platform/kernel/u-boot.git
15 years agoPrepare v2009.06 v2009.06
Wolfgang Denk [Sun, 14 Jun 2009 19:30:39 +0000 (21:30 +0200)]
Prepare v2009.06

Update CHANGELOG, fix minor coding stylke issue. Update Makefile.

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoRevert "SMC911x driver fixed for NFS boot"
Wolfgang Denk [Sun, 14 Jun 2009 18:31:36 +0000 (20:31 +0200)]
Revert "SMC911x driver fixed for NFS boot"

This reverts commit ca9c8a1e10fac01e6a1129f82a7ce18bd818fa43,
which causes compile warnings ("large integer implicitly truncated
to unsigned type") on all systems that use this driver. The warning
results from passing long constants (TX_CFG, RX_CFG) into
smc911x_set_mac_csr() which is declared to accept "unsigned
character" arguments only.

Being close to a release, with nobody available to actually test the
code or the suggested fixes, it seems better to revert the patch.

15 years agofsl-ddr: Fix handling of >4G of memory when !CONFIG_PHYS_64BIT
Kumar Gala [Fri, 12 Jun 2009 04:42:35 +0000 (23:42 -0500)]
fsl-ddr: Fix handling of >4G of memory when !CONFIG_PHYS_64BIT

The ddr code computes most things as 64-bit quantities and had some places
in the middle that it was using phy_addr_t and phys_size_t.

Instead we use unsigned long long through out and only at the last stage of
setting the LAWs and reporting the amount of memory to the board code do we
truncate down to what we can cover via phys_size_t.

This has the added benefit that the DDR controller itself is always setup
the same way regardless of how much memory we have.  Its only the LAW
setup that limits what is visible to the system.

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
15 years ago85xx: Use print_size to report amount of memory not mapped by TLBs
Kumar Gala [Fri, 12 Jun 2009 04:40:34 +0000 (23:40 -0500)]
85xx: Use print_size to report amount of memory not mapped by TLBs

Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
15 years ago85xx: Add README for MPC8569MDS
Haiying Wang [Wed, 20 May 2009 16:30:42 +0000 (12:30 -0400)]
85xx: Add README for MPC8569MDS

Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
15 years ago85xx: Add UART1 support for MPC8569MDS
Haiying Wang [Wed, 20 May 2009 16:30:33 +0000 (12:30 -0400)]
85xx: Add UART1 support for MPC8569MDS

MPC8569 UART1 signals are muxed with PortF bit[9-12], we need to define
those pins before using UART1.

Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
15 years ago85xx: Add PIB support at CS4/CS5 for MPC8569MDS
Haiying Wang [Wed, 20 May 2009 16:30:32 +0000 (12:30 -0400)]
85xx: Add PIB support at CS4/CS5 for MPC8569MDS

Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: Yu Liu <Yu.Liu@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
15 years ago85xx: Fix some settings for MPC8569MDS board
Haiying Wang [Thu, 4 Jun 2009 20:12:39 +0000 (16:12 -0400)]
85xx: Fix some settings for MPC8569MDS board

- Increase the size of malloc to 512KB because MPC8569MDS needs more memory for
malloc to support up to eight Ethernet interfaces.
- Move Environment address out of uboot thus the saved environment variables
will not be erased after u-boot is re-programmed.

Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
15 years ago85xx: Fix MURAM size for MPC8569
Haiying Wang [Wed, 20 May 2009 16:30:30 +0000 (12:30 -0400)]
85xx: Fix MURAM size for MPC8569

MPC8569 has 128K bytes MURAM.

Signed-off-by: Haiying Wang <Haiying.Wang@freescale.com>
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
15 years agoMerge branch 'master' of git://git.denx.de/u-boot-mpc83xx
Wolfgang Denk [Thu, 11 Jun 2009 21:53:15 +0000 (23:53 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-mpc83xx

15 years agoat91/cpu.c: add missing Copyright & GPL header
Jean-Christophe PLAGNIOL-VILLARD [Fri, 22 May 2009 18:23:51 +0000 (20:23 +0200)]
at91/cpu.c: add missing Copyright & GPL header

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
15 years agoapollon: Fix a OBJCFLAGS typo
Shinya Kuribayashi [Sun, 7 Jun 2009 12:45:16 +0000 (21:45 +0900)]
apollon: Fix a OBJCFLAGS typo

Signed-off-by: Shinya Kuribayashi <skuribay@pobox.com>
15 years agoPrepare 2009.06-rc3 v2009.06-rc3
Wolfgang Denk [Tue, 9 Jun 2009 22:19:28 +0000 (00:19 +0200)]
Prepare 2009.06-rc3

Update CHANGELOG

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agormu board: fix error: 'CONFIG_ENV_SECT_SIZE' undeclared
Wolfgang Denk [Tue, 9 Jun 2009 22:15:11 +0000 (00:15 +0200)]
rmu board: fix error: 'CONFIG_ENV_SECT_SIZE' undeclared

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoubifs: fix small error path mismatch
Daniel Mack [Thu, 4 Jun 2009 17:44:12 +0000 (19:44 +0200)]
ubifs: fix small error path mismatch

In do_readpage(), don't free 'dn' if its allocation failed.

Signed-off-by: Daniel Mack <daniel@caiaq.de>
15 years agoEP88x: fix broken linker script
Wolfgang Denk [Wed, 3 Jun 2009 22:31:07 +0000 (00:31 +0200)]
EP88x: fix broken linker script

Signed-off-by: Wolfgang Denk <wd@denx.de>
Tested-by: Mikhail Zaturenskiy <mzaturenskiy@shoppertrak.com>
15 years agoTQM85xx: minor config file cleanup
Wolfgang Denk [Thu, 14 May 2009 22:16:02 +0000 (00:16 +0200)]
TQM85xx: minor config file cleanup

Remove "saveenv" from "update" definition: the environment is outside
the U-Boot image on TQM85xx and therefor not affected by updates.

