[U-Boot-Users] s3c2440 -- serial_init

[U-Boot-Users] s3c2440 -- serial_init

[Tiju]

Hi all

I am trying to port u-boot to a s3c2440 samsung processor. 

I applied the smdk2440 boards patch (almost similar architecture) and trying to modify it for our board's requirement. But during the serial_init it goes to the _serial_putc and then after certain loops it goes to the hang function and hangs there.

What could be the problem? Or am I doing some wrong initializations?

Thanks in advance.

Tiju




      Now you can chat without downloading messenger. Go to http://in.messenger.yahoo.com/webmessengerpromo.php
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.denx.de/pipermail/u-boot/attachments/20080328/b635a576/attachment.htm 

[U-Boot-Users] s3c2440 -- serial_init

[michael]

Tiju wrote:
> Hi all
>
> I am trying to port u-boot to a s3c2440 samsung processor.
>
> I applied the smdk2440 boards patch (almost similar architecture) and 
> trying to modify it for our board's requirement. But during the 
> serial_init it goes to the _serial_putc and then after certain loops 
> it goes to the hang function and hangs there.
>
> What could be the problem? Or am I doing some wrong initializations?
>
>
Are you sure that the memory inizializzation is correct?

Regards Michael

[U-Boot-Users] s3c2440 -- serial_init

[Tiju]

 Hi Michael,

I have verified all the RAM initializations in  board/smdk2440/lowlevel_init.S and it seems to be fine. 
The other changes that I have done are

1. In the board/smdk2440/config.mk file I changed TEXT_BASE = 0x33F80000 to 
TEXT_BASE = 0x00000000 because I am booting from the NOR flash. The NOR flash address range is 0x00000000 - 0x08000000. The RAM address range is 0x30000000 - 0x38000000.

If I compile it with TEXT_BASE = 0x33F80000 it does not run at all!

2. When I changed this I found that it has a clash with the stack building in cpu/arm920t/start.S. Therefore I changed it to
.........
/* Set up the stack                         */
stack_setup:
/*  ldr r0, _TEXT_BASE  */  /* upper 128 KiB: relocated uboot   */
    mov r0, #0x33           /* make stack at 33F8*/
    mov r0, r0, lsl #24
    mov r1, #0xF8
    mov r1, r1, lsl #16
    orr r0, r0, r1
.......

Would these be a problem for the memory initializations?

Thanks in advance
Tiju



> Hi all
>
> I am trying to port u-boot to a s3c2440 samsung processor.
>
> I applied the smdk2440 boards patch (almost similar architecture) and 
> trying to modify it for our board's requirement. But during the 
> serial_init it goes to the _serial_putc and then after certain loops 
> it goes to the hang function and hangs there.
>
> What could be the problem? Or am I doing some wrong initializations?
>
>
Are you sure that the memory inizializzation is correct?

Regards Michael




      5, 50, 500, 5000 - Store N number of mails in your inbox. Go to http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.denx.de/pipermail/u-boot/attachments/20080331/aab20735/attachment.htm 

[U-Boot-Users] s3c2440 -- serial_init

[Tiju]

 Hi Michael,

I have verified all the RAM initializations in  board/smdk2440/lowlevel_init.S and it seems to be fine. 
The other changes that I have done are

1. In the board/smdk2440/config.mk file I changed TEXT_BASE = 0x33F80000 to 
TEXT_BASE = 0x00000000 because I am booting from the NOR flash. The NOR flash address range is 0x00000000 - 0x08000000. The RAM address range is 0x30000000 - 0x38000000.

If I compile it with TEXT_BASE = 0x33F80000 it does not run at all!

2. When I changed this I found that it has a clash with the stack building in cpu/arm920t/start.S. Therefore I changed it to
.........
/* Set up the stack                         */
stack_setup:
/*  ldr r0, _TEXT_BASE  */  /* upper 128 KiB: relocated uboot   */
    mov r0, #0x33           /* make stack at 33F8*/
    mov r0, r0, lsl #24
    mov r1, #0xF8
    mov r1, r1, lsl #16
    orr r0, r0, r1
.......

Would these be a problem for the memory initializations?

Thanks in advance
Tiju



> Hi all
>
> I am trying to port u-boot to a s3c2440 samsung processor.
>
> I applied the smdk2440 boards patch (almost similar architecture) and 
> trying to modify it for our board's requirement. But during the 
> serial_init it goes to the _serial_putc and then after certain loops 
> it goes to the hang function and hangs there.
>
> What could be the problem? Or am I doing some wrong initializations?
>
>
Are you sure that the memory inizializzation is correct?

