Nikita Kiryanov authored 10 years ago and Stefano Babic committed 10 years ago

The bmode command forces the SoC to use a specific boot device
by writing its boot mode into SRC_GPR9, and notifying the SoC of
the change using SRC_GPR10[28] bit: if the bit is on, bootROM
uses the value in SRC_GPR9 instead of SRC_SMBR1 to determine
the boot device.

SPL on the other hand is oblivious to this distinction, so once
the bootROM loads SPL from the device configured in SRC_GPR10,
SPL will attempt to load U-Boot from the device configured in
SRC_SMBR1, which is not updated by the bootROM to the value in
SRC_GPR9.

The result is that the selected boot device is not used across all
the boot stages.

Update spl_boot_device() to look at gpr9 when necessary.

Signed-off-by: Nikita Kiryanov <nikita@compulab.co.il>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Troy Kisky <troy.kisky@boundarydevices.com>
Cc: Tim Harvey <tharvey@gateworks.com>
Cc: Eric Nelson <eric.nelson@boundarydevices.com>
Cc: Fabio Estevam <fabio.estevam@freescale.com>
Cc: Heiko Schocher <hs@denx.de>

This is a MIME GnuPG-signed message.  If you see this text, it means that
your E-mail or Usenet software does not support MIME signed messages.
The Internet standard for MIME PGP messages, RFC 2015, was published in 1996.
To open this message correctly you will need to install E-mail or Usenet
software that supports modern Internet standards.

This reverts commit 1e96220a.

Remove duplicated vxworks.h header.
The same change was done by
"ARM: prevent compiler warnings from bootm.c"
(sha1: 8d196e52)

Signed-off-by: Michal Simek <michal.simek@xilinx.com>

CONFIG_UNIPHIER_SERIAL has been moved to Kconfig and
it is defined in ./.config but not in spl/.config,
so pin_init() should be called from the normal image
so that UART works correctly.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

This commit merges
  arch/arm/cpu/armv7/uniphier/ph1-*/board_postclk_init.c
to
  arch/arm/cpu/armv7/uniphier/board_postclk_init.c

Because PH1-Pro4 does not have the BCU block, add __weak to
bcu_init().

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

In the current implementation of the boot sequence of UniPhier
platform, 32KB temporary RAM is available before relocation.
The malloc area and the stack shares the 32KB area.

With CONFIG_SYS_MALLOC_F_LEN set to 0x7000 (28KB), only 0x1000 (4KB)
is left for the stack.  In some use cases, the system hangs up
with stack over-flow.

Even with driver-model UART enabled, the malloc area of 0x2000 (8KB)
should be enough.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

The function sg_set_pinsel is useful for switching I/O pins
but it can be only used in C code.  This commit adds a simple
macro that is available in asm code.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

FAT-formated USB storage device access is available.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

Support EHCI host driver used on Panasonic UniPhier platform.
Since Device Tree is not supported on UniPhier yet, the base address
of USB cores are passed from board files (platdevice.c).

TODO for me:
Move the base address to device trees.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Marek Vasut <marex@denx.de>

This commit adds register defines of MIO (Media I/O) block
of UniPhier platform.  This file is necessary to control
the reset signals of the USB cores.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Marek Vasut <marex@denx.de>

These IO pins are necessary for port power control and
over current detect.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

If a support card is attached to the main board, the on-board
SMSC9118 LAN controller is available.  It must be kept in reset
state for a while on start-up.

When the board is kicked via a debbuger rather than pushing the
hardware reset button, on-board chips are not reset; in this case
the reset signals should be asserted by software.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

The same bit-field macros are defined in include/linux/serial_reg.h
so let's include it and delete duplicated defines.

Also, remove unnecessary inclusion of <common.h>.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

Signed-off-by: Tom Rini <trini@ti.com>

This patch adds support for comments in the input to mkenvimage, i.e. in
the environment source: All lines starting with a # in the firs column
will be ignored.

Additionally empty lines will also be ignored.

Signed-off-by: Dominik Muth <dominik.muth@bkvibro.com>

In a message from Wolfgang Denk highlighting warnings from cppcheck,
the patch will address those that are correctly diagnosed. Some are
false-positives:

> [fs/zfs/zfs.c:937]: (error) Memory leak: l
dmu_read() allocates "l" if successful, so error-case should not free
it.
> [fs/zfs/zfs.c:1141]: (error) Memory leak: dnbuf
dmu_read() allocates "dnbuf" if successful, so error-case should not
free it.
> [fs/zfs/zfs.c:1372]: (error) Memory leak: osp
zio_read() allocates "osp" if successful, so error-case should
not free it.
> [fs/zfs/zfs.c:1726]: (error) Memory leak: nvlist
int_zfs_fetch_nvlist() allocates "nvlist" if successful, so error-case
should not free it.

