oweals/u-boot.git
16 years agoenvironment: fix bug introduced by commit a8409f4f1ac8
Wolfgang Denk [Wed, 14 May 2008 21:34:53 +0000 (23:34 +0200)]
environment: fix bug introduced by commit a8409f4f1ac8

env_get_char is not a function, but a pointer to one.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-ppc4xx
Wolfgang Denk [Wed, 14 May 2008 12:05:49 +0000 (14:05 +0200)]
Merge branch 'master' of git://denx.de/git/u-boot-ppc4xx

16 years agoFix config files for out-of-tree building
Wolfgang Denk [Wed, 14 May 2008 11:55:30 +0000 (13:55 +0200)]
Fix config files for out-of-tree building

Several board/<...>/config.mk files include dynamically built (by
the Makefile) config files but used the wrong file name of
$(TOPDIR)/board/$(BOARDDIR)/config.tmp
instead if the correct
$(OBJTREE)/board/$(BOARDDIR)/config.tmp

The bug is nasty because the build result is correct for the (normal)
in-tree builds, and because 'sinclude' is used no errors get raised
even for out-of-tree build tests. But out-of-tree builds use an
incomplete and thus usually incorrect configuration...

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoppc4xx: Fix bogus Canyonlands config.mk
Stefan Roese [Wed, 14 May 2008 11:40:03 +0000 (13:40 +0200)]
ppc4xx: Fix bogus Canyonlands config.mk

This patch fixes the canyonlands config.mk file to enable correct
out-of-tree builds. Thanks to Wolfgang Denk for spotting this.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Individual handling of ddr2_fixed.c for canyonlands_nand build
Stefan Roese [Wed, 14 May 2008 08:32:32 +0000 (10:32 +0200)]
ppc4xx: Individual handling of ddr2_fixed.c for canyonlands_nand build

Canyonlands has a file ddr2_fixed.c which needs special treatment when
building in separate directory. It has to be linked to build directory
otherwise it is not seen.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoenvironment: cleanup prototype declarations of env functions.
Wolfgang Denk [Wed, 14 May 2008 10:22:49 +0000 (12:22 +0200)]
environment: cleanup prototype declarations of env functions.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agolinkstation_HGLAN: Fix out of tree building.
Wolfgang Denk [Wed, 14 May 2008 10:21:48 +0000 (12:21 +0200)]
linkstation_HGLAN: Fix out of tree building.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoppc4xx: Individual handling of ddr2_fixed.c for canyonlands_nand build
Stefan Roese [Wed, 14 May 2008 08:32:32 +0000 (10:32 +0200)]
ppc4xx: Individual handling of ddr2_fixed.c for canyonlands_nand build

Canyonlands has a file ddr2_fixed.c which needs special treatment when
building in separate directory. It has to be linked to build directory
otherwise it is not seen.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoMakefile: fix "error: version_autogenerated.h: No such file or directory"
Wolfgang Denk [Tue, 13 May 2008 21:15:52 +0000 (23:15 +0200)]
Makefile: fix "error: version_autogenerated.h: No such file or directory"

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoppc4xx: Add 405EX(r) revision C PVR definitions and detection code
Stefan Roese [Tue, 13 May 2008 18:22:01 +0000 (20:22 +0200)]
ppc4xx: Add 405EX(r) revision C PVR definitions and detection code

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoRevert "pci: Add CONFIG_PCI_SKIP_HOST_BRIDGE config option"
Wolfgang Denk [Sun, 11 May 2008 23:11:21 +0000 (01:11 +0200)]
Revert "pci: Add CONFIG_PCI_SKIP_HOST_BRIDGE config option"

This reverts commit 55774b512fdf63c0516d441cc5da7c54bbffb7f2
which broke many PowerPC boards.

16 years agoMerge branch 'master' of /home/wd/git/u-boot/custodians
Wolfgang Denk [Sun, 11 May 2008 23:02:40 +0000 (01:02 +0200)]
Merge branch 'master' of /home/wd/git/u-boot/custodians

16 years agoRevert "Avoid initrd and logbuffer area overlaps"
Wolfgang Denk [Sun, 11 May 2008 22:56:28 +0000 (00:56 +0200)]
Revert "Avoid initrd and logbuffer area overlaps"

This reverts commit 1b5605ca57fbb364f4d78eeee28b974ed875e888
which breaks building on all PPC boards that don't use a log buffer.

16 years agoFix offset calculation for multi-type legacy images.
Nick Spence [Sat, 10 May 2008 21:02:04 +0000 (14:02 -0700)]
Fix offset calculation for multi-type legacy images.

Calculation of tail was incorrect when size % 4 == 0.

New code removes the conditional and does the same thing but with arithmetic

Signed-off-by: Nick Spence <nick.spence@freescale.com>
16 years agoRevert "Change env_get_char from a global function ptr to a function."
Wolfgang Denk [Sun, 11 May 2008 22:40:58 +0000 (00:40 +0200)]
Revert "Change env_get_char from a global function ptr to a function."

This reverts commit c0559be371b2a64b1a817088c3308688e2182f93
which is known to break booting from dataflash and NAND.

