oweals/u-boot.git
16 years agonet/eth.c: Fix env_enetaddr signed overflow
Shinya Kuribayashi [Mon, 19 Nov 2007 11:27:04 +0000 (20:27 +0900)]
net/eth.c: Fix env_enetaddr signed overflow

Assigning the output of simple_strtoul(CB:A9:87:65:43:21) to `char', we are
warned as below:

  U-Boot 1.2.0 (Aug 30 2007 - 08:27:37)

  DRAM:  256 MB
  Flash: 32 MB
  In:    serial
  Out:   serial
  Err:   serial
  Net:   NEC-Candy
  Warning: NEC-Candy MAC addresses don't match:
  Address in SROM is         00:00:4C:80:92:A2
  Address in environment is  FFFFFFCB:FFFFFFA9:FFFFFF87:65:43:21

This patch changes env_enetaddr type from `char' to `unsigned char'.

Cc: Masaki Ishikawa <ishikawa-masaki@cnt.mxe.nes.nec.co.jp>
Signed-off-by: Shinya Kuribayashi <shinya.kuribayashi@necel.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
16 years agoIntroduce new eth_receive routine
Rafal Jaworowski [Thu, 27 Dec 2007 17:19:02 +0000 (18:19 +0100)]
Introduce new eth_receive routine

The purpose of this routine is receiving a single network frame, outside of
U-Boot's NetLoop(). Exporting it to standalone programs that run on top of
U-Boot will let them utilise networking facilities. For sending a raw frame
the already existing eth_send() can be used.

The direct consumer of this routine is the newly introduced API layer for
external applications (enabled with CONFIG_API).

Signed-off-by: Rafal Jaworowski <raj@semihalf.com>
Signed-off-by: Piotr Kruszynski <ppk@semihalf.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
16 years ago8610: Move include of config.h earlier.
Jon Loeliger [Thu, 3 Jan 2008 16:41:04 +0000 (10:41 -0600)]
8610: Move include of config.h earlier.

Include config.h earlier in the set of #includes
so as to avoid a incidental and duplicate definition
of CFG_CACHELINE_SIZE.

Signed-off-by: Jon Loeliger
16 years agoDon't slam #undef DEBUG in the 8641HPCN config file.
Jon Loeliger [Tue, 4 Dec 2007 16:53:34 +0000 (10:53 -0600)]
Don't slam #undef DEBUG in the 8641HPCN config file.

Doing so prevents it from being individually set
and useful in other files.

Signed-off-by: Jon Loeliger <jdl@freescale.com>
16 years agoConvert MPC8641HPCN to use libfdt.
Jon Loeliger [Wed, 28 Nov 2007 20:47:18 +0000 (14:47 -0600)]
Convert MPC8641HPCN to use libfdt.

Assumes the presence of the aliases node in the DTS to
locate the ethernet, pci and serial nodes for fixups.

Use consistent fdtaddr and fdtfile in environment variables.

Signed-off-by: Jon Loeliger <jdl@freescale.com>
16 years agoMerge commit 'wd/master'
Jon Loeliger [Thu, 3 Jan 2008 15:46:55 +0000 (09:46 -0600)]
Merge commit 'wd/master'

16 years agocfi_flash: Fix bug in flash_isset() to use correct 32bit function
Stefan Roese [Wed, 2 Jan 2008 13:06:26 +0000 (14:06 +0100)]
cfi_flash: Fix bug in flash_isset() to use correct 32bit function

This bug was detected on the LWMON5 target which has 2 Intel 16bit wide
flash chips connected to a 32bit wide port.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoFix compile problem introduced by "cleanup" commit 3dfd708c
Wolfgang Denk [Wed, 2 Jan 2008 14:58:44 +0000 (15:58 +0100)]
Fix compile problem introduced by "cleanup" commit 3dfd708c

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMake scripts and Makefiles POSIX compliant
Wolfgang Denk [Wed, 2 Jan 2008 14:54:45 +0000 (15:54 +0100)]
Make scripts and Makefiles POSIX compliant

The bash builtin versions of the "test" (resp. "[") command allow
using "==" for string comparisons, but POSIX compatible implemen-
tations (like /usr/bin/test) insist on using "=" only. On such systems
you will see:

$ /usr/bin/test a == a && echo OK
/usr/bin/test: ==: binary operator expected

This patch fixes Makefiles and scripts to use POSIX style.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of /home/wd/git/u-boot/custodians
Wolfgang Denk [Wed, 2 Jan 2008 11:39:04 +0000 (12:39 +0100)]
Merge branch 'master' of /home/wd/git/u-boot/custodians