Regards Michael




      Unlimited freedom, unlimited storage. Get it now, on http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.denx.de/pipermail/u-boot/attachments/20080331/3df9bfd0/attachment.htm 

[U-Boot-Users] s3c2440 -- serial_init

[Tiju]

Hi Vishal,

I am stuck with the Kernel load address and entry point. It keeps on resetting. Output as shown below.
My ram range is from 30000000 to 38000000.


U-Boot 1.3.2 (Apr  4 2008 - 18:02:03)

DRAM:  128 MB
Flash: 32 MB
Using default environment

In:    serial
Out:   serial
Err:   serial
Hit any key to stop autoboot:  0
## Booting image at 34000000 ...
   Image Name:   TIJU Linux-2.6.24.4 Image
   Image Type:   ARM Linux Kernel Image (gzip compressed)
   Data Size:    4288104 Bytes =  4.1 MB
   Load Address: 36000000
   Entry Point:  36000000
   Verifying Checksum ... OK
   Uncompressing Kernel Image ... OK

Starting kernel ...

undefined instruction
pc : [<3790e004>]          lr : [<3101175c>]
sp : 30fcfaf8  ip : 30fcffb8     fp : 00000000
r10: 00000001  r9 : 30fcfe98     r8 : 30fcffdc
r7 : 31012e94  r6 : 34000040     r5 : 310194b4  r4 : 310194b1
r3 : 36000000  r2 : 30000100     r1 : 0000016a  r0 : 00000000
Flags: nzCv  IRQs off  FIQs off  Mode SVC_32
Resetting CPU ...

data abort
pc : [<31010ca8>]          lr : [<31010d40>]
sp : 30fcfa88  ip : 3101afdc     fp : 00000000
r10: 00000001  r9 : 30fcfe98     r8 : 30fcffdc
r7 : 31012e94  r6 : 34000040     r5 : 0065ba3e  r4 : 00036042
r3 : 3101afe0  r2 : 00036042     r1 : 00003707  r0 : 000036ff
Flags: nzCv  IRQs on  FIQs on  Mode SVC_32
Resetting CPU ...
                                                                                                              
What could be the problem?

Thanks,
Tiju



----- Original Message ----
From: Vishal Oliyil Kunnil <vishal.ok@gmail.com>
To: Tiju <tj_jac@yahoo.co.in>
Sent: Friday, 4 April, 2008 1:42:30 PM
Subject: Re: [U-Boot-Users] s3c2440 -- serial_init

Hi Tiju,
Never used the controller either, but I suggest you try enabling the
debug macro
in the ethernet driver ...
CONFIG_DM9000_DEBUG

-- snip --
#ifdef CONFIG_DM9000_DEBUG
#define DM9000_DBG(fmt,args....) printf(fmt ,##args)
#else                           /*  */
#define DM9000_DBG(fmt,args...)
#endif                          /*  */
--snip-