Also "beautify" code a bit (vertical alignment).

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoTQM85xx: adapt for new flash types
Wolfgang Denk [Thu, 14 May 2009 22:16:01 +0000 (00:16 +0200)]
TQM85xx: adapt for new flash types

Old TQM85xx boards had 'M' type Spansion Flashes from the S29GLxxxM
series while new boards have 'N' type Flashes from the S29GLxxxN
series, which have bigger sectors: 2 x 128 instead of 2 x 64 KB.

We now change the configuration to the new flash types for all
boards; this also works on old boards - we just waste two flash
sectors for the environment which could be smaller there.

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years ago85xx: Fix the wrong SYS_CLK_IN for 8569MDS
Dave Liu [Mon, 18 May 2009 09:49:23 +0000 (17:49 +0800)]
85xx: Fix the wrong SYS_CLK_IN for 8569MDS

The SYS_CLK_IN of MPC8569MDS is 66.66MHz,
The DDR_CLK_IN is same with SYS_CLK_IN in 8569 processor.
so, change the SYS_CLK_IN from 66MHz to 66.66MHz.

Signed-off-by: Dave Liu <daveliu@freescale.com>
15 years ago85xx: Fix the wrong BCSR address of 8569MDS
Dave Liu [Fri, 15 May 2009 02:27:44 +0000 (10:27 +0800)]
85xx: Fix the wrong BCSR address of 8569MDS

The BCSR17[7] = 1 will unlock the write protect of FLASH.
The WP# pin only controls the write protect of top/bottom sector,
That is why we can save env, but we can't write the first sector
before the patch.

Signed-off-by: Dave Liu <daveliu@freescale.com>
15 years ago85xx: bugfix for reading maximum TLB size on mpc85xx
Fredrik Arnerup [Tue, 2 Jun 2009 21:27:10 +0000 (16:27 -0500)]
85xx: bugfix for reading maximum TLB size on mpc85xx

The MAXSIZE field in the TLB1CFG register is 4 bits, not 8 bits.
This made setup_ddr_tlbs() try to set up a TLB larger than the e500 maximum
(256 MB)
which made u-boot hang in board_init_f() when trying to create a new stack
in RAM.
I have an mpc8540 with one 1GB dimm.

Signed-off-by: Fredrik Arnerup <fredrik.arnerup@edgeware.tv>
Signed-off-by: Andy Fleming <afleming@freescale.com>
Acked-by: Kumar Gala <galak@kernel.crashing.org>
15 years ago85xx: Fix the clock adjust of mpc8569mds board
Dave Liu [Fri, 27 Mar 2009 06:32:43 +0000 (14:32 +0800)]
85xx: Fix the clock adjust of mpc8569mds board

Currently the clk_adj is 6 (3/4 cycle), The settings will cause
the DDR controller hang at the data init. Change the clk_adj
from 6 to 4 (1/2 cycle), make the memory system stable.

Signed-off-by: Dave Liu <daveliu@freescale.com>
15 years agoMerge branch 'master' of git://git.denx.de/u-boot-usb
Wolfgang Denk [Tue, 9 Jun 2009 20:53:03 +0000 (22:53 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-usb

15 years agoat91: fix a USB problem for AT91SAM9261
RONETIX - Ilko Iliev [Fri, 5 Jun 2009 14:54:31 +0000 (16:54 +0200)]
at91: fix a USB problem for AT91SAM9261

This patch corrects the missing PLLB initialization in usb_cpu_init()
for AT91SAM9261.
Because of the missing PLLB initialization, the USB support for all
AT91SAM9261 based boards will work only if the PLLB is configured by a
precedent bootloader.

Signed-off-by: Ilko Iliev <iliev@ronetix.at>
Acked-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Remy Bohmer <linux@bohmer.net>
15 years agoppc4xx/net: Fix MDIO clock setup
Felix Radensky [Sun, 31 May 2009 17:44:15 +0000 (20:44 +0300)]
ppc4xx/net: Fix MDIO clock setup

This patch fixes MDIO clock setup in case when OPB frequency is 100MHz.
Current code assumes that the value of sysinfo.freqOPB is 100000000
when OPB frequency is 100MHz. In reality it is 100000001. As a result
MDIO clock is set to incorrect value, larger than 2.5MHz, thus violating
the standard. This in not a problem on boards equipped with Marvell PHYs
(e.g. Canyonlands), since those PHYs support MDIO clocks up to 8.3MHz,
but can be a problem for other PHYs (e.g. Realtek ones).

Signed-off-by: Felix Radensky <felix@embedded-sol.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
15 years agortl8169: fix PCI system memory address
Yoshihiro Shimoda [Wed, 25 Feb 2009 05:27:29 +0000 (14:27 +0900)]
rtl8169: fix PCI system memory address

When PCI device use system memory, some PCI host controller should be
set physical memory address.

Signed-off-by: Yoshihiro Shimoda <shimoda.yoshihiro@renesas.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
15 years agoSMC911x driver fixed for NFS boot
Manikandan Pillai [Wed, 8 Apr 2009 03:44:35 +0000 (09:14 +0530)]
SMC911x driver fixed for NFS boot

eth_halt() function in the smc911x drivers used to call the
smc911x_reset() function. eth_halt() used to be called after
tftp transfers. This used to put the ethernet chip in reset
while the linux boots up resulting in the ethernet driver
not coming up. NFS boot used to fail as a result.

This patch calls smc911x_shutdown() instead of smc911x_reset().
Some comments received has also been fixed.

Signed-off-by: Manikandan Pillai <mani.pillai@ti.com>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
15 years agoAdd config option for disabling DM9000-SROM support.
Remy Bohmer [Sun, 3 May 2009 10:11:40 +0000 (12:11 +0200)]
Add config option for disabling DM9000-SROM support.