16 years agoMinor coding style cleanup.
Wolfgang Denk [Wed, 2 Jan 2008 11:38:43 +0000 (12:38 +0100)]
Minor coding style cleanup.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-cfi-flash
Wolfgang Denk [Wed, 2 Jan 2008 10:30:58 +0000 (11:30 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-cfi-flash

16 years agoMerge branch 'for-1.3.2-ver2'
Stefan Roese [Sat, 29 Dec 2007 08:23:11 +0000 (09:23 +0100)]
Merge branch 'for-1.3.2-ver2'

Conflicts:

cpu/ppc4xx/fdt.c
include/configs/kilauea.h
include/configs/sequoia.h

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Coding style cleanup
Stefan Roese [Fri, 28 Dec 2007 16:29:56 +0000 (17:29 +0100)]
ppc4xx: Coding style cleanup

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Complete PMC440 board support
Matthias Fuchs [Fri, 28 Dec 2007 16:07:24 +0000 (17:07 +0100)]
ppc4xx: Complete PMC440 board support

This patch brings the PMC440 board configuration file.
Finally it enables the PMC440 board support.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Add FPGA support and BSP commands for PMC440 boards
Matthias Fuchs [Fri, 28 Dec 2007 16:07:18 +0000 (17:07 +0100)]
ppc4xx: Add FPGA support and BSP commands for PMC440 boards

This patch adds some BSP commands and FPGA booting support
for esd's PMC440 boards.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Add initial esd PMC440 board files
Matthias Fuchs [Fri, 28 Dec 2007 16:07:14 +0000 (17:07 +0100)]
ppc4xx: Add initial esd PMC440 board files

This patch adds the first files for the new esd PMC440 boards.
The next two patches will complete the PMC440 board support.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Add EEPROM write protection for PLU405 boards + misc. updates
Matthias Fuchs [Fri, 28 Dec 2007 16:10:36 +0000 (17:10 +0100)]
ppc4xx: Add EEPROM write protection for PLU405 boards + misc. updates

- add EEPROM write protection for esd PLU405 boards.
- initialize NAND GPIOs
- use correct io accessors
- cleanup

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: use correct io accessors for esd's LCD code
Matthias Fuchs [Fri, 28 Dec 2007 16:10:44 +0000 (17:10 +0100)]
ppc4xx: use correct io accessors for esd's LCD code

This patch fixes esd's LCD dectection code to work correctly with
newer gcc versions.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Maintenance patch for VOH405 boards
Matthias Fuchs [Fri, 28 Dec 2007 16:10:42 +0000 (17:10 +0100)]
ppc4xx: Maintenance patch for VOH405 boards

- add EEPROM write protection
- initialize NAND GPIOs
- use correct io accessors
- slow down I2C clock to 100kHz
- enable ext. I2C bus
- cleanup

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Enable 405EP PCI arbiter per default on all boards
Stefan Roese [Fri, 28 Dec 2007 15:08:08 +0000 (16:08 +0100)]
ppc4xx: Enable 405EP PCI arbiter per default on all boards

In an attmemt to clean up the 4xx start.S file, I removed the enabling
of the internal 405EP PCI arbiter. This is needed for multiple other
405EP platforms, like most of the esd 405EP. Now the internal PCI
arbiter is enabled again per default as it has been before.

Signed-off-by: Stefan Roese <sr@denx.de>
Acked-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Fix bug in cpu_init.c (405EP instead of 450EP)
Stefan Roese [Fri, 28 Dec 2007 14:53:46 +0000 (15:53 +0100)]
ppc4xx: Fix bug in cpu_init.c (405EP instead of 450EP)

Signed-off-by: Stefan Roese <sr@denx.de>
Acked-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: Fix compilation problem of kilauea/haleakala nand booting target
Stefan Roese [Fri, 28 Dec 2007 05:06:04 +0000 (06:06 +0100)]
ppc4xx: Fix compilation problem of kilauea/haleakala nand booting target

Use correct link to nand_ecc now located in drivers/mtd/nand/ for the
platforms mentioned above.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoMerge branch 'master' of /home/stefan/git/u-boot/u-boot into for-1.3.2-ver2
Stefan Roese [Thu, 27 Dec 2007 18:37:26 +0000 (19:37 +0100)]
Merge branch 'master' of /home/stefan/git/u-boot/u-boot into for-1.3.2-ver2

16 years agoRemove CPCI440 board
Matthias Fuchs [Thu, 27 Dec 2007 16:03:46 +0000 (17:03 +0100)]
Remove CPCI440 board

