[U-Boot-Users] Booting From hard disk.

[U-Boot-Users] Booting From hard disk.

[Sharad P. Katti]

Hi,

We are trying to boot linux from the harddisk, using uboot.
We need to know, how to access the kernel from the harddisk.
i.e, the address where the kernel will be residing in the harddisk,


We have copied the uImage to the target harddisk, by creating
a file system, to the /boot dir. (by connecting to another system)


We are able to get the harddisk info from the uboot, and the 
partition info, but we are not able to get the correct address,
to be given to diskboot.

Can someone help up in getting the correct address, 
or where to put the kernel in the harddisk.

TIA,

best regards,


-- 
Sharad P.Katti

"GOD is real, unless declared an int"

[U-Boot-Users] Booting From hard disk.

[Wolfgang Grandegger]

On 12/11/2004 05:04 PM Sharad P. Katti wrote:
> Hi,
> 
> We are trying to boot linux from the harddisk, using uboot.
> We need to know, how to access the kernel from the harddisk.
> i.e, the address where the kernel will be residing in the harddisk,
> 
> 
> We have copied the uImage to the target harddisk, by creating
> a file system, to the /boot dir. (by connecting to another system)
> 
> 
> We are able to get the harddisk info from the uboot, and the 
> partition info, but we are not able to get the correct address,
> to be given to diskboot.
> 
> Can someone help up in getting the correct address, 
> or where to put the kernel in the harddisk.

http://www.denx.de/twiki/bin/view/DULG/UsingPCCard may help.

Wolfgang.

[U-Boot-Users] Booting From hard disk.

[Sharad P. Katti]

Hi,

Thanks for your suggestions,

We are able to write the image in the raw partition of harddisk,
and we are able to see the u-boot header contents once we boot from the image.

But, it is restarting the target board after coming till "Starting
Kernel" & also we put some debug statements & it is restarting just
before  "theKernel" ( in armlinux.c)

We  are able to get the header information.. We are in doubt about what 
address we need to give for load address & entry point.( we have given
0000000 for both).

Whether the control is in the kernel or still in the Bootloader ?

regards,


-Sharad P.Katti
=============================
"GOD is real, unless declared as an int"

[U-Boot-Users] Booting From hard disk.

[Wolfgang Denk]

In message <eae046ff041212015556cbc541@mail.gmail.com> you wrote:
> 
> We are able to write the image in the raw partition of harddisk,
> and we are able to see the u-boot header contents once we boot from the image.
> 
> But, it is restarting the target board after coming till "Starting
> Kernel" & also we put some debug statements & it is restarting just
> before  "theKernel" ( in armlinux.c)

Are you sure that your image is working at all? Does it boot when you
download it over TFTP and boot then?

> We  are able to get the header information.. We are in doubt about what 
> address we need to give for load address & entry point.( we have given
> 0000000 for both).

You should know the memory map of your board better that we.  In  any
case it seems likely to me that there is flash or ROM at address 0 on
your  board,  and  not  RAM,  so  using  0x0000 as load & entry point
address is obviously plain wrong.

> Whether the control is in the kernel or still in the Bootloader ?

Neither one nor the  other.  U-Boot  is  _attempting_  to  start  the
kernel, which is not there where you pretend it is.

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Conceptual integrity in turn dictates that the  design  must  proceed
from  one  mind,  or  from  a  very small number of agreeing resonant
minds.               - Frederick Brooks Jr., "The Mythical Man Month"

[U-Boot-Users] Booting From hard disk.

[Sharad P. Katti]

Hi,


> Are you sure that your image is working at all? Does it boot when you
> download it over TFTP and boot then?

Since there is no net option in our board, we cant check this.


> You should know the memory map of your board better that we.  In  any
> case it seems likely to me that there is flash or ROM at address 0 on
> your  board,  and  not  RAM,  so  using  0x0000 as load & entry point
> address is obviously plain wrong.

in our board, the ram starts from 0xa0000000, and is of 32 MB.
while creating the uImage, the load address and entry point as 0xa0008000,
and we are loading the image from IDE device to the RAM as,

$diskboot 0xa0001000 0:1

and then,
 $bootm 0xa0001000 to boot from that location.

We are able to get the uImage header information properly,
but after the "Starting kernel", it is restarting, with "data abort"

> Neither one nor the  other.  U-Boot  is  _attempting_  to  start  the
> kernel, which is not there where you pretend it is.
 
