[U-Boot-Users] Booting with attached BDI but not alone
Hi,
I downloaded an u-boot.bin (1.1.4) on to my flash (a board with MPC875). If I connect BDI and type go I see output on my serial console. If I start the board without the attached BDI nothing appears.
Any hint?
I downloaded an u-boot.bin (1.1.4) on to my flash (a board with MPC875). If I connect BDI and type go I see output on my serial console. If I start the board without the attached BDI nothing appears.
Any hint?
Probably the BDI is configuring something correctly that your startup code is not.
Compare the register settings in the BDI configuration file to what is in the startup code.
Dave
DI BACCO ANTONIO - technolabs wrote:
Hi,
I downloaded an u-boot.bin (1.1.4) on to my flash (a board with MPC875). If I connect BDI and type go I see output on my serial console. If I start the board without the attached BDI nothing appears.
Any hint?
This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
U-Boot-Users mailing list U-Boot-Users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/u-boot-users
Maybe your BDI config file initializes something that U-boot doesn't. If you take everything out of the [INIT] section, do you get the same behavior?
regards, Ben
In message F1F6EC0C8B75034F9E3A79FC85122E8EA640E6@aquib01a you wrote:
I downloaded an u-boot.bin (1.1.4) on to my flash (a board with MPC875). If I connect BDI and type go I see output on my serial console. If I start the board without the attached BDI nothing appears.
Any hint?
Fix your board init code, i. e. check what the BDI init sequence is doing and what of this is missing in your initialization parts.
Maybe yous till include - against all our recommendations - include some setup of the memory controller in your BDI config file, and this is missing / broken in your code.
------_=_NextPart_001_01C7F3C2.C92E52CE Content-Type: text/html; charset="us-ascii"
And don't post HTML here, please.
Best regards,
Wolfgang Denk
Hi,
On 10 Sep 2007 at 17:53, DI BACCO ANTONIO - technolabs wrote:
Hi,
I downloaded an u-boot.bin (1.1.4)on to my flash (a board with MPC875). If I connect BDI and type go I see output on my serial console. If I start the board without the attached BDI nothing appears.
just a stupid idea: did you connect a pull-up resistor to BKPT on your board?
I heard rumours there are people who forgot this on their own MCF5373L board... ;-)
Regards, Wolfgang
just a stupid idea: did you connect a pull-up resistor to BKPT on your board?
My Board is an evaluation board and it worked till few months ago.
Now I removed any initialization from cgf file of BDI. Only special register 158 is set to 0x00000007 and I cannot remove this, otherwise BDI complains about halting the target.
Anyway the board continues to boot only using BDI not alone.
Now I don't understand what could be the problem. Should I set SPR 158 from within u-boot to 0x00000007?
Bye, Antonio.
Now I don't understand what could be the problem. Should I set SPR 158 from within u-boot to 0x00000007?
I saw that ICTRL is set in cpu/mpc8xx/start.S
Then I suppose BDI is doing other settings. But what?
This is my BDI cfg file (I removed all stuff except ICTRL):
; bdi configuration file for Adder/875 ; ---------------------------------------------- ; [INIT] ; init core register
WSPR 158 0x00000007 ;ICTRL: not serialized, no show cycles
[TARGET] BDIMODE AGENT ;the BDI working mode (LOADONLY | AGENT) BREAKMODE HARD ;SOFT or HARD, HARD uses PPC hardware breakpoints
[HOST] PROMPT adder875> IP 10.10.0.101 FILE u-boot.bin FORMAT BIN LOAD MANUAL
[FLASH] CHIPTYPE AM29BX16 ; Flash type (AM29F | AM29BX8 | AM29BX16 | I28BX8 | I28BX16) CHIPSIZE 0x800000 ; 8 Meg The size of one flash chip in bytes (e.g. AM29F010 = 0x20000) BUSWIDTH 16 ; The width of the flash memory bus in bits (8 | 16 | 32 ERASE 0xF8000000 ; FLASH sector #0 ERASE 0xF8010000 ; FLASH sector #1 ERASE 0xF8020000 ; FLASH sector #2 ERASE 0xF8030000 ; FLASH sector #3
FILE u-boot.bin FORMAT BIN
My Board is an evaluation board and it worked till few months ago.
Now I removed any initialization from cgf file of BDI. Only special register 158 is set to 0x00000007 and I cannot remove this, otherwise BDI complains about halting the target.
Anyway the board continues to boot only using BDI not alone.
Now I don't understand what could be the problem. Should I set SPR 158 from within u-boot to 0x00000007?
Is there any chance that the board is broken? I think I connected a wiggler to the BDM port in the wrong position for only a while. But why should it boot with attached BDI if is it broken?