This board never left prototyping state and it
became a millstone round my neck. So remove it.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoAdd support for Korat PPC440EPx board
Larry Johnson [Thu, 27 Dec 2007 16:28:51 +0000 (11:28 -0500)]
Add support for Korat PPC440EPx board

These patches add support for the PPC440EPx-based "Korat" board to
U-Boot.  They are based primarily on support for the Sequoia board.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoAdd configuration file for Korat board
Larry Johnson [Sat, 22 Dec 2007 20:16:25 +0000 (15:16 -0500)]
Add configuration file for Korat board

This patch supplies the configuration file for the Korat PPC440EPx-
processor board.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoAdd denali_data_eye.o and denali_spd_ddr2.o to PPC4xx Makefile
Larry Johnson [Sat, 22 Dec 2007 20:16:11 +0000 (15:16 -0500)]
Add denali_data_eye.o and denali_spd_ddr2.o to PPC4xx Makefile

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoAdd 440EPx DDR2 SPD DIMM support
Larry Johnson [Thu, 27 Dec 2007 15:54:48 +0000 (10:54 -0500)]
Add 440EPx DDR2 SPD DIMM support

This patch adds SPD DDR2 support for the 440EPx ("Denali") SDRAM
controller.  It should also work on the 440GRx.  It is based on the DDR2
SPD code for the 440EP/440EPx, but makes no provision for DDR1 support.

This code has been tested on prototype Korat boards with three Kingston
DIMMS: 512 MiB ECC (one rank), 512 MiB non-ECC (one rank) and 1 GiB ECC
(two ranks).  The Korat board has a single DIMM socket, but support has
been provided (though not tested) for boards with two DIMM sockets.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoCopy 440EPx/GRx SDRAM data-eye search to common directory
Larry Johnson [Sat, 22 Dec 2007 20:15:30 +0000 (15:15 -0500)]
Copy 440EPx/GRx SDRAM data-eye search to common directory

This patch creates a non-board-specific file for performing the SDRAM
data-eye search.  It also adds ECC error checking to the test of valid
data on readback when ECC is enabled.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoAdd definitions for 440EPx/GRx SDRAM controller to ppc440.h
Larry Johnson [Sat, 22 Dec 2007 20:15:13 +0000 (15:15 -0500)]
Add definitions for 440EPx/GRx SDRAM controller to ppc440.h

This patch adds the Denali SDRAM controller definitions to "ppc440.h".
It also fixes two typos in the definitions, so the board-specific
"sdram.h" files containing these definitions are also fixed to avoid
compiler warnings.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoAdd Ethernet 1000BASE-X support for PPC4xx
Larry Johnson [Thu, 27 Dec 2007 15:50:55 +0000 (10:50 -0500)]
Add Ethernet 1000BASE-X support for PPC4xx

This patch adds a new switch: "CONFIG_PHY_DYNAMIC_ANEG".  When this symbol
is defined, the PHY will advertise it's capabilities for autonegotiation
based on the capabilities shown in the PHY's status registers, including
1000BASE-X.  When "CONFIG_PHY_DYNAMIC_ANEG" is not defined, the PHY will
advertise hard-coded capabilities, as before.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoAdd driver for National Semiconductor LM73 temperature sensor
Larry Johnson [Thu, 27 Dec 2007 14:52:17 +0000 (09:52 -0500)]
Add driver for National Semiconductor LM73 temperature sensor

This driver is based on the driver for the LM75.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoAdd driver for STMicroelectronics M41T60 RTC
Larry Johnson [Sat, 22 Dec 2007 20:14:00 +0000 (15:14 -0500)]
Add driver for STMicroelectronics M41T60 RTC

This driver is based on the driver for the M41T11.  In the intended
application, the RTC will be powered by a large capacitor, rather than a
battery.  The driver therefore checks to see whether the RTC has lost
power.  The chip's OUT bit is normally reset from its power-up state.  If
the OUT bit is read as set, or if the date and time are not valid, then the
RTC is assumed to have lost power, and its date and time are reset to
1900-01-01 00:00:00.