We have checked the contents of RAM, using md, it is properly copied
at 0xa0001000.
then, after restarting, we cheked the contents at 0xa0008000, at the
both the locations,
the kernel is proper.

best regards,
-- 
Sharad P.Katti

"GOD is real, unless declared as an int"

[U-Boot-Users] Booting From hard disk.

[Wolfgang Denk]

In message <eae046ff041213082036dfdb35@mail.gmail.com> you wrote:
> 
> > Are you sure that your image is working at all? Does it boot when you
> > download it over TFTP and boot then?
> 
> Since there is no net option in our board, we cant check this.

Then load the kernel image using the BDi2000, or  over  serial  line.
TMTOWTDI.

> in our board, the ram starts from 0xa0000000, and is of 32 MB.
> while creating the uImage, the load address and entry point as 0xa0008000,
> and we are loading the image from IDE device to the RAM as,
> 
> $diskboot 0xa0001000 0:1

Start thinking.

How big is your kernel image?  Let's  assume  a  small  configuration
which fits into 500 kB (= 0x7d000).

You load it at address 0xa0001000, so it will take up the  area  from
0xa0001000 ... 0xa007e000.

Your load address is 0xa0008000 which is right in the middle  of  the
area where you stored the kernel image.

What do you think will happen when U-Boot  starts  uncompressing  the
kernel image and write it to the area starting at 0xa0008000?

Do you really expect anything elase  but  a  crash  because  you  are
overwriting the not-yet-uncompressed tail part of the image?

> We are able to get the uImage header information properly,
> but after the "Starting kernel", it is restarting, with "data abort"

Sounds as good a crash as any other to me...

> We have checked the contents of RAM, using md, it is properly copied
> at 0xa0001000.

Of course. U-Boot is working perfectly fine. It does exactly what you
tell it.

> then, after restarting, we cheked the contents at 0xa0008000, at the
> both the locations,
> the kernel is proper.

Don't try to outweird me.  I don't believe you.


Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
An age is called Dark not because  the  light  fails  to  shine,  but
because people refuse to see it.           -- James Michener, "Space"

[U-Boot-Users] Booting From hard disk.

[Sharad P. Katti]

hi,

thanks for your mail,

actually we did wrong calculations as u pointed to us, 

now, we are loading the kernel image at 0xa0100000, from harddisk.
which doesnot clash with the the address where the kernel will be loaded,
and executed.i,e at 0xa0008000.

But again it is restarting, giving the "data abort" error. We cross checked
with the contents of the RAM at 0xa0100000 through 0xa0157834, and
oxa0008000 through 0xa005f834 (the kernel size is 356 k), both are same.

What might be the problem to load the kernel..? is it related to the
kernel part.?

TIA,

best regards,

-- 
Sharad P.Katti

"GOD is real, unless declared as an int"

[U-Boot-Users] Re: One more step to conclude the u-boot port to pxa255 board

[Seido Nakanishi]

Hello,

We are just finnishing the port of u-boot to our customer (Dixtal) 
pxa255 board.

Everything is working fine, including support to ethernet interface.

We are using bdiGDB (BDI2000) tool for writing the u-boot binary and 
debugging.

Now that everything is working with BDI plugged and u-boot transferred 
to Flash Memory 0x00000000 address. With BDI we issue "reset" and 
"go"and u-boot comes up and boots the kernel(one more "go") and linux 
root file system from flash.

When we unplugged the BDI2000 and powered-up the board u-boot did not 
come up.

We found out that with BDI2000, after reset, the PC points to  
0x00000050. If we issue go 0x00000050 the u-boot works fine and if we 
issue go 0x00000000 (as per power-up sequence) it does not work. It 
seems that the u-boot is not prepared to be booted (0x00000000 entry 
point) at power-up.

We checked the u-boot configuration file (attached below) as well as 
u-boot.lds, as far as we understand u-boot should support 0x00000000 
entry point.

Are we missing anything ? Any suggestions to solve this problem ?

Thank You,
Seido Nakanishi


==================U-boot configuration file =================
#ifndef __CONFIG_H
#define __CONFIG_H

/*
 * If we are developing, we might want to start armboot from ram
 * so we MUST NOT initialize critical regs like mem-timing ...
 */
#define CONFIG_INIT_CRITICAL        /* undef for developing */
#define DEBUG

/*
 * High Level Configuration Options
 * (easy to change)
 */
#define CONFIG_PXA250        1    /* This is an PXA250 CPU    */
#define CONFIG_LUBBOCK        1    /* on an LUBBOCK Board        */
/*#define CONFIG_LCD        1
#define CONFIG_MMC        1 */
#undef CONFIG_MMC
#define BOARD_LATE_INIT        1
#define RTC            1   