Signed-off-by: Jorgen Lundman <lundman@lundman.net>

cppcheck reports:

[board/esd/common/auto_update.c:458]: (error) Uninitialized variable: cnt

The variable is not really used anywhere, so remove it.

Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Matthias Fuchs <matthias.fuchs@esd-electronics.com>
Acked-by: Matthias Fuchs <matthias.fuchs@esd.eu>

cppcheck reports:

[arch/arm/cpu/arm926ejs/mxs/timer.c:96]: (error) Uninitialized
variable: now

Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Marek Vasut <marex@denx.de>
Cc: Stefano Babic <sbabic@denx.de>

Nobuhiro Iwamatsu authored 10 years ago and Heiko Schocher committed 10 years ago

In case of repeated START condition, the restart has to be kicked
before clear status (MSR register). If it is kicked after clear status,
R-Car I2C may transfer data (TXD register) or receive data (RXD register)
instead of transferring slave address (MAR register).

Signed-off-by: Ryo Kataoka <ryo.kataoka.wt@renesas.com>
Signed-off-by: Nobuhiro Iwamatsu <nobuhiro.iwamatsu.yj@renesas.com>

Shaveta Leekha authored 10 years ago and Heiko Schocher committed 10 years ago

Some slow I2C devices like Power Monitor(ZM7304) at times
do not work well with low timeout value, so I2C bus get stuck
during read cycle with this device, changing it to 100ms from
10ms works fine

A lot of other i2c drivers like mxc and i2c drivers of BOOTROM
also use relax timeouts to give sufficient ticks to work well
with slower devices

Signed-off-by: Shaveta Leekha <shaveta@freescale.com>
Signed-off-by: Poonam Aggrwal <poonam.aggrwal@freescale.com>

Dirk Eibach authored 10 years ago and Heiko Schocher committed 10 years ago

ppc4xx i2c master gets stuck on errors while repeated start is
active. Can be easily reproduced by "i2c md" on an unpopulated
i2c address. There is not stop condition given, scl remains
pulled low.
The only way out seems to be doing a stop manually and then a
soft reset.

Signed-off-by: Dirk Eibach <dirk.eibach@gdsys.cc>
Reviewed-by: Stefan Roese <sr@denx.de>

Dirk Eibach authored 10 years ago and Heiko Schocher committed 10 years ago

Debugging some i2c trouble I saw on my scope that repeated
start is not working properply. The 4xx even held clock pulled down
after transfers. Having a look in the driver I realized
that IIC_CNTL_RPST is set on that part of the transfer that should
begin with a repeated start. But repeated start is about not sending a
stop condition, so IIC_CNTL_RPST has to be set on the last transfer
before the repeated start happens.

Signed-off-by: Dirk Eibach <dirk.eibach@gdsys.cc>
Reviewed-by: Stefan Roese <sr@denx.de>

Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>

Remove BFIN from the CPU macro in Makefile.

Signed-off-by: Sonic Zhang <sonic.zhang@analog.com>

Fix error detected by cppcheck:

[board/matrix_vision/mvblx/sys_eeprom.c:353]: (error) Buffer is
accessed out of bounds.

Signed-off-by: Wolfgang Denk <wd@denx.de>
cc: Michael Jones <michael.jones@matrix-vision.de>
Acked-by: Michael Jones <michael.jones@matrix-vision.de>

This file got corrupted by the automatic editin of commit 1a459660 "Add
GPL-2.0+ SPDX-License-Identifier to source files"; restore the
opiginal content and manually insert the SPDX ID.

The bug was detected by running cppcheck, which reported:
[common/cmd_fitupd.c:8]: (error) Invalid number of character ({) when
these macros are defined: 'CONFIG_UPDATE_TFTP'.

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

Fix error detected by cppcheck:

[drivers/net/uli526x.c:551]: (error) printf format string requires 3
parameters but only 2 are given.

Signed-off-by: Wolfgang Denk <wd@denx.de>
Cc: Roy Zang <tie-fei.zang@freescale.com>

Fix error detected by cppcheck:

[board/cogent/lcd.c:237]: (error) Invalid number of character (()
when these macros are defined:
'CONFIG_SHOW_ACTIVITY;CONFIG_STATUS_LED'.

