karo-tx-uboot.git
4 years agovideo: ipu: initialize g_ipu_clk, g_ldb_clk statically KARO-TX6UL-2015-10-23
Lothar Waßmann [Fri, 23 Oct 2015 10:46:03 +0000 (12:46 +0200)]
video: ipu: initialize g_ipu_clk, g_ldb_clk statically

When the FB driver has not been initialized, booting Linux will hang
the system when ipuv3_fb_shutdown() is called from arch_preboot_os()
because clk_enable(g_ipu_clk) is a NOP with g_ipu_clk == NULL.
There is no need to initialize these variables at runtime, so use
static initialization to prevent this problem.

4 years agoarm: mx6: fix the AIPS?_BASE_ADDR fsckup introduced by commit f5def95698f6
Lothar Waßmann [Thu, 22 Oct 2015 13:04:44 +0000 (15:04 +0200)]
arm: mx6: fix the AIPS?_BASE_ADDR fsckup introduced by commit f5def95698f6

commit f5def95698f6f3d8923f0e38706e3e5ebed81a10 ("imx: mx6: Fixed AIPS3 base address issue")
claimed to fix an AIPS3 base address issue, while it in fact messed up
the AIPS1 and AIPS2 base addresses too.

4 years agokaro: tx6: adapt romupdate command for i.MX6UL
Lothar Waßmann [Thu, 22 Oct 2015 09:28:29 +0000 (11:28 +0200)]
karo: tx6: adapt romupdate command for i.MX6UL

4 years agokaro: tx6: remove unused code from flash.c
Lothar Waßmann [Thu, 22 Oct 2015 09:28:05 +0000 (11:28 +0200)]
karo: tx6: remove unused code from flash.c

4 years agonand: mxs: convey return code from failed function rather than '0' in mxs_nand_ecc_wr...
Lothar Waßmann [Thu, 22 Oct 2015 09:27:22 +0000 (11:27 +0200)]
nand: mxs: convey return code from failed function rather than '0' in mxs_nand_ecc_write_page()

4 years agonand: mxs: make fake_ecc_layout static
Lothar Waßmann [Thu, 22 Oct 2015 09:26:36 +0000 (11:26 +0200)]
nand: mxs: make fake_ecc_layout static

4 years agonand: mxs: use CONFIG_ARCH_MX6 instead of CONFIG_SOC_MX6Q
Lothar Waßmann [Thu, 22 Oct 2015 09:26:22 +0000 (11:26 +0200)]
nand: mxs: use CONFIG_ARCH_MX6 instead of CONFIG_SOC_MX6Q

4 years agokaro: tx6ul: disable PMIC init
Lothar Waßmann [Tue, 20 Oct 2015 11:35:50 +0000 (13:35 +0200)]
karo: tx6ul: disable PMIC init

The I2C-GPIO driver requires configuration via DTB which is not
currently supported for Ka-Ro modules. So, disable the PMIC access for
now.

4 years agokaro: tx6: change DGCTRL init values depending on SoC type
Lothar Waßmann [Tue, 20 Oct 2015 11:28:03 +0000 (13:28 +0200)]
karo: tx6: change DGCTRL init values depending on SoC type

4 years agokaro: tx6: add TXUL support
Lothar Waßmann [Tue, 20 Oct 2015 11:27:51 +0000 (13:27 +0200)]
karo: tx6: add TXUL support

4 years agokaro: tx6: selectively set individual CCGR bits via DCD
Lothar Waßmann [Mon, 19 Oct 2015 11:49:04 +0000 (13:49 +0200)]
karo: tx6: selectively set individual CCGR bits via DCD

4 years agokaro: tx6: eliminate PHYS_SDRAM_1_WIDTH and use CONFIG_SYS_SDRAM_BUS_WIDTH instead
Lothar Waßmann [Mon, 19 Oct 2015 10:42:42 +0000 (12:42 +0200)]
karo: tx6: eliminate PHYS_SDRAM_1_WIDTH and use CONFIG_SYS_SDRAM_BUS_WIDTH instead

4 years agokaro: tx6: config.mk cleanup
Lothar Waßmann [Mon, 19 Oct 2015 10:37:29 +0000 (12:37 +0200)]
karo: tx6: config.mk cleanup

4 years agokaro: tx6: further cleanups
Lothar Waßmann [Mon, 19 Oct 2015 10:36:46 +0000 (12:36 +0200)]
karo: tx6: further cleanups