Regards,
Vishal
On Fri, Apr 4, 2008 at 12:35 PM, Tiju <tj_jac@yahoo.co.in> wrote:
>
> Hi Vishal,
>
> No, the ethernet controller is outside the processor. We are using DM9000A.
> The link led does not glow when LAN cable is connected. Instead when we ping
> it glows, but does not ping correctly.
>
> Thank You,
> Tiju
>
>
> ----- Original Message ----
> From: Vishal Oliyil Kunnil <vishal.ok@gmail.com>
> To: Tiju <tj_jac@yahoo.co.in>
>
> Sent: Thursday, 3 April, 2008 8:32:26 PM
> Subject: Re: [U-Boot-Users] s3c2440 -- serial_init
>
> Is the ethrnet core inside the processor ? (never used the smcx ...)
> Does ping work ? Does the link LED glow when you plug in the LAN cable ?
>
> Vishal
>
> On Thu, Apr 3, 2008 at 7:05 PM, Tiju <tj_jac@yahoo.co.in> wrote:
> >
> > Hi Vishal,
> >
> > Thank You so much for your help. The u-boot is up and running from
> yesterday
> > onwards. :)
> > The network driver doesn't seem to work properly. Trying to figure it out.
> > It says---
> >
> > DPB2440 # tftpboot
> > dm9000 i/o: 0x18000000, id: 0x90000a46
> > MAC: 00:e0:63:0e:56:78
> > could not establish link
> > TFTP from server 192.168.100.6; our IP address is 192.168.100.110
> > Filename 'zImage'.
> > Load address: 0x34000000
> > Loading: T T T #T T T T T T T
> > Retry count exceeded; starting again
> > TFTP from server 192.168..100.6; our IP address is 192.168.100.110
> > Filename 'zImage'.
> > Load address: 0x34000000
> > Loading: T T #T T T T T T T T
> > ....
> >
> > Any idea what the problem is?
> >
> > Regards,
> > Tiju
> >
> >
> > ----- Original Message ----
> > From: Vishal Oliyil Kunnil <vishal.ok@gmail.com>
> > To: Tiju <tj_jac@yahoo.co.in>
> >
> > Sent: Wednesday, 2 April, 2008 11:29:21 AM
> > Subject: Re: [U-Boot-Users] s3c2440 -- serial_init
> >
> > Hi Tiju,
> > You should not write the image to RAM. U-boot should re-locate by itself.
> > What is the debugger that you use ? Try putting a breakpoint at address
> > of start_armboot() -> you can get the address by
> > objdump -S u-boot.elf > u-boot.dump
> > A memory dump should show contents of 0x0 anf 0x31000000 to be same.
> > Also I suggest you use a stable release of u-boot and not  U-Boot
> > 1.3.2-00046-g23e20aa-dirty.
> >
> > Regards,
> > -Vishal
> >
> >
> >
> >
> > On Wed, Apr 2, 2008 at 8:58 AM, Tiju <tj_jac@yahoo.co.in> wrote:
> > >
> > > Hi Vishal,
> > >
> > > Thanks for the tips. After I compiling with TEXT_BASE=0x31000000 and
> > > commenting
> > > - CONFIG_SKIP_LOWLEVEL_INIT
> > > - CONFIG_SKIP_RELOCATE_UBOOT it does not work directly. After this I
> just
> > > wrote the same image to the NOR flash as well as the RAM (starting from
> > > 0x31000000) then on the u-boot the following appears.
> > >
> > > U-Boot 1.3.2-00046-g23e20aa-dirty (Apr  1 2008 - 19:33:51)
> > >
> > > DRAM:  64 MB
> > > *** Warning - bad CRC, using default environment
> > >
> > > In:    serial
> > > Out:  serial
> > > Err:  serial
> > > DPB2440 #
> > >
> > > After this it hangs. ie. no input or output.
> > >
> > > I think the initial problem to be taken care of is the relocation and
> then
> > > the serial. How do we go about these issues?
> > >
> > > Thanks in advance.
> > > Tiju
> > >
> > >
> > >
> > >
> > > ----- Original Message ----
> > > From: Vishal Oliyil Kunnil <vishal.ok@gmail.com>
> > > To: Tiju <tj_jac@yahoo.co.in>
> > >
> > > Sent: Monday, 31 March, 2008 6:31:43 PM
> > > Subject: Re: [U-Boot-Users] s3c2440 -- serial_init
> > >
> > > Tiju,
> > >
> > > Read the documentation:
> > > ----- snip - README -----
> > > - CONFIG_SKIP_LOWLEVEL_INIT
> > > - CONFIG_SKIP_RELOCATE_UBOOT
> > >                [ARM only] If these variables are defined, then
> > >                certain low level initializations (like setting up
> > >                the memory controller) are omitted and/or U-Boot does
> > >                not relocate itself into RAM.
> > >                Normally these variables MUST NOT be defined. The
> > >                only exception is when U-Boot is loaded (to RAM) by
> > >                some other boot loader or by a debugger which
> > >                performs these intializations itself.
> > > ----- snip - README -------
> > >
> > > I suggest you enable the RELOCATE and LOWLEVEL_INIT (by commenting the
> > > defines)
> > > - u-boot AFAIK cannot run lke the XIP kernel off NOR Flash; inspite of
> > > the fact the NOR
> > > flash can be directly read by the processor - unlike NAND. There is no
> > > noticeable performance
> > > hit or wastage of memory if you let u-boot relocate by itself.
> > >
> > > TEXT_BASE=0x31000000 in my opinion is correct.
> > >
> > > After relocating, a dump of flash contents and RAM (TEXT_BASE) contents
> > > should show you the same contents. If it does not - perhaps you have
> > > not configured
> > > the memory controller correctly. This can be verified by checking if
> > > you are able to
> > > write to RAM using the debugger after memory initializations (which is
> > > typically in LOWLEVEL_INIT, which you have skipped - I dont know if
> > > this is so for the s3c2440 ).
> > >
> > > Regards,
> > > Vishal
> > >
> > > On Mon, Mar 31, 2008 at 5:16 PM, Tiju <tj_jac@yahoo.co.in> wrote:
> > > >
> > > > Thanks Vishal.. I compiled with TEXT_BASE=0x31000000 and for
> > start_armboot
