platform/kernel/u-boot.git
16 years agoReworked FSL Book-E TLB macros to be more readable
Kumar Gala [Wed, 19 Dec 2007 07:18:15 +0000 (01:18 -0600)]
Reworked FSL Book-E TLB macros to be more readable

The old macros made it difficult to know what WIMGE and perm bits
were set for a TLB entry.  Actually use the bit masks for these items
since they are only a single bit.

Also moved the macros into mmu.h out of e500.h since they aren't specific
to e500.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoUse FSL Book-E MMU macros from Linux Kernel
Kumar Gala [Wed, 19 Dec 2007 05:21:51 +0000 (23:21 -0600)]
Use FSL Book-E MMU macros from Linux Kernel

Grab the FSL Book-E MAS register macros from Linux.  Also added
defines for page sizes up to 4TB and removed SHAREN since it doesnt
really exist.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-usb
Wolfgang Denk [Wed, 9 Jan 2008 21:41:02 +0000 (22:41 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-usb

16 years agotrivial: fix consequences of a bad merge
Guennadi Liakhovetski [Wed, 9 Jan 2008 00:15:25 +0000 (01:15 +0100)]
trivial: fix consequences of a bad merge

Fix what looks like a merge artifact.

Signed-off-by: Guennadi Liakhovetski <lg@denx.de>
16 years agoAdd Ctrl combo key support to usb keyboard driver.
Zhang Wei [Thu, 3 Jan 2008 02:51:15 +0000 (10:51 +0800)]
Add Ctrl combo key support to usb keyboard driver.

Ctrl combo key support is added, which is very useful to input Ctrl-C
for interrupt current job.
Also add usb_event_poll() calling to usb_kbd_testc(), which can get
key input when tstc() is called.

Signed-off-by: Zhang Wei <wei.zhang@freescale.com>
16 years agofix various comments
Marcel Ziswiler [Sun, 30 Dec 2007 02:30:56 +0000 (03:30 +0100)]
fix various comments

Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
16 years agofix comments with new drivers organization
Marcel Ziswiler [Sun, 30 Dec 2007 02:30:46 +0000 (03:30 +0100)]
fix comments with new drivers organization

Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
16 years agoRemove the obsolete terse version of do_mii()
Shinya Kuribayashi [Fri, 28 Dec 2007 03:50:59 +0000 (12:50 +0900)]
Remove the obsolete terse version of do_mii()

We now have more useful version of do_mii() and everybody use it.
Gerald Van Baren says:

> When I originally wrote the mii command 6(!) years ago, I wrote a
> verbose version that printed human readable decomposition of the flags,
> etc., and a terse one that didn't print as much stuff and thus had a
> smaller memory footprint.
>
> It sounds like the terse version has withered and died, apparently
> people are only using the verbose version (which is very understandable,
> I do myself).

Signed-off-by: Shinya Kuribayashi <shinya.kuribayashi@necel.com>
Signed-off-by: Gerald Van Baren <vanbaren@cideas.com>
16 years agoDo not reference sha1.c when building mkimage.
Mike Frysinger [Thu, 27 Dec 2007 18:42:56 +0000 (13:42 -0500)]
Do not reference sha1.c when building mkimage.

remove sha1.o from mkimage linking since it isn't actually used.

Signed-Off-By: Mike Frysinger <vapier@gentoo.org>
16 years agocommon/cmd_mii.c: Add sanity argc check
Shinya Kuribayashi [Thu, 27 Dec 2007 06:39:54 +0000 (15:39 +0900)]
common/cmd_mii.c: Add sanity argc check

If type mii command without arguments, we suffer from uninitialized argv[]
entries; for example we MIPS get stuck by TLB error.

Signed-off-by: Shinya Kuribayashi <shinya.kuribayashi@necel.com>
16 years agoMerge branch 'master' of /home/wd/git/u-boot/master/
Wolfgang Denk [Wed, 9 Jan 2008 20:34:46 +0000 (21:34 +0100)]
Merge branch 'master' of /home/wd/git/u-boot/master/

16 years agoCoding Style cleanup, update CHANGELOG
Wolfgang Denk [Wed, 9 Jan 2008 14:14:46 +0000 (15:14 +0100)]
Coding Style cleanup, update CHANGELOG

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agofix easylogo on big endian dev systems
Mike Frysinger [Tue, 18 Dec 2007 09:29:55 +0000 (04:29 -0500)]
fix easylogo on big endian dev systems

didnt realize how out of shape easylogo actually was until i tried using it.
this patch does byte swapping as need be on the input tga header since the tga
is in little endian but the host could just as well be big endian.  i didnt
bother using bswap macros or such stuff from system headers as nothing in
POSIX dictates byte swapping functionality.

