summaryrefslogtreecommitdiff
path: root/doc
diff options
context:
space:
mode:
Diffstat (limited to 'doc')
-rw-r--r--doc/README.bamboo62
-rw-r--r--doc/README.generic_usb_ohci57
-rw-r--r--doc/README.mpc8323erdb71
-rw-r--r--doc/README.mpc8360emds23
-rw-r--r--doc/README.mpc83xx.ddrecc (renamed from doc/README.mpc8349emds.ddrecc)60
-rw-r--r--doc/README.mpc8544ds122
-rw-r--r--doc/README.sbc8641d28
-rw-r--r--doc/README.zeus73
8 files changed, 465 insertions, 31 deletions
diff --git a/doc/README.bamboo b/doc/README.bamboo
index b50be01..e139c6d 100644
--- a/doc/README.bamboo
+++ b/doc/README.bamboo
@@ -1,3 +1,65 @@
+The 2 important dipswitches are configured as shown below:
+
+SW1 (for 33MHz SysClk)
+----------------------
+S1 S2 S3 S4 S5 S6 S7 S8
+OFF OFF OFF OFF OFF OFF OFF ON
+
+SW7 (for Op-Code Flash and Boot Option H)
+-----------------------------------------
+S1 S2 S3 S4 S5 S6 S7 S8
+OFF OFF OFF ON OFF OFF OFF OFF
+
+The EEPROM at location 0x52 is loaded with these 16 bytes:
+C47042A6 05D7A190 40082350 0d050000
+
+SDR0_SDSTP0[ENG]: 1 : PLL's VCO is the source for PLL forward divisors
+SDR0_SDSTP0[SRC]: 1 : Feedback originates from PLLOUTB
+SDR0_SDSTP0[SEL]: 0 : Feedback selection is PLL output
+SDR0_SDSTP0[TUNE]: 1000111000 : 10 <= M <= 22, 600MHz < VCO <= 900MHz
+SDR0_SDSTP0[FBDV]: 4 : PLL feedback divisor
+SDR0_SDSTP0[FBDVA]: 2 : PLL forward divisor A
+SDR0_SDSTP0[FBDVB]: 5 : PLL forward divisor B
+SDR0_SDSTP0[PRBDV0]: 1 : PLL primary divisor B
+SDR0_SDSTP0[OPBDV0]: 2 : OPB clock divisor
+SDR0_SDSTP0[LFBDV]: 1 : PLL local feedback divisor
+SDR0_SDSTP0[PERDV0]: 3 : Peripheral clock divisor 0
+SDR0_SDSTP0[MALDV0]: 2 : MAL clock divisor 0
+SDR0_SDSTP0[PCIDV0]: 2 : Sync PCI clock divisor 0
+SDR0_SDSTP0[PLLTIMER]: 7 : PLL locking timer
+SDR0_SDSTP0[RW]: 1 : EBC ROM width: 16-bit
+SDR0_SDSTP0[RL]: 0 : EBC ROM location: EBC
+SDR0_SDSTP0[PAE]: 0 : PCI internal arbiter: disabled
+SDR0_SDSTP0[PHCE]: 0 : PCI host configuration: disabled
+SDR0_SDSTP0[ZM]: 3 : ZMII mode: RMII mode 100
+SDR0_SDSTP0[CTE]: 0 : CPU trace: disabled
+SDR0_SDSTP0[Nto1]: 0 : CPU/PLB ratio N/P: not N to 1
+SDR0_SDSTP0[PAME]: 1 : PCI asynchronous mode: enabled
+SDR0_SDSTP0[MEM]: 1 : Multiplex: EMAC
+SDR0_SDSTP0[NE]: 0 : NDFC: disabled
+SDR0_SDSTP0[NBW]: 0 : NDFC boot width: 8-bit
+SDR0_SDSTP0[NBW]: 0 : NDFC boot page selection
+SDR0_SDSTP0[NBAC]: 0 : NDFC boot address selection cycle: 3 Addr. Cycles, 1 Col. + 2 Row (512 page size)
+SDR0_SDSTP0[NARE]: 0 : NDFC auto read : disabled
+SDR0_SDSTP0[NRB]: 0 : NDFC Ready/Busy : Ready
+SDR0_SDSTP0[NDRSC]: 33333 : NDFC device reset counter
+SDR0_SDSTP0[NCG0]: 0 : NDFC/EBC chip select gating CS0 : EBC
+SDR0_SDSTP0[NCG1]: 0 : NDFC/EBC chip select gating CS1 : EBC
+SDR0_SDSTP0[NCG2]: 0 : NDFC/EBC chip select gating CS2 : EBC
+SDR0_SDSTP0[NCG3]: 0 : NDFC/EBC chip select gating CS3 : EBC
+SDR0_SDSTP0[NCRDC]: 3333 : NDFC device read count
+
+PPC440EP Clocking Configuration
+
+SysClk is 33.0MHz, M is 20, VCO is 660.0MHz, CPU is 330.0MHz, PLB is 132.0MHz
+OPB is 66.0MHz, EBC is 44.0MHz, MAL is 66.0MHz, Sync PCI is 66.0MHz
+
+The above information is reported by Eugene O'Brien
+<Eugene.O'Brien@advantechamt.com>. Thanks a lot.