16 years agoAPI: remove duplicate syscall check
Jean-Christophe PLAGNIOL-VILLARD [Sun, 11 May 2008 21:13:57 +0000 (23:13 +0200)]
API: remove duplicate syscall check

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoppc4xx: Kilauea: Add CONFIG_BOOTP_SUBNETMASK to Kilauea board config
Markus Klotzbücher [Thu, 8 May 2008 14:00:55 +0000 (16:00 +0200)]
ppc4xx: Kilauea: Add CONFIG_BOOTP_SUBNETMASK to Kilauea board config

When using dhcp/bootp the "netmask" environment variable is not set
because CONFIG_BOOTP_SUBNETMASK is not defined. But usually this is
desireable, so the following patch adds this this option to the board
config.

Signed-off-by: Markus Klotzbuecher <mk@denx.de>
Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Update Makalu defconfig to use device-tree booting as default
Stefan Roese [Sat, 10 May 2008 08:30:36 +0000 (10:30 +0200)]
ppc4xx: Update Makalu defconfig to use device-tree booting as default

This patch reworks the default environment on Makalu. Now "net_nfs" for
example uses the device-tree style booting formerly know as "net_nfs_fdt".
Also the addresses in RAM were changed because of the new image booting
support, which check for image overwriting. So the addresses needed to
get adjusted.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoPPC: fix map_physmem build warning
Becky Bruce [Wed, 7 May 2008 18:28:16 +0000 (13:28 -0500)]
PPC: fix map_physmem build warning

map_physmem currently generates a warning when CONFIG_PHYS_64BIT is
enabled.  This quiets the warning.

Signed-off-by: Becky Bruce <Becky.Bruce@freescale.com>
16 years agoUpdate pci code to use phys_addr_t
Becky Bruce [Wed, 7 May 2008 18:24:57 +0000 (13:24 -0500)]
Update pci code to use phys_addr_t

Physical addrs need to be represented by phys_addr_t, not
unsigned long.  Otherwise, systems that use CONFIG_PHYS_64BIT
are going to fail mightily.

Signed-off-by: Becky Bruce <becky.bruce@freescale.com>
16 years agoSupport legacy multi-type images without FDT section.
Nick Spence [Fri, 9 May 2008 05:32:22 +0000 (22:32 -0700)]
Support legacy multi-type images without FDT section.

This patch enables legacy multi-type images containing only a Linux kernel
and root file system to be loaded, maintaining compatibility with previous
versions of u-boot.

This is required when using old image files such as a Linux 2.4 kernel /
filesystem.

Signed-off-by: Nick Spence <nick.spence@freescale.com>
Acked-by: Bartlomiej Sieka <tur@semihalf.com>
16 years agoUpdate CHANGELOG.
Wolfgang Denk [Fri, 9 May 2008 22:38:02 +0000 (00:38 +0200)]
Update CHANGELOG.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agopost/cpu/ppc4xx/Makefile: line length cleanup
Wolfgang Denk [Fri, 9 May 2008 22:36:09 +0000 (00:36 +0200)]
post/cpu/ppc4xx/Makefile: line length cleanup

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-mpc83xx
Wolfgang Denk [Fri, 9 May 2008 22:34:05 +0000 (00:34 +0200)]
Merge branch 'master' of git://denx.de/git/u-boot-mpc83xx

16 years agoFix @ -> <at> substitution
Stelian Pop [Thu, 8 May 2008 20:52:09 +0000 (22:52 +0200)]
Fix @ -> <at> substitution

When applying the AT91CAP9 patches upstream, something transformed
the '@' character into the ' <at> ' sequence.

The patch below restores the original form in all the places where
it has been modified (the AT91CAP9 files, the AT91SAM9260 files which
were copied from AT91CAP9, and a couple of other files where the
' <at> ' sequence was present).

Signed-off-by: Stelian Pop <stelian@popies.net>
16 years agoUpdate origin and copyright information in arch-at91sam9 header files
Stelian Pop [Thu, 8 May 2008 20:52:10 +0000 (22:52 +0200)]
Update origin and copyright information in arch-at91sam9 header files

When doing the AT91CAP9/AT91SAM9 port, a number of header files were
copied from the Linux kernel sources. This patch explicitly specifies
this origin for all the copied headers, and for those missing copyright
information, adds it.

Additionaly, the header file 'at91sam926x_mc.h' has been superceeded
in the latest kernel sources by 'at91sam9_smc.h'.

The copyright information has been confirmed by the AT91 Linux kernel
maintainer, Andrew Victor <avictor.za@gmail.com>.

Signed-off-by: Stelian Pop <stelian@popies.net>
16 years agoAdd copyright information in Atmel boards partition.c
Stelian Pop [Thu, 8 May 2008 20:52:11 +0000 (22:52 +0200)]
Add copyright information in Atmel boards partition.c

When Ulf did the dataflash.c cleanup, he didn't add his copyright on
the new created files. This patch fixes the problem.

Signed-off-by: Stelian Pop <stelian@popies.net>
16 years agomx31ads: fix 32kHz clock handling
Guennadi Liakhovetski [Thu, 8 May 2008 08:09:27 +0000 (10:09 +0200)]
mx31ads: fix 32kHz clock handling

According to schematics and to RedBoot sources, the MX31ADS uses a 32768Hz
oscillator as a SKIL source. Fix previously wrongly assumed 32000Hz value.
Also fix a typo when verifying a jumper configuration. While at it, make
two needlessly global functions static.