Support for adjusting the speed of the clock to improve accuracy is
provided through an environment variable.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoUse out_be32() and friends to access memory-mapped registers in sequoia.c
Larry Johnson [Sat, 22 Dec 2007 20:34:39 +0000 (15:34 -0500)]
Use out_be32() and friends to access memory-mapped registers in sequoia.c

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoUse definitions from "asm-ppc/mmu.h" in init.S for Sequoia
Larry Johnson [Sat, 22 Dec 2007 20:34:20 +0000 (15:34 -0500)]
Use definitions from "asm-ppc/mmu.h" in init.S for Sequoia

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoCosmetic changes to ECC POST for AMCC Denali core
Larry Johnson [Sat, 22 Dec 2007 20:23:50 +0000 (15:23 -0500)]
Cosmetic changes to ECC POST for AMCC Denali core

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoppc4xx: Fix compilation problem in 405 cache POST test
Stefan Roese [Wed, 26 Dec 2007 19:20:19 +0000 (20:20 +0100)]
ppc4xx: Fix compilation problem in 405 cache POST test

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Move virtual address of POST cache test to bigger address
Stefan Roese [Sat, 22 Dec 2007 11:20:09 +0000 (12:20 +0100)]
ppc4xx: Move virtual address of POST cache test to bigger address

On Sequoia & LWMON5 the virtual address of the POST cache test is now
moved to a bigger address. This enables usage of more memory on those
boards.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Fix problem in 44x cache POST routine
Stefan Roese [Sat, 22 Dec 2007 11:18:26 +0000 (12:18 +0100)]
ppc4xx: Fix problem in 44x cache POST routine

As repoted by Larry Johnson, running "diag run cache" caused a crash
in U-Boot. This problem was introduced by a patch that removed the
TLB entry for the cache test after the test has completed. Since this
TLB was only setup once, a 2nd attempt to run this cache test
failed with a crash. Now this TLB entry is created every time the
routine is called.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Update Makalu fdt support
Stefan Roese [Fri, 21 Dec 2007 06:51:29 +0000 (07:51 +0100)]
ppc4xx: Update Makalu fdt support

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Add fdt support to AMCC Katmai eval board
Stefan Roese [Wed, 19 Dec 2007 08:05:40 +0000 (09:05 +0100)]
ppc4xx: Add fdt support to AMCC Katmai eval board

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: fdt: Cleanup setup of cpu node setup
Stefan Roese [Tue, 18 Dec 2007 07:44:51 +0000 (08:44 +0100)]
ppc4xx: fdt: Cleanup setup of cpu node setup

Now the cpu node setup ("timebase-frequency" and "clock-frequency") is
without using the absolute path to the cpu node. This makes it possible
to use this U-Boot version with both versions of cpu-node naming
"cpu@0" and the former "PowerPC,440EPx@0".

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Fix lwmon5 compilation problem
Stefan Roese [Mon, 17 Dec 2007 16:26:21 +0000 (17:26 +0100)]
ppc4xx: Fix lwmon5 compilation problem

Now that the 440EPx ECC test is not board specific anymore
remove this Makefile.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoFix ppc4xx clear_bss() code
Anatolij Gustschin [Wed, 5 Dec 2007 16:43:20 +0000 (17:43 +0100)]
Fix ppc4xx clear_bss() code

ppc4xx clear_bss() fails if BSS segment size is not
divisible by 4 without remainder. This patch provides
fix for this problem.

Signed-off-by: Anatolij Gustschin <agust@denx.de>
16 years agoPPC4xx: Minimal changes to add vxWorks support
Niklaus Giger [Fri, 30 Nov 2007 17:35:11 +0000 (18:35 +0100)]
PPC4xx: Minimal changes to add vxWorks support

Signed-off-by: Niklaus Giger <niklaus.giger@netstal.com>
16 years agoppc4xx: Add CONFIG_BOOTP_SUBNETMASK to Sequoia board config
Markus Klotzbücher [Fri, 23 Nov 2007 12:09:18 +0000 (13:09 +0100)]
ppc4xx: Add CONFIG_BOOTP_SUBNETMASK to Sequoia 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 agoFix/enhance ECC POST for 440EPx/GRx
Larry Johnson [Sat, 27 Oct 2007 16:48:15 +0000 (12:48 -0400)]
Fix/enhance ECC POST for 440EPx/GRx

This patch allows the ECC POST to be used for different boards with the
PPC440 Denali SDRAM controller.  Modifications include skipping the test
if ECC is not enabled (as for non-ECC DIMMs) and adding synchronization
to prevent timing errors.

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoPPC4xx: Move/rename ECC POST for 440EPx/GRx
Larry Johnson [Sat, 27 Oct 2007 16:48:05 +0000 (12:48 -0400)]
PPC4xx: Move/rename ECC POST for 440EPx/GRx

Signed-off-by: Larry Johnson <lrj@acm.org>
16 years agoppc4xx: use correct io accessors for 4xx ethernet POST
Matthias Fuchs [Fri, 14 Dec 2007 10:20:33 +0000 (11:20 +0100)]
ppc4xx: use correct io accessors for 4xx ethernet POST

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: fix flush + invalidate_dcache_range arguments
Matthias Fuchs [Fri, 14 Dec 2007 10:19:56 +0000 (11:19 +0100)]
ppc4xx: fix flush + invalidate_dcache_range arguments

flush + invalidate_dcache_range() expect the start and stop+1 address.
So the stop address is the first address behind (!) the range.