+
+2007-08-06, Stefan Roese <sr@denx.de>
+---------------------------------------------------------------------
+
The configuration for the AMCC 440EP eval board "Bamboo" was changed
to only use 384 kbytes of FLASH for the U-Boot image. This way the
redundant environment can be saved in the remaining 2 sectors of the
diff --git a/doc/README.generic_usb_ohci b/doc/README.generic_usb_ohci
new file mode 100644
index 0000000..494dd1f
--- /dev/null
+++ b/doc/README.generic_usb_ohci
@@ -0,0 +1,57 @@
+Notes on the the generic USB-OHCI driver
+========================================
+
+This driver (drivers/usb_ohci.[ch]) is the result of the merge of
+various existing OHCI drivers that were basically identical beside
+cpu/board dependant initalization. This initalization has been moved
+into cpu/board directories and are called via the hooks below.
+
+Configuration options
+----------------------
+
+ CONFIG_USB_OHCI_NEW: enable the new OHCI driver
+
+ CFG_USB_OHCI_BOARD_INIT: call the board dependant hooks:
+
+ - extern int usb_board_init(void);
+ - extern int usb_board_stop(void);
+ - extern int usb_cpu_init_fail(void);
+
+ CFG_USB_OHCI_CPU_INIT: call the cpu dependant hooks:
+
+ - extern int usb_cpu_init(void);
+ - extern int usb_cpu_stop(void);
+ - extern int usb_cpu_init_fail(void);
+
+ CFG_USB_OHCI_REGS_BASE: defines the base address of the OHCI
+ registers
+
+ CFG_USB_OHCI_SLOT_NAME: slot name
+
+ CFG_USB_OHCI_MAX_ROOT_PORTS: maximal number of ports of the
+ root hub.
+
+
+Endianness issues
+------------------
+
+The USB bus operates in little endian, but unfortunately there are
+OHCI controllers that operate in big endian such as ppc4xx and
+mpc5xxx. For these the config option
+
+ CFG_OHCI_BE_CONTROLLER
+
+needs to be defined.
+
+
+PCI Controllers
+----------------
+
+You'll need to define
+
+ CONFIG_PCI_OHCI
+
+PCI Controllers need to do byte swapping on register accesses, so they
+should to define:
+
+ CFG_OHCI_SWAP_REG_ACCESS
diff --git a/doc/README.mpc8323erdb b/doc/README.mpc8323erdb
new file mode 100644
index 0000000..6f89829
--- /dev/null
+++ b/doc/README.mpc8323erdb
@@ -0,0 +1,71 @@
+Freescale MPC8323ERDB Board
+-----------------------------------------
+
+1. Memory Map
+ The memory map looks like this:
+
+ 0x0000_0000 0x03ff_ffff DDR 64M
+ 0x8000_0000 0x8fff_ffff PCI MEM 256M
+ 0x9000_0000 0x9fff_ffff PCI_MMIO 256M
+ 0xe000_0000 0xe00f_ffff IMMR 1M
+ 0xd000_0000 0xd3ff_ffff PCI IO 64M
+ 0xfe00_0000 0xfeff_ffff NOR FLASH (CS0) 16M
+
+2. Compilation
+
+ Assuming you're using BASH (or similar) as your shell:
+
+ export CROSS_COMPILE=your-cross-compiler-prefix-
+ make distclean
+ make MPC8323ERDB_config
+ make
+
+3. Downloading and Flashing Images
+
+3.1 Reflash U-boot Image using U-boot
+
+ N.b, have an alternate means of programming
+ the flash available if the new u-boot doesn't boot.