Signed-off-by: Guennadi Liakhovetski <lg@denx.de>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-mpc85xx
Wolfgang Denk [Fri, 9 May 2008 22:18:22 +0000 (00:18 +0200)]
Merge branch 'master' of git://denx.de/git/u-boot-mpc85xx

16 years agoAvoid initrd and logbuffer area overlaps
Marian Balakowicz [Wed, 7 May 2008 11:10:04 +0000 (13:10 +0200)]
Avoid initrd and logbuffer area overlaps

Add logbuffer to reserved LMB areas to prevent initrd allocation
from overlaping with it.

Make sure to use correct logbuffer base address.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years agoppc: Cleanup get_effective_memsize() use
Marian Balakowicz [Wed, 7 May 2008 11:08:54 +0000 (13:08 +0200)]
ppc: Cleanup get_effective_memsize() use

Removed duplicated effective memory size calculation code.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years agoFix build errors when CONFIG_LOGBUFFER and CONFIG_FIT are enabled
Marian Balakowicz [Wed, 7 May 2008 07:03:53 +0000 (09:03 +0200)]
Fix build errors when CONFIG_LOGBUFFER and CONFIG_FIT are enabled

Recent modifcations to LOGBUFFER handling code were incorrecly
introduced to fit_check_kernel() routine during
"Merge branch 'new-image' of git://www.denx.de/git/u-boot-testing",
commit 27f33e9f45ef7f9685cbdc65066a1828e85dde4f.

This patch cleans up this merge issue.

Signed-off-by: Marian Balakowicz <m8@semihalf.com>
16 years agoPropagate Error Status to the Shell on fw_printenv Errors
Grant Erickson [Wed, 7 May 2008 03:16:15 +0000 (20:16 -0700)]
Propagate Error Status to the Shell on fw_printenv Errors

Changed implementation such that fw_printenv returns failure status
when one or more specified variables do not exist or when incorrect
command syntax is used.

This aids scripting fw_printenv such that the script can key of the
return status rather than relying on standard error "scraping".

Signed-off-by: Grant Erickson <gerickson@nuovations.com>
Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoFix Compilation Errors with 'tools/env/fw_printenv'
Grant Erickson [Tue, 6 May 2008 23:18:00 +0000 (16:18 -0700)]
Fix Compilation Errors with 'tools/env/fw_printenv'

In the current top-of-tree, 1.3.3.-rc2, the optional tool
'tools/env/fw_printenv' fails to compile for two reasons:

1) The header watchdog.h cannot be found.
2) The header zlib.h is picked up from the tool chain rather than the
   project causing a prototype conflict for crc32.

This patch addresses both of these issues.

Platforms Tested On:
- AMCC "Kilauea"

Signed-off-by: Grant Erickson <gerickson@nuovations.com>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-mips
Wolfgang Denk [Fri, 9 May 2008 20:19:29 +0000 (22:19 +0200)]
Merge branch 'master' of git://denx.de/git/u-boot-mips

16 years agoMerge branch 'master' of /home/wd/git/u-boot/master/
Wolfgang Denk [Fri, 9 May 2008 20:19:16 +0000 (22:19 +0200)]
Merge branch 'master' of /home/wd/git/u-boot/master/

16 years agoMerge branch 'master' of /home/wd/git/u-boot/master/
Wolfgang Denk [Fri, 9 May 2008 20:18:58 +0000 (22:18 +0200)]
Merge branch 'master' of /home/wd/git/u-boot/master/

16 years agoFix readline_into_buffer() with CONFIG_CMDLINE_EDITING before relocating
James Yang [Mon, 5 May 2008 15:22:53 +0000 (10:22 -0500)]
Fix readline_into_buffer() with CONFIG_CMDLINE_EDITING before relocating

When CONFIG_CMDLINE_EDITING is enabled, readline_into_buffer() doesn't
work before relocating to RAM because command history is written into
a global array that is not writable before relocation.  This patch
defers to the no-editing and no-history code in readline_into_buffer()
if it is called before relocation.

Signed-off-by: James Yang <James.Yang@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agocosmetic: Adjust coding style for switch statements to be consistent
Detlev Zundel [Mon, 5 May 2008 14:11:22 +0000 (16:11 +0200)]
cosmetic: Adjust coding style for switch statements to be consistent

Signed-off-by: Detlev Zundel <dzu@denx.de>
16 years agoFix disk type output in disk/part.c
Detlev Zundel [Mon, 5 May 2008 14:11:21 +0000 (16:11 +0200)]
Fix disk type output in disk/part.c

Signed-off-by: Detlev Zundel <dzu@denx.de>
16 years agoMail address change, documentation modified
Vlad Lungu [Mon, 5 May 2008 11:20:03 +0000 (14:20 +0300)]
Mail address change, documentation modified

Signed-off-by: Vlad Lungu <vlad.lungu@windrvier.com>
16 years agomicroblaze: Repare intc handling
Michal Simek [Sun, 4 May 2008 13:42:41 +0000 (15:42 +0200)]
microblaze: Repare intc handling

Signed-off-by: Michal Simek <monstr@monstr.eu>
16 years agoinclude/gitignore: update to all architectures
Jean-Christophe PLAGNIOL-VILLARD [Sun, 4 May 2008 13:17:52 +0000 (15:17 +0200)]
include/gitignore: update to all architectures

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoide: Remove spurious second include of io.h
Marcel Ziswiler [Mon, 5 May 2008 00:12:06 +0000 (02:12 +0200)]
ide: Remove spurious second include of io.h

Removed the second include, with all the #ifdef around as suggested by Wolfgang.

Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
16 years agoFix some typos
Adrian Filipi [Tue, 6 May 2008 20:46:37 +0000 (16:46 -0400)]
Fix some typos

This patch fixes three typos.
The first is a repetition of CONFIG_CMD_BSP.
The second makes the #endif comment match its #if.
The third is a spelling error.

Signed-off-by: Adrian Filipi <adrian.filipi@eurotech.com>
16 years agoRecognize 'powerpc' As an Alias for IH_ARCH_PPC
Grant Erickson [Sun, 4 May 2008 23:45:01 +0000 (16:45 -0700)]
Recognize 'powerpc' As an Alias for IH_ARCH_PPC

Add support for the recognition of 'powerpc' as an alias for the PowerPC
architecture type since Linux is already trending in that direction,
preferring 'powerpc' to 'ppc'.

Signed-off-by: Grant Erickson <gerickson@nuovations.com>
16 years ago7450 and 86xx L2 cache invalidate bug corrections
Wheatley Travis [Fri, 2 May 2008 20:35:15 +0000 (13:35 -0700)]
7450 and 86xx L2 cache invalidate bug corrections

The 7610 and related parts have an L2IP bit in the L2CR that is
monitored to signal when the L2 cache invalidate is complete whereas the
7450 and related parts utilize L2I for this purpose. However, the
current code does not account for this difference. Additionally the 86xx
L2 cache invalidate code used an "andi" instruction where an "andis"
instruction should have been used.

This patch addresses both of these bugs.

Signed-off-by: Travis Wheatley <travis.wheatley@freescale.com>
Acked-By: Jon Loeliger <jdl@freescale.com>
16 years agoAvoid infinite loop "Generating include/autoconf.mk"
Wolfgang Denk [Fri, 9 May 2008 08:16:13 +0000 (10:16 +0200)]
Avoid infinite loop "Generating include/autoconf.mk"

Fix a bogus circular dependency that caused an infinite loop of
"Generating include/autoconf.mk" again and again.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoppc4xx: Kilauea: Fix incorrect FPGA FIFO address
Stefan Roese [Thu, 8 May 2008 09:10:46 +0000 (11:10 +0200)]
ppc4xx: Kilauea: Fix incorrect FPGA FIFO address

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Add fdt support to all remaining AMCC PPC4xx eval boards
Stefan Roese [Thu, 8 May 2008 09:05:15 +0000 (11:05 +0200)]
ppc4xx: Add fdt support to all remaining AMCC PPC4xx eval boards

This patch adds fdt (flattened device tree) support to all remaining AMCC
eval boards. Most newer boards already support device tree. With this patch,
all AMCC boards now enable device tree passing from U-Boot to Linux
arch/powerpc kernels.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Add weak default ft_board_setup() routine
Stefan Roese [Thu, 8 May 2008 09:01:09 +0000 (11:01 +0200)]
ppc4xx: Add weak default ft_board_setup() routine

This patch adds a default ft_board_setup() routine to the 4xx fdt code.
This routine is defined as weak and can be overwritten by a board specific
one if needed.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: acadia: Add fdt support and fix section overlap problem
Stefan Roese [Thu, 8 May 2008 08:48:58 +0000 (10:48 +0200)]
ppc4xx: acadia: Add fdt support and fix section overlap problem

This patch adds fdt (flattened device tree) support to the AMCC
Acadia eval board. This increases the image size and it doesn't
fit anymore into 256kByte. Since we didn't want to remove features
from the configuration, we decided to increase the U-Boot image size
(add one flash sector).

Also changed the default environment definition to make it
independent of such changes.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Add device tree support to AMCC Yosemite
Ira Snyder [Tue, 29 Apr 2008 18:18:54 +0000 (11:18 -0700)]
ppc4xx: Add device tree support to AMCC Yosemite

Add support for booting with a device tree blob. This is needed to boot
ARCH=powerpc kernels. Also add support for setting the eth0 mac address
via the ethaddr variable.

Signed-off-by: Ira W. Snyder <iws@ovro.caltech.edu>
Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Fix typos in 460GT/EX FBDV array
Dave Mitchell [Wed, 7 May 2008 16:00:23 +0000 (09:00 -0700)]
ppc4xx: Fix typos in 460GT/EX FBDV array

Corrected two typos in the 460GT/EX FBDV array.