Signed-Off-By: Mike Frysinger <vapier@gentoo.org>
16 years agocleanup easylogo
Mike Frysinger [Tue, 18 Dec 2007 08:23:25 +0000 (03:23 -0500)]
cleanup easylogo

- make the Makefile not suck
- include proper headers for prototypes
- fix obvious broken handling of strchr() when handling '.' in filenames

Signed-Off-By: Mike Frysinger <vapier@gentoo.org>
16 years agoFix build problems under Cygwin
raptorbrino@aim.com [Fri, 14 Dec 2007 02:23:28 +0000 (21:23 -0500)]
Fix build problems under Cygwin

This patch allows u-boot to build without error in a cygwin
environment.  Cygwin does not define __u64 in it's
include/asm/types.h file.  The -idirafter flag in the u-boot
build causes the inclusion of the cygwin types.h file as opposed
to u-bot/include/asm/types.h file which does define __u64.
Subsequently, sha1.c compile fails due to unknown symbol.

Signed-off-by: Brian Miller <raptorbrino@netscape.net>
16 years agocmd_bmp: Add support for displaying gzip compressed bmps
Hans-Christian Egtvedt [Fri, 30 Nov 2007 16:29:59 +0000 (17:29 +0100)]
cmd_bmp: Add support for displaying gzip compressed bmps

The existing code can show information about a gzip compressed BMP
image, but can't actually display it.

Therefore, move the decompression code out of bmp_info() and use it in
bmp_display() as well in order to display a compressed BMP image.

Also, clean things up a bit and fix a memory leak while we're at it.

[hskinnemoen@atmel.com: a bit of refactoring]
Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoFix and optimize MII operations on FEC (MPC8xx) controllers
Guennadi Liakhovetski [Thu, 29 Nov 2007 20:15:56 +0000 (21:15 +0100)]
Fix and optimize MII operations on FEC (MPC8xx) controllers

This patch fixes several issues at least on a MPC885 based system with two
FEC interfaces used in MII mode.

1. PHY discovery should first read PHY_PHYIDR2 register and only then
   PHY_PHYIDR1 like cpu/mpc8xx/fec.c::mii_discover_phy() does it,
   otherwise the values read are wrong. Also notice, that PHY discovery
   cannot work on MPC88x / MPC87x in setups with both FECs active at all
   in its present form, because for both interfaces the registers from FEC
   1 are used to communicate over MII.

2. Remove code duplication for resetting the FEC by isolating it into a
   separate function.

3. Initialize MII on FEC 1 when communicating over FEC 2 in fec_init().

4. Optimize mii_init() to only reset the FEC 1 controller once.

5. Fix a typo in mii_init() using index i instead of j thus potentially
   leading to unpredictable results.

Signed-off-by: Guennadi Liakhovetski <lg@denx.de>
16 years agoFix endianness conversions in rtl8169 driver
Guennadi Liakhovetski [Tue, 20 Nov 2007 12:14:20 +0000 (13:14 +0100)]
Fix endianness conversions in rtl8169 driver

It is unclear on what platforms this driver has been tested, since
noone up to now defines CONFIG_RTL8169 in the board configuration
header. Now it has been fixed for a big-endian mpc8241 based
linkstation platform. This patch presents the necessary endianness
conversion fixes.

Signed-off-by: Guennadi Liakhovetski <lg@denx.de>
16 years agoAdd Ctrl combo key support to usb keyboard driver.
Zhang Wei [Thu, 3 Jan 2008 02:51:15 +0000 (10:51 +0800)]
Add Ctrl combo key support to usb keyboard driver.

Ctrl combo key support is added, which is very useful to input Ctrl-C
for interrupt current job.
Also add usb_event_poll() calling to usb_kbd_testc(), which can get
key input when tstc() is called.

Signed-off-by: Zhang Wei <wei.zhang@freescale.com>
Signed-off-by: Markus Klotzbuecher <mk@denx.de>
16 years agoMerge git://www.denx.de/git/u-boot
Markus Klotzbuecher [Wed, 9 Jan 2008 12:57:10 +0000 (13:57 +0100)]
Merge git://www.denx.de/git/u-boot

Conflicts:

board/tqm5200/tqm5200.c

16 years agoCoding Style clenaup; update CHANGELOG
Wolfgang Denk [Wed, 9 Jan 2008 12:43:38 +0000 (13:43 +0100)]
Coding Style clenaup; update CHANGELOG

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoRemove bit swapping in Xilinx Spartan bitfile loading
Matthias Fuchs [Thu, 27 Dec 2007 16:13:11 +0000 (17:13 +0100)]
Remove bit swapping in Xilinx Spartan bitfile loading

This patch removes the unnecessary bit swapping when
booting .bit files with the 'fpga loadb' command.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoFix MSB check in Xilinx Spartan slave serial mode
Matthias Fuchs [Thu, 27 Dec 2007 16:13:05 +0000 (17:13 +0100)]
Fix MSB check in Xilinx Spartan slave serial mode