+
+ First try a:
+
+ tftpboot $loadaddr $uboot
+
+ to make sure that the TFTP load will succeed before
+ an erase goes ahead and wipes out your current firmware.
+ Then do a:
+
+ run tftpflash
+
+ which is a shorter version of the manual sequence:
+
+ tftp $loadaddr u-boot.bin
+ protect off fe000000 +$filesize
+ erase fe000000 +$filesize
+ cp.b $loadaddr fe000000 $filesize
+
+ To keep your old u-boot's environment variables, do a:
+
+ saveenv
+
+ prior to resetting the board.
+
+3.2 Downloading and Booting Linux Kernel
+
+ Ensure that all networking-related environment variables are set
+ properly (including ipaddr, serverip, gatewayip (if needed),
+ netmask, ethaddr, eth1addr, rootpath (if using NFS root),
+ fdtfile, and bootfile).
+
+ Then, do one of the following, depending on whether you
+ want an NFS root or a ramdisk root:
+
+ run nfsboot
+
+ or
+
+ run ramboot
+
+4 Notes
+
+ The console baudrate for MPC8323ERDB is 115200bps.
diff --git a/doc/README.mpc8360emds b/doc/README.mpc8360emds
index c87469f..5f20247 100644
--- a/doc/README.mpc8360emds
+++ b/doc/README.mpc8360emds
@@ -21,7 +21,13 @@ Freescale MPC8360EMDS Board
SW3[1:8]= 0000_0001 refers to bits labeled 1 through 6 is set as "On"
and bits labeled 8 is set as "Off".
-1.1 For the MPC8360E PB PROTO Board
+1.1 There are three type boards for MPC8360E silicon up to now, They are
+
+ * MPC8360E-MDS-PB PROTO (a.k.a 8360SYS PROTOTYPE)
+ * MPC8360E-MDS-PB PILOT (a.k.a 8360SYS PILOT)
+ * MPC8360EA-MDS-PB PROTO (a.k.a 8360SYS2 PROTOTYPE)
+
+1.2 For all the MPC8360EMDS Board
First, make sure the board default setting is consistent with the
document shipped with your board. Then apply the following setting:
@@ -33,6 +39,21 @@ Freescale MPC8360EMDS Board
JP6 1-2
on board Oscillator: 66M
+1.3 Since different board/chip rev. combinations have AC timing issues,
+ u-boot forces RGMII-ID (RGMII with Internal Delay) mode on by default
+ by the patch (mpc83xx: Disable G1TXCLK, G2TXCLK h/w buffers).
+
+ When the rev2.x silicon mount on these boards, and if you are using
+ u-boot version after this patch, to make the ethernet interfaces usable,
+ and to enable RGMII-ID on your board, you have to setup the jumpers
+ correctly.
+
+ * MPC8360E-MDS-PB PROTO
+ nothing to do
+ * MPC8360E-MDS-PB PILOT
+ JP9 and JP8 should be ON
+ * MPC8360EA-MDS-PB PROTO
+ JP2 and JP3 should be ON
2. Memory Map
diff --git a/doc/README.mpc8349emds.ddrecc b/doc/README.mpc83xx.ddrecc
index eb249c3..0029f08 100644
--- a/doc/README.mpc8349emds.ddrecc
+++ b/doc/README.mpc83xx.ddrecc
@@ -15,10 +15,10 @@ IMPORTANT NOTICE: enabling injecting multiple-bit errors is potentially
dangerous as such errors are NOT corrected by the controller. Therefore caution
should be taken when enabling the injection of multiple-bit errors: it is only
safe when used on a carefully selected memory area and used under control of
-the 'ecc test' command (see example 'Injecting Multiple-Bit Errors' below). In
-particular, when you simply set the multiple-bit errors in inject mask and
-enable injection, U-Boot is very likely to hang quickly as the errors will be
-injected when it accesses its code, data etc.