Some boards do not have SROM support for the DM9000 network adapter.
Instead of listing these board names in the driver code, make this
option configurable from the board config file.

It also removes a build warning for the at91sam9261ek board:
'dm9000x.c:545: warning: 'read_srom_word' defined but not used'

And it repaires the trizepsiv board build which was broken around the
same routines

Signed-off-by: Remy Bohmer <linux@bohmer.net>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
15 years agompc83xx: don't set SICRH_TSOBI1 to RMII/RTBI operation
Kim Phillips [Fri, 5 Jun 2009 19:11:33 +0000 (14:11 -0500)]
mpc83xx: don't set SICRH_TSOBI1 to RMII/RTBI operation

In GMII mode (which operates at 3.3V) both SICRH TSEC1/2 output buffer
impedance bits should be clear, i.e., SICRH[TSIOB1] = 0 and SICRH[TSIOB2] = 0.
SICRH[TSIOB1] was erroneously being set high.

U-Boot always operated this PHY interface in GMII mode.  It is assumed this
was missed in the clean up by the original board porters, and copied along
to the TQM and sbc boards.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
Acked-by: Ira Snyder <iws@ovro.caltech.edu>
Reviewed-by: David Hawkins <dwh@ovro.caltech.edu>
Tested-by: Paul Gortmaker <paul.gortmaker@windriver.com>
CC: Dave Liu <DaveLiu@freescale.com>
15 years agosmc911x: write back the manually set MAC address
Daniel Mack [Wed, 8 Apr 2009 11:23:37 +0000 (13:23 +0200)]
smc911x: write back the manually set MAC address

If the MAX address is given by the environment, write it back to the
hardware.

Signed-off-by: Daniel Mack <daniel@caiaq.de>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
15 years agoMerge branch 'master' of git://git.denx.de/u-boot-mmc
Wolfgang Denk [Thu, 4 Jun 2009 08:56:09 +0000 (10:56 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-mmc

15 years agoRedundant Environment: protect full sector size
Wolfgang Denk [Thu, 14 May 2009 22:16:03 +0000 (00:16 +0200)]
Redundant Environment: protect full sector size

Several boards used different ways to specify the size of the
protected area when enabling flash write protection for the sectors
holding the environment variables: some used CONFIG_ENV_SIZE and
CONFIG_ENV_SIZE_REDUND, some used CONFIG_ENV_SECT_SIZE, and some even
a mix of both for the "normal" and the "redundant" areas.

Normally, this makes no difference at all. However, things are
different when you have to deal with boards that can come with
different types of flash chips, which may have different sector
sizes.

Here we may have to chose CONFIG_ENV_SECT_SIZE such that it fits the
biggest sector size, which may include several sectors on boards using
the smaller sector flash types. In such a case, using CONFIG_ENV_SIZE
or CONFIG_ENV_SIZE_REDUND to enable the protection may lead to the
case that only the first of these sectors get protected, while the
following ones aren't.

This is no real problem, but it can be confusing for the user -
especially on boards that use CONFIG_ENV_SECT_SIZE to protect the
"normal" areas, while using CONFIG_ENV_SIZE_REDUND for the
"redundant" area.

To avoid such inconsistencies, I changed all sucn boards that I found
to consistently use CONFIG_ENV_SECT_SIZE for protection. This should
not cause any functional changes to the code.

Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Paul Ruhland
Cc: Pantelis Antoniou <panto@intracom.gr>
Cc: Stefan Roese <sr@denx.de>
Cc: Gary Jennejohn <garyj@denx.de>
Cc: Dave Ellis <DGE@sixnetio.com>
Acked-by: Stefan Roese <sr@denx.de>
15 years agommc: it's safe to ignore mmc_send_if_cond() return value
Ilya Yanok [Thu, 14 May 2009 10:03:09 +0000 (14:03 +0400)]
mmc: it's safe to ignore mmc_send_if_cond() return value

Return value of mmc_send_if_cond() can be safely ignored (as it is
done in Linux). This makes older cards work with MXC MCI controller.

Signed-off-by: Ilya Yanok <yanok@emcraft.com>
15 years agocfi_mtd: Fix bug in last sector detection
Stefan Roese [Mon, 11 May 2009 13:54:13 +0000 (15:54 +0200)]
cfi_mtd: Fix bug in last sector detection

This patch now enabled this cfi-mtd wrapper to correctly detect and
erase the last sector in an NOR FLASH device.

Signed-off-by: Stefan Roese <sr@denx.de>
15 years agommc: Fix decoding of SCR & function switch data on little-endian machines
Yauhen Kharuzhy [Wed, 6 May 2009 21:43:30 +0000 (00:43 +0300)]
mmc: Fix decoding of SCR & function switch data on little-endian machines

SCR & switch data are read from card as big-endian words and should be
converted to CPU byte order.

Signed-off-by: Yauhen Kharuzhy <jekhor@gmail.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
15 years agommc: Remove return from mmc_init for non SD 2.0 compatible cards.
Yauhen Kharuzhy [Thu, 7 May 2009 10:08:53 +0000 (13:08 +0300)]
mmc: Remove return from mmc_init for non SD 2.0 compatible cards.

Cards which are not compatible with SD 2.0 standard, may return response
for CMD8 command, but it will be invalid in terms of SD 2.0. We should
accept this case as admissible, just like Linux does.

Signed-off-by: Yauhen Kharuzhy <jekhor@gmail.com>
Signed-off-by: Andy Fleming <afleming@freescale.com>
15 years agommc: drop unnecessary casts
Rabin Vincent [Sun, 5 Apr 2009 08:00:56 +0000 (13:30 +0530)]
mmc: drop unnecessary casts

Now that response is a uint, we can drop all the casts.