Signed-off-by: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
16 years agoppc4xx: fdt: use fdt_fixup_ethernet()
Stefan Roese [Fri, 14 Dec 2007 07:41:29 +0000 (08:41 +0100)]
ppc4xx: fdt: use fdt_fixup_ethernet()

By using aliases in the dts file, the ethernet node fixup is
much easier with the recently added functions.

Please note that the dts file needs the aliases for this to work.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoppc4xx: Bring 4xx fdt support up-to-date
Stefan Roese [Thu, 13 Dec 2007 13:52:53 +0000 (14:52 +0100)]
ppc4xx: Bring 4xx fdt support up-to-date

This patch update the 4xx fdt support. It enabled fdt booting
on the AMCC Kilauea and Sequoia for now. More can follow later
quite easily.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agocfi_flash: Add missing check for erased dest to flash_write_cfibuffer()
Stefan Roese [Thu, 27 Dec 2007 06:50:54 +0000 (07:50 +0100)]
cfi_flash: Add missing check for erased dest to flash_write_cfibuffer()

The check for an sufficiently erased destination was missing in the
buffered write function of the cfi flash driver (when
CFG_FLASH_USE_BUFFER_WRITE is defined). This patch adds this check to that
writing to such a region will fail with the currect error message.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoMerge branch 'master' of /home/wd/git/u-boot/custodians
Wolfgang Denk [Thu, 27 Dec 2007 01:01:15 +0000 (02:01 +0100)]
Merge branch 'master' of /home/wd/git/u-boot/custodians

16 years agoSome configuration updates for the TQM5200 based TB5200 board:
Martin Krause [Mon, 12 Nov 2007 09:56:17 +0000 (10:56 +0100)]
Some configuration updates for the TQM5200 based TB5200 board:

- enable command line history
- increase malloc space (because of bigger flash sectors)

Signed-off-by: Martin Krause <martin.krause@tqs.de>
16 years agoTQM8xx: use the CFI flash driver on all TQM8xx boards
Martin Krause [Thu, 27 Sep 2007 09:10:08 +0000 (11:10 +0200)]
TQM8xx: use the CFI flash driver on all TQM8xx boards

Signed-off-by: Martin Krause <martin.krause@tqs.de>
16 years agoTQM885D: adjust for doubled flash sector size + some minor fixes
Martin Krause [Wed, 26 Sep 2007 15:55:56 +0000 (17:55 +0200)]
TQM885D: adjust for doubled flash sector size + some minor fixes

Signed-off-by: Martin Krause <martin.krause@tqs.de>
16 years agoTQM885D: Exchanged SDRAM timing by a more relaxed timing.
Jens Gehrlein [Wed, 26 Sep 2007 15:55:54 +0000 (17:55 +0200)]
TQM885D: Exchanged SDRAM timing by a more relaxed timing.

CAS-Latency=2, Write Recovery Time tWR=2
The max. supported bus frequency is 66 MHz. Therefore, changed
threshold to switch from 1:1 mode to 2:1 from 80 MHz to 66 MHz.

Signed-off-by: Martin Krause <martin.krause@tqs.de>
16 years agoTQM885D: use calculated cpuclk instead of measuring it
Martin Krause [Wed, 26 Sep 2007 15:55:56 +0000 (17:55 +0200)]
TQM885D: use calculated cpuclk instead of measuring it

On the TQM885D the measurement of cpuclk with the PIT reference
timer ist not necessary. Since all module variants use the same
external 10 MHz oscillator, the cpuclk only depends on the PLL
configuration - which is readable by software.

Signed-off-by: Martin Krause <martin.krause@tqs.de>
16 years agoTQM885D: fix SDRAM refresh
Jens Gehrlein [Thu, 27 Sep 2007 12:54:46 +0000 (14:54 +0200)]
TQM885D: fix SDRAM refresh

At 133 MHz the current SDRAM refresh rate is too fast
(measured 4 * 1.17 us).
CFG_MAMR_PTA changes from 39 to 128. This result
in a refresh rate of 4 * 7.8 us at the default clock
66 MHz. At 133 MHz the value will be then 4 * 3.8 us.
This is a compromise until a new method is found to
adjust the refresh rate.