#undef CONFIG_USE_IRQ            /* we don't need IRQ/FIQ stuff */

/*
 * Size of malloc() pool
 */
#define CFG_MALLOC_LEN        (CFG_ENV_SIZE + 128*1024)
#define CFG_GBL_DATA_SIZE    128    /* size in bytes reserved for 
initial data */

/*
 * SMSC91C111 Network Card
 */
#define CONFIG_DRIVER_SMC91111        1
#define CONFIG_SMC91111_BASE        0x08000300 /* chip select 2         */
#undef  CONFIG_SMC_USE_32_BIT                   /* 16 bit bus access     */
#undef  CONFIG_SMC_91111_EXT_PHY           /* we use internal phy   */
#undef  CONFIG_SHOW_ACTIVITY
#define CONFIG_NET_RETRY_COUNT        10       /* # of retries          */

/*
 * select serial console configuration
 */
#define CONFIG_FFUART           1       /* we use FFUART on LUBBOCK */
#define CONFIG_ENV_OVERWRITE
#define CONFIG_BAUDRATE        115200

/*#define CONFIG_COMMANDS        (CONFIG_CMD_DFL | CFG_CMD_MMC | 
CFG_CMD_FAT) */
#define CONFIG_COMMANDS        
(CFG_CMD_BDI|CFG_CMD_LCDDRAW|CFG_CMD_LCDTST|CFG_CMD_LOADS|CFG_CMD_LOADB|CFG_CMD_IMI|CFG_CMD_CACHE|CFG_CMD_FLASH|CFG_CMD_MEMORY|CFG_CMD_NET|CFG_CMD_ENV|CFG_CMD_BOOTD|CFG_CMD_CONSOLE|CFG_CMD_ASKENV|CFG_CMD_RUN|CFG_CMD_ECHO|CFG_CMD_REGINFO|CFG_CMD_IMMAP|CFG_CMD_DHCP|CFG_CMD_AUTOSCRIPT|CFG_CMD_ELF|CFG_CMD_MISC|CFG_CMD_SDRAM|CFG_CMD_DIAG|CFG_CMD_SAVES|CFG_CMD_PING|CFG_CMD_FAT|CFG_CMD_IMLS)

/* this must be included AFTER the definition of CONFIG_COMMANDS (if any) */
#include <cmd_confdefs.h>

/*
 * Configure Network Parameters
 */
#define CONFIG_BOOTDELAY    4
#define CONFIG_ETHADDR        00:d0:c9:60:87:aa
#define CONFIG_NETMASK        255.255.0.0
#define CONFIG_IPADDR        10.0.1.222
#define CONFIG_SERVERIP        10.0.1.200


/*
 * Configure Boot Parameters
 */
#define CONFIG_BOOTCOMMAND    "bootm 0xc00000" /* kernell address at 
Flash */
#define CONFIG_BOOTARGS        "root=/dev/mtdblock1 console=ttyS0,115200 
mem=32M rootfstype=jffs2"
#define CONFIG_CMDLINE_TAG

#if (CONFIG_COMMANDS & CFG_CMD_KGDB)
#define CONFIG_KGDB_BAUDRATE    230400        /* speed to run kgdb 
serial port */
#define CONFIG_KGDB_SER_INDEX    2        /* which serial port to use */
#endif

/*
 * Miscellaneous configurable options
 */
#define CFG_HUSH_PARSER        1
#define CFG_PROMPT_HUSH_PS2    "> "

#define CFG_LONGHELP                /* undef to save memory        */
#ifdef CFG_HUSH_PARSER
#define CFG_PROMPT        "uboot> "        /* Monitor Command Prompt */
#else
#define CFG_PROMPT        "uboot$ "        /* Monitor Command Prompt */
#endif
#define CFG_CBSIZE        256        /* Console I/O Buffer Size    */
#define CFG_PBSIZE (CFG_CBSIZE+sizeof(CFG_PROMPT)+16) /* Print Buffer 
Size */
#define CFG_MAXARGS        16        /* max number of command args    */
#define CFG_BARGSIZE        CFG_CBSIZE    /* Boot Argument Buffer Size    */
#define CFG_DEVICE_NULLDEV    1

#define CFG_MEMTEST_START    0xa0400000    /* memtest works on    */
#define CFG_MEMTEST_END        0xa0800000    /* 4 ... 8 MB in DRAM    */