Signed-off-by: Rabin Vincent <rabin@rab.in>
15 years agommc: fix response decoding on little endian
Rabin Vincent [Sun, 5 Apr 2009 08:00:55 +0000 (13:30 +0530)]
mmc: fix response decoding on little endian

The mmc code defines the response as an array of chars.  However, it
access the response bytes both as (i) an array of four uints (with
casts) and (ii) as individual chars.  The former case is used more
often, including by the driver when it assigns the response.

The char-wise accesses are broken on little endian systems because they
assume that the bytes in the uints are in big endian byte order.

This patch fixes this by changing the response to be an array of four
uints and replacing the char-wise accesses with equivalent uint-wise
accesses.

Signed-off-by: Rabin Vincent <rabin@rab.in>
15 years agommc: use lldiv to fix arm eabi build
Rabin Vincent [Sun, 5 Apr 2009 08:00:54 +0000 (13:30 +0530)]
mmc: use lldiv to fix arm eabi build

The generic MMC core uses direct long long divisions, which do not build
with ARM EABI toolchains.  Use lldiv() instead, which works everywhere.

Signed-off-by: Rabin Vincent <rabin@rab.in>
15 years agommc: check find_mmc_device return value
Rabin Vincent [Sun, 5 Apr 2009 08:00:53 +0000 (13:30 +0530)]
mmc: check find_mmc_device return value

find_mmc_device returns NULL if an invalid device number is specified.
Check for this to avoid dereferencing NULL pointers.

Signed-off-by: Rabin Vincent <rabin@rab.in>
15 years agommc: clean up help texts
Rabin Vincent [Sun, 5 Apr 2009 08:00:52 +0000 (13:30 +0530)]
mmc: clean up help texts

Remove some repeated words and superfluous newlines in the mmc command
help entries.

Signed-off-by: Rabin Vincent <rabin@rab.in>
15 years agoBlackfin: spi: fix pin handling of SPI0 SSEL4
Mike Frysinger [Fri, 29 May 2009 21:01:48 +0000 (17:01 -0400)]
Blackfin: spi: fix pin handling of SPI0 SSEL4

CS4 on SPI0 has a dedicated PH8 pin which needs to be enabled as a
peripheral in order to work.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agoBlackfin: fix if() logic in bootrom evt1 check
Mike Frysinger [Tue, 26 May 2009 06:51:57 +0000 (02:51 -0400)]
Blackfin: fix if() logic in bootrom evt1 check

A missing set of parenthesis caused the silicon revision to apply only to
the BF533 and not the BF531/BF532 variants.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agoMerge branch 'master' of git://git.denx.de/u-boot-blackfin
Wolfgang Denk [Thu, 28 May 2009 19:27:51 +0000 (21:27 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-blackfin

15 years agojffs2/mtdparts: Fix problem with usage from JFFS2 and MTDPARTS together
Stefan Roese [Sat, 16 May 2009 10:04:22 +0000 (12:04 +0200)]
jffs2/mtdparts: Fix problem with usage from JFFS2 and MTDPARTS together

Currently using JFFS2 with MTDPARTS enabled doesn't work. This is because
mtdparts_init() is available in both files, cmd_mtdparts.c and
cmd_jffs2.c. Please note that in the original cmd_jffs2.c file (before
the jffs2/mtdparts command/file split those 2 different versions
already existed. So this is nothing new. The main problem is that the
variables "current_dev" and "current_partnum" are declared in both
files now. This doesn't work.

This patch now changes the names of those variable to more specific
names: "current_mtd_dev" and "current_mtd_partnum". This is because
this patch also changes the declaration from static to global, so
that they can be used from both files.

Please note that my first tests were not successful. The MTD devices
selected via mtdparts are now accessed but I'm failing to see the
directory listed via the "ls" command. Nothing is displayed. Perhaps
I didn't generate the JFFS2 image correctly (I never used JFFS2 in
U-Boot before). Not sure. Perhaps somebody else could take a look at
this as well. I'll continue looking into this on Monday.

Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Wolfgang Denk <wd@denx.de>
Cc: Detlev Zundel <dzu@denx.de>
Cc: Ilya Yanok <yanok@emcraft.com>
Cc: Renaud barbier <renaud.barbier@ge.com>
15 years agoBlackfin: bf518f-ezbrd: setup portmux for async flash
Graf Yang [Sun, 24 May 2009 06:34:34 +0000 (02:34 -0400)]
Blackfin: bf518f-ezbrd: setup portmux for async flash

The pins for async memory where parallel flash lives are not enabled by
default, so make sure we mux them as needed.

Signed-off-by: Graf Yang <graf.yang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agonand: Fix problem with ECC ordering for PPC4xx NDFC platforms
Stefan Roese [Wed, 20 May 2009 08:58:03 +0000 (10:58 +0200)]
nand: Fix problem with ECC ordering for PPC4xx NDFC platforms

This patch enables Smart Media (SMC) ECC byte ordering which is used
on the PPC4xx NAND FLASH controller (NDFC). Without this patch we have
incompatible ECC byte ordering to the Linux kernel NDFC driver.

Signed-off-by: Stefan Roese <sr@denx.de>
Acked-by: Scott Wood <scottwood@freescale.com>
15 years agoppc4xx: Fix problem with ECC ordering for PPC4xx NDFC platforms
Stefan Roese [Wed, 20 May 2009 08:58:02 +0000 (10:58 +0200)]
ppc4xx: Fix problem with ECC ordering for PPC4xx NDFC platforms

This patch now uses the correct ECC byte order (Smart Media - SMC)
to be used on the 4xx NAND FLASH driver. Without this patch we have
incompatible ECC byte ordering to the Linux kernel NDFC driver.

Please note that we also have to enable CONFIG_MTD_NAND_ECC_SMC in
drivers/mtd/nand/nand_ecc.c for correct operation. This is done with
a seperate patch.

