Jtag Connection to QorIQ p1021: [CCS last error: ELF is not in expected HALT mode ](CCSProtocolPlugin)

alfonsokame · ‎03-15-2013

Hello,

I'm having trouble to connect to my Bareboard based on the P1021 processor using CodeWarrior for PA Version: 10.1.2. I checked the JTAG pins and they seem to be OK. While executing the initialization script (p1021RDB_init_core.tcl), the debugger seems not be able to stop and halt the processor. It hangs while executing this script.

Any idea what is wrong here? The goal is to flash the bootloader to the SPI Flash.

Here is the Log

ccs_open

ipaddr = 127.0.0.1

port = 41475

timeout = 15

serverh = 0

ccs_open; ccs_error = 0

ccs_get_connection_count

serverh = 0

count = 1

ccs_get_connection_count; ccs_error = 0

ccs_available_connections

serverh = 0

count = 0

ccs_available_connections; ccs_error = 0

ccs_available_connections

serverh = 0

count = 0

ccs_available_connections; ccs_error = 0

ccs_config_cc

serverh = 0

config_string = utap:0

ccs_config_cc; ccs_error = 0

...

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 2032

count = 2

size = 4

value: (size = 8)

FFFFFFF8 00000000

ccs_read_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 8

count = 2

size = 4

value: (size = 8)

00000000 00000000

ccs_read_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 1

count = 1

size = 4

value: (size = 4)

00000000

ccs_read_register; ccs_error = 0; duration=1 ms

ccs_read_mem

coreh = [serverh:0;cc_index:0;chain_pos:1]

addr = [space:0;size:4;address:0xffffff00]

data: (size = 256)

09000000 00000000 01000000 03000000

00000000 03000000 00000000 00000000

D4120000 00000000 00000000 00000000

00000000 00000000 00000000 00000000

ccs_read_mem; ccs_error = -2147418106; duration=2 ms

Error message: cannot start SAP transaction

Executing Initialization File: C:\Users\InstAdmin\workspace1\fuziuzu-core0/CFG/P1021RDB_init_core.tcl

ccs_stop_core

coreh = [serverh:0;cc_index:0;chain_pos:1]

radix x

cmdwin::eclipse::config hexprefix 0x

cmdwin::eclipse::config MemIdentifier v

cmdwin::eclipse::config MemWidth 32

cmdwin::eclipse::config MemAccess 32

cmdwin::eclipse::config MemSwap off

cmdwin::reg e500 Special Purpose Registers/PIR %d -np

ccs_stop_core; ccs_error = 0

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 2032

count = 1

size = 4

value: (size = 4)

FFFFFFFC

ccs_read_register; ccs_error = 0; duration=1 ms

ccs_read_reg

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 63

count = 1

value: (size = 1)

0x00000000

ccs_read_reg; ccs_error = 0; duration=1 ms

ccs_read_reg

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 406

count = 1

value: (size = 1)

0x00000000

ccs_read_reg; ccs_error = 0; duration=0 ms

ccs_read_reg

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 415

count = 1

value: (size = 1)

0x00000000

ccs_read_reg; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 286

count = 1

size = 4

value: (size = 4)

00000000

ccs_read_register; ccs_error = 0; duration=1 ms

ccs_read_mem

coreh = [serverh:0;cc_index:0;chain_pos:1]

addr = [space:0;size:4;address:0xffffff00]

data: (size = 256)

09000000 00000000 01000000 03000000

00000000 03000000 00000000 00000000

DF120000 00000000 00000000 00000000

00000000 00000000 00000000 00000000

ccs_read_mem; ccs_error = -2147418106; duration=2 ms

Error message: cannot start SAP transaction

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 285212672

count = 1

size = 4

cmdwin::stop

value: (size = 4)

66029B02

ccs_read_register; ccs_error = -2147418106; duration=3 ms

Error message: cannot start SAP transaction

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 2032

count = 2

size = 4

value: (size = 8)