Is there any chance that the board is broken? I think I connected a wiggler to the BDM port in the > wrong position for only a while. But why should it boot with attached BDI if is it broken?
I discovered that when the board is powered on, if I try to connect BDI connector trying to touch only pin 1 and 2 (vfls0 and sreset) the board starts. I suspect that bdi pulls up the sreset in some way.
Bye.
DI BACCO ANTONIO - technolabs wrote:
Is there any chance that the board is broken? I think I connected a
wiggler to the BDM port in the > wrong position for only a while. But why should it boot with attached BDI if is it broken?
I discovered that when the board is powered on, if I try to connect BDI connector trying to touch only pin 1 and 2 (vfls0 and sreset) the board starts. I suspect that bdi pulls up the sreset in some way.
The JTAG/COP header *should* have the appropriate pull-ups for HRESET#, and SRESET# *on the board*. These are open-drain signals.
However, there could be a mistake on your board, and perhaps the BDI2000 has weak pull-ups on the BDI 'just in case' the pull-ups are missing on the target.
Check your board schematic.
Dave
However, there could be a mistake on your board, and perhaps the BDI2000 has weak pull-ups on the BDI 'just in case' the pull-ups are missing on the target.
I verified that SRESET on the BDM connector is pulled up (on the schematic). I tried to pull up it with a wire but it doesn't work. It only works if I connect a 10 cm wire to SRESET pin on the BDM connector without connecting the other side of the wire (Probably it gets some electromagnetic noise). The board starts and remains working if I remove the wire, tehn, if I give a reset on the u-boot, the board hangs again.
Strange behaviour!! Bye, Antonio.
Hello!
DI BACCO ANTONIO - technolabs schrieb:
I verified that SRESET on the BDM connector is pulled up (on the schematic). I tried to pull up it with a wire but it doesn't work.
Have you used a voltage meter or oscilloscope to check the voltage level? Have you measured the resistance between SRESET and GND and between SRESET and supply? Have you checked if your processor pin is connected? Probably either the pull-up resistor isn't mounted or your PCB is defective.
With best regards Andreas Schweigstill
Have you used a voltage meter or oscilloscope to check the voltage level? Have you measured the resistance between SRESET and GND and between SRESET and supply? Have you checked if your processor pin is connected? Probably either the pull-up resistor isn't mounted or your PCB is defective.
Thank you for your suggestions but my board is an evaluation board named adder875. It worked and after connecting a wiggler with the connector upside down exibihts this strange behaviour. Anyway I will measure the voltage of the pins you suggested.
Bye, Antonio.
Hello!
DI BACCO ANTONIO - technolabs schrieb:
Thank you for your suggestions but my board is an evaluation board named adder875.
Many development boards do not have a quality which would be acceptable for industrial or even consumer products... Forgotten pull-up and series resistors or capacitors are quite common.
Keep in mind that many of those boards have been designed before the processor chip has been available. And marketing/sales departments force development to release such a board to customers long before it has been tested.
I have never seen a development board from a semiconductor manufacturer which doesn't contains any bugs.
With best regards Andreas Schweigstill
Andreas Schweigstill wrote:
Hello!
DI BACCO ANTONIO - technolabs schrieb:
Thank you for your suggestions but my board is an evaluation board named adder875.
Many development boards do not have a quality which would be acceptable for industrial or even consumer products... Forgotten pull-up and series resistors or capacitors are quite common.
Keep in mind that many of those boards have been designed before the processor chip has been available. And marketing/sales departments force development to release such a board to customers long before it has been tested.
I have never seen a development board from a semiconductor manufacturer which doesn't contains any bugs.
With best regards Andreas Schweigstill
You trimmed the critical sentence "It worked and ***after connecting a wiggler with the connector upside down*** exibihts this strange behaviour."
Sounds like his board was damaged by operator error (improperly connecting to the BDI connector). My guess is that the pullup resistor lost its smoke.
gvb
Hello!
Jerry Van Baren schrieb:
Sounds like his board was damaged by operator error (improperly connecting to the BDI connector). My guess is that the pullup resistor lost its smoke.
I agree that the board may have been damaged by the operator. But the pull-up resistor won't be damaged except from very high voltage.
Attaching a debug interface in the wrong direction probably may destroy it if it gets its power supply from the target. If it has its own supply (or gets it via USB from a PC) it can be possibly that either negative voltages (target GND connected to debugger VCC) or high voltages (target VCC connected to debugger GND) may be applied to the JTAG/BDM signals on the target. Such voltages may cause small defects to the silicon which show strange behaviour afterwards.
Regards Andreas Schweigstill
participants (7)
-
Andreas Schweigstill -
Ben Warren -
David Hawkins -
DI BACCO ANTONIO - technolabs -
Jerry Van Baren -
w.wegner＠astro-kom.de -
Wolfgang Denk