#undef    CFG_CLKS_IN_HZ        /* everything, incl board info, in Hz */

#define CFG_LOAD_ADDR        0xa2000000    /* default kernel load address */

#define CFG_HZ            3686400        /* incrementer freq: 3.6864 MHz */
#define CFG_CPUSPEED        0x161        /* set core clock to 
400/200/100 MHz */

                        /* valid baudrates */
#define CFG_BAUDRATE_TABLE    { 9600, 19200, 38400, 57600, 115200 }

#define CFG_MMC_BASE        0xF0000000

/*
 * Stack sizes
 *
 * The stack sizes are set up in start.S using the settings below
 */
#define CONFIG_STACKSIZE    (128*1024)    /* regular stack */
#ifdef CONFIG_USE_IRQ
#define CONFIG_STACKSIZE_IRQ    (4*1024)    /* IRQ stack */
#define CONFIG_STACKSIZE_FIQ    (4*1024)    /* FIQ stack */
#endif

/*
 * Physical Memory Map
 */
#define CFG_DRAM_BASE        0xa0000000
#define CFG_DRAM_SIZE        0x02000000

#define CONFIG_NR_DRAM_BANKS    4       /* we have 1 bank of DRAM */
#define PHYS_SDRAM_1        0xa0000000 /* SDRAM Bank #1 */
#define PHYS_SDRAM_1_SIZE    0x02000000 /* 32 MB */
#define PHYS_SDRAM_2        0xa4000000 /* SDRAM Bank #2 */
#define PHYS_SDRAM_2_SIZE    0x00000000 /* 0 MB */
#define PHYS_SDRAM_3        0xa8000000      /* SDRAM Bank #3             */
#define PHYS_SDRAM_3_SIZE    0x00000000      /* 0 MB                 */
#define PHYS_SDRAM_4        0xac000000      /* SDRAM Bank #4             */
#define PHYS_SDRAM_4_SIZE    0x00000000      /* 0 MB                 */

/*
 * FLASH and environment organization
 */
#define CFG_MAX_FLASH_BANKS    1        /* number of flash memory 
banks        */
#define CFG_MAX_FLASH_SECT    64        /* number of sectors in flash 
        */
#define PHYS_FLASH_1        0x00000000     /* Flash Bank #1 */
#define PHYS_FLASH_SIZE        0x01000000     /* 16 MB */
#define PHYS_FLASH_BANK_SIZE    0x01000000     /* 16 MB Banks */
#define PHYS_FLASH_SECT_SIZE    0x00040000    /* 256 KB sectors (x2)(2 
chips) */
#define CFG_MONITOR_LEN        0x00040000    /* set aside space for 
U-boot        */

/* Physical setup of Flash */
#define CFG_FLASH_BASE        PHYS_FLASH_1    /* Flash starts at the 
first bank     */
#define CFG_FLASH_USE_BUFFER_WRITE
#define CFG_MONITOR_BASE    CFG_FLASH_BASE    /* where does the monitor 
live?        */

#define FPGA_REGS_BASE_PHYSICAL 0x08000000

/*
 * GPIO settings
 */
#define CFG_GPSR0_VAL        0x00000000
#define CFG_GPSR1_VAL        0x00000000
#define CFG_GPSR2_VAL        0x00000000
#define CFG_GPCR0_VAL        0x00000000
#define CFG_GPCR1_VAL        0x00000000
#define CFG_GPCR2_VAL        0x00000000
#define CFG_GPDR0_VAL        0x08108000
#define CFG_GPDR1_VAL        0x00322b82
#define CFG_GPDR2_VAL        0x0001C000
#define CFG_GAFR0_L_VAL        0x80000000
#define CFG_GAFR0_U_VAL        0x80000010
#define CFG_GAFR1_L_VAL        0x099a9550
#define CFG_GAFR1_U_VAL        0x00000a08
#define CFG_GAFR2_L_VAL        0xA0000000
#define CFG_GAFR2_U_VAL        0x00000002

#define CFG_PSSR_VAL        0x00

/*
 * Memory settings
 */
#define CFG_MSC0_VAL        0x7ff124f2
#define CFG_MSC1_VAL        0x0000b8bc
#define CFG_MSC2_VAL        0x0000ffb9
#define CFG_MDCNFG_VAL        0x00001AC9
#define CFG_MDREFR_VAL        0x0009c017
#define CFG_MDMRS_VAL        0x00220032