Signed-off-by: Martin Krause <martin.krause@tqs.de>
16 years agoTQM860M: Support for 10col SDRAMs, max. 128 MiB
Jens Gehrlein [Thu, 27 Sep 2007 12:54:46 +0000 (14:54 +0200)]
TQM860M: Support for 10col SDRAMs, max. 128 MiB

Signed-off-by: Martin Krause <martin.krause@tqs.de>
16 years agoFix coding style issues; update CHANGELOG.
Wolfgang Denk [Thu, 27 Dec 2007 00:52:50 +0000 (01:52 +0100)]
Fix coding style issues; update CHANGELOG.

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of /home/wd/git/u-boot/custodians
Wolfgang Denk [Thu, 27 Dec 2007 00:13:05 +0000 (01:13 +0100)]
Merge branch 'master' of /home/wd/git/u-boot/custodians

16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-sh
Wolfgang Denk [Thu, 27 Dec 2007 00:12:56 +0000 (01:12 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-sh

Conflicts:

MAINTAINERS

Signed-off-by: Wolfgang Denk <wd@denx.de>
16 years agoMerge branch 'master' of /home/wd/git/u-boot/custodians
Wolfgang Denk [Wed, 26 Dec 2007 23:46:17 +0000 (00:46 +0100)]
Merge branch 'master' of /home/wd/git/u-boot/custodians

16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-avr32
Wolfgang Denk [Wed, 26 Dec 2007 23:46:08 +0000 (00:46 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-avr32

16 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-mpc85xx
Wolfgang Denk [Wed, 26 Dec 2007 23:35:03 +0000 (00:35 +0100)]
Merge branch 'master' of git://denx.de/git/u-boot-mpc85xx

16 years agoMerge branch 'testing' of git://www.denx.de/git/u-boot-fdt
Wolfgang Denk [Wed, 26 Dec 2007 23:22:24 +0000 (00:22 +0100)]
Merge branch 'testing' of git://denx.de/git/u-boot-fdt

16 years agocfi_flash: Add manufacturer-specific fixups
Haavard Skinnemoen [Fri, 14 Dec 2007 14:36:18 +0000 (15:36 +0100)]
cfi_flash: Add manufacturer-specific fixups

Run fixups based on the JEDEC manufacturer ID independent of the
command set ID.

This changes current behaviour: Previously, geometry reversal for AMD
chips were done based on the command set ID, while they are now done
based on the JEDEC manufacturer and device ID.

Also add fixup for top-boot Atmel chips. A fixup is needed for
AT49BV6416(T) too, but since u-boot currently only reads the low byte
of the device ID, there's no way to tell it apart from AT49BV642D,
which should not have this fixup. Since AT49BV642D support is
necessary to get ATNGW100 board support into mainline, I've commented
out the fixup for now.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agocfi_flash: Add cmdset-specific init functions
Haavard Skinnemoen [Fri, 14 Dec 2007 14:36:17 +0000 (15:36 +0100)]
cfi_flash: Add cmdset-specific init functions

Move things like reading JEDEC IDs and fixing up geometry reversal
into separate functions. The geometry reversal fixup is now performed
by altering the qry structure directly, which makes the sector init
code slightly cleaner.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agocfi_flash: Read whole QRY structure in one go
Haavard Skinnemoen [Fri, 14 Dec 2007 14:36:16 +0000 (15:36 +0100)]
cfi_flash: Read whole QRY structure in one go

Read out the whole CFI Standard Query structure after successful cfi
identification. This allows subsequent code to access this information
directly without having to go through flash_read_uchar() and friends.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoAVR32: Fix logic inversion in disable_interrupts()
Haavard Skinnemoen [Mon, 17 Dec 2007 10:02:44 +0000 (11:02 +0100)]
AVR32: Fix logic inversion in disable_interrupts()

disable_interrupts() should return nonzero if interrupts were
_enabled_ before, not disabled.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoAVR32: Enable interrupts at bootup
Haavard Skinnemoen [Fri, 14 Dec 2007 15:51:22 +0000 (16:51 +0100)]
AVR32: Enable interrupts at bootup

The timer code depends on the timer interrupt to keep track of the
upper 32 bits of the cycle counter. This obviously doesn't work when
interrupts are disabled the whole time.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoAVR32: Fix wrong pin setup for USART3
Haavard Skinnemoen [Thu, 15 Nov 2007 09:03:45 +0000 (10:03 +0100)]
AVR32: Fix wrong pin setup for USART3

As reported by Gerhard Berghofer:

in "gpio_enable_usart3" the correct pins for USART 3 are PB17 and PB18
instead of PB18 and PB19.

