setup Data Available db 0 ljmp ISR_Sof ; Start of Frame db 0 ljmp ISR_Sutok ; Setup Data Loading db 0 ljmp ISR_Susp ; Global Suspend db 0 ljmp ISR_Ures ; USB Reset db 0 ljmp ISR_IBN ; IN Bulk NAK interrupt db 0 ljmp ISR_Ep0in ; End Point 0 In db 0 ljmp ISR_Ep0out ; End Point 0 Out db 0 ljmp ISR_Ep1in ; End Point 1 In db 0 ljmp ISR_Ep1out ; End Point 1 Out db 0 ljmp ISR_Ep2in db 0 ljmp ISR_Ep2out db 0 ljmp ISR_Ep3in db 0 ljmp ISR_Ep3out db 0 ljmp ISR_Ep4in db 0 ljmp ISR_Ep4out db 0 ljmp ISR_Ep5in db 0 ljmp ISR_Ep5out db 0 ljmp ISR_Ep6in db 0 ljmp ISR_Ep6out db 0 ljmp ISR_Ep7in db 0 ljmp ISR_Ep7out db 0 ; end
LOL. I got fed up trying to set up the jump table in C soooo I decided to code it in Assembly instead. (How I orginally had it) I've tried using the EXACT code same code that you're using (that code is floating around on the Keil Site) and yet the MON51 does NOT like the LJMP USB_Jump_Table instruction since it attempts to write code at location 45h. The kicker is, that I have disabled ALL of my interrupts EXCEPT for the Suspend interrupt. Unless I'm missing something, there should be no reason for the offset of 8 since the AVEC/IVEC register is 0 by default on reset and since I have not enabled the SUTOK interrupt. Also interesting, when I disable the autovectoring (AVEN = 0) even then, a LJMP 4100H will turn into LJMP 4108H when using the MON51 debugger. Btw, thanks for your help thus far =)