Linux on custom Xilinx board with PPC405 hangs on boot

Linux on custom Xilinx board with PPC405 hangs on boot

[Peter N. Andreasen]

[-- Attachment #1: Type: text/plain, Size: 1570 bytes --]

I have an FPGA board which is based on the Xilinx ML300 board but without
disk and display. It has 4MB flash, and 32MB SDRAM, and uses the uartlite
serial port.
The RAM is located at 0x0000'0000 - 0x01FF'FFFF
I downloaded the kernel from Montavista, built a board support package in
Xilinx Platform Studio, and managed to configure and build the kernel.
However when I transfer the resulting zImage.elf file to the board, the
attached terminal only shows:

loaded at:     00040000 00D4140E
board data at: 00D41183 00D41105
relocated to:  0004258F 00043501
zimage at:     00048550 0074E751
initrd at:     00740800 00D420FF
avail ram:     00D40500 20000000

Linux/PPC load: root=/dev/ram rw console=/dev/ttyS0
Uncompressing Linux...done.
Now booting the kernel

and then it dies.

In the debug window I have this:

XMD% dow c:/Hardi/zImage.initrd.elf
        section, .text: 0x00400000-0x004047b8
        section, .data: 0x00405000-0x004d1000
        section, .bss: 0x004d1000-0x004d41e0

Downloading ELF File c:/Hardi/zImage.initrd.elf
Program header record #0, Size = 0xD1000.
Downloading program section : Start = 0x00400000, length = 0x000d1000
Downloaded Program c:/Hardi/zImage.initrd.elf
Setting PC with program start addr = 0x00400000
PC reset to 0x00400000, Clearing MSR Register
XMD% con
Processor started. Type "stop" to stop processor
RUNNING> stop
XMD%
Processor stopped at PC: 0x6fbf0700

XMD% dis 0x6fbf0700
6FBF0700:   00000000
XMD%


so it looks like the processor somewhere is sent out in space -
Does anyone have an idea how to find the problem?

Peter

[-- Attachment #2: Type: text/html, Size: 1934 bytes --]

Re: Linux on custom Xilinx board with PPC405 hangs on boot

[Andrei Konovalov]

Peter N. Andreasen wrote:
> I have an FPGA board which is based on the Xilinx ML300 board but 
> without disk and display. It has 4MB flash, and 32MB SDRAM, and uses the 
> uartlite serial port.
> The RAM is located at 0x0000'0000 - 0x01FF'FFFF
> I downloaded the kernel from Montavista, built a board support package 
> in Xilinx Platform Studio, and managed to configure and build the kernel.
> However when I transfer the resulting zImage.elf file to the board, the 
> attached terminal only shows:
> 
> loaded at:     00040000 00D4140E
> board data at: 00D41183 00D41105
> relocated to:  0004258F 00043501
> zimage at:     00048550 0074E751
> initrd at:     00740800 00D420FF
> avail ram:     00D40500 20000000
> 
> Linux/PPC load: root=/dev/ram rw console=/dev/ttyS0
...

"console=ttyS0" (not "/dev/ttyS0" BTW) is for 16x50 compatible UART.
UartLite needs something like "console=ttl0".

Re: Linux on custom Xilinx board with PPC405 hangs on boot

[Peter N. Andreasen]

[-- Attachment #1: Type: text/plain, Size: 687 bytes --]

Thank you for the advice.
Now I have been able to insert instructions to display a number on an
attached display which means that I can now see that the code runs until the
file head_4xx.S in the line where the comment is "Now turn on the MMU for
real" and then calls start_kernel by way of an rfi instruction. The
processor then ends up in the Instruction storage exception where I have
placed another write to the display.
Do you have any idea how to find out what happened here?
cheers
Peter


On 9/12/06, Andrei Konovalov <akonovalov@ru.mvista.com> wrote:
>
>
> "console=ttyS0" (not "/dev/ttyS0" BTW) is for 16x50 compatible UART.
> UartLite needs something like "console=ttl0".
>
>

[-- Attachment #2: Type: text/html, Size: 1027 bytes --]

Re: Linux on custom Xilinx board with PPC405 hangs on boot

[Peter N. Andreasen]

[-- Attachment #1: Type: text/plain, Size: 3822 bytes --]