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoAdd new Xilinx Spartan FPGA types
Matthias Fuchs [Thu, 27 Dec 2007 16:12:56 +0000 (17:12 +0100)]
Add new Xilinx Spartan FPGA types

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoAdd pre and post configuration callbacks for Spartan FPGAs
Matthias Fuchs [Thu, 27 Dec 2007 16:12:43 +0000 (17:12 +0100)]
Add pre and post configuration callbacks for Spartan FPGAs

This patch adds a post configuration callback for Spartan2/3 FPGAs.
pre and post configuration callback are now optional and
not called when the function pointer is set to NULL.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoImprove configuration of FPGA subsystem
Matthias Fuchs [Thu, 27 Dec 2007 16:12:34 +0000 (17:12 +0100)]
Improve configuration of FPGA subsystem

This patch removes the FPGA subsystem configuration through
the CONFIG_FPGA bitmask configuration option.

See README for the new options:

CONFIG_FPGA,
CONFIG_FPGA_<vendor>,
CONFIG_FPGA_<family>

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoAdd Epson RX8025 RTC support
Matthias Fuchs [Thu, 27 Dec 2007 15:55:17 +0000 (16:55 +0100)]
Add Epson RX8025 RTC support

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoserial: Make default_serial_console() a weak function
Matthias Fuchs [Thu, 27 Dec 2007 15:57:23 +0000 (16:57 +0100)]
serial: Make default_serial_console() a weak function

With this patch it is possible to reimplement default_serial_console()
in board specific code. This will be done in the upcomming PMC440
U-Boot port. This also allows the lwmon board maintainer to
remove the '#if !defined(CONFIG_LWMON) ...' from common/serial.c.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoadd 'terminal program' functionality
Harald Welte [Wed, 19 Dec 2007 13:14:47 +0000 (14:14 +0100)]
add 'terminal program' functionality

This patch adds a 'cu' like serial terminal command to u-boot
using which you can access other serial ports from the system console.

OpenMoko uses this in their Neo1973 phones to get access to the GSM
Modem and GPS chip from the bootloader.

Signed-off-by: Harald Welte <laforge@openmoko.org>
16 years agoRe-introduce the 'nand read.oob' and 'nand write.oob' commands
Harald Welte [Wed, 19 Dec 2007 13:12:53 +0000 (14:12 +0100)]
Re-introduce the 'nand read.oob' and 'nand write.oob' commands
that used to exist with the legacy NAND code

Signed-off-by: Harald Welte <laforge@openmoko.org>
16 years agoFix building with CRAMFS but not JFFS2 support
Harald Welte [Wed, 19 Dec 2007 13:09:58 +0000 (14:09 +0100)]
Fix building with CRAMFS but not JFFS2 support

Signed-off-by: Harald Welte <laforge@openmoko.org>
16 years agoAllow CONFIG_AUTO_COMPLETE and command history CONFIG_CMDLINE_EDITING at the sametime
Jean-Christophe PLAGNIOL-VILLARD [Sat, 22 Dec 2007 14:52:58 +0000 (15:52 +0100)]
Allow CONFIG_AUTO_COMPLETE and command history CONFIG_CMDLINE_EDITING at the sametime

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoARM: support board-specific ethernet PHY init
Jean-Christophe PLAGNIOL-VILLARD [Tue, 11 Dec 2007 09:53:12 +0000 (10:53 +0100)]
ARM: support board-specific ethernet PHY init

Add until the new phylib will be arrived

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoIXP: Add full baud-rate support for ixp42x, ixp45x and ixp46x
Jean-Christophe PLAGNIOL-VILLARD [Sat, 8 Dec 2007 15:34:08 +0000 (16:34 +0100)]
IXP: Add full baud-rate support for ixp42x, ixp45x and ixp46x

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoAdd missing file in gitignore and comments
Jean-Christophe PLAGNIOL-VILLARD [Sun, 9 Dec 2007 10:01:10 +0000 (11:01 +0100)]
Add missing file in gitignore and comments

based on Linux source tree's .gitignore files

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoCoding Style cleanup, update CHANGELOG
Wolfgang Denk [Wed, 9 Jan 2008 10:36:21 +0000 (11:36 +0100)]
Coding Style cleanup, update CHANGELOG

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-arm
Wolfgang Denk [Wed, 9 Jan 2008 10:30:15 +0000 (11:30 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-arm

16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-fdt
Wolfgang Denk [Wed, 9 Jan 2008 10:28:56 +0000 (11:28 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-fdt

16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-mpc83xx
Wolfgang Denk [Wed, 9 Jan 2008 10:27:02 +0000 (11:27 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-mpc83xx

16 years agoMerge branch 'master' of /home/wd/git/u-boot/master/
Wolfgang Denk [Wed, 9 Jan 2008 10:17:59 +0000 (11:17 +0100)]
Merge branch 'master' of /home/wd/git/u-boot/master/

