oweals/u-boot.git
12 years agodisk: part_efi: add new partition attribute definitions
Stephen Warren [Mon, 8 Oct 2012 08:14:35 +0000 (08:14 +0000)]
disk: part_efi: add new partition attribute definitions

Add no_block_io_protocol and legacy_bios_bootable attribute definitions.
These are sourced from UEFI Spec 2.3, page 105, table 19. Credits to the
libparted source for the specification pointer.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
12 years agodisk: part_efi: print partition UUIDs
Stephen Warren [Mon, 8 Oct 2012 08:14:34 +0000 (08:14 +0000)]
disk: part_efi: print partition UUIDs

When printing the partition table, print the partition type UUID and the
individual partition UUID. Do this unconditionally, since partition UUIDs
are useful.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
12 years agodisk: part_efi: re-order partition list printf, change case
Stephen Warren [Mon, 8 Oct 2012 08:14:33 +0000 (08:14 +0000)]
disk: part_efi: re-order partition list printf, change case

The partition name is a long variable-length string. Move it last on
the line to ensure consistent layout and that the entries align with
the "header" line. Also, surround it in quotes, so if it's empty, it's
obvious that something is still being printed.

Also, change the case of the LBA numbers; lower-case looks nicer in my
opinion, and will be more consistent with the UUID printing that is
added later in this series.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
12 years agodisk: part_efi: remove indent level from loop
Stephen Warren [Mon, 8 Oct 2012 08:14:32 +0000 (08:14 +0000)]
disk: part_efi: remove indent level from loop

Simplify the partition printing loop in print_part_efi() to bail out
early when the first invalid partition is found, rather than indenting
the whole body of the loop. This simplifies later patches.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
12 years agodisk: get_device_and_partition() return value fixes
Stephen Warren [Mon, 8 Oct 2012 07:45:54 +0000 (07:45 +0000)]
disk: get_device_and_partition() return value fixes

When no valid partitions are found, guarantee that we return -1. This
most likely already happens, since the most recent get_partition_info()
will have returned an error. However, it's best to be explicit.

Remove an unnecessary assignment of ret=0 in the success case; this value
is over-written with the processed partition ID later.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
12 years agoremove unnecessary includes from cmd_ide.c
Pavel Herrmann [Sun, 7 Oct 2012 05:56:14 +0000 (05:56 +0000)]
remove unnecessary includes from cmd_ide.c

mpc8xx and mpc5xxx specific includes in cmd_ide.c are not required, remove them.

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agosplit PCS440EP specific code from cmd_ide.c
Pavel Herrmann [Sun, 7 Oct 2012 05:56:13 +0000 (05:56 +0000)]
split PCS440EP specific code from cmd_ide.c

Move specific ide_input_data and friends to board-specific file.

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agosplit AU1X00 specific code from cmd_ide.c
Pavel Herrmann [Tue, 9 Oct 2012 07:10:08 +0000 (07:10 +0000)]
split AU1X00 specific code from cmd_ide.c

move special case of ide_swap_read() for AU1X00 SoC into SoC-specific directory.

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agomove CPC45 ide_led to the same file as other IDE hooks
Pavel Herrmann [Tue, 9 Oct 2012 07:06:25 +0000 (07:06 +0000)]
move CPC45 ide_led to the same file as other IDE hooks

Keep all IDE-related hooks and overrides in a single file, to avoid confusion

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agomake ide_led() a weak alias
Pavel Herrmann [Sun, 7 Oct 2012 05:56:10 +0000 (05:56 +0000)]
make ide_led() a weak alias

Make ide_led() a weak alias instead of global/local function/empty macro
based on CONFIG_IDE_LED value and/or board-specific CONFIGs, to get rid of
board-specific code in cmd_ide.c
Define dummy values to get rid of compoler errors in case where ide_led()
used to be an empty macro

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agosplit CPC45 board-specific IDE functions from cmd_ide.c
Pavel Herrmann [Sun, 7 Oct 2012 05:56:09 +0000 (05:56 +0000)]
split CPC45 board-specific IDE functions from cmd_ide.c

Move input_data() and friends to board/cpc45/ide.c, as overrides for weak
aliases in cmd_ide.c

note: checkpatch emits warnings about using volatile

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agochange all versions of input_data() and output_data() to global weak aliases
Pavel Herrmann [Tue, 9 Oct 2012 07:04:39 +0000 (07:04 +0000)]
change all versions of input_data() and output_data() to global weak aliases

This changes input_data() and friends from static function to global symbols
under weak alias, to enable board specific overrides (and therefore get rid of
board-specific code in cmd_ide.c)
Also declare ide_bus_offset in the header file, so other files can use
ATA_CURR_BASE as well.

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agosplit IVM power hooks from cmd_ide.c
Pavel Herrmann [Sun, 7 Oct 2012 05:56:07 +0000 (05:56 +0000)]
split IVM power hooks from cmd_ide.c

Move power control code from ide_reset() into IVM-specific IDE reset code.