I have a custom Xilinx FPGA board which is similar to ML300 but uses
Uartllite and does not have disk or display.
When I start the Linux kernel I end up with an exception during the probe
for Flash - I think.
I have added a bunch of printk lines in order to see what is happening but I
am a little stuck at the moment.
The Flash chip is an AM29LV320D 4Mbyte in 16-bit access mode, and it sits on
address 0xFF80'0000
This is what I get on my terminal. I would be happy to hear your advice.
Peter



loaded at:     00400000 004CF1E0
board data at: 004CC138 004CC150
relocated to:  004052F4 0040530C
zimage at:     00405801 00472CFB
initrd at:     00473000 004CB2FF
avail ram:     004D0000 02000000

Linux/PPC load: root=/dev/ram rw console=ttl0
Uncompressing Linux...done.
Now booting the kernel

id mach(): done
MMU:enter
MMU:hw init
MMU:mapin
MMU:mapin_ram done
MMU:setio
MMU:exit

Linux version 2.4.26 (pna@glazedevel) (gcc version 3.4.1) #50 Wed Sep 20
11:06:12 CEST 2006
setup_arch: enter
setup_arch: bootmem
Xilinx Virtex-II Pro port (C) 2002 MontaVista Software, Inc. (
source@mvista.com)
arch: exit
On node 0 totalpages: 8192
zone(0): 8192 pages.
zone(1): 0 pages.
zone(2): 0 pages.
Kernel command line: root=/dev/ram rw console=ttl0
(start) init/main.c: parse_options
(end) parse_options
(start) arch/ppc/mm/pgtable.c: __ioremap
(end) arch/ppc/mm/pgtable.c: __ioremap
Xilinx INTC #0 at 0x81200000 mapped to 0xFDFFF000
(start) arch/ppc/mm/pgtable.c: __ioremap
(end) arch/ppc/mm/pgtable.c: __ioremap
Calibrating delay loop... 287.53 BogoMIPS
Memory: 30924k available (776k kernel code, 276k data, 44k init, 0k highmem)
Dentry cache hash table entries: 4096 (order: 3, 32768 bytes)
Inode cache hash table entries: 2048 (order: 2, 16384 bytes)
Mount cache hash table entries: 512 (order: 0, 4096 bytes)
Buffer cache hash table entries: 1024 (order: 0, 4096 bytes)
Page-cache hash table entries: 8192 (order: 3, 32768 bytes)
POSIX conformance testing by UNIFIX
(start) init/main.c: init
(start) init/main.c: do_basic_setup

Linux NET4.0 for Linux 2.4
Based upon Swansea University Computer Society NET3.039
Initializing RT netlink socket
(start) init/main.c: do_initcalls
Starting kswapd
devfs: v1.12c (20020818) Richard Gooch (rgooch@atnf.csiro.au)
devfs: boot_options: 0x1
JFFS version 1.0, (C) 1999, 2000  Axis Communications AB
JFFS2 version 2.1. (C) 2001 Red Hat, Inc., designed by Axis Communications
AB.
Generic RTC Driver v1.07
RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize
loop: loaded (max 8 devices)
(start) drivers/mtd/maps/physmap.c: init_physmap
physmap flash device: 4000000 at ff800000
(start) arch/ppc/mm/pgtable.c: __ioremap
(end) arch/ppc/mm/pgtable.c: __ioremap
(start) drivers/mtd/chips/chipreg.c: do_map_probe
(start) drivers/mtd/chips/chipreg.c: get_mtd_chip_driver
(end) drivers/mtd/chips/chipreg.c: get_mtd_chip_driver
(after get_mtd_chip_driver) drivers/mtd/chips/chipreg.c: do_map_probe
(start) drivers/mtd/chips/gen_probe.c: mtd_do_chip_probe
(start) drivers/mtd/chips/gen_probe.c: genprobe_ident_chips
Instruction machine check in kernel mode.
Oops: machine check, sig: 7
NIP: C00A2960 XER: 40000000 LR: C009CBD8 SP: C04D9D90 REGS: c04d9ce0 TRAP:
0200    Not tainted
MSR: 00009030 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
TASK = c04d8000[1] 'swapper' Last syscall: 120
last math 00000000 last altivec 00000000
GPR00: C00A2958 C04D9D90 C04D8000 C00DC1C8 000000F0 00400000 C04D9DD0
C00DB387
GPR08: 00000002 C307D000 C0176360 00000001 C00F0000 000057E0 80001040
80000820
GPR16: 00001020 00000840 10021000 08040800 C00D0000 C00EAC18 C00D0000
00000000
GPR24: C00C0000 00000001 C04D9DD0 C04D9E10 C04D9E10 00400000 C04D9DD0
C00DC1C8
Call backtrace:
C009CE80 C00A26D0 C009D3A0 C009C750 C00E9BB0 C00024F4 C0005ED4
Kernel panic: Attempted to kill init!
 <0>Rebooting in 180 seconds..