Signed-off-by: Dave Mitchell <dmitchell@amcc.com>
Signed-off-by: Stefan Roese <sr@denx.de>
16 years ago85xx: Limit CPU2 workaround to parts that have the errata
Andy Fleming [Wed, 7 May 2008 21:54:31 +0000 (16:54 -0500)]
85xx: Limit CPU2 workaround to parts that have the errata

Signed-off-by: Ebony Zhu <ebony.zhu@freescale.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
16 years agompc83xx: system performance settings for MPC8349EMDS.
Lee Nipper [Fri, 25 Apr 2008 20:44:45 +0000 (15:44 -0500)]
mpc83xx: system performance settings for MPC8349EMDS.

These same settings are used on MPC8349ITX, and
improve performance on MPC8349EMDS.

Signed-off-by: Lee Nipper <lee.nipper@freescale.com>
Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
16 years ago[MIPS] cpu/mips/cache.S: Fix build warning
Shinya Kuribayashi [Tue, 6 May 2008 04:22:52 +0000 (13:22 +0900)]
[MIPS] cpu/mips/cache.S: Fix build warning

Some old GNU assemblers, such as v2.14 (ELDK 3.1.1), v2.16 (ELDK 4.1.0),
warns illegal global symbol references by bal (and jal also) instruction.
This does not happen with the latest binutils v2.18.

Here's an example on gth2_config:

mips_4KC-gcc  -D__ASSEMBLY__ -g  -Os   -D__KERNEL__ -DTEXT_BASE=0x90000000 -I/home/skuribay/devel/u-boot.git/include -fno-builtin -ffreestanding -nostdinc -isy
stem /opt/eldk311/usr/bin/../lib/gcc-lib/mips-linux/3.3.3/include -pipe  -DCONFIG_MIPS -D__MIPS__ -G 0 -mabicalls -fpic -pipe -msoft-float -march=4kc -mtune=4k
c -EB -c -o cache.o cache.S
cache.S: Assembler messages:
cache.S:243: Warning: Pretending global symbol used as branch target is local.
cache.S:250: Warning: Pretending global symbol used as branch target is local.

In principle, gas might be sensitive to global symbol references in PIC
code because they should be processed through GOT (global offset table).
But if `bal' instruction is used, it results in PC-based offset jump.
This is the cause of this warning.

In practice, we know it doesn't matter whether PC-based reference or GOT-
based. As for this case, both will work before/after relocation. But let's
fix the code.

This patch explicitly sets up a target address, then jump there.
Here's an example of disassembled code with/without this patch.

 90000668:       1485ffef        bne     a0,a1,90000628 <mips_cache_reset+0x20>
 9000066c:       ac80fffc        sw      zero,-4(a0)
 90000670:       01402821        move    a1,t2
-90000674:       0411ffba        bal     90000560 <mips_init_icache>
-90000678:       01803021        move    a2,t4
-9000067c:       01602821        move    a1,t3
-90000680:       0411ffcc        bal     900005b4 <mips_init_dcache>
-90000684:       01a03021        move    a2,t5
-90000688:       03000008        jr      t8
-9000068c:       00000000        nop
+90000674:       01803021        move    a2,t4
+90000678:       8f8f83ec        lw      t7,-31764(gp)
+9000067c:       01e0f809        jalr    t7
+90000680:       00000000        nop
+90000684:       01602821        move    a1,t3
+90000688:       01a03021        move    a2,t5
+9000068c:       8f8f81e0        lw      t7,-32288(gp)
+90000690:       01e0f809        jalr    t7
+90000694:       00000000        nop
+90000698:       03000008        jr      t8
+9000069c:       00000000        nop

Signed-off-by: Shinya Kuribayashi <skuribay@ruby.dti.ne.jp>
16 years agoMerge branch 'master' of git+ssh://10.10.0.7/home/wd/git/u-boot/master
Wolfgang Denk [Mon, 5 May 2008 11:25:04 +0000 (13:25 +0200)]
Merge branch 'master' of git+ssh://10.10.0.7/home/wd/git/u-boot/master

16 years agoAllow building mips versions with ELDK 3.1.1
Vlad Lungu [Mon, 5 May 2008 11:04:00 +0000 (14:04 +0300)]
Allow building mips versions with ELDK 3.1.1

.gpword works only with local symbols on certain binutils versions

Signed-off-by: Vlad Lungu <vlad.lungu@windrvier.com>
16 years agoMAKEALL: add inka4x0 board
Wolfgang Denk [Mon, 5 May 2008 10:52:36 +0000 (12:52 +0200)]
MAKEALL: add inka4x0 board

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agokb9202 board: fix build problem.
Wolfgang Denk [Sun, 4 May 2008 19:34:23 +0000 (21:34 +0200)]
kb9202 board: fix build problem.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoPrepare for v1.3.3-rc3 v1.3.3-rc3
Wolfgang Denk [Sun, 4 May 2008 10:10:33 +0000 (12:10 +0200)]
Prepare for v1.3.3-rc3

Update ChNAGELOG, minor white space cleanup.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of /home/wd/git/u-boot/master/
Wolfgang Denk [Sat, 3 May 2008 23:03:30 +0000 (01:03 +0200)]
Merge branch 'master' of /home/wd/git/u-boot/master/