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agosplit mpc8xx hooks from cmd_ide.c
Pavel Herrmann [Tue, 9 Oct 2012 07:01:56 +0000 (07:01 +0000)]
split mpc8xx hooks from cmd_ide.c

move most of mpc8xx hooks from cmd_ide.c into ide_preinit() and newly created
ide_init_postreset() (invoked after calling ide_reset after ide_preinit),
some cleanup to make checkpatch happy, enable IDE init hooks in configs of
affected boards.
confusingly, these hooks are used by more than just mpc8xx-based boards, and
therefore are placed in arch/ppc/lib/

note: checkpatch still emits warnings about using volatile

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agoremove CONFIG_SC3 from cmd_ide.c
Pavel Herrmann [Sun, 7 Oct 2012 05:56:05 +0000 (05:56 +0000)]
remove CONFIG_SC3 from cmd_ide.c

There is no difference in codepath with CONFIG_SC3 enabled, so just remove it

Signed-off-by: Pavel Herrmann <morpheus.ibis@gmail.com>
12 years agokerneldoc: stdio: tmpl: Add stdio template
Marek Vasut [Sat, 6 Oct 2012 14:07:06 +0000 (14:07 +0000)]
kerneldoc: stdio: tmpl: Add stdio template

Add STDIO documentation template.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Marek Vasut <marek.vasut@gmail.com>
Cc: Tom Rini <trini@ti.com>
[trini: Fix DOCBOOKS line as we don't have linker_lists yet]
Signed-off-by: Tom Rini <trini@ti.com>
12 years agokerneldoc: Annotate drivers/serial/serial.c
Marek Vasut [Mon, 8 Oct 2012 11:36:39 +0000 (11:36 +0000)]
kerneldoc: Annotate drivers/serial/serial.c

Add kerneldoc annotations into serial core.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Marek Vasut <marek.vasut@gmail.com>
Cc: Tom Rini <trini@ti.com>
12 years agoserial: Reorder get_current()
Marek Vasut [Sat, 6 Oct 2012 14:07:04 +0000 (14:07 +0000)]
serial: Reorder get_current()

Reorder the get_current() function to make it a bit more readable.
The code does not grow and there is minor change in the code logic,
where dev != NULL is now checked in any case.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Marek Vasut <marek.vasut@gmail.com>
Cc: Tom Rini <trini@ti.com>
12 years agoserial: Reorder serial_assign()
Marek Vasut [Sat, 6 Oct 2012 14:07:03 +0000 (14:07 +0000)]
serial: Reorder serial_assign()

Reorder serial_assign() function to get rid of the extra level of
indentation. Also, adjust the return value to be -EINVAL instead of
positive one to be more consistent.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Marek Vasut <marek.vasut@gmail.com>
Cc: Tom Rini <trini@ti.com>
12 years agoserial: Use default_serial_puts() in drivers
Marek Vasut [Sat, 6 Oct 2012 14:07:02 +0000 (14:07 +0000)]
serial: Use default_serial_puts() in drivers

Replace the in-place ad-hoc implementation of serial_puts() within
the drivers with default_serial_puts() call. This cuts down on the
code duplication quite a bit.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Marek Vasut <marek.vasut@gmail.com>
Cc: Tom Rini <trini@ti.com>
12 years agoserial: Implement default_serial_puts()
Marek Vasut [Sat, 6 Oct 2012 14:07:01 +0000 (14:07 +0000)]
serial: Implement default_serial_puts()

U-Boot contains a lot of duplicit implementations of serial_puts()
call which just pipes single characters into the port in loop. Implement
function that does this behavior into common code, so others can make
easy use of it.

This function is called default_serial_puts() and it's sole purpose
is to call putc() in loop on the whole string passed to it.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Marek Vasut <marek.vasut@gmail.com>
Cc: Tom Rini <trini@ti.com>
12 years agompc82xx: Remove BMW board port
Stefan Roese [Wed, 19 Sep 2012 03:18:53 +0000 (03:18 +0000)]
mpc82xx: Remove BMW board port

As the board seems to be unmaintained for some time, lets remove
the support in mainline completely.

Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Marek Vasut <marek.vasut@gmail.com>
Cc: James F. Dougherty <jfd@broadcom.com>
Cc: Wolfgang Denk <wd@denx.de>
Acked-by: Marek Vasut <marex@denx.de>
12 years agonds32: Change macro from BOARD_LATE_INIT to CONFIG_BOARD_LATE_INIT
Nobuhiro Iwamatsu [Tue, 17 Apr 2012 16:41:14 +0000 (16:41 +0000)]
nds32: Change macro from BOARD_LATE_INIT to CONFIG_BOARD_LATE_INIT

With almost all the architecture and board BOARD_LATE_INIT does not use.
CONFIG_BOARD_LATE_INIT is used instead.
This changed CONFIG_BOARD_LATE_INIT from BOARD_LATE_INIT.

Signed-off-by: Nobuhiro Iwamatsu <nobuhiro.iwamatsu.yj@renesas.com>
CC: Macpaul Lin <macpaul@andestech.com>
[trini: Fixup for context changes]
Signed-off-by: Tom Rini <trini@ti.com>
12 years agotools/env: Improve debug prints
Joe Hershberger [Mon, 15 Oct 2012 15:29:25 +0000 (15:29 +0000)]
tools/env: Improve debug prints

Provide more information when using redundant environments
Consistently print debug info to stderr

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
12 years agotools/env: Fix variable delete operation
Joe Hershberger [Mon, 15 Oct 2012 15:29:24 +0000 (15:29 +0000)]
tools/env: Fix variable delete operation

Fix crash introduced by a073d63a36524453a817ab029fad5b188f46127e
when attempting to delete a variable.

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
12 years agoMerge branch 'agust@denx.de-next' of git://git.denx.de/u-boot-staging
Tom Rini [Mon, 15 Oct 2012 20:37:22 +0000 (13:37 -0700)]
Merge branch 'agust@denx.de-next' of git://git.denx.de/u-boot-staging

12 years agotools/env: Fix build failure from missing header include
Joe Hershberger [Fri, 12 Oct 2012 10:23:37 +0000 (10:23 +0000)]
tools/env: Fix build failure from missing header include

This was introduced in:
8679d0ffdcc0beafea8e6942c0c67cf859afa18e -
COMMON: Use __stringify() instead of MK_STR()

The header is now needed since common.h is not included in this tool.

