project/bcm63xx/u-boot.git
10 years agoRevert "ARM: SPL: do not set gd again"
Tom Rini [Tue, 16 Sep 2014 15:08:46 +0000 (11:08 -0400)]
Revert "ARM: SPL: do not set gd again"

At the high level, the problem is that we set gd multiple times (and
still do, even after the commit we're reverting).  We set important
parts of gd to the copy which is not above stack but rather in the data
section.  For the release, we're going to revert this change and for the
next release we shall correct things to only, really, set gd once to an
appropriate location and ensure that comments about it are correct too.

This reverts commit f0c3a6c4ad09210d5d4aeafe87685ee75e5683d6.

Acked-by: Albert Aribaud <albert.u.boot@aribaud.net>
Signed-off-by: Tom Rini <trini@ti.com>
10 years agokconfig: armv8: move CONFIG_ARM64 to Kconfig
Masahiro Yamada [Sat, 13 Sep 2014 18:01:51 +0000 (03:01 +0900)]
kconfig: armv8: move CONFIG_ARM64 to Kconfig

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
10 years agovexpress64: kconfig: consolidate CONFIG_TARGET_VEXPRESS_AEMV8A_SEMI
Masahiro Yamada [Sat, 13 Sep 2014 18:01:50 +0000 (03:01 +0900)]
vexpress64: kconfig: consolidate CONFIG_TARGET_VEXPRESS_AEMV8A_SEMI

We do not have to distinguish CONFIG_TARGET_VEXPRESS_AEMV8A_SEMI
from CONFIG_TARGET_VEXPRESS_AEMV8A.  Rename the former to the latter.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Reviewed-by: Steve Rae <srae@broadcom.com>
Cc: David Feng <fenghua@phytium.com.cn>
10 years agonet: dns: fix for DNS queries sent to the wrong MAC address
Gerhard Sittig [Fri, 12 Sep 2014 06:48:15 +0000 (08:48 +0200)]
net: dns: fix for DNS queries sent to the wrong MAC address

When a DNS query is sent out, the ethernet packet can get directed to
the MAC address of a server that was communicated to before.  This is
wrong when the previously stored MAC address corresponds to a different
server's IP address, i.e. when the IP address of the previous and the
current communication are different.

The error can get reproduced by running a sequence of e.g. a TFTP
download and a DNS query, where the TFTP and DNS servers reside on
individual machines.

The fix is to clear the server's MAC address that might be left from a
previous operation, and to fetch the peer's MAC address in a new ARP
lookup, before the DNS query is sent.  This is the approach taken in
other network services, like 8e52533d1095 ("net: tftpsrv: Get correct
client MAC address").

Reported-by: Dirk Zimoch <dirk.zimoch@psi.ch>
Signed-off-by: Gerhard Sittig <gsi@denx.de>
10 years agokconfig: add sanity checks for SPL configuration
Masahiro Yamada [Wed, 10 Sep 2014 09:13:10 +0000 (18:13 +0900)]
kconfig: add sanity checks for SPL configuration

For the SPL configuration, "make <dir>/<target>" is used.
Here,
  <dir> is either "spl" or "tpl"
  <target> is one of "config", "menuconfig", "xconfig", etc.

This commit adds two checks:

[1] If <dir> is given an unsupported subimage, the configuration
    should error out like this:

  $ make qpl/menuconfig
  ***
  *** "make qpl/menuconfig" is not supported.
  ***

[2] Make sure that "CONFIG_SPL" is enabled in the ".config" before
    running "make spl/menuconfig.  Otherwise, the SPL image
    is not built at all.  Having "spl/.config" makes no sense.
    In such a case, the configuration should exit with a message:

  $ make spl/menuconfig
  ***
  *** Create ".config" with "CONFIG_SPL" enabled
  *** before "make spl/menuconfig".
  ***

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Suggested-by: Simon Glass <sjg@chromium.org>
10 years agonet: usb: Add SMSC copyright to smsc95xx driver
Simon Glass [Mon, 8 Sep 2014 19:44:14 +0000 (13:44 -0600)]
net: usb: Add SMSC copyright to smsc95xx driver

This driver was upstreamed without an SMSC copyright, even thought it seems
that SMSC was the original author.