4 years agoarm: mx6: use imx6_thermal driver
Lothar Waßmann [Mon, 19 Oct 2015 10:28:12 +0000 (12:28 +0200)]
arm: mx6: use imx6_thermal driver

4 years agoarm: mx6: use fuse_read() for OCOTP fuse access
Lothar Waßmann [Mon, 19 Oct 2015 10:27:05 +0000 (12:27 +0200)]
arm: mx6: use fuse_read() for OCOTP fuse access

4 years agoarm: mxs: ocotp: disable OCOTP clock after access is finished
Lothar Waßmann [Mon, 12 Oct 2015 12:04:28 +0000 (14:04 +0200)]
arm: mxs: ocotp: disable OCOTP clock after access is finished

4 years agoremove obsolete functions ocotp_clk_{en,dis}able
Lothar Waßmann [Mon, 12 Oct 2015 09:28:33 +0000 (11:28 +0200)]
remove obsolete functions ocotp_clk_{en,dis}able

4 years agokconfig: arm: remove duplicate definitions of SOC_MX*
Lothar Waßmann [Mon, 12 Oct 2015 12:02:36 +0000 (14:02 +0200)]
kconfig: arm: remove duplicate definitions of SOC_MX*

4 years agokconfig: arm: move ARCH_MX? out of choice which selects TARGETs that depend on those
Lothar Waßmann [Mon, 12 Oct 2015 12:01:23 +0000 (14:01 +0200)]
kconfig: arm: move ARCH_MX? out of choice which selects TARGETs that depend on those

4 years agokconfig: arm: make cache configurable via Kconfig
Lothar Waßmann [Mon, 12 Oct 2015 08:42:36 +0000 (10:42 +0200)]
kconfig: arm: make cache configurable via Kconfig

4 years agoarm: mx6: remove CONFIG_ macros that are now included in Kconfig from header file
Lothar Waßmann [Mon, 12 Oct 2015 08:41:54 +0000 (10:41 +0200)]
arm: mx6: remove CONFIG_ macros that are now included in Kconfig from header file

4 years agoarm: mx6: rename some missed instances of CONFIG_MX6* to CONFIG_SOC_MX6*
Lothar Waßmann [Mon, 12 Oct 2015 08:40:51 +0000 (10:40 +0200)]
arm: mx6: rename some missed instances of CONFIG_MX6* to CONFIG_SOC_MX6*

4 years agodma: apbh: let APBH_DMA be selected, instead of enabling it by default
Lothar Waßmann [Mon, 12 Oct 2015 08:37:22 +0000 (10:37 +0200)]
dma: apbh: let APBH_DMA be selected, instead of enabling it by default

4 years agonet: fec_mxc: move CONFIG_FEC_MXC_PHYADDR from Kconfig to include/configs/*.h
Lothar Waßmann [Mon, 12 Oct 2015 08:21:52 +0000 (10:21 +0200)]
net: fec_mxc: move CONFIG_FEC_MXC_PHYADDR from Kconfig to include/configs/*.h

4 years agoarm: imx: prevent build error when CONFIG_FEC_MXC_PHYADDR is undefined
Lothar Waßmann [Mon, 12 Oct 2015 08:15:31 +0000 (10:15 +0200)]
arm: imx: prevent build error when CONFIG_FEC_MXC_PHYADDR is undefined

4 years agoimx: remove duplicated prototype for fecmxc_initialize() which is already defined...
Lothar Waßmann [Fri, 9 Oct 2015 10:32:18 +0000 (12:32 +0200)]
imx: remove duplicated prototype for fecmxc_initialize() which is already defined in include/netdev.h

4 years agokconfig: remove redundant 'default n'
Lothar Waßmann [Fri, 9 Oct 2015 09:45:53 +0000 (11:45 +0200)]
kconfig: remove redundant 'default n'

4 years agogpio: use appropriate return values for cmd handler
Lothar Waßmann [Fri, 9 Oct 2015 09:45:21 +0000 (11:45 +0200)]
gpio: use appropriate return values for cmd handler

4 years agonet: fec_mxc: use some more appropriate return values
Lothar Waßmann [Fri, 9 Oct 2015 09:43:47 +0000 (11:43 +0200)]
net: fec_mxc: use some more appropriate return values

4 years agonet: fec_mxc: consolidate data types and remove useless type casts
Lothar Waßmann [Fri, 9 Oct 2015 09:43:17 +0000 (11:43 +0200)]
net: fec_mxc: consolidate data types and remove useless type casts

4 years agonet: fec_mxc: don't write to reserved register on i.MX6UL
Lothar Waßmann [Fri, 9 Oct 2015 09:42:36 +0000 (11:42 +0200)]
net: fec_mxc: don't write to reserved register on i.MX6UL

4 years agonet: fec_mxc: don't try to clear read-only registers
Lothar Waßmann [Fri, 9 Oct 2015 08:53:38 +0000 (10:53 +0200)]
net: fec_mxc: don't try to clear read-only registers

The registers in the affected address range are all read only. There
is no use in trying to clear them.