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
12 years agokerneldoc: Add myself to the git-mailrc for kerneldoc
Marek Vasut [Sat, 6 Oct 2012 14:05:01 +0000 (14:05 +0000)]
kerneldoc: Add myself to the git-mailrc for kerneldoc

Add entry for kerneldoc into the git-mailrc pointing to the U-Boot ML
and myself.

Signed-off-by: Marek Vasut <marex@denx.de>
12 years agokerneldoc: Implement "Example" section handling
Marek Vasut [Sat, 6 Oct 2012 14:05:00 +0000 (14:05 +0000)]
kerneldoc: Implement "Example" section handling

The default kernel-doc strips starting spaces from every single
line in the Example section. This makes the code look bad. Thus,
implement special handling for this section.

Signed-off-by: Marek Vasut <marex@denx.de>
12 years agokerneldoc: Add nicer CSS stylesheet for HTML docs
Marek Vasut [Sat, 6 Oct 2012 14:04:59 +0000 (14:04 +0000)]
kerneldoc: Add nicer CSS stylesheet for HTML docs

Import basic CSS stylesheet for the HTML documentation. The base for
the stylesheet is taken from:

http://ds9a.nl/docbook/minimal-page.html

I customized the CSS a bit further, for example to add curvy corners
to example section and change the tint of gray. The HTML documentation
does not look that crude anymore.

Signed-off-by: Marek Vasut <marex@denx.de>
12 years agokerneldoc: Implant DocBook from Linux kernel
Marek Vasut [Sat, 6 Oct 2012 14:04:58 +0000 (14:04 +0000)]
kerneldoc: Implant DocBook from Linux kernel

Pull slightly modified version of Documentation/DocBook, the related perl
script scripts/kernel-doc and the scripts/docproc.c from Linux kernel and
implant it into U-Boot. This will allow smooth generation of kerneldoc
style documentation.

It was necessary to modify the DocBook/Makefile to work with U-Boot build
system. The changes were only minor though and involved replacing the kbuild
specific parts.

It was also necessary to replace use of variables like KERNEL_VERSION with
U_BOOT_VERSION, strings like Linux kernel with U-Boot Bootloader etc. so
the generated result actually matches.

Finally, it was necessary to adjust docproc.c, since the documentation in
U-Boot is located in doc/DocBook instead of Documentation/DocBook as is in
case of the Linux kernel.

Some parts of the DocBook Makefile are unused, but to allow easier sync with
Linux kernel, these parts are still left in. The targets enabled now are
"htmldocs" "pdfdocs" "psdocs" "xmldocs" and "cleandocs" to remove the results
of documentation build.

Linux scripts/docproc.c:
commit f0f3ca8d967462dafb815412b14ca3339b9817a6
Date:   Wed Jun 15 11:53:13 2011 +0200

Linux scripts/kernel-doc:
commit 1b40c1944db445c1de1c47ffd8cd426167f488e8
Date:   Sun Aug 12 10:46:15 2012 +0200

Linux Documentation/DocBook:
commit bb8187d35f820671d6dd76700d77a6b55f95e2c5
Date:   Thu May 17 19:06:13 2012 -0400

Signed-off-by: Marek Vasut <marex@denx.de>
12 years agoconfigs: mx6qsabre_common.h: Use default clock definitions
Fabio Estevam [Sun, 30 Sep 2012 05:07:10 +0000 (05:07 +0000)]
configs: mx6qsabre_common.h: Use default clock definitions

Since commit 50d4a707f0 (mx5/6: Define default SoC input clock frequencies)
we can use the default clock values.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
12 years agomx6qsabreauto: Add Ethernet support
Fabio Estevam [Tue, 25 Sep 2012 08:43:57 +0000 (08:43 +0000)]
mx6qsabreauto: Add Ethernet support

mx6qsabreauto has a AR8031 Gigabit PHY.

Add support for it.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
12 years agomx25: Clean up lowlevel_init
Benoît Thébaudeau [Mon, 20 Aug 2012 09:00:57 +0000 (09:00 +0000)]
mx25: Clean up lowlevel_init

Clean up mx25 lowlevel_init:
 - Add comments.
 - Do not use write32 repeatedly with the same value in order not to increase
   code size.
 - Make register values configurable.
 - Use macro parameters with default values instead of literal constants.
 - Use defined macros instead of duplicating code.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: John Rigby <jcrigby@gmail.com>
Cc: Matthias Weisser <weisserm@arcor.de>
12 years agomx31: Fix PDR0_CSI_PODF
Benoît Thébaudeau [Tue, 14 Aug 2012 08:43:07 +0000 (08:43 +0000)]
mx31: Fix PDR0_CSI_PODF

The CSI PODF bit-field used by the previous code for the i.MX31 CCM PDR0
register is actually composed of two bit-fields: one pre-divider and one
post-divider. This patch fixes the CCM access macros and the code using them
accordingly.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx35: Define MAX and AIPS registers
Benoît Thébaudeau [Tue, 14 Aug 2012 11:03:59 +0000 (11:03 +0000)]
mx35: Define MAX and AIPS registers

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx31: Add more CCM access macros
Benoît Thébaudeau [Tue, 14 Aug 2012 08:43:29 +0000 (08:43 +0000)]
mx31: Add more CCM access macros