See the kernel version for a code comparison:

http://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/?id=2f7ca802bdae2ca41022618391c70c2876d92190

It's not clear who actually moved this code, or whether the kernel was the
original source, or somewhere else, but it probably should still have the
SMSC copyright.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agomtdcore: Fix a build error with CONFIG_CMD_MTDPARTS_SPREAD
maxin.john@enea.com [Mon, 8 Sep 2014 17:04:16 +0000 (19:04 +0200)]
mtdcore: Fix a build error with CONFIG_CMD_MTDPARTS_SPREAD

This patch fixes the build error for CONFIG_CMD_MTDPARTS_SPREAD

Signed-off-by: Maxin B. John <maxin.john@enea.com>
10 years agomtd: nand: davinci_nand: update write_page function for keystone RBL
Khoronzhuk, Ivan [Sat, 6 Sep 2014 19:17:07 +0000 (22:17 +0300)]
mtd: nand: davinci_nand: update write_page function for keystone RBL

After mtd was synced with Linux 3.14
(ff94bc40af3481d47546595ba73c136de6af6929)
the number of parameters for write_page function of nand_chip was
changed. The additional two var were needed for subpage write.
As keystone has no supbage write they are not needed. So correct
only function definition by upgrading it's parameter list.
That helps to get ritd of compilation warning.

Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@ti.com>
10 years agogeneric_board: do not set gd->fdt_blob unless CONFIG_OF_CONTROL=y
Masahiro Yamada [Sat, 6 Sep 2014 14:39:00 +0000 (23:39 +0900)]
generic_board: do not set gd->fdt_blob unless CONFIG_OF_CONTROL=y

gd->fdt_blob is used for FDT control of U-Boot.
If CONFIG_OF_CONTROL is not defined, it is useless.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Simon Glass <sjg@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
10 years agokbuild: standalone: simplify clean-files
Masahiro Yamada [Thu, 4 Sep 2014 16:56:52 +0000 (01:56 +0900)]
kbuild: standalone: simplify clean-files

Files added $(extra-) are removed by "make clean".
Besides, wildcard "*.srec *.bin" is simpler.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
10 years agoscripts/Makefile.clean: clean also $(extra-m) and $(extra-)
Masahiro Yamada [Thu, 4 Sep 2014 16:56:51 +0000 (01:56 +0900)]
scripts/Makefile.clean: clean also $(extra-m) and $(extra-)

This commit is a backport from Linux Kernel,
commit 9d5db8949f1ecf4019785b04d8986835d3c0e99e,
written by me.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
10 years agokconfig: show an error message when defconfig is not found
Masahiro Yamada [Thu, 4 Sep 2014 13:16:15 +0000 (22:16 +0900)]
kconfig: show an error message when defconfig is not found

When a non-existing defconfig is specified,
display an easy-to-understand message
(fake the error message on Linux Kernel):

  $ make foo_defconfig
  ***
  *** Can't find default configuration "confis/foo_defconfig"!
  ***

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Stephen Warren <swarren@nvidia.com>
10 years agocommon: fix include guards for CONFIG_MP
Gabriel Huau [Wed, 3 Sep 2014 20:57:54 +0000 (13:57 -0700)]
common: fix include guards for CONFIG_MP

This was breaking the build for some boards:
MPC8536DS MPC8536DS_36BIT MPC8536DS_SDCARD MPC8536DS_SPIFLASH qemu-ppce500

Include only these features for some PPC boards if the configuration for MultiProcessor
is enabled.

Signed-off-by: Gabriel Huau <contact@huau-gabriel.fr>
Cc: Tom Rini <trini@ti.com>
Cc: York Sun <yorksun@freescale.com>
Acked-by: York Sun <yorksun@freescale.com>
10 years agokconfig: fix a bug of "make config"
Masahiro Yamada [Wed, 3 Sep 2014 20:41:34 +0000 (05:41 +0900)]
kconfig: fix a bug of "make config"

Since 3ff291f371fa9858426774f3732924bacb61ed1c
(kconfig: convert Kconfig helper script into a shell script),
"make config" is not working because of a missing '$' before '(Q)'.