> > > > jumps to 0x310004d0 but there are junk values present there.
> > > >
> > > > By default these are  enabled:
> > > > #define CONFIG_SKIP_RELOCATE_UBOOT  1
> > > > #define CONFIG_SKIP_LOWLEVEL_INIT  1
> > > > so u-boot does not relocate.
> > > >
> > > > Still now what should the value of TEXT_BASE be? I dont want to
> relocate
> > > > u-boot because from NOR flash byte by byte read can be done.
> > > >
> > > > Regards,
> > > > Tiju
> > > >
> > > >
> > > > ----- Original Message ----
> > > > From: Vishal Oliyil Kunnil <vishal.ok@gmail.com>
> > > > To: Tiju <tj_jac@yahoo.co.in>
> > > >
> > > > Sent: Monday, 31 March, 2008 2:58:12 PM
> > > > Subject: Re: [U-Boot-Users] s3c2440 -- serial_init
> > > >
> > > > TEXT_BASE is the address for which u-boot is linked for.If you take an
> > > > objdump of u-boot
> > > > elf, you will see that it links for address beginning with that
> > > > specified by TEXT_BASE.
> > > > Meaning, you link for the address thus specified.
> > > > Typically  the binary will be run from the reset vector of the
> > > > processor, which is not necessarily
> > > > TEXT_BASE : say 0x0 flash address. U-boot starts executing from the
> > > > reset vector,
> > > > relocates to RAM  and since it is linked for TEXT_BASE, the  ldr pc,
> > > > _start_armboot
> > > > will branch to the start_armboot which is in RAM.
> > > > -------snip - start.S --------------
> > > > ldr    pc, _start_armboot
> > > > _start_armboot: .word start_armboot
> > > > -------snip - start.S --------------
> > > > Regards,
> > > > Vishal
> > > >
> > > > On Mon, Mar 31, 2008 at 1:41 PM, Tiju <tj_jac@yahoo.co.in> wrote:
> > > > >
> > > > > Hi Vishal,
> > > > >
> > > > > What does TEXT_BASE mean? Is this the place where the UBOOT gets
> > > > relocated?
> > > > >
> > > > > Thank You,
> > > > > Tiju
> > > > >
> > > > >
> > > > >
> > > > > ----- Original Message ----
> > > > > From: "vishal.ok at gmail.com" <vishal..ok@gmail.com>
> > > > > To: Tiju <tj_jac@yahoo.co.in>
> > > > > Sent: Monday, 31 March, 2008 1:35:47 PM
> > > > > Subject: Re: [U-Boot-Users] s3c2440 -- serial_init
> > > > >
> > > > > Hi Tiju,
> > > > >
> > > > > Try with TEXT_BASE=0x31000000, and leave start.s untouched !
> > > > > I think you have misinterpreted the meaning of TEXT_BASE.
> > > > >
> > > > > Regards,
> > > > > Vishal
> > > > >
> > > > > On 3/31/08, Tiju <tj_jac@yahoo.co.in> wrote:
> > > > > > Hi Michael,
> > > > > >
> > > > > > I have verified all the RAM initializations in
> > > > > > board/smdk2440/lowlevel_init.S and it seems to be fine.
> > > > > > The other changes that I have done are
> > > > > >
> > > > > > 1.. In the board/smdk2440/config.mk file I changed TEXT_BASE =
> > > > 0x33F80000
> > > > > to
> > > > > > TEXT_BASE = 0x00000000 because I am booting from the NOR flash.
> The
> > > NOR
> > > > > > flash address range is 0x00000000 - 0x08000000. The RAM address
> > range
> > > is
> > > > > > 0x30000000 - 0x38000000.
> > > > > >
> > > > > > If I compile it with TEXT_BASE = 0x33F80000 it does not run at
> all!
> > > > > >
> > > > > > 2. When I changed this I found that it has a clash with the stack
> > > > building
> > > > > > in cpu/arm920t/start.S. Therefore I changed it to
> > > > > > ..........
> > > > > > /* Set up the stack */
> > > > > > stack_setup:
> > > > > > /* ldr r0, _TEXT_BASE */ /* upper 128 KiB: relocated uboot */
> > > > > > mov r0, #0x33 /* make stack at 33F8*/
> > > > > > mov r0, r0, lsl #24
> > > > > > mov r1, #0xF8
> > > > > > mov r1, r1, lsl #16
> > > > > > orr r0, r0, r1
> > > > > > ........
> > > > > >
> > > > > > Would these be a problem for the memory initializations?
> > > > > >
> > > > > > Thanks in advance
> > > > > > Tiju
> > > > > >
> > > > > >
> > > > > >
> > > > > > > Hi all
> > > > > > >
> > > > > > > I am trying to port u-boot to a s3c2440 samsung processor.
> > > > > > >
> > > > > > > I applied the smdk2440 boards patch (almost similar
> architecture)
> > > and
> > > > > > > trying to modify it for our board's requirement. But during the
> > > > > > > serial_init it goes to the _serial_putc and then after certain
> > loops
> > > > > > > it goes to the hang function and hangs there.
> > > > > > >
> > > > > > > What could be the problem? Or am I doing some wrong
> > initializations?
> > > > > > >
> > > > > > >
> > > > > > Are you sure that the memory inizializzation is correct?
> > > > > >
> > > > > > Regards Michael
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > 5, 50, 500, 5000 - Store N number of mails in your inbox. Go to
> > > > > >
> http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html
> > > > >
> > > > >
> > > > >  ________________________________
> > > > >  Get your domain and website for less than Rs.100/month*. Click
> here.
> > > >
> > > >
> > > >
> > > >  ________________________________
> > > >  Get your domain and website for less than Rs.100/month*.. Click here.
> > >
> > >
> > >  ________________________________
> > >  Bring your gang together - do your thing. Start your group.
> >
> >
> >
> >  ________________________________
> >  Bring your gang together - do your thing. Start your group.
>
>
>
>  ________________________________
>  Get your domain and website for less than Rs.100/month*. Click here.