16 years agodrivers/net/tsec.c: Fix typo.
Wolfgang Denk [Sat, 3 May 2008 22:35:15 +0000 (00:35 +0200)]
drivers/net/tsec.c: Fix typo.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoversion_autogenerated.h: use printf rather than echo -n
Mike Frysinger [Sat, 3 May 2008 01:45:12 +0000 (21:45 -0400)]
version_autogenerated.h: use printf rather than echo -n

Some systems are dumb and do not implement the -n flag to echo (like OS X).
Convert the Makefile to use printf as this should work everywhere.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agofix building when saveenv is disabled in some setups
Mike Frysinger [Fri, 2 May 2008 22:17:50 +0000 (18:17 -0400)]
fix building when saveenv is disabled in some setups

If you enable environment in the flash, but disable the embedded
option, and you disable the saveenv command, then the #if nested
logic will trigger a compile failure:
env_flash.c: In function 'env_relocate_spec':
env_flash.c:399: error: 'flash_addr' undeclared (first use in this function)
The fix is to add CMD_SAVEENV ifdef protection like everywhere else.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
16 years agoSBC8548: fix address mask to allow 64M flash
Jeremy McNicoll [Fri, 2 May 2008 20:10:04 +0000 (16:10 -0400)]
SBC8548: fix address mask to allow 64M flash

Fix incorrect mask to enable all 64MB of onboard flash.
Previously U-Boot incorrectly mapped only 8MB of flash, this
patch correctly maps all the available flash.

Signed-off-by: Jeremy McNicoll <jeremy.mcnicoll@windriver.com>
16 years agoqemu_mips: add README
Jean-Christophe PLAGNIOL-VILLARD [Fri, 2 May 2008 17:48:56 +0000 (19:48 +0200)]
qemu_mips: add README

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-mips
Wolfgang Denk [Sat, 3 May 2008 22:02:29 +0000 (00:02 +0200)]
Merge branch 'master' of git://denx.de/git/u-boot-mips

16 years agoFix misspelled comment
Marcel Ziswiler [Fri, 2 May 2008 00:35:59 +0000 (02:35 +0200)]
Fix misspelled comment

Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
16 years agomkimage: make mmap() checks consistent
Mike Frysinger [Thu, 1 May 2008 08:13:05 +0000 (04:13 -0400)]
mkimage: make mmap() checks consistent

The mmap() related code is full of inconsistent casts/constants when
it comes to error checking, and may break when building on some
systems (like ones that do not implicitly define the caddr_t type).
Let's just avoid the whole mess by writing the code nice and clean in
the first place.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
16 years agoFix defined but not used build warning
Marcel Ziswiler [Thu, 1 May 2008 07:05:34 +0000 (09:05 +0200)]
Fix defined but not used build warning

- warning: 'srom' defined but not used

Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
16 years agoFix implicit declaration build warnings
Marcel Ziswiler [Thu, 1 May 2008 07:05:26 +0000 (09:05 +0200)]
Fix implicit declaration build warnings

- warning: implicit declaration of function ‘serial_initialize’

Signed-off-by: Marcel Ziswiler <marcel@ziswiler.com>
16 years agoTSEC: add config options for VSC8601 RGMII PHY
Andre Schwarz [Tue, 29 Apr 2008 17:18:32 +0000 (19:18 +0200)]
TSEC: add config options for VSC8601 RGMII PHY

The Vitesse VSC8601 RGMII PHY has internal delay for both Rx
and Tx clock lines. They are configured using 2 bits in extended
register 0x17.
Therefore CFG_VSC8601_SKEW_TX and CFG_VSC8601_SKEW_RX have
been introduced with valid values 0-3 giving 0.0, 1.4,1.7 and 2.0ns delay.

Signed-off-by: Andre Schwarz <andre.schwarz@matrix-vision.de>
Acked-by: Andy Fleming <afleming@freescale.com>
Acked-by: Ben Warren <biggerbadderben@gmail.com>
--

 drivers/net/tsec.c |    6 ++++++
 drivers/net/tsec.h |    3 +++
 2 files changed, 9 insertions(+), 0 deletions(-)

16 years agoRevert "ColdFire: Get information from the correct GCC"
Wolfgang Denk [Sat, 3 May 2008 21:07:15 +0000 (23:07 +0200)]
Revert "ColdFire: Get information from the correct GCC"

This reverts commit b7166e05a513c0806b63b9dfb6f1d77645cede2a
(replaced by commit c4e5f52a58d278eebb87f476e353972c5dacea40).

16 years agoconfig.mk: use correct (cross) compiler
Wolfgang Denk [Sat, 3 May 2008 20:25:00 +0000 (22:25 +0200)]
config.mk: use correct (cross) compiler

Some config.mk files reference $(CC) to test for specific tool chain
features, so make sure $(CC) gets set before including any such
config files.

This patch replaces commit b7166e05a5 ("ColdFire: Get information from
the correct GCC").

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agopxa: fix previous definition on cpu init
Jean-Christophe PLAGNIOL-VILLARD [Thu, 1 May 2008 00:13:44 +0000 (02:13 +0200)]
pxa: fix previous definition on cpu init

start.S:183:1: warning: "ICMR" redefined
In file included from start.S:33:
include/asm/arch/pxa-regs.h:935:1: warning: this is the location of the previous definition
start.S:187:1: warning: "RCSR" redefined
...