+the 'ecc testdw' 'ecc testword' command (see example 'Injecting Multiple-Bit
+Errors' below). In particular, when you simply set the multiple-bit errors in
+inject mask and enable injection, U-Boot is very likely to hang quickly as the
+errors will be injected when it accesses its code, data etc.
Use cases for DDR 'ecc' command:
@@ -40,7 +40,7 @@ Injecting Single-Bit Errors
2. Run test over some memory region
-=> ecc test 200000 10
+=> ecc testdw 200000 10
3. Check ECC status
@@ -61,57 +61,57 @@ Memory Error Detect:
16 errors were generated, Single-Bit Error flag was not set as Single Bit Error
Counter did not reach Single-Bit Error Threshold.
-4. Make sure used memory region got re-initialized with 0xcafecafe pattern
+4. Make sure used memory region got re-initialized with 0x0123456789abcdef
=> md 200000
-00200000: cafecafe cafecafe cafecafe cafecafe ................
-00200010: cafecafe cafecafe cafecafe cafecafe ................
-00200020: cafecafe cafecafe cafecafe cafecafe ................
-00200030: cafecafe cafecafe cafecafe cafecafe ................
-00200040: cafecafe cafecafe cafecafe cafecafe ................
-00200050: cafecafe cafecafe cafecafe cafecafe ................
-00200060: cafecafe cafecafe cafecafe cafecafe ................
-00200070: cafecafe cafecafe cafecafe cafecafe ................
+00200000: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200010: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200020: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200030: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200040: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200050: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200060: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200070: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
00200080: deadbeef deadbeef deadbeef deadbeef ................
00200090: deadbeef deadbeef deadbeef deadbeef ................
-
Injecting Multiple-Bit Errors
-----------------------------
1. Set more than 1 bit in Data Path Error Inject Mask
-=> ecc injectdatahi 5
+=> ecc injectdatahi 1
+=> ecc injectdatalo 1
2. Run test over some memory region
-=> ecc test 200000 10
+=> ecc testword 200000 1
3. Check ECC status
=> ecc status
...
-Memory Data Path Error Injection Mask High/Low: 00000005 00000000
+Memory Data Path Error Injection Mask High/Low: 00000001 00000001
...
Memory Error Detect:
- Multiple Memory Errors: 1
+ Multiple Memory Errors: 0
Multiple-Bit Error: 1
Single-Bit Error: 0
...
-Observe that both Multiple Memory Errors and Multiple-Bit Error flags are set.
+The Multiple Memory Errors flags not set and Multiple-Bit Error flags are set.
-4. Make sure used memory region got re-initialized with 0xcafecafe pattern
+4. Make sure used memory region got re-initialized with 0x0123456789abcdef
=> md 200000
-00200000: cafecafe cafecafe cafecafe cafecafe ................
-00200010: cafecafe cafecafe cafecafe cafecafe ................
-00200020: cafecafe cafecafe cafecafe cafecafe ................
-00200030: cafecafe cafecafe cafecafe cafecafe ................
-00200040: cafecafe cafecafe cafecafe cafecafe ................
-00200050: cafecafe cafecafe cafecafe cafecafe ................
-00200060: cafecafe cafecafe cafecafe cafecafe ................
-00200070: cafecafe cafecafe cafecafe cafecafe ................
+00200000: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200010: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200020: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200030: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200040: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200050: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200060: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
+00200070: 01234567 89abcdef 01234567 89abcdef .#Eg.....#Eg....
00200080: deadbeef deadbeef deadbeef deadbeef ................
00200090: deadbeef deadbeef deadbeef deadbeef ................
@@ -140,7 +140,7 @@ Test Single-Bit Error Counter and Threshold
...