16 years agoMerge branch 'lwmon5-no-ocm'
Stefan Roese [Wed, 9 Jan 2008 09:43:47 +0000 (10:43 +0100)]
Merge branch 'lwmon5-no-ocm'

16 years agoPOST: Execute SPR test after relocation
Stefan Roese [Wed, 9 Jan 2008 09:38:58 +0000 (10:38 +0100)]
POST: Execute SPR test after relocation

On LWMON5 we now use d-cache as init-ram and stack. The SPR POST test uses
self modifying code and this doesn't work with stack in d-cache, since
I can't move the code from d-cache to i-cache. We move the SPR test to
be executed a little later, after relocation. Then stack is located in
SDRAM and this self-modifying code is no problem anymore.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Change LWMON5 to not use OCM for init-ram and POST anymore
Stefan Roese [Wed, 9 Jan 2008 09:28:20 +0000 (10:28 +0100)]
ppc4xx: Change LWMON5 to not use OCM for init-ram and POST anymore

This patch configures the LWMON5 port to use d-cache as init-ram and
the unused GPT0_COMP6 as POST WORD storage.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Add CFG_POST_ALT_WORD_ADDR to support non OCM POST WORD storage
Stefan Roese [Wed, 9 Jan 2008 09:25:46 +0000 (10:25 +0100)]
ppc4xx: Add CFG_POST_ALT_WORD_ADDR to support non OCM POST WORD storage

The privious 4xx POST implementation only supported storing the POST
WORD in OCM. Since we need to reserve the OCM on LWMON5 for the logbuffer
we need to store the POST WORD in some other non volatile location.
This patch adds CFG_POST_ALT_WORD_ADDR to specify an address for such
a location.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Add 44x cache locking to better support init-ram in d-cache
Stefan Roese [Wed, 9 Jan 2008 09:23:16 +0000 (10:23 +0100)]
ppc4xx: Add 44x cache locking to better support init-ram in d-cache

This patch adds support for locking the init-ram/stack in d-cache,
so that other regions may use d-cache as well

Note, that this current implementation locks exactly 4k of d-cache,
so please make sure that you don't define a bigger init-ram area. Take
a look at the lwmon5 440EPx implementation as a reference.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoFix memset bug in ext2fs_read_file()
Wolfgang Denk [Wed, 9 Jan 2008 09:16:33 +0000 (10:16 +0100)]
Fix memset bug in ext2fs_read_file()

ext2fs_read_file() had the function arguments swapped.

Pointed out by Mike Montour, 19 Dec 2007 22:34:25 -0800

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoFix problems with usb storage devices on MPC5200 /TQM5200
Markus Klotzbücher [Tue, 5 Jan 1988 07:17:15 +0000 (08:17 +0100)]
Fix problems with usb storage devices on MPC5200 /TQM5200

The MPC5200 OHCI controller operates in big endian, so
CFG_OHCI_BE_CONTROLLER must be defined for it to work properly.

Signed-off-by: Markus Klotzbuecher <mk@denx.de>
16 years agoFix compile problem with new env code.
Wolfgang Denk [Tue, 8 Jan 2008 21:58:27 +0000 (22:58 +0100)]
Fix compile problem with new env code.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agotools: fix fw_printenv tool to compile again
Markus Klotzbücher [Tue, 27 Nov 2007 09:23:20 +0000 (10:23 +0100)]
tools: fix fw_printenv tool to compile again

This patch updates the fw_printenv/fw_setenv userspace tool to include
the correct MTD header in order to compile against current kernel
headers. Backward compatibility is preserved by introducing an option
MTD_VERSION which can be set to "old" for compilation using the old MTD
headers. Along with this a number of warnings are fixed.

Signed-off-by: Markus Klotzbuecher <mk@denx.de>
16 years agoppc4xx: assign PCI interrupts on seuqoia boards
Matthias Fuchs [Tue, 8 Jan 2008 14:40:09 +0000 (15:40 +0100)]
ppc4xx: assign PCI interrupts on seuqoia boards

Some operating systems rely on assigned PCI interrupts.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Move cpu/ppc4xx/vecnum.h into include path
Matthias Fuchs [Tue, 8 Jan 2008 14:50:49 +0000 (15:50 +0100)]
ppc4xx: Move cpu/ppc4xx/vecnum.h into include path

This patch allows the use of 4xx interrupt vector number defines
in board specific code outside cpu/ppc4xx.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Fix UIC2 vector number base
Matthias Fuchs [Tue, 8 Jan 2008 14:39:01 +0000 (15:39 +0100)]
ppc4xx: Fix UIC2 vector number base

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Update PLB/PCI divider for PMC440 board
Matthias Fuchs [Tue, 8 Jan 2008 11:49:58 +0000 (12:49 +0100)]
ppc4xx: Update PLB/PCI divider for PMC440 board

This patch updates the PLB/PCI divider when running at
400MHz CPU frequency from 4 to 3 which results in 44MHz PCI sync clock.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Disable error message when no NAND chip is installed on PMC440
Matthias Fuchs [Tue, 8 Jan 2008 10:13:09 +0000 (11:13 +0100)]
ppc4xx: Disable error message when no NAND chip is installed on PMC440

Add CFG_NAND_QUIET_TEST option to disable error message when
no NAND chip is installed on PMC440 boards.