which is obviously correct. There's currently no code that uses
USART3, but custom boards may run into problems.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoREADME: Remove ATSTK1000 daughterboard list
Haavard Skinnemoen [Thu, 1 Nov 2007 11:44:20 +0000 (12:44 +0100)]
README: Remove ATSTK1000 daughterboard list

As noted by Kim Phillips, these lists tend to become out of date.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoAdd ATSTK100[234] to MAINTAINERS
Haavard Skinnemoen [Tue, 30 Oct 2007 13:56:36 +0000 (14:56 +0100)]
Add ATSTK100[234] to MAINTAINERS

Add all the ATSTK1000 daughterboards to MAINTAINERS along with their
"mother". Also update the entry for ATSTK1000 to be not only about the
AP7000 CPU; it's intended to handle all CPUs in the AT32AP family.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoAVR32: Add support for the ATSTK1004 board
Haavard Skinnemoen [Mon, 29 Oct 2007 12:02:54 +0000 (13:02 +0100)]
AVR32: Add support for the ATSTK1004 board

ATSTK1004 is a daughterboard for ATSTK1000 with the AT32AP7002 CPU,
which is a derivative of AT32AP7000.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoAVR32: Add support for the ATSTK1003 board
Haavard Skinnemoen [Mon, 29 Oct 2007 12:02:54 +0000 (13:02 +0100)]
AVR32: Add support for the ATSTK1003 board

ATSTK1003 is a daughterboard for ATSTK1000 with the AT32AP7001 CPU,
which is a derivative of AT32AP7000.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoAVR32: Make some AT32AP700x peripherals optional
Haavard Skinnemoen [Mon, 29 Oct 2007 12:23:33 +0000 (13:23 +0100)]
AVR32: Make some AT32AP700x peripherals optional

Add a chip-features file providing definitions of the form

AT32AP700x_CHIP_HAS_<peripheral>

to indicate the availability of the given peripheral on the currently
selected chip.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoAVR32: Rename at32ap7000 -> at32ap700x
Haavard Skinnemoen [Mon, 29 Oct 2007 12:09:56 +0000 (13:09 +0100)]
AVR32: Rename at32ap7000 -> at32ap700x

The SoC-specific code for all the AT32AP700x CPUs is practically
identical; the only difference is that some chips have less features
than others. By doing this rename, we can add support for the AP7000
derivatives simply by making some features conditional.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoatmel_mci: Show SR when block read fails
Haavard Skinnemoen [Fri, 29 Jun 2007 16:22:34 +0000 (18:22 +0200)]
atmel_mci: Show SR when block read fails

Show controller status as well as card status when an error occurs
during block read.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoppc4xx: Bring 4xx fdt support up-to-date
Stefan Roese [Thu, 13 Dec 2007 13:52:53 +0000 (14:52 +0100)]
ppc4xx: Bring 4xx fdt support up-to-date

This patch update the 4xx fdt support. It enabled fdt booting
on the AMCC Kilauea and Sequoia for now. More can follow later
quite easily.

Signed-off-by: Stefan Roese <sr@denx.de>
16 years agoMerge commit 'u-boot-fdt/testing'
Stefan Roese [Thu, 13 Dec 2007 13:02:42 +0000 (14:02 +0100)]
Merge commit 'u-boot-fdt/testing'

16 years agocfi_flash: Use map_physmem() and unmap_physmem()
Haavard Skinnemoen [Thu, 13 Dec 2007 11:56:34 +0000 (12:56 +0100)]
cfi_flash: Use map_physmem() and unmap_physmem()

Use map_physmem() and unmap_physmem() to convert from physical to
virtual addresses. This gives the arch a chance to provide an uncached
mapping for flash accesses.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoIntroduce map_physmem() and unmap_physmem()
Haavard Skinnemoen [Thu, 13 Dec 2007 11:56:33 +0000 (12:56 +0100)]
Introduce map_physmem() and unmap_physmem()

map_physmem() returns a virtual address which can be used to access a
given physical address without involving the cache. unmap_physmem()
should be called when the virtual address returned by map_physmem() is
no longer needed.