Besides, "make config" should be invoked via scripts/multiconfig.sh
to avoid a warning message:
Kconfig:11:warning: environment variable KCONFIG_OBJDIR undefined

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
10 years agokconfig: fix savedefconfig to handle TPL correctly
Masahiro Yamada [Wed, 3 Sep 2014 20:41:33 +0000 (05:41 +0900)]
kconfig: fix savedefconfig to handle TPL correctly

Since 3ff291f371fa9858426774f3732924bacb61ed1c
(kconfig: convert Kconfig helper script into a shell script),
"make savedefconfig" of TPL boards has not been working.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.o.rg
Acked-by: Simon Glass <sjg@chromium.org>
10 years agokconfig: fix whitespace handling bug of savedefconfig
Masahiro Yamada [Wed, 3 Sep 2014 20:41:32 +0000 (05:41 +0900)]
kconfig: fix whitespace handling bug of savedefconfig

Commit 3ff291f371fa9858426774f3732924bacb61ed1c
(kconfig: convert Kconfig helper script into a shell script)
introduced another regression.

Shell usually handles whitespaces as separators,
so "make saveconfig" outputs

  # CONFIG_FOO is not set

into:

  #
  CONFIG_FOO
  is
  not
  set

Whitespaces should not be treated as separators here.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
10 years agokconfig: fix savedefconfig to output empty defconfig
Masahiro Yamada [Wed, 3 Sep 2014 20:41:31 +0000 (05:41 +0900)]
kconfig: fix savedefconfig to output empty defconfig

Commit 3ff291f371fa9858426774f3732924bacb61ed1c
(kconfig: convert Kconfig helper script into a shell script)
introduced a minor regression.

make alldefconfig; make savedefconfig
should create an empty 'defconfig'.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
10 years agostandalone: use GCC_VERSION defined in compiler-gcc.h
Masahiro Yamada [Wed, 3 Sep 2014 17:40:59 +0000 (02:40 +0900)]
standalone: use GCC_VERSION defined in compiler-gcc.h

Now GCC_VERSION is defined in include/linux/compiler-gcc.h
(with a little different definition).
Use it and delete the one in examples/standlone/stub.c.

This should work on Clang too because __GNUC__, __GNUC_MINOR__,
__GNUC_PATCHLEVEL__ are also defined on Clang.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agocompiler*.h: sync include/linux/compiler*.h with Linux 3.16
Masahiro Yamada [Wed, 3 Sep 2014 17:40:58 +0000 (02:40 +0900)]
compiler*.h: sync include/linux/compiler*.h with Linux 3.16

Copy them from Linux v3.16 tag.
My main motivation of this commit is to add compiler-clang.h.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agoimplement the Android sparse image format
Steve Rae [Wed, 3 Sep 2014 17:05:54 +0000 (10:05 -0700)]
implement the Android sparse image format

update to provide usable implementation to U-Boot

Signed-off-by: Steve Rae <srae@broadcom.com>
10 years agocleanup code which handles the Android sparse image format
Steve Rae [Wed, 3 Sep 2014 17:05:53 +0000 (10:05 -0700)]
cleanup code which handles the Android sparse image format

- port dprintf() to debug()
- update formatting

Signed-off-by: Steve Rae <srae@broadcom.com>
10 years agoupdate code which handles Android sparse image format
Steve Rae [Wed, 3 Sep 2014 17:05:52 +0000 (10:05 -0700)]
update code which handles Android sparse image format

- remove unnecessary functions

Signed-off-by: Steve Rae <srae@broadcom.com>
10 years agoadd code to handle Android sparse image format
Steve Rae [Wed, 3 Sep 2014 17:05:51 +0000 (10:05 -0700)]
add code to handle Android sparse image format