Disable a couple of config defines that are only used for NAND_U_BOOT.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoFix compile problem with new env code.
Wolfgang Denk [Tue, 8 Jan 2008 21:58:27 +0000 (22:58 +0100)]
Fix compile problem with new env code.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agotools: fix fw_printenv tool to compile again
Markus Klotzbücher [Tue, 27 Nov 2007 09:23:20 +0000 (10:23 +0100)]
tools: fix fw_printenv tool to compile again

This patch updates the fw_printenv/fw_setenv userspace tool to include
the correct MTD header in order to compile against current kernel
headers. Backward compatibility is preserved by introducing an option
MTD_VERSION which can be set to "old" for compilation using the old MTD
headers. Along with this a number of warnings are fixed.

Signed-off-by: Markus Klotzbuecher <mk@denx.de>
16 years agoLIBFDT: use memmove() instead of memcpy()
Gerald Van Baren [Tue, 8 Jan 2008 04:47:32 +0000 (23:47 -0500)]
LIBFDT: use memmove() instead of memcpy()

This is partial patch from the DTC/libfdt
commit  67b6b33b9b413a450a72135b5dc59c0a1e33e647
Author: David Gibson <david@gibson.dropbear.id.au>
Date:   Wed Nov 21 11:56:14 2007 +1100

    The patch also fixes one genuine bug caught by valgrind -
    _packblocks() in fdt_rw.c was using memcpy() where it should have been
    using memmove().

Signed-off-by: Gerald Van Baren <vanbaren@cideas.com>
16 years agolibfdt: Add more documentation (patch the seventh)
David Gibson [Mon, 17 Dec 2007 03:42:07 +0000 (14:42 +1100)]
libfdt: Add more documentation (patch the seventh)

This patch adds more documenting comments to libfdt.h.  Specifically,
these document the read/write functions (not including fdt_open_into()
and fdt_pack(), for now).

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
16 years agolibfdt: Add more documentation (patch the sixth)
David Gibson [Mon, 17 Dec 2007 03:41:52 +0000 (14:41 +1100)]
libfdt: Add more documentation (patch the sixth)

This patch adds some more documenting comments to libfdt.h.
Specifically this documents all the write-in-place functions.

Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
16 years agoFine grained per property /chosen updating.
Gerald Van Baren [Sun, 30 Dec 2007 03:45:27 +0000 (22:45 -0500)]
Fine grained per property /chosen updating.

Implement a suggestion by Scott Wood to make the /chosen handling fine
grained.  Don't overwrite pre-existing properties on a per-property basis,
so if /chosen exists but a necessary /chosen/property doesn't, it gets
created.  If a /chosen property exists, it is NOT overwritten unless the
"force" flag is true.

Signed-off-by: Gerald Van Baren <vanbaren@cideas.com>
16 years agoImprove the FDT help message.
Gerald Van Baren [Sat, 5 Jan 2008 20:33:29 +0000 (15:33 -0500)]
Improve the FDT help message.

Add a note that "fdt copy" makes the new address active.
Remove most of the extra hints at the end of the fdt help.

Signed-off-by: Gerald Van Baren <vanbaren@cideas.com>
16 years agoSupport setting FDT properties with optional values.
Gerald Van Baren [Sat, 5 Jan 2008 19:52:04 +0000 (14:52 -0500)]
Support setting FDT properties with optional values.

Fix a bug found and documented by Bartlomiej Sieka where the optional
value on "fdt set <path> <prop> [<val>]" wasn't optional.

=> fdt mknode / testnode
=> fdt print /testnode
testnode {
};
=> fdt set /testnode testprop
=> fdt print /testnode
testnode {
        testprop;
};

Signed-off-by: Gerald Van Baren <vanbaren@cideas.com>
16 years agoAdd fdt_find_and_setprop() to fdt_support.h
Matthias Fuchs [Fri, 28 Dec 2007 10:56:30 +0000 (11:56 +0100)]
Add fdt_find_and_setprop() to fdt_support.h

fdt_find_and_setprop() is used by several 4xx boards and it's
missing in the appropriate header. This patch eliminates a
warning when building U-Boot for such boards.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
Acked-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Return 0 on success in 4xx ethernet driver
Stefan Roese [Tue, 8 Jan 2008 17:39:30 +0000 (18:39 +0100)]
ppc4xx: Return 0 on success in 4xx ethernet driver

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoMerge branch 'inka4x0-ng' of /home/m8/git/u-boot/
Wolfgang Denk [Tue, 8 Jan 2008 16:15:18 +0000 (17:15 +0100)]
Merge branch 'inka4x0-ng' of /home/m8/git/u-boot/