FFFFFFF8 00000000

ccs_read_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 1

count = 1

size = 4

value: (size = 4)

00000000

ccs_read_register; ccs_error = 0; duration=0 ms

thread break: Stopped, 0x0, 0x0, cpuPowerPCBig, Connected (state, tid, pid, cpu, target)

cmdwin::reg e500 Special Purpose Registers/IVPR = 0xFFFF0000

ccs_write_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 63

count = 1

size = 4

value: (size = 4)

FFFF0000

ccs_write_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 63

count = 1

size = 4

value: (size = 4)

FFFF0000

ccs_read_register; ccs_error = 0; duration=1 ms

cmdwin::reg e500 Special Purpose Registers/IVOR15 = 0x0000F000

ccs_write_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 415

count = 1

size = 4

value: (size = 4)

0000F000

ccs_write_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 415

count = 1

size = 4

value: (size = 4)

0000F000

ccs_read_register; ccs_error = 0; duration=1 ms

cmdwin::reg e500 Special Purpose Registers/IVOR6 = 0x0000F700

ccs_write_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 406

count = 1

size = 4

value: (size = 4)

0000F700

ccs_write_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 406

count = 1

size = 4

value: (size = 4)

0000F700

ccs_read_register; ccs_error = 0; duration=1 ms

cmdwin::reg e500 Special Purpose Registers/IAC1 = 0xfffffffc

ccs_write_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 312

count = 1

size = 4

value: (size = 4)

FFFFFFFC

ccs_write_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 312

count = 1

size = 4

value: (size = 4)

FFFFFFFC

ccs_read_register; ccs_error = 0; duration=0 ms

cmdwin::reg e500 Special Purpose Registers/DBCR0 = 0x40800000

ccs_write_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 308

count = 1

size = 4

value: (size = 4)

40800000

ccs_write_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 308

count = 1

size = 4

value: (size = 4)

40800000

ccs_read_register; ccs_error = 0; duration=1 ms

cmdwin::reg e500 Special Purpose Registers/DBCR1 = 0x00000000

ccs_write_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 309

count = 1

size = 4

value: (size = 4)

00000000

ccs_write_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 309

count = 1

size = 4

value: (size = 4)

00000000

ccs_read_register; ccs_error = 0; duration=1 ms

cmdwin::reg Standard Special Purpose Registers/MSR = 0x02000200

ccs_write_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 2034

count = 1

size = 4

value: (size = 4)

02000200

ccs_write_register; ccs_error = 0; duration=1 ms

ccs_read_register

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 2034

count = 1

size = 4

value: (size = 4)

02000200

ccs_read_register; ccs_error = 0; duration=0 ms

cmdwin::eclipse::config runcontrolsync off

cmdwin::go

ccs_read_reg

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 308

count = 1

value: (size = 1)

0x40800000

ccs_read_reg; ccs_error = 0; duration=1 ms

ccs_read_reg

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 2034

count = 1

value: (size = 1)

0x02000200

ccs_read_reg; ccs_error = 0; duration=1 ms

ccs_write_reg

coreh = [serverh:0;cc_index:0;chain_pos:1]

index = 308

count = 1

value: (size = 1)

0x41800000

ccs_write_reg; ccs_error = 0; duration=0 ms

ccs_out

coreh = [serverh:0;cc_index:0;chain_pos:1]

ccs_io addr: (size = 4)

0x00000000 0x00007531 0x00000000 0x00000000

value: (size = 8)

00000000 00000200

ccs_out; ccs_error = 0; duration=0 ms

ccs_run_core

coreh = [serverh:0;cc_index:0;chain_pos:1]

ccs_run_core; ccs_error = 0

cmdwin::wait 50

cmdwin::eclipse::config runcontrolsync on

cmdwin::stop

ccs_stop_core

coreh = [serverh:0;cc_index:0;chain_pos:1]

ccs_stop_core; ccs_error = -2147418101

Error message: ELF is not in expected HALT mode

Any idea??

Thanks a lot!