4 years agonet: fec_mxc: use a definite delay when waiting for TDAR/RDAR to clear
Lothar Waßmann [Fri, 9 Oct 2015 08:58:56 +0000 (10:58 +0200)]
net: fec_mxc: use a definite delay when waiting for TDAR/RDAR to clear

4 years agonet: fec_mxc: autodetect PHY address if phy_id < 0
Lothar Waßmann [Fri, 9 Oct 2015 09:00:32 +0000 (11:00 +0200)]
net: fec_mxc: autodetect PHY address if phy_id < 0

4 years agoRevert "video: ipu_disp: Fix clock polarity logic"
Lothar Waßmann [Fri, 9 Oct 2015 08:37:42 +0000 (10:37 +0200)]
Revert "video: ipu_disp: Fix clock polarity logic"

This reverts commit 2740e5de4f3cd0aa36efcfe1a995fb6e3858cc97.

The patch made the U-Boot driver behave different from the Linux
driver, so that for U-Boot and Linux different settings in the DTB
would be required!

4 years agokaro: configs: set 'baseboard' variable in 'noenv' U-Boot
Lothar Waßmann [Tue, 22 Sep 2015 11:57:24 +0000 (13:57 +0200)]
karo: configs: set 'baseboard' variable in 'noenv' U-Boot

4 years agokaro: use lldiv() for 64bit division
Lothar Waßmann [Tue, 22 Sep 2015 11:56:37 +0000 (13:56 +0200)]
karo: use lldiv() for 64bit division

4 years agokaro: tx6: don't perform DQS gating calibration KARO-TX6-2015-09-18
Lothar Waßmann [Fri, 18 Sep 2015 14:31:32 +0000 (16:31 +0200)]
karo: tx6: don't perform DQS gating calibration

DQS Gating calibration may yield incorrect results due to a bug in the
ROM code. Also preset all relevant timing registers to sane values
before doing the DDR calibration.

4 years agokaro: tx6: don't position CSF at fixed offset
Lothar Waßmann [Fri, 18 Sep 2015 12:34:49 +0000 (14:34 +0200)]
karo: tx6: don't position CSF at fixed offset

4 years agokaro: tx6: remove usesless symbol _image_binary_end
Lothar Waßmann [Fri, 18 Sep 2015 12:34:27 +0000 (14:34 +0200)]
karo: tx6: remove usesless symbol _image_binary_end

4 years agokaro: tx6: cleanup defconfig files
Lothar Waßmann [Fri, 18 Sep 2015 12:04:24 +0000 (14:04 +0200)]
karo: tx6: cleanup defconfig files

4 years agokaro: cleanup after merge of v2015.10-rc2
Lothar Waßmann [Thu, 17 Sep 2015 05:31:36 +0000 (07:31 +0200)]
karo: cleanup after merge of v2015.10-rc2

4 years agosunxi: increase SYS_MONITOR_LEN
Boris Brezillon [Mon, 27 Jul 2015 14:21:26 +0000 (16:21 +0200)]
sunxi: increase SYS_MONITOR_LEN

Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
4 years agodts: fix dependency of OF_SPL_REMOVE_PROPS
Masahiro Yamada [Fri, 28 Aug 2015 11:28:42 +0000 (20:28 +0900)]
dts: fix dependency of OF_SPL_REMOVE_PROPS

This should depend on SPL_OF_CONTROL (it is not equivalent to
SPL && OF_CONTROL).

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agopinctrl: sandbox: add sandbox pinctrl driver
Masahiro Yamada [Thu, 27 Aug 2015 03:44:30 +0000 (12:44 +0900)]
pinctrl: sandbox: add sandbox pinctrl driver

This driver actually does nothing but test pinctrl uclass, and
demonstrate how things work.

To try this driver, uncomment /* #define DEBUG */ in the
drivers/pinctrl/pinctrl-sandbox.c, and debug messages will be
displayed.