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5: Optimize lowlevel_init code size
Benoît Thébaudeau [Tue, 14 Aug 2012 05:19:12 +0000 (05:19 +0000)]
mx5: Optimize lowlevel_init code size

Optimize mx5 lowlevel_init.S code size:
 - Compute values at compile time rather than at runtime where possible.
 - Assign r4 to hold the zero value rather than setting registers to 0 again and
   again.
 - Associate a function to setup_pll rather than expanding its large macro code
   multiple times.
 - Allocate constant values in section only if used.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
Tested-by: Stefano Babic <sbabic@denx.de>
Acked-by: Stefano Babic <sbabic@denx.de>
12 years agomx6qsabreauto: Use ttymxc3 as console
Otavio Salvador [Wed, 26 Sep 2012 11:37:01 +0000 (11:37 +0000)]
mx6qsabreauto: Use ttymxc3 as console

The mx6qsabreauto console is different than mx6qsabresd so the console
configuration is now set in the board file.

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Acked-by: Stefano Babic <sbabic@denx.de>
12 years agomx25: Fix eSDHC support
Benoît Thébaudeau [Thu, 27 Sep 2012 10:28:29 +0000 (10:28 +0000)]
mx25: Fix eSDHC support

The MMC driver appropriate for the i.MX25 is fsl_esdhc, which has nothing to do
with mxcmmc.

Also, each eSDHC instance has a dedicated clock, so gd->sdhc_clk must be set
accordingly. This is good for the case only a single SDHC instance is used
(initialization made with fsl_esdhc_mmc_init()). A future patch will fix the
multi-instance use case (initialization made directly with
fsl_esdhc_initialize()).

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Eric Bénard <eric@eukrea.com>
Cc: Otavio Salvador <otavio@ossystems.com.br>
12 years agomx25: Define cpu_eth_init() only if needed
Benoît Thébaudeau [Thu, 27 Sep 2012 10:28:09 +0000 (10:28 +0000)]
mx25: Define cpu_eth_init() only if needed

The FEC is the only SoC Ethernet support available on i.MX25, so define
cpu_eth_init() only for it instead of returning a misleading success code.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx25: Clean up clocks API
Benoît Thébaudeau [Thu, 27 Sep 2012 10:27:57 +0000 (10:27 +0000)]
mx25: Clean up clocks API

Use the standard mxc_get_clock() instead of exporting internal functions and
using literal constant values.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx25 clocks: Fix MXC_FEC_CLK
Benoît Thébaudeau [Thu, 27 Sep 2012 10:27:44 +0000 (10:27 +0000)]
mx25 clocks: Fix MXC_FEC_CLK

mxc_get_clock(MXC_FEC_CLK) should return the IPG clock, not the AHB clock.

Also, imx_get_fecclk() was correct but reimplemented the calculation of the IPG
clock, so remove the duplicated code.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx25: Define more standard clocks
Benoît Thébaudeau [Thu, 27 Sep 2012 10:27:28 +0000 (10:27 +0000)]
mx25: Define more standard clocks

Define AHB, IPG and CSPI clocks.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx25: Clean up clock calculations
Benoît Thébaudeau [Thu, 27 Sep 2012 10:27:14 +0000 (10:27 +0000)]
mx25: Clean up clock calculations

Avoid possible overflow in clock calculations, and do not waste calls to lldiv()
to divide simple ulongs.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx25: Fix decode_pll
Benoît Thébaudeau [Thu, 27 Sep 2012 10:26:54 +0000 (10:26 +0000)]
mx25: Fix decode_pll

The MFN bit-field of the PLL registers represents a signed value. See the
reference manual.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5/6 clocks: Fix SDHC clocks
Benoît Thébaudeau [Thu, 27 Sep 2012 10:24:37 +0000 (10:24 +0000)]
mx5/6 clocks: Fix SDHC clocks

The i.MX5 eSDHC clocks were considered as coming from the IPG clock although
they have dedicated clock paths.

Also, on i.MX5/6, each SDHC instance has a dedicated clock, so gd->sdhc_clk must
be set accordingly. This is good for the case only a single SDHC instance is
used (initialization made with fsl_esdhc_mmc_init()). A future patch will fix
the multi-instance use case (initialization made directly with
fsl_esdhc_initialize()).

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Eric Bénard <eric@eukrea.com>
Cc: Otavio Salvador <otavio@ossystems.com.br>
12 years agomx51: Fix I2C clock ID check
Benoît Thébaudeau [Thu, 27 Sep 2012 10:24:13 +0000 (10:24 +0000)]
mx51: Fix I2C clock ID check

There are only 2 I²C instances on i.MX51, but 3 on i.MX53.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5 clocks: Fix MXC_FEC_CLK
Benoît Thébaudeau [Thu, 27 Sep 2012 10:23:58 +0000 (10:23 +0000)]
mx5 clocks: Fix MXC_FEC_CLK

The FEC clock does not come from PLL1, but from the IPG clock. The previous code
was even inconsistent with itself, returning the IPG clock as expected for
imx_get_fecclk(), but the PLL1 clock for mxc_get_clock(MXC_FEC_CLK).

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5 clocks: Simplify imx_get_cspiclk()
Benoît Thébaudeau [Thu, 27 Sep 2012 10:23:42 +0000 (10:23 +0000)]
mx5 clocks: Simplify imx_get_cspiclk()

The code handling the dividers was duplicated for each possible input clock, and
this function can benefit from the newly introduced get_standard_pll_sel_clk()
function instead of duplicating this mux handling code.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5 clocks: Fix get_uart_clk()
Benoît Thébaudeau [Thu, 27 Sep 2012 10:23:23 +0000 (10:23 +0000)]
mx5 clocks: Fix get_uart_clk()

This function returned 66500000 instead of the correct lp_apm clock frequency if
the CCM.CSCMR1.uart_clk_sel mux is set to 3.