16 years agompc83xx: fix missed pci_hose -> hose conversion for new libfdt code
Kim Phillips [Tue, 8 Jan 2008 15:59:49 +0000 (09:59 -0600)]
mpc83xx: fix missed pci_hose -> hose conversion for new libfdt code

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agompc83xx: rm remaining FLAT_TREE code
Kim Phillips [Thu, 20 Dec 2007 22:28:34 +0000 (16:28 -0600)]
mpc83xx: rm remaining FLAT_TREE code

..in board pci.c files

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agompc83xx: remove FLAT_TREE code
Kim Phillips [Thu, 20 Dec 2007 21:57:28 +0000 (15:57 -0600)]
mpc83xx: remove FLAT_TREE code

need to rm it from pci code, too!

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agompc83xx: convert to using do_fixup_*()
Kim Phillips [Thu, 20 Dec 2007 20:09:22 +0000 (14:09 -0600)]
mpc83xx: convert to using do_fixup_*()

convert to using simpler mpc85xx style fdt update code; streamline by
eliminating macros OF_SOC, OF_CPU, etc. which allows us to rm
the old school FLAT_TREE code from 83xx (since the sbc8349 was just
converted over to using libfdt).

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agosbc8349: enable libfdt by default on WRS SBC8349 board.
Paul Gortmaker [Thu, 20 Dec 2007 17:58:51 +0000 (12:58 -0500)]
sbc8349: enable libfdt by default on WRS SBC8349 board.

Make libfdt the default for the WRS SBC8349 board.
Parallel of commit 35cc4e4823668e8745854899cfaedd4489beb0ef
done for the other 83xx based boards.  Also fix a typo in CONFIG_PCI.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
16 years agosbc8349: migrate board to libfdt
Paul Gortmaker [Thu, 20 Dec 2007 17:58:16 +0000 (12:58 -0500)]
sbc8349: migrate board to libfdt

This adds libfdt support code for the Wind River sbc8349 board.

Parallel of commit 3fde9e8b22cfbd7af489214758f9839a206576cb for
the other Freescale 83xx boards.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
16 years agosbc8349: Remove board specific ECC code
Paul Gortmaker [Thu, 20 Dec 2007 17:56:19 +0000 (12:56 -0500)]
sbc8349: Remove board specific ECC code

ECC code is now shared for all 83xx boards, so remove board specific one.
See commit daab8c67d2defef73dc26ab07f0c3afd1b05d019 for reference.

Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
16 years agompc83xx: Remove CONFIG options related to OF that we dont use (on 837x)
Kim Phillips [Thu, 20 Dec 2007 07:30:48 +0000 (01:30 -0600)]
mpc83xx: Remove CONFIG options related to OF that we dont use (on 837x)

continuation of commit 37395fa2b0d9d617f28d44ca11592260ef16105a to 837x

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agoAdd support CONFIG_UEC_ETH3 in MPC83xx
Joakim Tjernlund [Thu, 6 Dec 2007 15:43:40 +0000 (16:43 +0100)]
Add support CONFIG_UEC_ETH3 in MPC83xx

Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
16 years agoRemove CONFIG options related to OF that we dont use
Kumar Gala [Mon, 26 Nov 2007 17:00:54 +0000 (11:00 -0600)]
Remove CONFIG options related to OF that we dont use

The MPC8360E MDS config defined:
CONFIG_OF_HAS_BD_T
CONFIG_OF_HAS_UBOOT_ENV

Which we don't use or ever needed.  This seems like copy-paste feature creep.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agompc83xx: supress compiler warning
Kim Phillips [Mon, 10 Dec 2007 20:16:22 +0000 (14:16 -0600)]
mpc83xx: supress compiler warning

mpc8360emds.c: In function ‘ft_board_setup’:
mpc8360emds.c:335: warning: assignment discards qualifiers from pointer target type
mpc8360emds.c:345: warning: assignment discards qualifiers from pointer target type

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agompc83xx: fix remaining fdt_find_node_by_path references
Kim Phillips [Tue, 27 Nov 2007 20:17:29 +0000 (14:17 -0600)]
mpc83xx: fix remaining fdt_find_node_by_path references

rename to fdt_path_offset

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agompc83xx: fix CFG_ENV_ADDR and CFG_ENV_SECT_SIZE definitions for 837x
Kim Phillips [Mon, 19 Nov 2007 18:30:09 +0000 (12:30 -0600)]
mpc83xx: fix CFG_ENV_ADDR and CFG_ENV_SECT_SIZE definitions for 837x