  DRAM:  128 MiB
  sandbox pinmux: group = 1 (serial_a), function = 1 (serial)
  Using default environment

  In:    cros-ec-keyb
  Out:   lcd
  Err:   lcd
  Net:   Net Initialization Skipped
  eth0: eth@10002000, eth1: eth@80000000, eth5: eth@90000000
  => i2c dev 0
  Setting bus to 0
  sandbox pinmux: group = 0 (i2c), function = 0 (i2c)
  sandbox pinconf: group = 0 (i2c), param = 3, arg = 1

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agopinctrl: add pin control uclass support
Masahiro Yamada [Thu, 27 Aug 2015 03:44:29 +0000 (12:44 +0900)]
pinctrl: add pin control uclass support

This creates a new framework for handling of pin control devices,
i.e. devices that control different aspects of package pins.

This uclass handles pinmuxing and pin configuration; pinmuxing
controls switching among silicon blocks that share certain physical
pins, pin configuration handles electronic properties such as pin-
biasing, load capacitance etc.

This framework can support the same device tree bindings, but if you
do not need full interface support, you can disable some features to
reduce memory foot print.  Typically around 1.5KB is necessary to
include full-featured uclass support on ARM board (CONFIG_PINCTRL +
CONFIG_PINCTRL_FULL + CONFIG_PINCTRL_GENERIC + CONFIG_PINCTRL_PINMUX),
for example.

We are often limited on code size for SPL.  Besides, we still have
many boards that do not support device tree configuration.  The full
pinctrl, which requires OF_CONTROL, does not make sense for those
boards.  So, this framework also has a Do-It-Yourself (let's say
simple pinctrl) interface.  With CONFIG_PINCTRL_FULL disabled, the
uclass itself provides no systematic mechanism for identifying the
peripheral device, applying pinctrl settings, etc.  They must be
done in each low-level driver.  In return, you can save much memory
footprint and it might be useful especially for SPL.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agodm: core: allow device_bind() to not return a device pointer
Masahiro Yamada [Thu, 27 Aug 2015 03:44:28 +0000 (12:44 +0900)]
dm: core: allow device_bind() to not return a device pointer

This is useful when we want to bind a device, but do not need the
pointer to the device.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agotegra: nyan: Enable TPM command and driver
Simon Glass [Sun, 23 Aug 2015 00:31:44 +0000 (18:31 -0600)]
tegra: nyan: Enable TPM command and driver

The TPM is listed in the device tree. Enable the driver and 'tpm' command
so that it can be used.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Enable 'tpmtest' command for Chrome OS boards with TPMs
Simon Glass [Sun, 23 Aug 2015 00:31:43 +0000 (18:31 -0600)]
tpm: Enable 'tpmtest' command for Chrome OS boards with TPMs

This command provides a few useful tests so enable it for common boards.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agodm: tpm: Add a 'tpmtest' command
Simon Glass [Sun, 23 Aug 2015 00:31:42 +0000 (18:31 -0600)]
dm: tpm: Add a 'tpmtest' command

These tests come from Chrome OS code. They are not particularly tidy but can
be useful for checking that the TPM is behaving correctly. Some knowledge of
TPM operation is required to use these.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Add functions to access flags and permissions
Simon Glass [Sun, 23 Aug 2015 00:31:41 +0000 (18:31 -0600)]
tpm: Add functions to access flags and permissions

Add a few new functions which will be used by the test command in a future
patch.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Add a 'tpm info' command
Simon Glass [Sun, 23 Aug 2015 00:31:40 +0000 (18:31 -0600)]
tpm: Add a 'tpm info' command

Add a command to display basic information about a TPM such as the model and
open/close state. This can be useful for debugging.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agodm: tpm: Convert LPC driver to driver model
Simon Glass [Sun, 23 Aug 2015 00:31:39 +0000 (18:31 -0600)]
dm: tpm: Convert LPC driver to driver model

Convert the tpm_tis_lpc driver to use driver model and update boards which
use it.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agodm: tpm: Convert I2C driver to driver model
Simon Glass [Sun, 23 Aug 2015 00:31:38 +0000 (18:31 -0600)]
dm: tpm: Convert I2C driver to driver model

Convert the tpm_tis_i2c driver to use driver model and update boards which
use it.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agoexynos: x86: dts: Add tpm nodes to the device tree for Chrome OS devices
Simon Glass [Sun, 23 Aug 2015 00:31:37 +0000 (18:31 -0600)]
exynos: x86: dts: Add tpm nodes to the device tree for Chrome OS devices

Add a TPM node to the various Chromebooks so that driver can be converted to
driver model.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Check that parse_byte_string() has data to parse
Simon Glass [Sun, 23 Aug 2015 00:31:36 +0000 (18:31 -0600)]
tpm: Check that parse_byte_string() has data to parse

Rather then crashing when there is no data, print an error. The error is
printed by the caller to parse_byte_string().

Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agodm: tpm: sandbox: Convert TPM driver to driver model
Simon Glass [Sun, 23 Aug 2015 00:31:35 +0000 (18:31 -0600)]
dm: tpm: sandbox: Convert TPM driver to driver model

Convert the sandbox TPM driver to use driver model. Add it to the device
tree so that it can be found on start-up.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Report tpm errors on the command line
Simon Glass [Sun, 23 Aug 2015 00:31:34 +0000 (18:31 -0600)]
tpm: Report tpm errors on the command line

When a 'tpm' command fails, we set the return code but give no indication
of failure. This can be confusing.

Add an error message when any tpm command fails.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agodm: i2c: Add a command to adjust the offset length
Simon Glass [Sun, 23 Aug 2015 00:31:33 +0000 (18:31 -0600)]
dm: i2c: Add a command to adjust the offset length

I2C chips can support a register offset, with registers accessible by
sending this offset as the first part of any read or write transaction.
Most I2C chips have a single byte offset, thus the offset length is 1.
This provides access for up 256 registers.

However other offset lengths are supported, including 0.

Add a command to provide access to the offset length from the command
line. This allows the offset length to be read or written.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agodm: tpm: Convert the TPM command and library to driver model
Simon Glass [Sun, 23 Aug 2015 00:31:32 +0000 (18:31 -0600)]
dm: tpm: Convert the TPM command and library to driver model

Add driver model support to the TPM command and the TPM library. Both
support only a single TPM at present.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agodm: tpm: Add a uclass for Trusted Platform Modules
Simon Glass [Sun, 23 Aug 2015 00:31:31 +0000 (18:31 -0600)]
dm: tpm: Add a uclass for Trusted Platform Modules

Add a new uclass for TPMs which uses almost the same TIS (TPM Interface
Specification) as is currently implemented. Since init() is handled by the
normal driver model probe() method, we don't need to implement that. Also
rename the transfer method to xfer() which is a less clumbsy name.

Once all drivers and users are converted to driver model we can remove the
old code.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: tpm_tis_i2c: Tidy up delays
Simon Glass [Sun, 23 Aug 2015 00:31:30 +0000 (18:31 -0600)]
tpm: tpm_tis_i2c: Tidy up delays

Use a _US suffix for microseconds and a _MS suffic for milliseconds. Move
all timeouts and delays into one place. Use mdelay() instead of udelay()
where appropriate.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: tpm_tis_i2c: Use a consistent tpm_tis_i2c_ prefix
Simon Glass [Sun, 23 Aug 2015 00:31:29 +0000 (18:31 -0600)]
tpm: tpm_tis_i2c: Use a consistent tpm_tis_i2c_ prefix

Use the same prefix on each function for consistency.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: tpm_tis_i2c: Simplify init code
Simon Glass [Sun, 23 Aug 2015 00:31:28 +0000 (18:31 -0600)]
tpm: tpm_tis_i2c: Simplify init code

Move all the init and uninit code into one place.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: tpm_tis_i2c: Move definitions into the header file
Simon Glass [Sun, 23 Aug 2015 00:31:27 +0000 (18:31 -0600)]
tpm: tpm_tis_i2c: Move definitions into the header file

Some definitions are in the C file and some are in the header file. Move
everything into the header file for consistency and to reduce clutter.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: tpm_tis_i2c: Merge struct tpm into tpm_chip
Simon Glass [Sun, 23 Aug 2015 00:31:26 +0000 (18:31 -0600)]
tpm: tpm_tis_i2c: Merge struct tpm into tpm_chip

There are too many structures storing the same sort of information. Move the
fields from struct tpm into struct tpm_chip and remove the former struct.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: tpm_tis_i2c: Merge struct tpm_dev into tpm_chip
Simon Glass [Sun, 23 Aug 2015 00:31:25 +0000 (18:31 -0600)]
tpm: tpm_tis_i2c: Merge struct tpm_dev into tpm_chip

There are too many structures storing the same sort of information. Move the
fields from struct tpm_dev into struct tpm_chip and remove the former
struct.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: tpm_tis_i2c: Drop struct tpm_vendor_specific
Simon Glass [Sun, 23 Aug 2015 00:31:24 +0000 (18:31 -0600)]
tpm: tpm_tis_i2c: Drop struct tpm_vendor_specific