[-- Attachment #2: Type: text/html, Size: 4398 bytes --]

Re: Linux on custom Xilinx board with PPC405 hangs on boot

[Linas Vepstas]

On Wed, Sep 20, 2006 at 11:16:53AM +0200, Peter N. Andreasen wrote:
> I have a custom Xilinx FPGA board which is similar to ML300 but uses
> Uartllite and does not have disk or display.
> When I start the Linux kernel I end up with an exception during the probe
> for Flash - I think.

Well, its not just "an exception" its a machine check.

> (after get_mtd_chip_driver) drivers/mtd/chips/chipreg.c: do_map_probe
> (start) drivers/mtd/chips/gen_probe.c: mtd_do_chip_probe
> (start) drivers/mtd/chips/gen_probe.c: genprobe_ident_chips
> Instruction machine check in kernel mode.
> Oops: machine check, sig: 7
> NIP: C00A2960 XER: 40000000 LR: C009CBD8 SP: C04D9D90 REGS: c04d9ce0 TRAP:
> 0200    Not tainted
> MSR: 00009030 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11

Machine checks happen when some hunk of hardware is wired to the 
machine-check pin of the cpu chip, and that bit of hardware decides to
raise the wire.  I'd say the first step is to figure ou what hardware
is wired up this way, and what would make it unhappy enough to assert a
machine check.

SRR1 has bits that state what caused he machine check. -- e.g partity
error on data or address bus, "transfer error", or MC signal.

--linas

RE: Linux on custom Xilinx board with PPC405 hangs on boot

[Liu Dave-r63238]

[-- Attachment #1: Type: text/plain, Size: 4316 bytes --]


 

	I have a custom Xilinx FPGA board which is similar to ML300 but
uses Uartllite and does not have disk or display.
	When I start the Linux kernel I end up with an exception during
the probe for Flash - I think. 
	I have added a bunch of printk lines in order to see what is
happening but I am a little stuck at the moment. 
	The Flash chip is an AM29LV320D 4Mbyte in 16-bit access mode,
and it sits on address 0xFF80'0000
	This is what I get on my terminal. I would be happy to hear your
advice.
	Peter
	
	
	
	loaded at:     00400000 004CF1E0 
	board data at: 004CC138 004CC150
	relocated to:  004052F4 0040530C
	zimage at:     00405801 00472CFB
	initrd at:     00473000 004CB2FF
	avail ram:     004D0000 02000000
	
	Linux/PPC load: root=/dev/ram rw console=ttl0 
	Uncompressing Linux...done.
	Now booting the kernel
	
	id mach(): done
	MMU:enter
	MMU:hw init
	MMU:mapin
	MMU:mapin_ram done
	MMU:setio
	MMU:exit
	
	Linux version 2.4.26 (pna@glazedevel) (gcc version 3.4.1) #50
Wed Sep 20 11:06:12 CEST 2006
	setup_arch: enter
	setup_arch: bootmem
	Xilinx Virtex-II Pro port (C) 2002 MontaVista Software, Inc.
(source@mvista.com)
	arch: exit 
	On node 0 totalpages: 8192
	zone(0): 8192 pages.
	zone(1): 0 pages.
	zone(2): 0 pages.
	Kernel command line: root=/dev/ram rw console=ttl0
	(start) init/main.c: parse_options
	(end) parse_options
	(start) arch/ppc/mm/pgtable.c: __ioremap 
	(end) arch/ppc/mm/pgtable.c: __ioremap
	Xilinx INTC #0 at 0x81200000 mapped to 0xFDFFF000
	(start) arch/ppc/mm/pgtable.c: __ioremap
	(end) arch/ppc/mm/pgtable.c: __ioremap
	Calibrating delay loop... 287.53 BogoMIPS 
	Memory: 30924k available (776k kernel code, 276k data, 44k init,
0k highmem)
	Dentry cache hash table entries: 4096 (order: 3, 32768 bytes)
	Inode cache hash table entries: 2048 (order: 2, 16384 bytes)
	Mount cache hash table entries: 512 (order: 0, 4096 bytes) 
	Buffer cache hash table entries: 1024 (order: 0, 4096 bytes)
	Page-cache hash table entries: 8192 (order: 3, 32768 bytes)
	POSIX conformance testing by UNIFIX
	(start) init/main.c: init
	(start) init/main.c: do_basic_setup 
	             