Fix the definitions of CFG_ENV_ADDR and CFG_ENV_SECT_SIZE for 837x.
This change guarantees that the environment will be located on the
first flash sector after the U-Boot image.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agompc83xx: mpc8360 rev.2.1 erratum 2: replace rgmii-id with rgmii-rxid
Kim Phillips [Fri, 9 Nov 2007 20:28:08 +0000 (14:28 -0600)]
mpc83xx: mpc8360 rev.2.1 erratum 2: replace rgmii-id with rgmii-rxid

u-boot itself uses GMII mode on the 8360.  Fix up UCC phy-connection-type
properties in the device tree so the PHY gets configured for internal delay on
RX only by the OS, as prescribed by mpc8360 rev. 2.1 pb mds erratum #2.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agompc83xx: update the CREDITS and MAINTAINERS
Dave Liu [Tue, 18 Sep 2007 04:41:15 +0000 (12:41 +0800)]
mpc83xx: update the CREDITS and MAINTAINERS

update the CREDITS and MAINTAINERS.

Signed-off-by: Dave Liu <daveliu@freescale.com>
16 years agompc83xx: add MAINTAINER and MAKEALL entries for the mpc837xemds
Dave Liu [Tue, 18 Sep 2007 04:40:21 +0000 (12:40 +0800)]
mpc83xx: add MAINTAINER and MAKEALL entries for the mpc837xemds

Add the MAINTAINER and MAKEALL entries for mpc837xemds

Signed-off-by: Dave Liu <daveliu@freescale.com>
16 years agompc83xx: Add the MPC837xEMDS board readme
Dave Liu [Tue, 18 Sep 2007 04:38:53 +0000 (12:38 +0800)]
mpc83xx: Add the MPC837xEMDS board readme

Add the README.mpc837xemds to /doc

Signed-off-by: Dave Liu <daveliu@freescale.com>
16 years agompc83xx: Add the support of MPC837xEMDS board
Dave Liu [Tue, 18 Sep 2007 04:37:57 +0000 (12:37 +0800)]
mpc83xx: Add the support of MPC837xEMDS board

The MPC837xEMDS board support:
* DDR2 400MHz hardcoded and SPD init
* Local bus NOR Flash
* I2C, UART, MII and RTC
* eTSEC RGMII
* PCI host

Signed-off-by: Dave Liu <daveliu@freescale.com>
16 years agompc83xx: Add the support of MPC8315E SoC
Dave Liu [Tue, 18 Sep 2007 04:36:58 +0000 (12:36 +0800)]
mpc83xx: Add the support of MPC8315E SoC

The MPC8315E SoC including e300c3 core and new IP blocks,
such as TDM, PCI Express and SATA controller.

Signed-off-by: Dave Liu <daveliu@freescale.com>
16 years agompc83xx: Add the support of MPC837x SoC
Dave Liu [Tue, 18 Sep 2007 04:36:11 +0000 (12:36 +0800)]
mpc83xx: Add the support of MPC837x SoC

The MPC837x SoC including e300c4 core and new IP blocks,
such as SDHC, PCI Express and SATA controller.

Signed-off-by: Dave Liu <daveliu@freescale.com>
16 years agoMPC8360E-MDS: configure and enable second UART
Anton Vorontsov [Wed, 14 Nov 2007 15:54:53 +0000 (18:54 +0300)]
MPC8360E-MDS: configure and enable second UART

Despite user manual, BCSR9.7 is negated (high) on HRST, so
UART2 is disabled. Fix that and configure QE pins properly.

Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
16 years ago83xx: fix CFG_ENV_ADDR and CFG_ENV_SECT_SIZE definitions
Timur Tabi [Mon, 5 Nov 2007 15:34:06 +0000 (09:34 -0600)]
83xx: fix CFG_ENV_ADDR and CFG_ENV_SECT_SIZE definitions

Fix the definitions of CFG_ENV_ADDR and CFG_ENV_SECT_SIZE for all of the
currently-defined 83xx boards.  This change guarantees that the environment
will be located on the first flash sector after the U-Boot image.

Signed-off-by: Timur Tabi <timur@freescale.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-ppc4xx
Wolfgang Denk [Tue, 8 Jan 2008 12:41:42 +0000 (13:41 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-ppc4xx

16 years agoMerge with git://www.denx.de/git/u-boot.git
Peter Pearse [Mon, 7 Jan 2008 15:34:22 +0000 (15:34 +0000)]
Merge with git://www.denx.de/git/u-boot.git

16 years agoppc4xx: Remove weak binding from common Denali data-eye search code
Larry Johnson [Fri, 4 Jan 2008 18:27:02 +0000 (13:27 -0500)]
ppc4xx: Remove weak binding from common Denali data-eye search code

Now that there are no board-specific versions of
"denali_core_search_data_eye()", the weak binding on the common version
can be removed.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoMerge branch 'katmai-ddr-gda'
Stefan Roese [Sat, 5 Jan 2008 09:13:40 +0000 (10:13 +0100)]
Merge branch 'katmai-ddr-gda'