This patch adds a stub implementation which simply returns the
physical address cast to a uchar * for all architectures except AVR32,
which converts the physical address to an uncached virtual mapping.
unmap_physmem() is a no-op on all architectures, but if any
architecture needs to do such mappings through the TLB, this is the
hook where those TLB entries can be invalidated.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agocfi_flash: Introduce read and write accessors
Haavard Skinnemoen [Thu, 13 Dec 2007 11:56:32 +0000 (12:56 +0100)]
cfi_flash: Introduce read and write accessors

Introduce flash_read{8,16,32,64) and flash_write{8,16,32,64} and use
them to access the flash memory. This makes it clearer when the flash
is actually being accessed; merely dereferencing a volatile pointer
looks just like any other kind of access.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoImplement __raw_{read,write}[bwl] on all architectures
Haavard Skinnemoen [Thu, 13 Dec 2007 11:56:31 +0000 (12:56 +0100)]
Implement __raw_{read,write}[bwl] on all architectures

This adds implementations of __raw_read[bwl] and __raw_write[bwl] to
m68k, ppc, nios and nios2. The m68k and ppc implementations were taken
from Linux.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agocfi_flash: Reorder functions and eliminate extra prototypes
Haavard Skinnemoen [Sat, 6 Oct 2007 16:55:36 +0000 (18:55 +0200)]
cfi_flash: Reorder functions and eliminate extra prototypes

Reorder the functions in cfi_flash.c so that each function only uses
functions that have been defined before it. This allows the static
prototype declarations near the top to be eliminated and might allow
gcc to do a better job inlining functions.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agocfi_flash: Make some needlessly global functions static
Haavard Skinnemoen [Thu, 13 Dec 2007 11:56:29 +0000 (12:56 +0100)]
cfi_flash: Make some needlessly global functions static

Make functions not declared in any header file static.

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agocfi_flash: Break long lines
Haavard Skinnemoen [Thu, 13 Dec 2007 11:56:28 +0000 (12:56 +0100)]
cfi_flash: Break long lines

This patch tries to keep all lines in the cfi_flash driver below 80
columns. There are a few lines left which don't fit this requirement
because I couldn't find any trivial way to break them (i.e. it would
take some restructuring, which I intend to do in a later patch.)

Signed-off-by: Haavard Skinnemoen <hskinnemoen@atmel.com>
16 years agoCFI: synchronize command offsets with Linux CFI driver
Bartlomiej Sieka [Tue, 11 Dec 2007 12:59:57 +0000 (13:59 +0100)]
CFI: synchronize command offsets with Linux CFI driver

Fixes non-working CFI Flash on the Inka4x0 board.

Signed-off-by: Bartlomiej Sieka <tur@semihalf.com>
16 years agoHandle MPC85xx PCIe reset errata (PCI-Ex 38)
Kumar Gala [Fri, 7 Dec 2007 18:17:34 +0000 (12:17 -0600)]
Handle MPC85xx PCIe reset errata (PCI-Ex 38)

On the MPC85xx boards that have PCIe enable the PCIe errata fix.
(MPC8544DS, MPC8548CDS, MPC8568MDS).

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoUpdate Freescale MPC85xx ADS/CDS/MDS board config
Kumar Gala [Fri, 7 Dec 2007 18:04:30 +0000 (12:04 -0600)]
Update Freescale MPC85xx ADS/CDS/MDS board config

* Enabled CONFIG_CMD_ELF

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoHandle Asynchronous DDR clock on 85xx
Kumar Gala [Fri, 7 Dec 2007 10:59:26 +0000 (04:59 -0600)]
Handle Asynchronous DDR clock on 85xx

The MPC8572 introduces the concept of an asynchronous DDR clock with
regards to the platform clock.

Introduce get_ddr_freq() to report the DDR freq regardless of sync/async
mode.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoUpdate Freescale MPC85xx ADS/CDS/MDS board config
Kumar Gala [Thu, 29 Nov 2007 16:34:28 +0000 (10:34 -0600)]
Update Freescale MPC85xx ADS/CDS/MDS board config

* Removed some misc environment setup
* Enabled CONFIG_CMDLINE_EDITING

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoMove the MPC8541/MPC8555/MPC8548 CDS board under board/freescale.
Kumar Gala [Thu, 29 Nov 2007 16:47:44 +0000 (10:47 -0600)]
Move the MPC8541/MPC8555/MPC8548 CDS board under board/freescale.

Minor path corrections needed to ensure buildability.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoMove the MPC8540 ADS board under board/freescale.
Kumar Gala [Thu, 29 Nov 2007 16:16:18 +0000 (10:16 -0600)]
Move the MPC8540 ADS board under board/freescale.

Minor path corrections needed to ensure buildability.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoMove the MPC8560 ADS board under board/freescale.
Kumar Gala [Thu, 29 Nov 2007 16:14:50 +0000 (10:14 -0600)]
Move the MPC8560 ADS board under board/freescale.

Minor path corrections needed to ensure buildability.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
16 years agoMove the MPC8568 MDS board under board/freescale.
Kumar Gala [Thu, 29 Nov 2007 16:13:47 +0000 (10:13 -0600)]
Move the MPC8568 MDS board under board/freescale.

Minor path corrections needed to ensure buildability.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>