Add original file (pristine) from :
  https://www.codeaurora.org/cgit/quic/la/kernel/lk/plain/app/aboot/aboot.c?h=master
[3b5092d20bd15a7a2879c13e9f64acc48d04af2d]

Signed-off-by: Steve Rae <srae@broadcom.com>
10 years agocmd_mem: add static to internally used functions
Masahiro Yamada [Wed, 3 Sep 2014 15:38:26 +0000 (00:38 +0900)]
cmd_mem: add static to internally used functions

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
10 years agokgdb: Remove first_entry for kgdb
Peng Fan [Mon, 1 Sep 2014 13:48:07 +0000 (21:48 +0800)]
kgdb: Remove first_entry for kgdb

There are two ways to run into handle_exception, run command 'kgdb' and
encounter a breakpoint which triggers exception handling.

The origin source code only saves regs when first run command 'kgdb'.
Take the following for example, When run 'kgdb', regs is saved to entry_regs.
When run 'bootz', regs is not saved. However, if we set a breakpoint, then
continue. When breakpoint is reached, run `quit`, and Now return to the
instruction which follows kgdb, but not bootz.This may cause errors. So,
save regs for each handle_exception call to return to the correct place.
Example:
Target      |    Host
=>kgdb      |    (gdb)b bootz
            |    (gdb)c
=>bootz     |
            |    (gdb)Here stop because of breakpoint
            |    (gdb)q

Signed-off-by: Peng Fan <van.freenix@gmail.com>
10 years agotools/genboardscfg.py: improve performance more with Kconfiglib
Masahiro Yamada [Mon, 1 Sep 2014 10:57:38 +0000 (19:57 +0900)]
tools/genboardscfg.py: improve performance more with Kconfiglib

The idea of using Kconfiglib was given by Tom Rini.
It allows us to scan lots of defconfigs very quickly.
This commit also uses multiprocessing for further acceleration.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Suggested-by: Tom Rini <trini@ti.com>
Acked-by: Simon Glass <sjg@chromium.org>
10 years agotools: Import Kconfiglib
Masahiro Yamada [Mon, 1 Sep 2014 10:57:37 +0000 (19:57 +0900)]
tools: Import Kconfiglib

Kconfiglib is the flexible Python Kconfig parser and library
created by Ulf Magnusson.
(https://github.com/ulfalizer/Kconfiglib)

This commit imports kconfiglib.py from
commit ce84c22e58fa59cb93679d4ead03c3cd1387965e,
with ISC SPDX-License-Identifier.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Signed-off-by: Ulf Magnusson <ulfalizer@gmail.com>
Cc: Ulf Magnusson <ulfalizer@gmail.com>
Cc: Wolfgang Denk <wd@denx.de>
10 years agoSPDX: Add ISC SPDX-License-Identifier
Masahiro Yamada [Mon, 1 Sep 2014 10:57:36 +0000 (19:57 +0900)]
SPDX: Add ISC SPDX-License-Identifier

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Wolfgang Denk <wd@denx.de>
10 years agokbuild: force to define __UBOOT__ in all the C sources
Masahiro Yamada [Sun, 31 Aug 2014 06:16:53 +0000 (15:16 +0900)]
kbuild: force to define __UBOOT__ in all the C sources

U-Boot has imported various source files from other projects,
mostly Linux.

Something like

  #ifdef __UBOOT__
    [ modification for U-Boot ]
  #else
    [ original code ]
  #endif

is an often used strategy for clarification of adjusted parts,
that is, easier re-sync in future.

Instead of defining __UBOOT__ in each source file,
passing it from the top Makefile would be easier.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Marek Vasut <marex@denx.de>
Acked-by: Heiko Schocher <hs@denx.de>
10 years agortl8169: Defer network packet processing
Thierry Reding [Thu, 28 Aug 2014 10:38:03 +0000 (12:38 +0200)]
rtl8169: Defer network packet processing

When network protocol errors occur (such as a file not being found on a
TFTP server), the processing done by the NetReceive() function will end
up calling the driver's .halt() implementation. However, after that the
device no longer has access to the memory buffers and will cause errors
such as this in the rtl_recv() function when trying to hand descriptors
back to the device:

pci_hose_bus_to_phys: invalid physical address

This can be fixed by deferring processing of network packets until the
descriptors have been handed back. That way rtl_halt() tearing down
network buffers is not going to prevent access to the buffers.