This patch fixes this issue by introducing the get_standard_pll_sel_clk()
function that will be used by future patches to handle identical muxes used by
many other clocks.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5 clocks: Fix get_ipg_per_clk()
Benoît Thébaudeau [Thu, 27 Sep 2012 10:23:08 +0000 (10:23 +0000)]
mx5 clocks: Fix get_ipg_per_clk()

This fixes the "IPG PERCLK" frequency printed by the clocks command. The issue
was that get_ipg_per_clk() used periph_clk instead of lp_apm in the case
CCM.CBCMR.perclk_lp_apm_sel is set.

It also fixes I²C support.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5 clocks: Fix get_periph_clk()
Benoît Thébaudeau [Thu, 27 Sep 2012 10:22:51 +0000 (10:22 +0000)]
mx5 clocks: Fix get_periph_clk()

In the case periph_clk comes from periph_apm_clk, the latter is selected by the
CCM.CBCMR.periph_apm_sel mux, which can source the lp_apm clock from its
input ♯2. get_periph_clk() returned 0 instead of the lp_apm clock frequency in
this case.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5 clocks: Fix get_lp_apm()
Benoît Thébaudeau [Thu, 27 Sep 2012 10:22:37 +0000 (10:22 +0000)]
mx5 clocks: Fix get_lp_apm()

If CCM.CCSR.lp_apm is set, the lp_apm clock is not necessarily 32768 Hz x 1024.
In that case:
 - on i.MX51, this clock comes from the output of the FPM,
 - on i.MX53, this clock comes from the output of PLL4.

This patch fixes the code accordingly.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5 clocks: Add and use CCSR definitions
Benoît Thébaudeau [Thu, 27 Sep 2012 10:22:22 +0000 (10:22 +0000)]
mx5 clocks: Add and use CCSR definitions

This fixes config_pll_clk(), which used 0x20 instead of 0x200 for PLL4_CLOCK.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx51: Fix USB PHY clocks
Benoît Thébaudeau [Fri, 28 Sep 2012 07:09:03 +0000 (07:09 +0000)]
mx51: Fix USB PHY clocks

The i.MX51 has a single USB PHY clock, while the i.MX53 has two. These 3 clocks
have different clock gate control bit-fields.

The existing code was correct only for i.MX53, so this patch fixes the i.MX51
use case.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Marek Vasut <marex@denx.de>
Cc: Jana Rapava <fermata7@gmail.com>
Cc: Wolfgang Grandegger <wg@denx.de>
Cc: Igor Grinberg <grinberg@compulab.co.il>
12 years agomx5: Fix clock gate values
Benoît Thébaudeau [Thu, 27 Sep 2012 10:21:22 +0000 (10:21 +0000)]
mx5: Fix clock gate values

The clock gate values are 2-bit bit-fields. Hence, setting or clearing only one
of these bits like what was done is wrong and can lead to unpredictable behavior
depending on the original value of these bit-fields.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5: Use explicit clock gate names
Benoît Thébaudeau [Thu, 27 Sep 2012 10:21:00 +0000 (10:21 +0000)]
mx5: Use explicit clock gate names

Use clock gate definitions having names showing clearly the gated clock instead
of names giving only a register field index.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5 clocks: Cleanup
Benoît Thébaudeau [Thu, 27 Sep 2012 10:20:33 +0000 (10:20 +0000)]
mx5 clocks: Cleanup

Clean up the i.MX5 clock driver:
 - Use readl() and writel() instead of their __raw_ counterparts.
 - Use the clr/setbits_le32() family of macros rather than expanding code.
 - Use accessor macros for bit-fields instead of _MASK and _OFFSET.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
12 years agomx5/6: Define default SoC input clock frequencies
Benoît Thébaudeau [Thu, 27 Sep 2012 10:19:58 +0000 (10:19 +0000)]
mx5/6: Define default SoC input clock frequencies

Define default SoC input clock frequencies for i.MX5/6 in order to get rid of
duplicated definitions.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Jason Liu <r64343@freescale.com>
Cc: Matt Sealey <matt@genesi-usa.com>
Cc: Fabio Estevam <fabio.estevam@freescale.com>
12 years agoimx: Use MXC_I2C_CLK in imx i2c driver
Matthias Weisser [Mon, 24 Sep 2012 02:46:53 +0000 (02:46 +0000)]
imx: Use MXC_I2C_CLK in imx i2c driver

i2c didn't work on imx25 due to missing MXC_IPG_PERCLK. Now using
MXC_I2C_CLK on all imx systems using i2c.