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agopxa: fix pcmcia operation on 'i' may be undefined
Jean-Christophe PLAGNIOL-VILLARD [Thu, 1 May 2008 00:13:43 +0000 (02:13 +0200)]
pxa: fix pcmcia operation on 'i' may be undefined

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMPC8610HPCD: Drop -O2 from the build flags
Kumar Gala [Wed, 30 Apr 2008 21:24:35 +0000 (16:24 -0500)]
MPC8610HPCD: Drop -O2 from the build flags

Make the flags use -Os like all other boards

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-ppc4xx
Wolfgang Denk [Sat, 3 May 2008 18:46:40 +0000 (20:46 +0200)]
Merge branch 'master' of git://denx.de/git/u-boot-ppc4xx

16 years agoRTC: Fix month offset by one problem in M41T62 RTC driver
Stefan Roese [Wed, 30 Apr 2008 13:50:39 +0000 (15:50 +0200)]
RTC: Fix month offset by one problem in M41T62 RTC driver

This patch fixes a problem with the month being read and written
incorrectly (offset by one). This only gets visible by also using
the Linux driver (rtc-m41t80).

Tested on AMCC Canyonlands.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years ago[MIPS] cpu/mips/config.mk: Fix GNU assembler minor version picker
Shinya Kuribayashi [Sat, 3 May 2008 04:51:44 +0000 (13:51 +0900)]
[MIPS] cpu/mips/config.mk: Fix GNU assembler minor version picker

Current trick to pick up GNU assembler minor version uses a dot(.) as a
delimiter, and take the second field to obtain minor version number. But
as can be expected, this doesn't work with a version string which has
dots more than needs.

Here's an example:

$ mips-linux-gnu-as --version | grep 'GNU assembler'
GNU assembler (Sourcery G++ Lite 4.2-129) 2.18.50.20080215
$ mips-linux-gnu-as --version | grep 'GNU assembler' | cut -d. -f2
2-129) 2
$

This patch restricts the version format to 2.XX.XX... This will work
in most cases.

$ mips-linux-gnu-as --version | grep 'GNU assembler' | egrep -o '2\.[0-9\.]+'
2.18.50.20080215
$ mips-linux-gnu-as --version | grep 'GNU assembler' | egrep -o '2\.[0-9\.]+' | cut -d. -f2
18
$

Signed-off-by: Shinya Kuribayashi <skuribay@ruby.dti.ne.jp>
16 years ago[MIPS] cpu/mips/cache.S: Add dcache_enable
Shinya Kuribayashi [Sat, 3 May 2008 04:51:28 +0000 (13:51 +0900)]
[MIPS] cpu/mips/cache.S: Add dcache_enable

Recent bootelf command fixes (017e9b7925f74878d0e9475388cca9bda5ef9482,
"allow ports to override bootelf behavior") requires ports to have this
function.

Signed-off-by: Shinya Kuribayashi <skuribay@ruby.dti.ne.jp>
Acked-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoMerge branch 'master' of git+ssh://10.10.0.7/home/wd/git/u-boot/master
Wolfgang Denk [Thu, 1 May 2008 19:39:34 +0000 (21:39 +0200)]
Merge branch 'master' of git+ssh://10.10.0.7/home/wd/git/u-boot/master

16 years agoonenand: rename 16 bit memory copy into memcpy_16() to avoid conflicts
Wolfgang Denk [Thu, 1 May 2008 19:30:16 +0000 (21:30 +0200)]
onenand: rename 16 bit memory copy into memcpy_16() to avoid conflicts

Onenand needs a version of memcpy() which performs 16 bit accesses
only; make sure the name does not conflict with the standard
function.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of git+ssh://10.10.0.7/home/wd/git/u-boot/master
Wolfgang Denk [Wed, 30 Apr 2008 21:04:51 +0000 (23:04 +0200)]
Merge branch 'master' of git+ssh://10.10.0.7/home/wd/git/u-boot/master

16 years agocmd_nand: fix warning: str2long ncompatible pointer type
Jean-Christophe PLAGNIOL-VILLARD [Wed, 30 Apr 2008 20:38:17 +0000 (22:38 +0200)]
cmd_nand: fix warning: str2long ncompatible pointer type

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
16 years agoFix calculation of I2C clock for some 86xx chips
Timur Tabi [Fri, 4 Apr 2008 16:16:11 +0000 (11:16 -0500)]
Fix calculation of I2C clock for some 86xx chips

Some 86xx chips use CCB as the base clock for the I2C, and others used CCB/2.
There is no pattern that can be used to determine which chips use which
frequency, so the only way to determine is to look up the actual SOC
designation and use the right value for that SOC.

Signed-off-by: Timur Tabi <timur@freescale.com>
16 years agoColdFire: Fix ethernet hang issue for mcf547x_8x
TsiChung Liew [Wed, 30 Apr 2008 17:11:19 +0000 (12:11 -0500)]
ColdFire: Fix ethernet hang issue for mcf547x_8x

The ethernet hang is caused by receiving buffer in DRAM is not
yet ready due to access cycles require longer time in DRAM.
Relocate DMA buffer descriptors from DRAM to internal SRAM.