This function is misnamed since it only applies to a single driver. Merge
its fields into its parent.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: tpm_tis_i2c: Drop unnecessary methods
Simon Glass [Sun, 23 Aug 2015 00:31:23 +0000 (18:31 -0600)]
tpm: tpm_tis_i2c: Drop unnecessary methods

The function methods in struct tpm_vendor_specific just call local functions.
Change the code to use a direct call.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Move the I2C TPM code into one file
Simon Glass [Sun, 23 Aug 2015 00:31:22 +0000 (18:31 -0600)]
tpm: Move the I2C TPM code into one file

The current Infineon I2C TPM driver is written in two parts, intended to
support use with other I2C devices. However we don't have any users and the
Atmel I2C TPM device does not use this file.

We should simplify this and remove the unused abstration. As a first step,
move the code into one file.

Also the name tpm_private.h suggests that the header file is generic to all
TPMs but it is not. Rename it indicate that it relates only to this driver

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Convert drivers to use SPDX
Simon Glass [Sun, 23 Aug 2015 00:31:21 +0000 (18:31 -0600)]
tpm: Convert drivers to use SPDX

Add an SPDX header to two drivers that don't have it.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Convert board config TPM options to Kconfig
Simon Glass [Sun, 23 Aug 2015 00:31:20 +0000 (18:31 -0600)]
tpm: Convert board config TPM options to Kconfig

Convert all TPM options to Kconfig and tidy up.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Add Kconfig options for TPMs
Simon Glass [Sun, 23 Aug 2015 00:31:19 +0000 (18:31 -0600)]
tpm: Add Kconfig options for TPMs

Add new Kconfig options for TPMs in preparation for moving boards to use
Kconfig for TPM configuration.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard<christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Drop two unused options
Simon Glass [Sun, 23 Aug 2015 00:31:18 +0000 (18:31 -0600)]
tpm: Drop two unused options

The address of the I2C TPM is now defined in the device tree so there is no
need for the CONFIG options.

Remove them from the README and board config to avoid confusion.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agotpm: Remove old pre-driver-model I2C code
Simon Glass [Sun, 23 Aug 2015 00:31:17 +0000 (18:31 -0600)]
tpm: Remove old pre-driver-model I2C code

This is not used anymore by any board so drop it.

Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Christophe Ricard <christophe-h.ricard@st.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
4 years agoexynos: Rise ARM voltage to 1.1V for chained bootloaders
Misha Komarovskiy [Tue, 25 Aug 2015 08:53:26 +0000 (11:53 +0300)]
exynos: Rise ARM voltage to 1.1V for chained bootloaders

If board uses downstream Chrome OS U-Boot as first stage
bootloader and upstream version is chained second stage,
1.1V is minimum voltage borderline.

Signed-off-by: Misha Komarovskiy <zombah@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agobuildman: Correct 'Series-cover-cc' detection logic
Simon Glass [Sun, 23 Aug 2015 00:28:01 +0000 (18:28 -0600)]
buildman: Correct 'Series-cover-cc' detection logic

This requires 'Series-cover_cc' at present which is incorrect. Fix it.

Signed-off-by: Simon Glass <sjg@chromium.org>
4 years agodrivers: kconfig: Sort driver menu in alphabetical order
Bin Meng [Fri, 21 Aug 2015 05:44:16 +0000 (22:44 -0700)]
drivers: kconfig: Sort driver menu in alphabetical order

Sort different types of drivers in alphabetical order.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agodrivers: kconfig: Move PHYS_TO_BUS to "Device Drivers" menu
Bin Meng [Fri, 21 Aug 2015 05:44:15 +0000 (22:44 -0700)]
drivers: kconfig: Move PHYS_TO_BUS to "Device Drivers" menu

Right now PHYS_TO_BUS shows in the Kconfig main menu, move it.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agodrivers: kconfig: Move "Generic Driver Options" menu to the top
Bin Meng [Fri, 21 Aug 2015 05:44:14 +0000 (22:44 -0700)]
drivers: kconfig: Move "Generic Driver Options" menu to the top

Make "Generic Driver Options" menu show on the top in the Kconfig.

Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agodm: core: Add Kconfig for simple bus driver
Marek Vasut [Sun, 2 Aug 2015 23:15:48 +0000 (01:15 +0200)]
dm: core: Add Kconfig for simple bus driver

Add Kconfig entries for the simple-bus driver, both for U-Boot
and for SPL. The simple-bus is enabled by default in U-Boot and
disabled by default in SPL to preserve the original behavior.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Simon Glass <sjg@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
Modified to fit on top of Masahiro's $(SPL) setup:
Signed-off-by: Simon Glass <sjg@chromium.org>
4 years agodm: Use dev_get_addr() where possible
Simon Glass [Tue, 11 Aug 2015 14:33:29 +0000 (08:33 -0600)]
dm: Use dev_get_addr() where possible

This is a convenient way for a driver to get the hardware address of a
device, when regmap or syscon are not being used. Change existing callers
to use it as an example to others.

Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Acked-by: Stephen Warren <swarren@wwwdotorg.org>
4 years agodm: simplify uclass_foreach_dev() implementation
Masahiro Yamada [Mon, 10 Aug 2015 16:09:43 +0000 (01:09 +0900)]
dm: simplify uclass_foreach_dev() implementation

This can be simply written with list_for_each_entry(), maybe
this macro was not necessary in the first place.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Simon Glass <sjg@chromium.org>
4 years agosunxi: mmc: set transfer timeout according to byte_cnt.
Yousong Zhou [Sat, 29 Aug 2015 13:26:11 +0000 (21:26 +0800)]
sunxi: mmc: set transfer timeout according to byte_cnt.

Originally a timeout value of 2 seconds was used regardless of the size
of data to be transfered.  This prevented slow devices from working
correctly while there was no much gain for faster devices, e.g. it takes
3708ms for a transfer of uImage of size 1899008 bytes.

Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
4 years agosunxi: Enable non-secure access to RTC on sun6i (A31s)
Chen-Yu Tsai [Tue, 25 Aug 2015 02:49:19 +0000 (10:49 +0800)]
sunxi: Enable non-secure access to RTC on sun6i (A31s)

On the A31s the RTC is by default secured. Thus when u-boot
loads the kernel in non-secure world, the RTC is unavailable. The
SoC has a TrustZone Protection Controller, which can be used to
enable non-secure access to the RTC.

On the A31 the TZPC doesn't seem to do anything, i.e. changes to
its register contents do not affect access to the RTC.

Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
4 years agosunxi: Fix MAINTAINERS board sorting
Hans de Goede [Thu, 27 Aug 2015 18:07:28 +0000 (20:07 +0200)]
sunxi: Fix MAINTAINERS board sorting

The boards are sorted by SoC, move the Mele_A1000G_quad entry to the list
of sun6i boards where it belongs.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi: Add support for the Olimex A20 EVB
Marcus Cooper [Wed, 26 Aug 2015 18:38:33 +0000 (20:38 +0200)]
sunxi: Add support for the Olimex A20 EVB

Signed-off-by: Marcus Cooper <codekipper@gmail.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
4 years agosunxi: Add inet98v_rev2 defconfig and dts file
Hans de Goede [Tue, 25 Aug 2015 12:20:49 +0000 (14:20 +0200)]
sunxi: Add inet98v_rev2 defconfig and dts file

The inet98v_rev2 is a pcb used in generic A13 based tablets. It features
volume buttons, a power barrel, micro-usb otg, headphone connector and
a power button.

The dts file is identical to the one submitted to the upstream kernel.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi: Add inet97fv2_defconfig
Hans de Goede [Sat, 22 Aug 2015 18:06:02 +0000 (20:06 +0200)]
sunxi: Add inet97fv2_defconfig

The inet97fv2 is a board found in the first generation of cheap allwinner
A10 based 7" tablets.

Note that this patch does not add a dts file as we already have one from
our dts syncs with the kernel.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi: Ampe_A76_defconfig: Add otg id pin configuration
Hans de Goede [Tue, 25 Aug 2015 12:19:06 +0000 (14:19 +0200)]
sunxi: Ampe_A76_defconfig: Add otg id pin configuration

Add otg id pin configuration, this speeds up bootup when no host cable
is plugged into the otg port.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agomtd: nand: Make CONFIG_SYS_NAND_U_BOOT_OFFS configurable through Kconfig
Hans de Goede [Fri, 21 Aug 2015 19:49:51 +0000 (21:49 +0200)]
mtd: nand: Make CONFIG_SYS_NAND_U_BOOT_OFFS configurable through Kconfig

Make CONFIG_SYS_NAND_U_BOOT_OFFS configurable through Kconfig, just like
SYS_NAND_BUSWIDTH_16BIT this is only enabled on some SoCs using depends,
to avoid double defining it for SoCs which have not yet moved to Kconfig
for this.