Reported-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Thierry Reding <treding@nvidia.com>
10 years agoopenrisc: Fix a few type cast related warnings
Vasili Galka [Tue, 26 Aug 2014 10:46:17 +0000 (13:46 +0300)]
openrisc: Fix a few type cast related warnings

Use size_t type for positive offsets instead of the loff_t type. The
later is defined as long long, which is larger than the pointer type
on OpenRISC architecture and therefore the following warning was
generated:

"warning: cast to pointer from integer of different size"

Signed-off-by: Vasili Galka <vvv444@gmail.com>
10 years agoFix a few printf argument verification warnings
Vasili Galka [Tue, 26 Aug 2014 10:45:48 +0000 (13:45 +0300)]
Fix a few printf argument verification warnings

The parameters of size_t type shall be formatted using "%zu" and not
using "%d".

Precision argument for the "%.*s" parameters shall be of int type.

Signed-off-by: Vasili Galka <vvv444@gmail.com>
10 years agocosmetic: README.scrapyard: add NIOS2 boards
Masahiro Yamada [Mon, 25 Aug 2014 01:37:49 +0000 (10:37 +0900)]
cosmetic: README.scrapyard: add NIOS2 boards

Commit 70fbc461 removed obsolete PCI5441 and PK1C20 boards.
This commit adds them to README.scrapyard and also fills
some commit IDs and dates for removed boards.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Thomas Chou <thomas@wytron.com.tw>
Acked-by: Thomas Chou <thomas@wytron.com.tw>
10 years agokconfig: remove redundant "string" type in arch and board Kconfigs
Masahiro Yamada [Sat, 13 Sep 2014 18:01:49 +0000 (03:01 +0900)]
kconfig: remove redundant "string" type in arch and board Kconfigs

Now the types of CONFIG_SYS_{ARCH, CPU, SOC, VENDOR, BOARD, CONFIG_NAME}
are specified in arch/Kconfig.

We can delete the ones in arch and board Kconfig files.

This commit can be easily reproduced by the following command:

find . -name Kconfig -a ! -path ./arch/Kconfig | xargs sed -i -e '
/config[[:space:]]SYS_\(ARCH\|CPU\|SOC\|\VENDOR\|BOARD\|CONFIG_NAME\)/ {
    N
    s/\n[[:space:]]*string//
}
'

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
10 years agokconfig: add CONFIG_SYS_{ARCH, CPU, ...} type def to arch/Kconfig
Masahiro Yamada [Sat, 13 Sep 2014 18:01:48 +0000 (03:01 +0900)]
kconfig: add CONFIG_SYS_{ARCH, CPU, ...} type def to arch/Kconfig

This commit adds the type definitions (+ help messages) of
CONFIG_SYS_{ARCH, CPU, SOC, VENDOR, BOARD, CONFIG_NAME} to arch/Kconfig,
which would save lots of type defs for taget boards.
(See also the next commit.)

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
10 years agoMerge branch 'master' of git://git.denx.de/u-boot-dm
Tom Rini [Sat, 13 Sep 2014 20:32:52 +0000 (16:32 -0400)]
Merge branch 'master' of git://git.denx.de/u-boot-dm

10 years agoMerge branch 'master' of git://git.denx.de/u-boot-arm
Tom Rini [Fri, 12 Sep 2014 15:27:46 +0000 (11:27 -0400)]
Merge branch 'master' of git://git.denx.de/u-boot-arm

10 years agoarm: Fix _start for CONFIG_SYS_DV_NOR_BOOT_CFG
Benoît Thébaudeau [Wed, 3 Sep 2014 21:32:34 +0000 (23:32 +0200)]
arm: Fix _start for CONFIG_SYS_DV_NOR_BOOT_CFG