Signed-off-by: Stefan Roese <sr@denx.de>
Acked-by: Scott Wood <scottwood@freescale.com>
15 years agoppc4xx: Move definition for PPC4xx NAND FLASH controller to header
Stefan Roese [Wed, 20 May 2009 08:58:01 +0000 (10:58 +0200)]
ppc4xx: Move definition for PPC4xx NAND FLASH controller to header

This patch moves the definition for the PPC4xx NAND FLASH controller
(NDFC) CONFIG_NAND_NDFC into include/ppc4xx.h. This is needed for the
upcoming fix for the ECC byte ordering of the NDFC driver.

Signed-off-by: Stefan Roese <sr@denx.de>
Acked-by: Scott Wood <scottwood@freescale.com>
15 years agoMerge branch 'master' of git://git.denx.de/u-boot-ubi
Wolfgang Denk [Wed, 20 May 2009 20:42:04 +0000 (22:42 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-ubi

15 years agoMerge branch 'master' of git://git.denx.de/u-boot-blackfin
Wolfgang Denk [Wed, 20 May 2009 20:40:44 +0000 (22:40 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-blackfin

15 years agocommon: fix inline--weak error spotted by gcc 4.4
Kim Phillips [Tue, 19 May 2009 17:53:36 +0000 (12:53 -0500)]
common: fix inline--weak error spotted by gcc 4.4

cmd_ide.c:547: error: inline function 'ide_inb' cannot be declared weak

removing the inline attribute fixes it.

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
15 years agoTQM834x: remove defines causing gcc4.4 warnings
Kim Phillips [Tue, 19 May 2009 17:53:32 +0000 (12:53 -0500)]
TQM834x: remove defines causing gcc4.4 warnings

Configuring for TQM834x board...
cpu_init.c: In function 'cpu_init_f':
cpu_init.c:262: warning: array subscript is above array bounds
cpu_init.c:263: warning: array subscript is above array bounds
cpu_init.c:270: warning: array subscript is above array bounds
...

Signed-off-by: Kim Phillips <kim.phillips@freescale.com>
15 years agoUBI: fix return code in ubi_volume_read
Andreas Huber [Tue, 19 May 2009 09:06:30 +0000 (11:06 +0200)]
UBI: fix return code in ubi_volume_read

Return -ENODEV instead of 0 when trying to read from a non existing volume.

Signed-off-by: Andreas Huber <andreas.huber@keymile.com>
Signed-off-by: Stefan Roese <sr@denx.de>
15 years agoBlackfin: fix timer_init()/timer_reset()
Graf Yang [Tue, 19 May 2009 08:40:08 +0000 (04:40 -0400)]
Blackfin: fix timer_init()/timer_reset()

The timer_init() function was not using the right csync instruction, nor
was it doing it right after disabling the core timer.

The timer_reset() function would reset the timestamp, but not the actual
timer, so there was a common edge case where get_timer() return a jump of
one timestamp (couple milliseconds) right after resetting.  This caused
many functions to improperly timeout right away.

Signed-off-by: Graf Yang <graf.yang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agoMIPS: lib_mips/board.c: Remove unused variables
Shinya Kuribayashi [Sat, 16 May 2009 00:12:09 +0000 (09:12 +0900)]
MIPS: lib_mips/board.c: Remove unused variables

This fixes the following build warnings:

board.c: In function 'board_init_r':
board.c:328: warning: unused variable 'i'
board.c:326: warning: unused variable 'e'

Signed-off-by: Shinya Kuribayashi <skuribay@pobox.com>
15 years agoMIPS: Make all extern-ed functions in bitops.h static
Shinya Kuribayashi [Sat, 16 May 2009 00:12:09 +0000 (09:12 +0900)]
MIPS: Make all extern-ed functions in bitops.h static

All these functions are expected to be static inline-ed.
This patch also fixes the following build warnings on MIPS targets:

include/asm/bitops.h: In function 'ext2_find_next_zero_bit':
include/asm/bitops.h:862: warning: '__fswab32' is static but used in inline function 'ext2_find_next_zero_bit' which is not static
include/asm/bitops.h:885: warning: '__fswab32' is static but used in inline function 'ext2_find_next_zero_bit' which is not static
include/asm/bitops.h:887: warning: '__fswab32' is static but used in inline function 'ext2_find_next_zero_bit' which is not static

Signed-off-by: Shinya Kuribayashi <skuribay@pobox.com>
15 years agoMIPS: Implement ethernet halt for au1x00
Thomas Lange [Fri, 24 Apr 2009 14:22:16 +0000 (16:22 +0200)]
MIPS: Implement ethernet halt for au1x00

Implement ethernet halt() by putting MAC0 in reset.
If we do not do this, we will get memory corruption
when ethernet frames are received during early OS boot.

Signed-off-by: Thomas Lange <thomas@corelatus.se>
Signed-off-by: Shinya Kuribayashi <skuribay@pobox.com>
15 years agoPrepare v2009.06-rc2 v2009.06-rc2
Wolfgang Denk [Fri, 15 May 2009 21:29:23 +0000 (23:29 +0200)]
Prepare v2009.06-rc2

Update CHANGELOG.

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoARM: fix PXA build by defining UP2OCR
Daniel Mack [Tue, 5 May 2009 10:48:29 +0000 (12:48 +0200)]
ARM: fix PXA build by defining UP2OCR

U-Boot does not currently build for PXA platforms with USB support
enabled:

usb.c:46: error: 'UP2OCR' undeclared (first use in this function)

Signed-off-by: Daniel Mack <daniel@caiaq.de>
Cc: Markus Klotzbuecher <mk@denx.de>
Edited commit message.
Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoMinor Coding Style fix; update CHANGELOG.
Wolfgang Denk [Fri, 15 May 2009 20:32:57 +0000 (22:32 +0200)]
Minor Coding Style fix; update CHANGELOG.