Having this in Kconfig is useful because this is something which may
differ from one board to the other even when using the same SoC.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
Acked-by: Scott Wood <scottwood@freescale.com>
4 years agosunxi_nand_spl: clear status flags in SPL implementation
Boris Brezillon [Sat, 29 Aug 2015 10:29:38 +0000 (12:29 +0200)]
sunxi_nand_spl: clear status flags in SPL implementation

Some status flags remain set until you explicetly clear the bit
in the status register.
Fix the SPL implementation to avoid false positive.

Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
[hdegoede@redhat.com: Port from v2015.07 to v2015.10]
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
4 years agosunxi_nand_spl: Remove NAND_SUNXI_SPL_SYNDROME_PARTITIONS_END
Hans de Goede [Fri, 21 Aug 2015 13:23:57 +0000 (15:23 +0200)]
sunxi_nand_spl: Remove NAND_SUNXI_SPL_SYNDROME_PARTITIONS_END

We only ever use syndrome mode for the partitions which contain the SPL,
as that is required for the BROM to be able to read the SPL.

Instead of using some arbritray limit for deciding whether or not to
use syndrome, be smart and check if u-boot-dtb.bin is directly behind
the SPL, if it is not then it is on its own partition and we should not
use syndrome.

Note the reason why we only use syndrome mode for the SPL is because it
comeswith weaker randomization, introducing a risk for more bit errors,
so we want to avoid it when possible.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi_nand_spl: Rename SPL_NAND_SUNXI to NAND_SUNXI
Hans de Goede [Sun, 16 Aug 2015 12:48:22 +0000 (14:48 +0200)]
sunxi_nand_spl: Rename SPL_NAND_SUNXI to NAND_SUNXI

We eventually want to add full nand support, since it makes no sense
to build SPL with nand support and u-boot without, or the other way
around, a single option will suffice.

Renaming the Kconfig option now makes things easier when we add full
nand support in the future.

The "obj-$(CONFIG_NAND_SUNXI) += sunxi_nand_spl.o" is moved to an
"ifdef CONFIG_SPL_BUILD" block in the Makefile.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi_nand_spl: Add support for backup boot partitions
Hans de Goede [Sat, 15 Aug 2015 19:51:33 +0000 (21:51 +0200)]
sunxi_nand_spl: Add support for backup boot partitions

The BROM does not care / use bad page markings, instead it deals with
any bad pages in the first erase-block by simply trying to load "boot0"
from the next erase-block.

This commit implements the same strategy for the sunxi spl nand code,
allowing it to boot from the backup boot partition when the main boot
partition is bad (tested by erasing the main boot partition).

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi_nand_spl: Auto detect nand configuration parameters
Hans de Goede [Sat, 15 Aug 2015 19:23:08 +0000 (21:23 +0200)]
sunxi_nand_spl: Auto detect nand configuration parameters

Auto detect the nand configuration parameters, like the BROM does.

This allows us to get rid of various Kconfig settings, and is
necessary to support generic boards like the mk802 which have seen
many production runs with different nands.

The full blown u-boot/kernel nand driver uses the nand id to determine
this info, for the SPL we do as the BROM does and simply try a few
standard configs.

Note the table only contains configs which are known to actually be used,
rather then all the configs the BROM tries. This means that it may need
to be updated in the future as we add support for nand on more boards.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi_nand_spl: Parametrize lowlevel read functions
Hans de Goede [Sat, 15 Aug 2015 18:51:53 +0000 (20:51 +0200)]
sunxi_nand_spl: Parametrize lowlevel read functions

Parametrize the lowlevel nand_read_page function, instead of directly
using the CONFIG_foo settings for page-size, etc. there and add a few
wrappers / helper functions for calling it.

This is a preparation patch for adding auto-detecting of the nand
parameters like the BROM does.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi_nand_spl: Properly config page-size in the nand ctl register
Hans de Goede [Sat, 15 Aug 2015 18:05:13 +0000 (20:05 +0200)]
sunxi_nand_spl: Properly config page-size in the nand ctl register

Properly config page-size in the nand ctl register, it seems that things
work fine without doing this, but still lets play it safe and properly
set the page-size.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
4 years agosunxi_nand_spl: Add support for sun4i and sun5i SoCs
Hans de Goede [Sat, 15 Aug 2015 11:17:49 +0000 (13:17 +0200)]
sunxi_nand_spl: Add support for sun4i and sun5i SoCs

Other then having a few less chip-select lines the nand controller
on sun4i, sun5i and sun7i is identical.

Note this patch also muxes GPC7 to the NAND on sun7i where as before
it was not muxed this way. GPC7 is a standard NAND pin, so it should
always be muxed to the NAND when in use.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>