Signed-off-by: Matthias Weisser <weisserm@arcor.de>
Acked-by: Stefano Babic <sbabic@denx.de>
12 years agomx25: add CPU revision 1.2
Eric Benard [Sun, 23 Sep 2012 02:03:05 +0000 (02:03 +0000)]
mx25: add CPU revision 1.2

tested on a MCIMX257CJM4A which now reports :
CPU:   Freescale i.MX25 rev1.2 at 399 MHz

Signed-off-by: Eric Bénard <eric@eukrea.com>
Acked-by: Otavio Salvador <otavio@ossystems.com.br>
12 years agomx6q: Add basic support for mx6qsabreauto
Fabio Estevam [Mon, 24 Sep 2012 08:09:33 +0000 (08:09 +0000)]
mx6q: Add basic support for mx6qsabreauto

mx6qsabreauto is a board based on mx6q SoC with the following features:
- 2GB of DDR3
- 2 USB ports
- 1 HDMI output port
- SPI NOR
- 2 LVDS LCD ports
- Gigabit Ethernet
- Camera
- eMMC and SD card slot
- Multichannel Audio
- CAN
- SATA
- NAND
- PCIE
- Video Input

Add very basic support for it.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
12 years agoconfigs: mx6: Add a common config file
Fabio Estevam [Mon, 24 Sep 2012 08:09:32 +0000 (08:09 +0000)]
configs: mx6: Add a common config file

Add a common mx6 config file that can be shared between some mx6 boards.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
12 years agoi.MX6: get rid of redundant struct src_regs (dupe of struct src)
Eric Nelson [Tue, 18 Sep 2012 15:26:32 +0000 (15:26 +0000)]
i.MX6: get rid of redundant struct src_regs (dupe of struct src)

Signed-off-by: Eric Nelson <eric.nelson@boundarydevices.com>
12 years agoi.MX6: define struct iomuxc and IOMUX_GPR2 register bitfields
Eric Nelson [Wed, 19 Sep 2012 08:32:31 +0000 (08:32 +0000)]
i.MX6: define struct iomuxc and IOMUX_GPR2 register bitfields

Signed-off-by: Eric Nelson <eric.nelson@boundarydevices.com>
Acked-by: Stefano Babic <sbabic@denx.de>
12 years agoi.MX6: Add ANATOP_PFD_480 bitfield constants
Eric Nelson [Wed, 19 Sep 2012 08:29:46 +0000 (08:29 +0000)]
i.MX6: Add ANATOP_PFD_480 bitfield constants

Signed-off-by: Eric Nelson <eric.nelson@boundarydevices.com>
Acked-by: Stefano Babic <sbabic@denx.de>
12 years agoi.MX6: define IOMUX_GPR3 register bitfields
Eric Nelson [Fri, 21 Sep 2012 11:41:42 +0000 (11:41 +0000)]
i.MX6: define IOMUX_GPR3 register bitfields

Signed-off-by: Eric Nelson <eric.nelson@boundarydevices.com>
12 years agoi.MX6: define bitfields for CHSCCDR register
Eric Nelson [Wed, 19 Sep 2012 08:33:50 +0000 (08:33 +0000)]
i.MX6: define bitfields for CHSCCDR register

Signed-off-by: Eric Nelson <eric.nelson@boundarydevices.com>
12 years agoi.MX6: change register name for CCM_CHSCCDR to match ref. manual
Eric Nelson [Mon, 17 Sep 2012 10:20:50 +0000 (10:20 +0000)]
i.MX6: change register name for CCM_CHSCCDR to match ref. manual

Register CCM_CHSCCDR (offset 0x34 in CCM) is named CCM_CHSCCDR in
reference manual, but was named chscdr in struct mxc_ccm_reg.

Signed-off-by: Eric Nelson <eric.nelson@boundarydevices.com>
12 years agoi.MX6: provide functional names for CCM_CCGR0-CCGR6 bit fields
Eric Nelson [Fri, 21 Sep 2012 07:33:51 +0000 (07:33 +0000)]
i.MX6: provide functional names for CCM_CCGR0-CCGR6 bit fields

Add meaningful constants for each clock channels and use them for
enabling and disabling i.MX6 clocks.

Includes an update to enable/disable the IPU1 clock in
drivers/video/ipu_common to remove IMX5x register access
when used on i.MX6 as discussed in V1:

     http://patchwork.ozlabs.org/patch/185129/

Signed-off-by: Eric Nelson <eric.nelson@boundarydevices.com>
12 years agomx6qsabresd: Add 8-bit USDHC support
Fabio Estevam [Tue, 18 Sep 2012 09:27:49 +0000 (09:27 +0000)]
mx6qsabresd: Add 8-bit USDHC support

USDHC3 has 8 pins wired in mx6qsabresd. Configure the extra pins.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
12 years agomx6qsabresd: Add Ethernet support
Fabio Estevam [Tue, 18 Sep 2012 17:24:23 +0000 (17:24 +0000)]
mx6qsabresd: Add Ethernet support

mx6qsabresd has a AR8031 Gigabit PHY.

Add support for it.

Also increase CONFIG_SYS_MALLOC_LEN so that FEC buffer allocation does not fail.

Tested on 1Gbp and 100Mbps networks.

Suggested-by: Kim Phillips <kim.phillips@freescale.com>
Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Acked-by: Kim Phillips <kim.phillips@freescale.com>
12 years agomx6: Add basic support for mx6qsabresd board.
Fabio Estevam [Thu, 13 Sep 2012 03:18:20 +0000 (03:18 +0000)]
mx6: Add basic support for mx6qsabresd board.

mx6qsabresd is a board based on mx6q SoC with the following features:
- 1GB of DDR3
- 1 USB OTG port
- 1 HDMI output port
- SPI NOR
- LVDS panel
- Gigabit Ethernet
- Camera Connector
- eMMC and SD card slot
- Audio

Add very basic support for it.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
12 years agomx6q: Factor out common DDR3 init code
Fabio Estevam [Thu, 13 Sep 2012 03:18:19 +0000 (03:18 +0000)]
mx6q: Factor out common DDR3 init code

Factor out common DDR3 initialization code, allowing easier maintainance of such
scripts.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
12 years agopxa: Add code to examine cpu model and revision
Lukasz Dalek [Tue, 2 Oct 2012 22:51:06 +0000 (00:51 +0200)]
pxa: Add code to examine cpu model and revision

Add function which return CPU model and revision which can be used for
cpu detection.