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoIDE: bail out of dev_print() for unknown device types
Wolfgang Denk [Fri, 15 May 2009 07:27:58 +0000 (09:27 +0200)]
IDE: bail out of dev_print() for unknown device types

Commit 574b319512 introduced a subtle bug by mixing a list of tests
for "dev_desc->type" and "dev_desc->if_type" into one switch(), which
then mostly did not work because "dev_desc->type" cannot take any
"IF_*" type values. A later fix in commit 8ec6e332ea changed the
switch() into testing "dev_desc->if_type", but at this point the
initial test for unknown device types was completely lost, which
resulted in output like that for IDE ports without device attached:

  Device 1: Model:  Firm:  Ser#:
            Type: # 1F #
            Capacity: not available

This patch re-introduces the missing test for unknown device types.

Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Stefan Roese <sr@denx.de>
Cc: Detlev Zundel <dzu@denx.de>
Tested-by: Stefan Roese <sr@denx.de>
15 years ago74xx_7xx: Fix rounding problem in CPU frequency calculation
Stefan Roese [Thu, 14 May 2009 05:25:13 +0000 (07:25 +0200)]
74xx_7xx: Fix rounding problem in CPU frequency calculation

This patch fixes a problem in the CPU frequency calculation. Without it
a 798MHz CPU is displayed as 368.503 MHz. And with it it's 798 MHz.

Signed-off-by: Stefan Roese <sr@denx.de>
15 years agoconsole.h: remove unused prototype 'console_realloc'
Jean-Christophe PLAGNIOL-VILLARD [Wed, 13 May 2009 20:24:12 +0000 (22:24 +0200)]
console.h: remove unused prototype 'console_realloc'

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
15 years agoapi: remove un-needed ifdef CONFIG_API already handle by the Makefile
Jean-Christophe PLAGNIOL-VILLARD [Wed, 13 May 2009 20:16:31 +0000 (22:16 +0200)]
api: remove un-needed ifdef CONFIG_API already handle by the Makefile

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
15 years agoFix e-mail address of Gary Jennejohn.
Detlev Zundel [Wed, 13 May 2009 08:54:10 +0000 (10:54 +0200)]
Fix e-mail address of Gary Jennejohn.

Signed-off-by: Detlev Zundel <dzu@denx.de>
15 years agoRemove inline qualifier from show_boot_progress()
Emil Medve [Tue, 12 May 2009 18:48:32 +0000 (13:48 -0500)]
Remove inline qualifier from show_boot_progress()

The 'inline' is conflicting with the semantic of 'weak' attribute and with the
way the show_boot_progress() function is used.

Also gcc 4.4 is complaining about it:

main.c:51: error: inline function 'show_boot_progress' cannot be declared weak

Signed-off-by: Emil Medve <Emilian.Medve@Freescale.com>
15 years agoMPC8260: fixup device tree by property instead of path
Wolfgang Denk [Tue, 12 May 2009 13:17:35 +0000 (15:17 +0200)]
MPC8260: fixup device tree by property instead of path

cpu/mpc8260/cpu.c used to use do_fixup_by_path_u32() to update the
clock frequencies in the device tree, using a CPU path
"/cpus/OF_CPU", with OF_CPU beind defined in the board config file.

However, this does not work when one board config file (here:
MPC8260ADS.h) is intended to be used for several diffrent CPUs and
therefor contains a generic definition like "cpu@0", as the device
trees that will then be loaded will contain specific names like
"PowerPC,8272@0".

We switch to using do_fixup_by_prop_u32() instead, so we can search
for device_type="cpu", as it is done in other architectures, too.

Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Heiko Schocher <hs@denx.de>
Acked-by: Heiko Schocher <hs@denx.de>
Tested-by: Heiko Schocher <hs@denx.de>
15 years agoFix OneNAND ipl to read CONFIG_SYS_MONITOR_LEN
Rohit Hagargundgi [Mon, 9 Mar 2009 14:15:46 +0000 (19:45 +0530)]
Fix OneNAND ipl to read CONFIG_SYS_MONITOR_LEN

Add CONFIG_SYS_MONITOR_LEN macro to apollon board config.
CONFIG_SYS_MONITOR_LEN defines the U-Boot image size.
and is used by OneNAND ipl when reading U-Boot image.

Signed-off-by: Rohit Hagargundgi <h.rohit at samsung.com>
Acked-by: Kyungmin Park <kyungmin.park@samsung.com>
15 years agoSchedule removal of non-CONFIG_NET_MULTI net driver API
Ben Warren [Wed, 29 Apr 2009 00:04:51 +0000 (17:04 -0700)]
Schedule removal of non-CONFIG_NET_MULTI net driver API

This will make CONFIG_NET_MULTI the only net driver configuration and
we'll be able to remove this option.

Signed-off-by: Ben Warren <biggerbadderben@gmail.com>
15 years agopowerpc/inka4x0: Remove left-over ide reset code.
Detlev Zundel [Tue, 5 May 2009 14:04:06 +0000 (16:04 +0200)]
powerpc/inka4x0: Remove left-over ide reset code.

The pin which was used in preliminary versions of the board for ide
reset is really connected to the rtc clock.

Signed-off-by: Detlev Zundel <dzu@denx.de>
15 years agoAdd imls utility command
Marco Stornelli [Tue, 28 Apr 2009 17:04:02 +0000 (19:04 +0200)]
Add imls utility command

This patch adds, under tools folder, a new command called imls. Its
goal is the same of UBoot's imls but it can be used as Linux shell
command. It reads from raw mtd partition and prints the list of the
stored images.