Memory Single-Bit Error Management (0..255):
Single-Bit Error Threshold: 255
- Single Bit Error Counter: 60
+ Single Bit Error Counter: 199
Memory Error Detect:
Multiple Memory Errors: 1
diff --git a/doc/README.mpc8544ds b/doc/README.mpc8544ds
new file mode 100644
index 0000000..bf257a0
--- /dev/null
+++ b/doc/README.mpc8544ds
@@ -0,0 +1,122 @@
+Overview
+--------
+The MPC8544DS system is similar to the 85xx CDS systems such
+as the MPC8548CDS due to the similar E500 core. However, it
+is placed on the same board as the 8641 HPCN system.
+
+
+Flash Banks
+-----------
+Like the 85xx CDS systems, the 8544 DS board has two flash banks.
+They are both present on boot, but there locations can be swapped
+using the dip-switch SW10, bit 2.
+
+However, unlike the CDS systems, but similar to the 8641 HPCN
+board, a runtime reset through the FPGA can also affect a swap
+on the flash bank mappings for the next reset cycle.
+
+Irrespective of the switch SW10[2], booting is always from the
+boot bank at 0xfff8_0000.
+
+
+Memory Map
+----------
+
+0xff80_0000 - 0xffbf_ffff Alernate bank 4MB
+0xffc0_0000 - 0xffff_ffff Boot bank 4MB
+
+0xffb8_0000 Alternate image start 512KB
+0xfff8_0000 Boot image start 512KB
+
+
+Flashing Images
+---------------
+
+For example, to place a new image in the alternate flash bank
+and then reset with that new image temporarily, use this:
+
+ tftp 1000000 u-boot.bin.8544ds
+ erase ffb80000 ffbfffff
+ cp.b 1000000 ffb80000 80000
+ pixis_reset altbank
+
+
+To overwrite the image in the boot flash bank:
+
+ tftp 1000000 u-boot.bin.8544ds
+ protect off all
+ erase fff80000 ffffffff
+ cp.b 1000000 fff80000 80000
+
+Other example U-Boot image and flash manipulations examples
+can be found in the README.mpc85xxcds file as well.
+
+
+The pixis_reset command
+-----------------------
+A new command, "pixis_reset", is introduced to reset mpc8641hpcn board
+using the FPGA sequencer. When the board restarts, it has the option
+of using either the current or alternate flash bank as the boot
+image, with or without the watchdog timer enabled, and finally with
+or without frequency changes.
+
+Usage is;
+
+ pixis_reset
+ pixis_reset altbank
+ pixis_reset altbank wd
+ pixis_reset altbank cf <SYSCLK freq> <COREPLL ratio> <MPXPLL ratio>
+ pixis_reset cf <SYSCLK freq> <COREPLL ratio> <MPXPLL ratio>
+
+Examples;
+
+ /* reset to current bank, like "reset" command */
+ pixis_reset
+
+ /* reset board but use the to alternate flash bank */
+ pixis_reset altbank
+
+ /* reset board, use alternate flash bank with watchdog timer enabled*/
+ pixis_reset altbank wd
+
+ /* reset board to alternate bank with frequency changed.
+ * 40 is SYSCLK, 2.5 is COREPLL ratio, 10 is MPXPLL ratio
+ */
+ pixis-reset altbank cf 40 2.5 10
+
+Valid clock choices are in the 8641 Reference Manuals.
+
+
+Using the Device Tree Source File
+---------------------------------
+To create the DTB (Device Tree Binary) image file,
+use a command similar to this:
+
+ dtc -b 0 -f -I dts -O dtb mpc8544ds.dts > mpc8544ds.dtb
+
+Likely, that .dts file will come from here;
+
+ linux-2.6/arch/powerpc/boot/dts/mpc8544ds.dts
+
+After placing the DTB file in your TFTP disk area,
+you can download that dtb file using a command like:
+
+ tftp 900000 mpc8544ds.dtb
+
+Burn it to flash if you want.
+
+
+Booting Linux
+-------------
+
+Place a linux uImage in the TFTP disk area too.