16 years agoppc4xx: Remove unused CONFIG_ECC_ERROR_RESET from 44x_spd_ddr2.c
Stefan Roese [Sat, 5 Jan 2008 08:13:46 +0000 (09:13 +0100)]
ppc4xx: Remove unused CONFIG_ECC_ERROR_RESET from 44x_spd_ddr2.c

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Update Katmai/44x_spd_ddr2.c code for optimal DDR2 setup
Stefan Roese [Sat, 5 Jan 2008 08:12:41 +0000 (09:12 +0100)]
ppc4xx: Update Katmai/44x_spd_ddr2.c code for optimal DDR2 setup

On Katmai the complete auto-calibration somehow doesn't seem to
produce the best results, meaning optimal values for RQFD/RFFD.
This was discovered by GDA using a high bandwidth scope,
analyzing the DDR2 signals. GDA provided a fixed value for RQFD,
so now on Katmai "only" RFFD is auto-calibrated.

This patch also adds RDCC calibration as mentioned on page 7 of
the AMCC PowerPC440SP/SPe DDR2 application note:
"DDR1/DDR2 Initialization Sequence and Dynamic Tuning"

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Remove sdram.h from PMC440 board
Matthias Fuchs [Wed, 2 Jan 2008 15:48:42 +0000 (16:48 +0100)]
ppc4xx: Remove sdram.h from PMC440 board

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: use common denali core defines and data eye search code for PMC440
Matthias Fuchs [Wed, 2 Jan 2008 15:48:34 +0000 (16:48 +0100)]
ppc4xx: use common denali core defines and data eye search code for PMC440

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: More cleanup for esd's LCD code
Matthias Fuchs [Wed, 2 Jan 2008 11:05:14 +0000 (12:05 +0100)]
ppc4xx: More cleanup for esd's LCD code

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Fix Sequoia NAND booting target
Stefan Roese [Fri, 4 Jan 2008 11:00:01 +0000 (12:00 +0100)]
ppc4xx: Fix Sequoia NAND booting target

The Sequoia NAND booting target now uses the recently extracted
cpu/ppc4xx/denali_data_eye.c file too.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Use CFG_4xx_GPIO_TABLE to configure Korat board
Lawrence R. Johnson [Thu, 3 Jan 2008 20:02:02 +0000 (15:02 -0500)]
ppc4xx: Use CFG_4xx_GPIO_TABLE to configure Korat board

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoppc4xx: Use CFG_4xx_GPIO_TABLE to configure Sequoia board
Lawrence R. Johnson [Fri, 4 Jan 2008 07:11:56 +0000 (02:11 -0500)]
ppc4xx: Use CFG_4xx_GPIO_TABLE to configure Sequoia board

Note: this patch changes the configuration of some GPIO registers:

   Register      Old Value   New Value
---------------  ----------  ----------
DCR GPIO0_TCR    0x0000000F  0x0000F0CF
DCR GPIO0_TSRH   0x55005000  0x00000000
DCR GPIO1_TCR    0xC2000000  0xE2000000
DCR GPIO1_TSRL   0x0C000000  0x00200000
DCR GPIO1_ISR2L  0x00050000  0x00110000

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoppc4xx: Add functionality to GPIO support
Lawrence R. Johnson [Thu, 3 Jan 2008 23:54:00 +0000 (18:54 -0500)]
ppc4xx: Add functionality to GPIO support

This patch makes two additions to GPIO support:

First, it adds function gpio_read_in_bit() to read the a bit from the
GPIO Input Register (GPIOx_IR) in the same way that function
gpio_read_out_bit() reads a bit from the GPIO Output Register
(GPIOx_OR).

Second, it modifies function gpio_set_chip_configuration() to provide
an additional option for configuring the GPIO from the
"CFG_4xx_GPIO_TABLE".

According to the 440EPx User's Manual, when an alternate output is used,
the three-state control is configured in one of two ways, depending on
the particular output.  The first option is to select the corresponding
alternate three-state control in the GPIOx_TRSH/L registers.  The second
option is to select the GPIO Three-State Control Register (GPIOx_TCR) in
the GPIOx_TRSH/L registers, and set the corresponding bit in the
GPIOx_TCR register to enable the output.  For example, the Manual
specifies configuring the GPIO00 Alternate 1 Signal (PreAddr07) to use
the alternate three-state control (first option), and specifies
configuring the GPIO32 Alternate 1 Signal (USB2OM0) with the output
enabled in the GPIOx_TCR register (second option).

Currently, gpio_set_chip_configuration() configures all alternate signal
outputs to use the first option.  This patch allow the second option to
be selected by setting the "out_val" element in the table entry to
"GPIO_OUT_1".  The first option is used when the "out_val" element is
set to "GPIO_OUT_0".  Because "out_val" is not currently used when an
alternate signal is selected, and because all current GPIO tables set
"out_val" to "GPIO_OUT_0" for all alternate signals, this patch should
not change any existing configurations.

Signed-off-by: Larry Johnson <lrj@acm.org>