Signed-off-by: Marco Stornelli <marco.stornelli@gmail.com>
15 years agonetloop: updates for NetLoop
Heiko Schocher [Tue, 28 Apr 2009 06:36:11 +0000 (08:36 +0200)]
netloop: updates for NetLoop

Fix some issues introduced from commit:
2f70c49e5b9813635ad73666aa30f304c7fdeda9
suggested by Mike Frysinger.

- added some comment for the env_id variable in common_cmd_nvedit.c
- moved some variables in fn scope instead of file scope
- NetInitLoop now static void

Signed-off-by: Heiko Schocher <hs@denx.de>
Acked-by: Ben Warren <biggerbadderben@gmail.com>
15 years ago82xx, ids8247: added ids8247 board to MAKEALL script
Heiko Schocher [Tue, 28 Apr 2009 05:48:39 +0000 (07:48 +0200)]
82xx, ids8247: added ids8247 board to MAKEALL script

Signed-off-by: Heiko Schocher <hs@denx.de>
15 years agoarm/imx31_phycore: Fix bi_arch_number
Detlev Zundel [Thu, 7 May 2009 11:08:55 +0000 (13:08 +0200)]
arm/imx31_phycore: Fix bi_arch_number

Signed-off-by: Detlev Zundel <dzu@denx.de>
Cc: Sascha Hauer <s.hauer@pengutronix.de>
15 years agoremove myself as phycore/litekit Maintainer
Sascha Hauer [Fri, 15 May 2009 08:21:33 +0000 (10:21 +0200)]
remove myself as phycore/litekit Maintainer

I never acked a patch that adds me as phycore i.MX31 maintainer nor was
it me who pushed the patches, so remove myself from the maintainer list
so that other people do not longer wait for my ack.

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
15 years agoAFEB9260 network fix
Sergey Lapin [Tue, 12 May 2009 08:25:14 +0000 (12:25 +0400)]
AFEB9260 network fix

AFEB9260 uses PA10, PA11 for ETX2 and ETX3.
Also, due to extarnal pull-up on IRQ line, Micrel PHY ID is 1 after reset sequence,
not 0.

Signed-off-by: Sergey Lapin <slapin@ossfans.org>
15 years agoBlackfin: bf518f-ezbrd: reset ethernet PHY during init
Graf Yang [Tue, 5 May 2009 06:26:27 +0000 (02:26 -0400)]
Blackfin: bf518f-ezbrd: reset ethernet PHY during init

We don't know what state the ethernet PHY is in when starting up, so make
sure we set it to a sane state.  This fixes troubles seen when Linux boots
up, configures the PHY is a non-default state, and then the system reboots
into U-Boot which previously expected a reset state only.

Signed-off-by: Graf Yang <graf.yang@analog.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agoBlackfin: bf527-ezkit: fix SPI flash env params
Mike Frysinger [Tue, 5 May 2009 05:35:41 +0000 (01:35 -0400)]
Blackfin: bf527-ezkit: fix SPI flash env params

The BF527-EZKIT settings for storing the environment in SPI flash wasn't
using the correct sector settings for the SPI flash part that is actually
on the board.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agoBlackfin: avoid get_sclk() with early serial debug
Mike Frysinger [Sat, 25 Apr 2009 03:54:19 +0000 (23:54 -0400)]
Blackfin: avoid get_sclk() with early serial debug

When the clock functions were changed to use cached values (and thereby
avoiding expensive math functions), early serial debug broke because the
baud programming is called before external memory is available.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agoBlackfin: fix booting with older bootroms (no EVT1)
Mike Frysinger [Sat, 25 Apr 2009 03:39:41 +0000 (23:39 -0400)]
Blackfin: fix booting with older bootroms (no EVT1)

When dropping jump block support, the assumption was that all bootroms
supported entry point redirection via the EVT1 register.  Unfortunately,
this turned out to be incorrect for the oldest Blackfin parts (BF533-0.2
and older and BF561).  No one really noticed earlier because these parts
usually are booted by bypassing the bootrom entirely, and older BF533
parts are not supported at all (too many anomalies).

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agoBlackfin: recurse with early serial initcode
Mike Frysinger [Sat, 25 Apr 2009 03:22:48 +0000 (23:22 -0400)]
Blackfin: recurse with early serial initcode

Make sure we recurse through serial_putc() rather than bang on the UART
transmit register directly to avoid hardware overflows when using \n.

Signed-off-by: Mike Frysinger <vapier@gentoo.org>
15 years agoPrepare v2009.06-rc1 v2009.06-rc1
Wolfgang Denk [Fri, 1 May 2009 19:59:12 +0000 (21:59 +0200)]
Prepare v2009.06-rc1

Update CHANGELOG.

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoat91: remove lowlevel_init.S
Jean-Christophe PLAGNIOL-VILLARD [Fri, 1 May 2009 13:38:06 +0000 (15:38 +0200)]
at91: remove lowlevel_init.S

lowlevel_init.S is not used any more so remove it.
As consequence, we also don't have to generate u-boot.lds
but can use a static version as before.

This also fixes the out-of-tree build problem introduced
with commit f0a2c7b4 "at91: add support for the PM9263 board"

Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoUpdate CHANGELOG; minor coding style cleanup.
Wolfgang Denk [Thu, 30 Apr 2009 22:16:11 +0000 (00:16 +0200)]
Update CHANGELOG; minor coding style cleanup.