      Share files, take polls, and make new friends - all under one roof. Go to http://in.promos.yahoo.com/groups/
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.denx.de/pipermail/u-boot/attachments/20080404/24131031/attachment.htm 

[U-Boot-Users] s3c2440 -- serial_init

[Harald Welte]

On Fri, Mar 28, 2008 at 04:58:37PM +0530, Tiju wrote:
> Hi all
> 
> I am trying to port u-boot to a s3c2440 samsung processor. 
> 
> I applied the smdk2440 boards patch (almost similar architecture) and
> trying to modify it for our board's requirement. But during the
> serial_init it goes to the _serial_putc and then after certain loops
> it goes to the hang function and hangs there.

this sounds like a flow control issue.  The s3c24xx boards support
hardware flow control handling, i.e. their TX FIFO will not transmit any
more characters unless the RTS/CTS handshake allows it to.

Since the u-boot serial code does an endless loop waiting for the next
character, it will busy-wait there.

Please note that this is true even for UARTs that don't even have
RTS/CTS.  Don't use hw flow control on them.

Another possible cause could be lack of proper GPIO configuration for
the RTS/CTS pins, missing connection on the board, broken serial cable.

Basically anthing that affects RTS/CTS from within the s3c24xx and the
PC on the other side.

-- 
- Harald Welte <laforge@gnumonks.org>           http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
                                                  (ETSI EN 300 175-7 Ch. A6)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.denx.de/pipermail/u-boot/attachments/20080406/ba4eb1bf/attachment.pgp 

[U-Boot-Users] s3c2440 -- serial_init

[Harald Welte]

On Fri, Apr 04, 2008 at 08:29:48PM +0530, Tiju wrote:
> Hi Vishal,
> 
> I am stuck with the Kernel load address and entry point. It keeps on
> resetting. Output as shown below.

are you sure

1) your PLL's are configured correctly, i.e.
* the memory bus doesn't exceed the maximum bus frequency allowed given
  your capacitive bus load (see samsung manual)
* the core CPU speed doesn't exceed the max speed
* the core CPU voltage is configued correct for the core CPU clock rate

2) Are you sure your linux kernel is compiled for the right ARM variant
   with no strange compiler flags?

I suggest running extensive memory testing routines (u-boot has memory
testing built-in) before trying to boot the kernel, just to make sure
the entire hardware and PLL / power config is stable at all.

Another interesting question would be to compare the memory contents at
the program counter with the content it should have according to the
uImage/zImage/vmlinux file of th kernel.

-- 
- Harald Welte <laforge@gnumonks.org>           http://laforge.gnumonks.org/
============================================================================
"Privacy in residential applications is a desirable marketing option."
                                                  (ETSI EN 300 175-7 Ch. A6)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
Url : http://lists.denx.de/pipermail/u-boot/attachments/20080406/3b44af5b/attachment.pgp