	Linux NET4.0 for Linux 2.4
	Based upon Swansea University Computer Society NET3.039
	Initializing RT netlink socket
	(start) init/main.c: do_initcalls
	Starting kswapd
	devfs: v1.12c (20020818) Richard Gooch ( rgooch@atnf.csiro.au)
	devfs: boot_options: 0x1
	JFFS version 1.0, (C) 1999, 2000  Axis Communications AB
	JFFS2 version 2.1. (C) 2001 Red Hat, Inc., designed by Axis
Communications AB. 
	Generic RTC Driver v1.07
	RAMDISK driver initialized: 16 RAM disks of 4096K size 1024
blocksize
	loop: loaded (max 8 devices)
	(start) drivers/mtd/maps/physmap.c: init_physmap
	physmap flash device: 4000000 at ff800000 
	(start) arch/ppc/mm/pgtable.c: __ioremap
	(end) arch/ppc/mm/pgtable.c: __ioremap
	(start) drivers/mtd/chips/chipreg.c: do_map_probe
	(start) drivers/mtd/chips/chipreg.c: get_mtd_chip_driver
	(end) drivers/mtd/chips/chipreg.c: get_mtd_chip_driver 
	(after get_mtd_chip_driver) drivers/mtd/chips/chipreg.c:
do_map_probe
	(start) drivers/mtd/chips/gen_probe.c: mtd_do_chip_probe
	(start) drivers/mtd/chips/gen_probe.c: genprobe_ident_chips
	Instruction machine check in kernel mode.  
	^^^^^^^^^
	You should make sure if it is instruction or data machinecheck.
	The kernel you using is very old.
	 
	Oops: machine check, sig: 7
	NIP: C00A2960 XER: 40000000 LR: C009CBD8 SP: C04D9D90 REGS:
c04d9ce0 TRAP: 0200    Not tainted 
	
^^^^^^^^^^^
	It has really machinecheck.
	 
	MSR: 00009030 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
	TASK = c04d8000[1] 'swapper' Last syscall: 120 
	last math 00000000 last altivec 00000000
	GPR00: C00A2958 C04D9D90 C04D8000 C00DC1C8 000000F0 00400000
C04D9DD0 C00DB387 
	GPR08: 00000002 C307D000 C0176360 00000001 C00F0000 000057E0
80001040 80000820 
	GPR16: 00001020 00000840 10021000 08040800 C00D0000 C00EAC18
C00D0000 00000000 
	GPR24: C00C0000 00000001 C04D9DD0 C04D9E10 C04D9E10 00400000
C04D9DD0 C00DC1C8 
	Call backtrace: 
	C009CE80 C00A26D0 C009D3A0 C009C750 C00E9BB0 C00024F4 C0005ED4  
	 
	You need dump more register to find the cause. eg. ESR. MCSR...
	 
	Kernel panic: Attempted to kill init!
	 <0>Rebooting in 180 seconds.. 
	
	


[-- Attachment #2: Type: text/html, Size: 7040 bytes --]

RE: Linux on custom Xilinx board with PPC405 hangs on boot

[Liu Dave-r63238]

<snip>
> > Instruction machine check in kernel mode.
> > Oops: machine check, sig: 7
> > NIP: C00A2960 XER: 40000000 LR: C009CBD8 SP: C04D9D90 REGS:=20
> c04d9ce0 TRAP:
> > 0200    Not tainted
> > MSR: 00009030 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
>=20
> Machine checks happen when some hunk of hardware is wired to=20
> the machine-check pin of the cpu chip, and that bit of=20
> hardware decides to raise the wire.  I'd say the first step=20
> is to figure ou what hardware is wired up this way, and what=20
> would make it unhappy enough to assert a machine check.
>=20
> SRR1 has bits that state what caused he machine check. -- e.g=20
> partity error on data or address bus, "transfer error", or MC signal.

The PPC405 has different registers for machine check. Unlike classic
powerpc.