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

There are a number of places where U-Boot intentionally and legally
accesses physical address 0x0000, for example when installing
exception vectors on systems where these are located in low memory.

Add "cppcheck-suppress nullPointer" comments to silence cppcheck
where this is intentional and legal.

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

Correct environment variable for output directory is KBUILD_OUTPUT.

Signed-off-by: Timo Ketola <timo@exertus.fi>

On architectures where 'long' is 64 bit, the u-boot environment
as seen by the fw_env tools was missing 4 bytes.
This patch fixes getenvsize(), and thus also ensures that the
environment's CRC32 checksum is calculated correctly.

Signed-off-by: Dominic Sacré <dominic.sacre@gmx.de>
Cc: Wolfgang Denk <wd@denx.de>
Cc: Tom Rini <trini@ti.com>

All the other files in drivers/mtd/* are not really CFI related. So lets
only include the files that are CFI specific.

Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Heiko Schocher <hs@denx.de>
Cc: Scott Wood <scottwood@freescale.com>
Cc: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Jagan Teki <jagannadh.teki@gmail.com>
Acked-by: Jagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Acked-by: Heiko Schocher <hs@denx.de>

This commit imports some updates in misc scripts from Linux 3.18-rc1.

[1] commit cbb4d3e6510b99522719c5ef0 by Horia Geanta
scripts/kernel-doc: handle object-like macros

[2] commit 164f0d2efaaef83 by Michal Marek
kbuild: Fix handling of backslashes in *.cmd files

[3] commit 270a00963cd367214e by Randy Dunlap
scripts/kernel-doc: recognize __meminit

[4] commit a4954fd7724c0f55361eb5 by Masahiro Yamada
  kbuild: remove obj-n and lib-n handling

[5] commit 5b2389b45d1a9c12b9f by Masahiro Yamada
kbuild: simplify build, clean, modbuiltin shorthands

Signed-off-by: Horia Geanta <horia.geanta@freescale.com>
Signed-off-by: Michal Marek <mmarek@suse.cz>
Signed-off-by: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

This commit imports Kconfig updates from Linux 3.18-rc1.

'kvmconfig' and 'tinyconfig' help message have been commented out
since they are unavailable at least now;  in the future perhaps
we can implement 'tinyconfig' to disable most of CONFIG_CMD_* to
create a very small U-Boot image.

[1] commit 3aaefce10351 by Josh Triplett
x86, platform, kconfig: move kvmconfig functionality to a helper

[2] commit 0da1d4a0b951 by Josh Triplett
x86: Add "make tinyconfig" to configure the tiniest possible kernel

[3] commit c40724d3f381 by Brian Norris
  kconfig: lxdialog: fix spelling

[4] commit 7285996aa000 by Brian Norris
  kconfig: nconfig: fix multi-byte UTF handling

Signed-off-by: Josh Triplett <josh@joshtriplett.org>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>

Since Linux 3.15, relative path feature and related fixes,
cleanups have been merged to the top Makefile.

The relative path feature looks stable enough, so let's import it
to U-Boot along with various cleanups.

Commits imported from Linux (some need adjustment) are:

[1] commit 7e1c04779efd by Michal Marek
  kbuild: Use relative path for $(objtree)

[2] commit 890676c65d69 by Michal Marek
  kbuild: Use relative path when building in the source tree

[3] commit 9da0763bdd82 by Michal Marek
  kbuild: Use relative path when building in a subdir of the source tree

[4] commit c2e28dc975ea by Michal Marek
  kbuild: Print the name of the build directory

[5] commit 066b7ed95580 by Michal Marek
  kbuild: Do not print the build directory with make -s

[6] commit 3f1d9a6cec01 by Michal Marek
  kbuild: make -s should be used with kernelrelease/kernelversion/image_name

[7] commit 7ff525712acf by Masahiro Yamada
  kbuild: fake the "Entering directory ..." message more simply

[8] commit 745a254322c8 by Masahiro Yamada
  kbuild: use $(Q) for sub-make target

[9] commit aa55c8e2f7a3 by Masahiro Yamada
  kbuild: handle C=... and M=... after entering into build directory

[10] commit ab7474ea5361 by Borislav Petkov
  Kbuild: Ignore GREP_OPTIONS env variable

To use relative path feature, tools/Makefile and scripts/Makefile.autoconf
must be tweaked.

Signed-off-by: Michal Marek <mmarek@suse.cz>
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>