The boards using CONFIG_SYS_DV_NOR_BOOT_CFG (i.e. calimain,
da850evm_direct_nor and enbw_cmc) had the _start symbol defined after
the CONFIG_SYS_DV_NOR_BOOT_CFG word rather than before it in
arch/arm/lib/vectors.S. Because of that, if by lack of luck
'gd->mon_len = (ulong)&__bss_end - (ulong)_start' (see setup_mon_len())
was a multiple of 4 kiB (see reserve_uboot()), then the last BSS word
overlapped the first word of the following reserved RAM area (or went
beyond the top of RAM without such an area) after relocation because
__image_copy_start did not match _start (see relocate_code()).

This was broken by commit 41623c9 'arm: move exception handling out of
start.S files', which defined _start twice (before and after the
CONFIG_SYS_DV_NOR_BOOT_CFG word), then by commit 0a26e1d 'arm: fix a
double-definition error of _start symbol', which kept the definition of
the _start symbol after the CONFIG_SYS_DV_NOR_BOOT_CFG word. This new
commit fixes this issue by restoring the original behavior, i.e. by
defining the _start symbol before the CONFIG_SYS_DV_NOR_BOOT_CFG word.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau.dev@gmail.com>
Cc: Albert Aribaud <albert.u.boot@aribaud.net>
Cc: Manfred Rudigier <manfred.rudigier@omicron.at>
Cc: Christian Riesch <christian.riesch@omicron.at>
Cc: Sudhakar Rajashekhara <sudhakar.raj@ti.com>
Cc: Heiko Schocher <hs@denx.de>
10 years agoarm: Make reset position-independent
Benoît Thébaudeau [Wed, 3 Sep 2014 21:32:33 +0000 (23:32 +0200)]
arm: Make reset position-independent

Some boards, like mx31pdk and tx25, require the beginning of the SPL
code to be position-independent. For these two boards, this is because
they use the i.MX external NAND boot, which starts by executing the
first NAND Flash page from the NFC page buffer. The SPL then needs to
copy itself to its actual link address in order to free the NFC page
buffer and use it to load the non-SPL image from Flash before running
it. This means that the SPL runtime address differs from its link
address between the reset and the initial copy performed by
board_init_f(), so this part of the SPL binary must be
position-independent.

This requirement was broken by commit 41623c9 'arm: move exception
handling out of start.S files', which used an absolute address to branch
to the reset routine. This new commit restores the original behavior,
which just performed a relative branch. This fixes the boot of mx31pdk
and tx25.

Signed-off-by: Benoît Thébaudeau <benoit.thebaudeau.dev@gmail.com>
Reported-by: Helmut Raiger <helmut.raiger@hale.at>
Cc: Albert Aribaud <albert.u.boot@aribaud.net>
Cc: Magnus Lilja <lilja.magnus@gmail.com>
Cc: John Rigby <jcrigby@gmail.com>
Tested-by: Magnus Lilja <lilja.magnus@gmail.com>
10 years agoREADME.clang: build command with clang
Jeroen Hofstee [Wed, 10 Sep 2014 18:08:52 +0000 (20:08 +0200)]
README.clang: build command with clang

Cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agoMakefile: default to cc for host compiler
Jeroen Hofstee [Wed, 10 Sep 2014 18:08:51 +0000 (20:08 +0200)]
Makefile: default to cc for host compiler

Since the host compiler might not be gcc but e.g. clang
default to cc/c++ to invoke it.

cc: Masahiro Yamada <yamada.m@jp.panasonic.com>
cc: Tom Rini <trini@ti.com>
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agodm: tegra: Enable driver model for serial
Simon Glass [Thu, 4 Sep 2014 22:27:36 +0000 (16:27 -0600)]
dm: tegra: Enable driver model for serial

Use driver model for serial ports.

Since Tegra now uses driver model for serial, adjust the definition of
V_NS16550_CLK so that it is clear that this is only used for SPL.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agotegra: dts: Add serial port details
Simon Glass [Thu, 4 Sep 2014 22:27:35 +0000 (16:27 -0600)]
tegra: dts: Add serial port details

Some Tegra device tree files do not include information about the serial
ports. Add this and also add information about the input clock speed.