-Dave

Re: Linux on custom Xilinx board with PPC405 hangs on boot

[Peter N. Andreasen]

[-- Attachment #1: Type: text/plain, Size: 1435 bytes --]

Liu, Linas, list
Thanks a lot for the answers. It helped me to review the hardware
configuration again.
The problem was the definition of the Flash size.
It turns out that the Xilinx generated header files define the size of the
Flash. But the kernel also has a configuration option: Memory Technology
Devices / Mapping Drivers for Chip Access / Physical Length of Chip Mapping
My arch/ppc/platforms/xilinx_ocp/xparameters_ml300.h defines this as 4Mbyte,
but the option in the kernel was set to 64 Mbyte (an extra 0 which I had not
seen)
thanks again
Peter




On 9/21/06, Liu Dave-r63238 <DaveLiu@freescale.com> wrote:
>
> <snip>
> > > Instruction machine check in kernel mode.
> > > Oops: machine check, sig: 7
> > > NIP: C00A2960 XER: 40000000 LR: C009CBD8 SP: C04D9D90 REGS:
> > c04d9ce0 TRAP:
> > > 0200    Not tainted
> > > MSR: 00009030 EE: 1 PR: 0 FP: 0 ME: 1 IR/DR: 11
> >
> > Machine checks happen when some hunk of hardware is wired to
> > the machine-check pin of the cpu chip, and that bit of
> > hardware decides to raise the wire.  I'd say the first step
> > is to figure ou what hardware is wired up this way, and what
> > would make it unhappy enough to assert a machine check.
> >
> > SRR1 has bits that state what caused he machine check. -- e.g
> > partity error on data or address bus, "transfer error", or MC signal.
>
> The PPC405 has different registers for machine check. Unlike classic
> powerpc.
>
> -Dave
>

[-- Attachment #2: Type: text/html, Size: 1870 bytes --]

Re: Linux on custom Xilinx board with PPC405 hangs on boot

[Peter N. Andreasen]

On 9/21/06, Peter N. Andreasen <peterarbejde@gmail.com> wrote:
> Liu, Linas, list
> Thanks a lot for the answers. It helped me to review the hardware configuration again.
> The problem was the definition of the Flash size.
> It turns out that the Xilinx generated header files define the size of the Flash. But the kernel also has a configuration option: Memory Technology Devices / Mapping Drivers for Chip Access / Physical Length of Chip Mapping
> My arch/ppc/platforms/xilinx_ocp/xparameters_ml300.h defines this as 4Mbyte, but the option in the kernel was set to 64 Mbyte (an extra 0 which I had not seen)

Hello list,
I have successfully booted the kernel, but the uartlite device seems
to give me a lot of headache. I am using Busybox v. 1.0 and made some
changes to the libb.h:
# define CURRENT_VC "/dev/ttl0"
# define VC_1 "/dev/ttl1"
# define VC_2 "/dev/ttl2"
# define VC_3 "/dev/ttl3"
# define VC_4 "/dev/ttl4"
# define VC_5 "/dev/ttl5"

#  define SC_0 "/dev/ttl0"
#  define SC_1 "/dev/ttl1"
#  define SC_FORMAT "/dev/ttl%d"
# define VC_FORMAT "/dev/ttl%d"

To make sure I get the correct device node I also changed this in
xuartllite_serial.h:
#define XULITE_MINOR_START	187
#define UARTLITE_TTY_NAME	"ttl"
#define UARTLITE_TTY_DEVFS_NAME	"ttl/%d"
#define UARTLITE_CU_NAME	"cul"
#define UARTLITE_CU_DEVFS_NAME	"cul/%d"
#define XULITE_MAJOR            204
#define XULITE_AUX_MAJOR        205

The system stops when init launches the busybox init process. This is
the message I get just after "Freeing unused kernel memory: 44k init"

First I got an assert() from the Xilinx Uartlite driver:
...
init started:  Bu
<2>Xilinx OS Independent Code XAssert:  xuartlite.c:194
Code may crash due to unhandled errors.
...
and the system stops.
The check made in the driver is for a 0 length string. If I change it
to ignore that, the console show this instead:
init started:  Bu
init started:  B
trying to run init_process
...
(the "trying to run init process" is a message I put in
run_init_process in init.c in the kernel, and the message it tries to
output is "init started: Busybox version 1.0 .....")

Now my question is:
What exactly are the device nodes I have to make in /dev/ on the
ramdisk on target? I made a ttl, ttl0, and tty  which have major
204/minor 187, and ttl1, ttl2 minor 188 and 189. I also
Hope someone can help me on this

Peter