Signed-off-by: TsiChung Liew <Tsi-Chung.Liew@freescale.com>
16 years agoColdFire: Fix compilation issue caused by new changes in fsl_i2c.c
TsiChung Liew [Wed, 30 Apr 2008 17:10:47 +0000 (12:10 -0500)]
ColdFire: Fix compilation issue caused by new changes in fsl_i2c.c

Signed-off-by: Luigi Comio Mantellini <luigi.mantellini@idf-hit.com>
Signed-off-by: TsiChung Liew <Tsi-Chung.Liew@freescale.com>
16 years agoColdFire: Get information from the correct GCC
TsiChung Liew [Wed, 30 Apr 2008 17:10:23 +0000 (12:10 -0500)]
ColdFire: Get information from the correct GCC

Signed-off-by: Kurt Mahan <kmahan@freescale.com>
Signed-off-by: TsiChung Liew <Tsi-Chung.Liew@freescale.com>
16 years agoFix warning in env_nand.c if compiled for DaVinci Schmoogie
dirk.behme@googlemail.com [Wed, 30 Apr 2008 16:02:59 +0000 (18:02 +0200)]
Fix warning in env_nand.c if compiled for DaVinci Schmoogie

Fix warnings

nv_nand.c: In function 'saveenv':
env_nand.c:200: warning: passing argument 3 of 'nand_write' from incompatible pointer type
env_nand.c: In function 'env_relocate_spec':
env_nand.c:275: warning: passing argument 3 of 'nand_read' from incompatible pointer type

if compiled for davinci_schmoogie_config.

Signed-off-by: Dirk Behme <dirk.behme@gmail.com>
Ack by: Sergey Kubushyn <ksi@koi8.net>

16 years agoFix warnings while compiling net/net.c for MPC8610HPCD board
Anatolij Gustschin [Wed, 30 Apr 2008 11:34:40 +0000 (13:34 +0200)]
Fix warnings while compiling net/net.c for MPC8610HPCD board

MPC8610HPCD board adds -O2 gcc option to PLATFORM_CPPFLAGS
causing overriding default -Os option. New gcc (ver. 4.2.2)
produces warnings while compiling net/net.c file with -O2
option. The patch is an attempt to fix this.

Signed-off-by: Anatolij Gustschin <agust@denx.de>
16 years agolwmon5: fix offset error in sysmon0 POST
Sascha Laue [Wed, 30 Apr 2008 13:23:38 +0000 (15:23 +0200)]
lwmon5: fix offset error in sysmon0 POST

Signed-off-by: Sascha Laue <sascha.laue@liebherr.com>
Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agolwmon5: fix manual merge error in POST
Sascha Laue [Wed, 30 Apr 2008 13:16:35 +0000 (15:16 +0200)]
lwmon5: fix manual merge error in POST

Signed-off-by: Sascha Laue <sascha.laue@liebherr.com>
16 years agocmd_nand.c: fix another 'incompatible pointer type' warning.
Wolfgang Denk [Wed, 30 Apr 2008 15:46:26 +0000 (17:46 +0200)]
cmd_nand.c: fix another 'incompatible pointer type' warning.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMakefile: fix parallel builds
Wolfgang Denk [Wed, 30 Apr 2008 15:25:07 +0000 (17:25 +0200)]
Makefile: fix parallel builds

This problem shows up with parallel builds only; it results in
somewhat cryptic error messages like

$ JOBS=-j6 MAKEALL netstar
Configuring for netstar board...
arm-linux-ld: cannot find -lgeneric
make[1]: *** [eeprom.srec] Error 1

A few boards (like netstar and voiceblue) need some libraries for
building; however, the board Makefile does not contain any such
dependencies which may cause problems with parallel builds. Adding
such dependencies is difficult as we would also have to provide build
rules, which already exist in the respective library Makefiles.

To solve this, we make sure that all libraries get built before the
board code.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoppc4xx: Adapt Canyonlands fixed DDR2 setup to new DIMM module
Stefan Roese [Wed, 30 Apr 2008 12:51:36 +0000 (14:51 +0200)]
ppc4xx: Adapt Canyonlands fixed DDR2 setup to new DIMM module

This patch changes the Canyonlands/Glacier fixed DDR2 controller setup
used for NAND booting to match the values needed for the new 512MB
DIMM modules shipped with the productions boards:

Crucial: CT6464AC667.8FB

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Fix problem with DIMMs with 8 banks in 44x_spd_ddr2.c
Stefan Roese [Wed, 30 Apr 2008 08:49:43 +0000 (10:49 +0200)]
ppc4xx: Fix problem with DIMMs with 8 banks in 44x_spd_ddr2.c

This patch fixes a problem with DIMMs that have 8 banks. Now the
MCIF0_MBxCF register will be setup correctly for this setup too.

This was noticed with the 512MB DIMM on Canyonlands/Glacier.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoPrepare v1.3.3-rc2, again. v1.3.3-rc2
Wolfgang Denk [Tue, 29 Apr 2008 21:41:06 +0000 (23:41 +0200)]
Prepare v1.3.3-rc2, again.

Signed-off-by: Wolfgang Denk <wd@denx.de>