The console alias needs to be set up to indicate which port is used for
the console.

Also add a binding file since this is missing.

Series-changes; 5
- Add full serial port nodes from Linux tree (commit fc9d4dbe)
- Use /chosen/stdout-path instead of /aliases/console to specify the console

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: serial: Add driver model support for ns16550
Simon Glass [Thu, 4 Sep 2014 22:27:34 +0000 (16:27 -0600)]
dm: serial: Add driver model support for ns16550

Add driver model support so that ns16550 can support operation both with
and without driver model.

The driver needs a clock frequency so cannot stand alone unfortunately. The
clock frequency must be provided by a separate driver.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: serial: Collect common baud rate code in ns16550
Simon Glass [Thu, 4 Sep 2014 22:27:33 +0000 (16:27 -0600)]
dm: serial: Collect common baud rate code in ns16550

The same sequence is used in several places, so move it into a function.
Note that UART_LCR_BKSE is an alias for UART_LCR_DLAB.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: serial: Move baud rate calculation to ns16550.c
Simon Glass [Thu, 4 Sep 2014 22:27:32 +0000 (16:27 -0600)]
dm: serial: Move baud rate calculation to ns16550.c

Move the function that calculates the baud rate divisor into ns16550.c so
it can be used by that file.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agosandbox: dts: Add a serial console node
Simon Glass [Thu, 4 Sep 2014 22:27:29 +0000 (16:27 -0600)]
sandbox: dts: Add a serial console node

If the sandbox device tree is provided to U-Boot (with the -d flag) then it
will use the device tree version in preference to the built-in device. The
only difference is the colour.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agosandbox: serial: Support a coloured console
Simon Glass [Thu, 4 Sep 2014 22:27:28 +0000 (16:27 -0600)]
sandbox: serial: Support a coloured console

The current sandbox serial driver is a pretty trivial example and does not
have the featues that might be needed for other board serial drivers. To
help provide a better example, add a text colour property to the device
tree for sandbox. This uses platform data, a device tree node, driver
private data and a remove() method.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agosandbox: Convert serial driver to use driver model
Simon Glass [Thu, 4 Sep 2014 22:27:27 +0000 (16:27 -0600)]
sandbox: Convert serial driver to use driver model

Adjust the sandbox serial driver to use the new driver model uclass. The
driver works much as before, but within the new framework.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: Add a uclass for serial devices
Simon Glass [Thu, 4 Sep 2014 22:27:26 +0000 (16:27 -0600)]
dm: Add a uclass for serial devices

Serial devices support simple byte input/output and a few operations to find
out whether data is available. Add a basic uclass for serial devices to be
used by drivers that are converted to driver model.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: Adjust lists_bind_fdt() to return the bound device
Simon Glass [Thu, 4 Sep 2014 22:27:25 +0000 (16:27 -0600)]
dm: Adjust lists_bind_fdt() to return the bound device

Allow the caller to find out the device that was bound in response to this
call.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: fdt: Add a function to look up a chosen node
Simon Glass [Thu, 4 Sep 2014 22:27:24 +0000 (16:27 -0600)]
dm: fdt: Add a function to look up a chosen node

Within /chosen we may have a node which points to another node, similar
to how /aliases works. Add a helper function to do this lookup.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agoserial: Set up the 'priv' pointer when creating a serial device
Simon Glass [Thu, 4 Sep 2014 22:27:23 +0000 (16:27 -0600)]
serial: Set up the 'priv' pointer when creating a serial device

The stdio_dev structure has a private pointer for its creator, but it is
not set up by the serial system. Set it to point to the serial device so
that it can be found by code called by stdio.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agotegra: Convert tegra GPIO driver to use driver model
Simon Glass [Wed, 3 Sep 2014 23:37:03 +0000 (17:37 -0600)]
tegra: Convert tegra GPIO driver to use driver model

This is an implementation of GPIOs for Tegra that uses driver model. It has
been tested on trimslice and also using the new iotrace feature.