Signed-off-by: Lukasz Dalek <luk0104@gmail.com>
12 years agotegra: nand: add board pinmux
Lucas Stach [Sat, 29 Sep 2012 10:02:09 +0000 (10:02 +0000)]
tegra: nand: add board pinmux

Boards may require a different pinmux setup for NAND than the default one.
Add a way to call into board specific code to set this up.

Signed-off-by: Lucas Stach <dev@lynxeye.de>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotegra: clean up board include hell
Lucas Stach [Sat, 29 Sep 2012 10:02:08 +0000 (10:02 +0000)]
tegra: clean up board include hell

The prototypes used in board files were all scattered out, which lead to
code duplication between SPL and normal U-Boot and some prototypes not actually
being used. Consolidate this in a common board header.

Signed-off-by: Lucas Stach <dev@lynxeye.de>
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotegra: add funcmux entry for NAND attached to KBC
Lucas Stach [Thu, 27 Sep 2012 13:04:27 +0000 (13:04 +0000)]
tegra: add funcmux entry for NAND attached to KBC

Secondary config for the Flash attachment.

Signed-off-by: Lucas Stach <dev@lynxeye.de>
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotegra20: rework UART GPIO handling
Lucas Stach [Tue, 25 Sep 2012 20:21:14 +0000 (20:21 +0000)]
tegra20: rework UART GPIO handling

Rename board provided gpio_config_uart() to
gpio_early_init_uart() as it does the same thing as the equally
called function provided by the uart-switch code. This allows
to simply call this function in early board init whether or not
we are building with CONFIG_UART_SWITCH defined.

Also provide a weak symbol for this function, to avoid the
need to provide this function for boards that don't need any
fixup.

This patch supersedes the earlier posted
"tegra: convert gpio_config_uart to weak symbol".
Build tested with MAKEALL -s tegra20

Signed-off-by: Lucas Stach <dev@lynxeye.de>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotegra20: add clock_set_pllout function
Lucas Stach [Tue, 25 Sep 2012 20:21:13 +0000 (20:21 +0000)]
tegra20: add clock_set_pllout function

Common practice on Tegra 2 boards is to use the pllp_out4 FO
to generate the ULPI reference clock. For this to work we have
to override the default hardware generated output divider.

This function adds a clean way to do so.

Signed-off-by: Lucas Stach <dev@lynxeye.de>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotegra20: complete periph_id enum
Lucas Stach [Tue, 25 Sep 2012 09:59:12 +0000 (09:59 +0000)]
tegra20: complete periph_id enum

Most Tegra boards output the ULPI reference clock on pad DEV2.

Complete the periph_id enum so that we are able to enable this
clock output circuit.

Signed-off-by: Lucas Stach <dev@lynxeye.de>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotegra: enable CONFIG_CMD_PART
Stephen Warren [Tue, 25 Sep 2012 13:32:26 +0000 (13:32 +0000)]
tegra: enable CONFIG_CMD_PART

This is extremely likely to be used from the boot.scr that Tegra's default
bootcmd locates and executes.

Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agoTegra20: Move some include files to arch-tegra for sharing with Tegra30
Tom Warren [Wed, 19 Sep 2012 22:50:56 +0000 (15:50 -0700)]
Tegra20: Move some include files to arch-tegra for sharing with Tegra30

The move is pretty straight-forward. ap20.h and tegra20.h were renamed to ap.h and tegra.h.
Some files remain in arch-tegra20 but 'include' a file in 'arch-tegra' with #defines & structs
that will be common between T20 and T30 HW. HW-specific #defines, etc. stay in the 'arch-tegra20'
'root' file.

All boards build OK w/MAKEALL -s tegra20. Checkpatch.pl runs clean. Seaboard works OK.

Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agoTegra20: Move some code files to common directories for upcoming Tegra30 patches.
Tom Warren [Wed, 19 Sep 2012 21:08:52 +0000 (14:08 -0700)]
Tegra20: Move some code files to common directories for upcoming Tegra30 patches.

Move files that are going to be common between T20 and T30 into 'tegra-common'
subdirs in AVP (arm720t), CPU (armv7), and shared (arch/arm/cpu/.) areas. Any
files that are left behind in '/tegra20' will be copied to '/tegra30' subdirs
and modified for that SoC. The 'common' files should need only minor changes.

Include files (arch/arm/include/asm/arch-tegra/tegra20) will be done in a
follow-on patch.

Builds fine w/MAKEALL -s tegra20. Checkpatch.pl is clean.

Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotegra: Rename Medcom to Medcom-Wide
Thierry Reding [Wed, 19 Sep 2012 00:37:21 +0000 (00:37 +0000)]
tegra: Rename Medcom to Medcom-Wide

Medcom is the marketing name for an older, PXA-based version of the same
device. In order to avoid confusion, rename the Tegra-based version to
the new marketing name.

Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotegra: Update Avionic Design vendor prefix
Thierry Reding [Wed, 19 Sep 2012 00:37:20 +0000 (00:37 +0000)]
tegra: Update Avionic Design vendor prefix

The official vendor prefix for Avionic Design is now "ad". Update the
board DTS files accordingly.

Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de>
Acked-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
12 years agotools: Add a README note about fw_printenv lock file
Joe Hershberger [Thu, 4 Oct 2012 08:31:00 +0000 (08:31 +0000)]
tools: Add a README note about fw_printenv lock file

Add a mention of the lock file to the README for the fw_printenv tool.

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
Reported-by: Luka Perkov <uboot@lukaperkov.net>
12 years agoenv: Check for NULL pointer in envmatch()
Joe Hershberger [Wed, 3 Oct 2012 09:38:50 +0000 (09:38 +0000)]
env: Check for NULL pointer in envmatch()

If the pointer passed into envmatch() is NULL, return -1 instead of
crashing.

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
12 years agotools/env: Serialize calls to fw_*env
Joe Hershberger [Wed, 3 Oct 2012 09:38:49 +0000 (09:38 +0000)]
tools/env: Serialize calls to fw_*env

Use a lock file at /var/lock/fw_printenv.lock.
Avoids seriously confusing the MTD driver.

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
12 years agotools/env: Don't call env_init() in fw_getenv()
Joe Hershberger [Wed, 3 Oct 2012 09:38:48 +0000 (09:38 +0000)]
tools/env: Don't call env_init() in fw_getenv()

We will only call fw_getenv when the env has already been initialized.

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
12 years agotools/env: Remove unneeded complexity
Joe Hershberger [Wed, 3 Oct 2012 09:38:47 +0000 (09:38 +0000)]
tools/env: Remove unneeded complexity

The length included the name length, and then it was subtracted back
out on each use.  Now we don't include it in the first place.  Also
realloc as we process arguments and eliminate memset.  Use memcpy
instead of manually copying each byte.

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
12 years agotools/env: Use a board-specific default env
Joe Hershberger [Wed, 3 Oct 2012 09:38:46 +0000 (09:38 +0000)]
tools/env: Use a board-specific default env

Originally added in aa701b94336b358798d676eef12a7b90bdac23f5

Before this patch, there was a hard-coded env that was used as default
if the env in flash is detected as invalid.  Now this tool (compiled
for a given board) will share the default env with the u-boot for the
board.

Fix include of config.h

Need to define "TEXT_BASE" when building the fw_env tool so that the
default env will be correct for environments which use it.

Define __ASSEMBLY__ when calling #include <config.h> so that we only
get #defines (all we're interested in).

Signed-off-by: Joe Hershberger <joe.hershberger@ni.com>
12 years agocommon: cmd_bdinfo: fix type of value in print_lnum
Daniel Schwierzeck [Wed, 3 Oct 2012 08:36:11 +0000 (08:36 +0000)]
common: cmd_bdinfo: fix type of value in print_lnum

This fixes a warning when compiling with ELDK-5.2.1 for MIPS64:

cmd_bdinfo.c: In function 'print_lnum':
cmd_bdinfo.c:56:2: warning: format '%llX' expects argument of type 'long long unsigned int', but argument 3 has type 'u64' [-Wformat]

Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
12 years agoinput: Add debugging for key matrix key codes
Simon Glass [Thu, 27 Sep 2012 15:18:43 +0000 (15:18 +0000)]
input: Add debugging for key matrix key codes

These are read from the fdt - add a debug feature to display the mapping
on start-up.

See that we get debug output listing the keycodes

Signed-off-by: Simon Glass <sjg@chromium.org>
12 years agoinput: Allow key ghosting filter to be disabled
Simon Glass [Thu, 27 Sep 2012 15:18:42 +0000 (15:18 +0000)]
input: Allow key ghosting filter to be disabled

Some keyboards will not need a key ghosting filter, so make this feature
optional.

Signed-off-by: Simon Glass <sjg@chromium.org>
12 years agoinput: Separate out keyboard repeat/delay from init
Simon Glass [Thu, 27 Sep 2012 15:18:41 +0000 (15:18 +0000)]
input: Separate out keyboard repeat/delay from init

It is inconvenient to have to specify the keyboard repeat and delay at
init time if it is not yet available, so move this into a separate
function.

Some drivers will want to do this when their keyboard init routine
is actually called.

Signed-off-by: Simon Glass <sjg@chromium.org>
12 years agoinput: Correct key_matrix fdt decoding
Simon Glass [Thu, 27 Sep 2012 15:18:40 +0000 (15:18 +0000)]
input: Correct key_matrix fdt decoding

Some issues with this were not addressed in the previous series. Fix up
the binding decoding to deal with what is actually expected in the fdt.

This corrects the broken keyboard on seaboard.

Signed-off-by: Simon Glass <sjg@chromium.org>
12 years agotools: Add cleanpatch
Fabio Estevam [Fri, 28 Sep 2012 03:36:14 +0000 (03:36 +0000)]
tools: Add cleanpatch

There are some errors reported by checkpatch.pl that can be easily cleaned up by
using the cleanpatch tool.

Import the cleanpatch script from linux kernel 3.5.4 stable version as from the
following commit:

commit cb3ed5b7e09c6c0462e396d55e3fecc0980a333a
Author: H. Peter Anvin <hpa@zytor.com>
Date:   Fri May 25 17:58:26 2007 -0700

    scripts: Make cleanfile/cleanpatch warn about long lines

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
12 years agopatman: Handle checkpatch.pl not providing file/line info
Simon Glass [Thu, 27 Sep 2012 15:33:46 +0000 (15:33 +0000)]
patman: Handle checkpatch.pl not providing file/line info

Sometimes we don't get a valid filename or line number from checkpatch.pl,
for example if the patch is in a bad format. Deal with this by using a
default value, rather than a stack trace.

Signed-off-by: Simon Glass <sjg@chromium.org>