/*
 * PCMCIA and CF Interfaces
 */
#define CFG_MECR_VAL        0x00000000
#define CFG_MCMEM0_VAL        0x00000000
#define CFG_MCMEM1_VAL        0x00000000
#define CFG_MCATT0_VAL        0x00000000
#define CFG_MCATT1_VAL        0x00000000
#define CFG_MCIO0_VAL        0x00000000
#define CFG_MCIO1_VAL        0x00000000

/* timeout values are in ticks */
#define CFG_FLASH_ERASE_TOUT    (25*CFG_HZ) /* Timeout for Flash Erase */
#define CFG_FLASH_WRITE_TOUT    (25*CFG_HZ) /* Timeout for Flash Write */

/* FIXME */
#define CFG_ENV_IS_IN_FLASH    1
#define CFG_ENV_ADDR        (PHYS_FLASH_1 + 0x3C000)    /* Addr of 
Environment Sector    */
#define CFG_ENV_SIZE        0x4000    /* Total Size of Environment 
Sector    */
#endif    /* __CONFIG_H */

[U-Boot-Users] Re: One more step to conclude the u-boot port to pxa255 board

[Wolfgang Denk]

In message <41C0A901.6080308@topcomm.com.br> you wrote:
> 
> Now that everything is working with BDI plugged and u-boot transferred 
> to Flash Memory 0x00000000 address. With BDI we issue "reset" and 
> "go"and u-boot comes up and boots the kernel(one more "go") and linux 
> root file system from flash.
> 
> When we unplugged the BDI2000 and powered-up the board u-boot did not 
> come up.

This means that your BDI config file contains certain  initialization
sequences which are missing in U-Boot.

> seems that the u-boot is not prepared to be booted (0x00000000 entry 
> point) at power-up.

No. It is.

> Are we missing anything ? Any suggestions to solve this problem ?

See below.

> #define CONFIG_PXA250        1    /* This is an PXA250 CPU    */
> #define CONFIG_LUBBOCK        1    /* on an LUBBOCK Board        */

You said that this is  a  PXA255  on  a  custom  board.  Why  do  you
configure for a PXA250 on a Lubbock board then?

> #define CONFIG_COMMANDS        
> (CFG_CMD_BDI|CFG_CMD_LCDDRAW|CFG_CMD_LCDTST|CFG_CMD_LOADS|CFG_CMD_LOADB|CFG_CMD_IMI|CFG_CMD_CACHE|CFG_CMD_FLASH|CFG_CMD_MEMORY|CFG_CMD_NET|CFG_CMD_ENV|CFG_CMD_BOOTD|CFG_CMD_CONSOLE|CFG_CMD_ASKENV|CFG_CMD_RUN|CFG_CMD_ECHO|CFG_CMD_REGINFO|CFG_CMD_IMMAP|CF
> G_CMD_DHCP|CFG_CMD_AUTOSCRIPT|CFG_CMD_ELF|CFG_CMD_MISC|CFG_CMD_SDRAM|CFG_CMD_DIAG|CFG_CMD_SAVES|CFG_CMD_PING|CFG_CMD_FAT|CFG_CMD_IMLS)

Do you really think anybody can read this code?

> #define CONFIG_ETHADDR        00:d0:c9:60:87:aa

It is an extremely bad idea to hard-wire the MAC address in U-Boot.

> #define CFG_DRAM_BASE        0xa0000000
> #define CFG_DRAM_SIZE        0x02000000
> 
> #define CONFIG_NR_DRAM_BANKS    4       /* we have 1 bank of DRAM */


Please read this line. There is at least  one  serious  bug  in  this
line.

> #define PHYS_SDRAM_1        0xa0000000 /* SDRAM Bank #1 */
> #define PHYS_SDRAM_1_SIZE    0x02000000 /* 32 MB */
> #define PHYS_SDRAM_2        0xa4000000 /* SDRAM Bank #2 */
> #define PHYS_SDRAM_2_SIZE    0x00000000 /* 0 MB */
> #define PHYS_SDRAM_3        0xa8000000      /* SDRAM Bank #3             */
> #define PHYS_SDRAM_3_SIZE    0x00000000      /* 0 MB                 */
> #define PHYS_SDRAM_4        0xac000000      /* SDRAM Bank #4             */
> #define PHYS_SDRAM_4_SIZE    0x00000000      /* 0 MB                 */

Adding  these  numbers  makes   me   believe   your   definition   of
CFG_DRAM_SIZE might be completely bogus.

Best regards,

Wolfgang Denk

-- 
Software Engineering:  Embedded and Realtime Systems,  Embedded Linux
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Prof:        So the American government went to IBM to come up with a
             data encryption standard and they came up with ...
Student:     EBCDIC!