Signed-off-by: Wolfgang Denk <wd@denx.de>
15 years agoOMAP3: Fix timer handling to 1ms and CONFIG_SYS_HZ to 1000
Manikandan Pillai [Tue, 21 Apr 2009 15:29:05 +0000 (17:29 +0200)]
OMAP3: Fix timer handling to 1ms and CONFIG_SYS_HZ to 1000

Signed-off-by: Dirk Behme <dirk.behme@googlemail.com>
Signed-off-by: Manikandan Pillai <mani.pillai@ti.com>
15 years agoMerge branch 'master' of /home/wd/git/u-boot/custodians
Wolfgang Denk [Thu, 30 Apr 2009 21:54:07 +0000 (23:54 +0200)]
Merge branch 'master' of /home/wd/git/u-boot/custodians

15 years agoOMAP3: Beagle: Set pinmux conditionally for Rev C boards
Dirk Behme [Fri, 17 Apr 2009 12:27:56 +0000 (14:27 +0200)]
OMAP3: Beagle: Set pinmux conditionally for Rev C boards

The Beagle Rev C boards pull UART2 from an alternate set of balls.

Signed-off-by: Steve Sakoman <steve@sakoman.com>
Signed-off-by: Dirk Behme <dirk.behme@googlemail.com>
15 years agoubifs: BUG: Blocks commpressed with zlib
Ricardo Ribalda Delgado [Mon, 27 Apr 2009 16:33:33 +0000 (18:33 +0200)]
ubifs: BUG: Blocks commpressed with zlib

Blocks compressed with zlib dont have the full gzip header.

Without this patch, block compressed with zlib cannot be readed!

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@uam.es>
15 years agolib_generic: gunzip: New function zunzip
Ricardo Ribalda Delgado [Mon, 27 Apr 2009 16:33:32 +0000 (18:33 +0200)]
lib_generic: gunzip: New function zunzip

Separate gunzip in

gunzip: Find the end of the header and call zunzip.
zunzip: Inflate gunzip block without header.

UBI fs blocks can be compresed in lzo, zlib or no-compression. The
current implementation of u-boot supported all the compressions but
there was a bug in the implementation of the zlib blocks.

UBIFS's Zlib blocks do not have header but they were compressed using
gunzip, a function used to decompress gunzip files/sectors with a
header.

This patch adds a new function zunzip that uncompress a zlib block with
no header.

Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda@uam.es>
15 years agoMerge branch 'master' of git://git.denx.de/u-boot-arm
Wolfgang Denk [Thu, 30 Apr 2009 21:01:08 +0000 (23:01 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-arm

15 years agoMerge branch 'master' of git://git.denx.de/u-boot-cfi-flash
Wolfgang Denk [Thu, 30 Apr 2009 20:59:45 +0000 (22:59 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-cfi-flash

15 years agoMerge branch 'master' of git://git.denx.de/u-boot-video
Wolfgang Denk [Thu, 30 Apr 2009 20:58:17 +0000 (22:58 +0200)]
Merge branch 'master' of git://git.denx.de/u-boot-video

15 years agoat91: fixed plla calc when no USB support is active
Daniel Gorsulowski [Thu, 23 Apr 2009 13:37:16 +0000 (15:37 +0200)]
at91: fixed plla calc when no USB support is active

Signed-off-by: Daniel Gorsulowski <Daniel.Gorsulowski@esd.eu>
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
15 years agoarm925t: Fix CONFIG_SYS_HZ to 1000
Ladislav Michl [Tue, 21 Apr 2009 23:12:04 +0000 (01:12 +0200)]
arm925t: Fix CONFIG_SYS_HZ to 1000

Let CONFIG_SYS_HZ to have value of 1000 effectively fixing all users of
get_timer.

Changes since original version:
* Set PTV=2 (divisor 8) for boards using 12MHz timer clock source to
  improve timer resolution.

Signed-off-by: Ladislav Michl <ladis@linux-mips.org>
15 years agoOMAP3: Remove legacy NAND defines
Dirk Behme [Tue, 14 Apr 2009 18:15:17 +0000 (20:15 +0200)]
OMAP3: Remove legacy NAND defines

Remove remaining legacy NAND defines for Beagle, EVM, Overo and Pandora.

Signed-off-by: Dirk Behme <dirk.behme@googlemail.com>
15 years agoOMAP3: Print correct silicon revision
Sanjeev Premi [Mon, 27 Apr 2009 15:57:54 +0000 (21:27 +0530)]
OMAP3: Print correct silicon revision

The function display_board_info() displays incorrect
silicon revision - based on the return value from
function get_cpu_rev().

This patch fixes the problem.

Signed-off-by: Sanjeev Premi <premi@ti.com>
15 years agoOMAP3: Remove unused board-types
Sanjeev Premi [Mon, 27 Apr 2009 15:57:44 +0000 (21:27 +0530)]
OMAP3: Remove unused board-types

The board-types defined in struct omap3_sysinfo seem to be
unused. The function display_board_info() is passed
board type as an argument; which is ignored.

This patch removes all uses of board-type, related definitions
and functions.

Signed-off-by: Sanjeev Premi <premi@ti.com>
15 years agoOMAP3: Use functions print_cpuinfo() and checkboard()
Sanjeev Premi [Mon, 27 Apr 2009 15:57:27 +0000 (21:27 +0530)]
OMAP3: Use functions print_cpuinfo() and checkboard()

Use the functions print_cpuinfo() and checkboard() to
display the cpu and board specific information.

These functions reuse content from the existing function
display_board_info() - which has been removed.

Also, updated the existig OMAP3 configurations to
define:
 - CONFIG_DISPLAY_CPUINFO
 - CONFIG_DISPLAY_BOARDINFO

Signed-off-by: Sanjeev Premi <premi@ti.com>
15 years agoMTD: Change cfi-mtd to accept non-uniform sector sizes
Stefan Roese [Wed, 18 Mar 2009 10:17:37 +0000 (11:17 +0100)]
MTD: Change cfi-mtd to accept non-uniform sector sizes

With this patch non-uniform NOR FLASH chips (chips with multiple erase
regions) can be exported via the cfi-mtd layer and therefor used by UBI.
We select the largest sector size as erasesize. The cfi driver will make
sure that the smaller sectors are handled correctly.

Signed-off-by: Stefan Roese <sr@denx.de>