.

hwrobel · ‎03-18-2013

Hi,

The typical reset vector is 0xffff_fffc. So if you see an access to 0xffff_ff00, I would assume that the debugger tries to read in a page of your boot flash to be able to show you the instructions correctly.

Any NOR flash would not lead to bad behavior as a boot chip select will always complete the transaction due to the GPCM being used.

Do not only check the cfg_ signals, BTW. The EC doc describes several signals with footnotes which are not standard cfg_, but are still quite relevant for proper operation.

If you can’t figure it out at all, then you could turn CCS logging and start a tech support request on the freescale web with the debugger log, your init scripts and maybe even schematic excerpts.

Heinz

View solution in original post

hwrobel · ‎03-15-2013

Your description seems to indicate that this is about your board design and not our RDB.

Go through the EC document and check every pin with a footnote thoroughly with a scope during reset. I mean it.

Most common cause for such things are missing pullups/pulldowns on pins listed in the EC with a footnote, or external logic overdriving pullups/downs during reset, thereby messing up the reset state of the chip.

If you have a BSDL scan tool, then this might help to verify pin states during reset. Note that via BSDL you will only see externally driven pin state. Internal resistors on otherwise open pins may not be visible.

Also, don't use an RDB script "as is" on a custom board. Very likely that won't work as expected. If in doubt, start with a script with minimalistic content just to see if you can "Connect" and stop at a reset vector.

Hope this helps,

Heinz

alfonsokame · ‎03-18-2013

Hello Heinz,

thanks for your reply!

Im still struggling in order to get the p1021 based bareboard run!

All I checked all the reset configuration signals (cfg_..). They all seem to match their Reset state that is described in the reference manual. While trying to connect to the BareBoard, I notice that one step before the init_core.tcl is being executed, A memory or IO is trying to be read and there is an error:

ccs_read_mem

coreh = [serverh:0;cc_index:0;chain_pos:1]

addr = [space:0;size:4;address:0xffffff00]

data: (size = 256)

09000000 00000000 01000000 03000000

00000000 03000000 00000000 00000000

D4120000 00000000 00000000 00000000

00000000 00000000 00000000 00000000

ccs_read_mem; ccs_error = -2147418106; duration=2 ms :smileyminus:

Error message: cannot start SAP transaction :smileyminus:

Which Moduleis he trying to read on the address ffffff00? Is it Memory? How to configure it?

Many Thanks for your reply!!

hwrobel · ‎03-18-2013

Hi,

The typical reset vector is 0xffff_fffc. So if you see an access to 0xffff_ff00, I would assume that the debugger tries to read in a page of your boot flash to be able to show you the instructions correctly.

Any NOR flash would not lead to bad behavior as a boot chip select will always complete the transaction due to the GPCM being used.

Do not only check the cfg_ signals, BTW. The EC doc describes several signals with footnotes which are not standard cfg_, but are still quite relevant for proper operation.

If you can’t figure it out at all, then you could turn CCS logging and start a tech support request on the freescale web with the debugger log, your init scripts and maybe even schematic excerpts.

Heinz

alfonsokame · ‎04-03-2013

Thanks very much Heinz,

We found the problem here, It was due to 2 too weak pull-up resistances on the spi-clock and the MOSI. It had an influence on the JTAG connection

Thanks very much for your support once again

esode · ‎05-07-2013

Hello Alfonsokame,

We are using P2010 chip with CW and CW-TAP on a custom board, which has 2 more chips on the JTAG. We have a very similar problem to yours, like "Error message: cannot start SAP transaction". However, we are not using any SPI protocol on the board. I wonder whether you are using SPI then it brings out the problem you mentioned. I want to get a clue whether SPI hardware connections on the board are related to the problem.

I appreciate your comments.

Eser

Jtag Connection to QorIQ p1021: [CCS last error: ELF is not in expected HALT mode ](CCSProtocolPlugin)

Jtag Connection to QorIQ p1021: [CCS last error: ELF is not in expected HALT mode ](CCSProtocolPlugin)

General