+
+ tftp 1000000 uImage.8544
+ tftp 900000 mpc8544ds.dtb
+ bootm 1000000 - 900000
+
+Watch your ethact, netdev and bootargs U-Boot environment variables.
+You may want to do something like this too:
+
+ setenv ethact eTSEC3
+ setenv netdev eth1
diff --git a/doc/README.sbc8641d b/doc/README.sbc8641d
new file mode 100644
index 0000000..a051466
--- /dev/null
+++ b/doc/README.sbc8641d
@@ -0,0 +1,28 @@
+Wind River SBC8641D reference board
+===========================
+
+Created 06/14/2007 Joe Hamman
+Copyright 2007, Embedded Specialties, Inc.
+Copyright 2007 Wind River Systemes, Inc.
+-----------------------------
+
+1. Building U-Boot
+------------------
+The SBC8641D code is known to build using ELDK 4.1.
+
+ $ make sbc8641d_config
+ Configuring for sbc8641d board...
+
+ $ make
+
+
+2. Switch and Jumper Settings
+-----------------------------
+All Jumpers & Switches are in their default positions. Please refer to
+the board documentation for details. Some settings control CPU voltages
+and settings may change with board revisions.
+
+3. Known limitations
+--------------------
+PCI:
+ The PCI command may hang if no boards are present in either slot.
diff --git a/doc/README.zeus b/doc/README.zeus
new file mode 100644
index 0000000..1848d8c
--- /dev/null
+++ b/doc/README.zeus
@@ -0,0 +1,73 @@
+
+Storage of the board specific values (ethaddr...)
+-------------------------------------------------
+
+The board specific environment variables that should be unique
+for each individual board, can be stored in the I2C EEPROM. This
+will be done from offset 0x80 with the length of 0x80 bytes. The
+following command can be used to store the values here:
+
+=> setdef de:20:6a:ed:e2:72 de:20:6a:ed:e2:73 AB0001
+
+ ethaddr eth1addr serial#
+
+Now those 3 values are stored into the I2C EEPROM. A CRC is added
+to make sure that the values get not corrupted.
+
+
+SW-Reset Pushbutton handling:
+-----------------------------
+
+The SW-reset push button is connected to a GPIO input too. This
+way U-Boot can "see" how long the SW-reset was pressed, and a
+specific action can be taken. Two different actions are supported:
+
+a) Release after more than 5 seconds and less then 10 seconds:
+ -> Run POST
+
+ Please note, that the POST test will take a while (approx. 1 min
+ on the 128MByte board). This is mainly due to the system memory
+ test.
+
+b) Release after more than 10 seconds:
+ -> Restore factory default settings
+
+ The factory default values are restored. The default environment
+ variables are restored (ipaddr, serverip...) and the board
+ specific values (ethaddr, eth1addr and serial#) are restored
+ to the environment from the I2C EEPROM. Also a bootline parameter
+ is added to the Linux bootline to signal the Linux kernel upon
+ the next startup, that the factory defaults should be restored.
+
+The command to check this sw-reset status and act accordingly is
+
+=> chkreset
+
+This command is added to the default "bootcmd", so that it is called
+automatically upon startup.
+
+Also, the 2 LED's are used to indicate the current status of this
+command (time passed since pushing the button). When the POST test
+will be run, the green LED will be switched off, and when the
+factory restore will be initiated, the reg LED will be switched off.
+
+
+Loggin of POST results:
+-----------------------
+
+The results of the POST tests are logged in a logbuffer located at the end
+of the onboard memory. It can be accessed with the U-Boot command "log":
+
+=> log show
+<4>POST memory PASSED
+<4>POST cache PASSED
+<4>POST cpu PASSED
+<4>POST uart PASSED
+<4>POST ethernet PASSED
+
+The DENX Linux kernel tree has support for this log buffer included. Exactly
+this buffer is used for logging of all kernel messages too. By enabling the
+compile time option "CONFIG_LOGBUFFER" this support is enabled. This way you
+can access the U-Boot log messages from Linux too.
+
+2007-08-10, Stefan Roese <sr@denx.de>