The implementation uses a top-level GPIO device (which has no actual GPIOS).
Under this all the banks are created as separate GPIO devices.

The GPIOs are named as per the Tegra datasheet/header files: A0..A7, B0..B7,
..., Z0..Z7, AA0..AA7, etc.

Since driver model is not yet available before relocation, or in SPL, a
special function is provided for seaboard's SPL code.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: tegra: Set up a pre-reloc malloc()
Simon Glass [Wed, 3 Sep 2014 23:37:02 +0000 (17:37 -0600)]
dm: tegra: Set up a pre-reloc malloc()

Allocate 1KB so that driver model can operate before relocation.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: Make driver model available before board_init()
Simon Glass [Wed, 3 Sep 2014 23:37:01 +0000 (17:37 -0600)]
dm: Make driver model available before board_init()

For some boards board_init() will change GPIOs, so we need to have driver
model available before then. Adjust the board init to arrange this, but
enable it for driver model only, just to be safe.

This does create additional #ifdef logic, but it is safer than trying to
make a pervasive change which may cause some boards to break.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agoSet up stdio earlier when using driver model
Simon Glass [Wed, 3 Sep 2014 23:37:00 +0000 (17:37 -0600)]
Set up stdio earlier when using driver model

Since driver model registers itself with the stdio subsystem, and we
want to avoid delayed registration and other complexity associated with
the current serial console, move the stdio subsystem init earlier when
driver model is used for serial.

This simplifies the implementation. Should there be any problems with
this approach they can be dealt with as boards are converted over to
use driver model for serial.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agodm: Move pre-reloc init earlier to cope with board_early_init_f()
Simon Glass [Wed, 3 Sep 2014 23:36:59 +0000 (17:36 -0600)]
dm: Move pre-reloc init earlier to cope with board_early_init_f()

In order to support GPIO access in board_early_init_f() we must set up
driver model before this function is called. In any case, earlier is
better since driver model is (or will become) a key function for most
init.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agoMerge branch 'master' of git://www.denx.de/git/u-boot-microblaze
Tom Rini [Wed, 10 Sep 2014 10:59:49 +0000 (06:59 -0400)]
Merge branch 'master' of git://denx.de/git/u-boot-microblaze

10 years agomicroblaze: remove #ident directive to fix build error
Masahiro Yamada [Mon, 8 Sep 2014 09:05:26 +0000 (18:05 +0900)]
microblaze: remove #ident directive to fix build error

The microblaze-generic board fails to build at least
with the kernel.org crosstool:
https://www.kernel.org/pub/tools/crosstool/files/bin/x86_64/4.9.0/
x86_64-gcc-4.9.0-nolibc_microblaze-linux.tar.xz

$ make CROSS_COMPILE=microblaze-linux- microblaze-generic_defconfig all
  [ snip ]
  CC      disk/part.o
  CC      disk/part_dos.o
  LD      disk/built-in.o
  CC      drivers/block/systemace.o
{standard input}: Assembler messages:
{standard input}:2495: Error: operation combines symbols in different segments
{standard input}:2496: Error: operation combines symbols in different segments
{standard input}:2499: Error: operation combines symbols in different segments
{standard input}:2500: Error: operation combines symbols in different segments
{standard input}:2505: Error: operation combines symbols in different segments
{standard input}:2506: Error: operation combines symbols in different segments
{standard input}:2515: Error: operation combines symbols in different segments
{standard input}:2516: Error: operation combines symbols in different segments
{standard input}:2519: Error: operation combines symbols in different segments
{standard input}:2520: Error: operation combines symbols in different segments
{standard input}:2529: Error: operation combines symbols in different segments
{standard input}:2530: Error: operation combines symbols in different segments
{standard input}:2533: Error: operation combines symbols in different segments
{standard input}:2534: Error: operation combines symbols in different segments
{standard input}:2539: Error: operation combines symbols in different segments
{standard input}:2540: Error: operation combines symbols in different segments
{standard input}:2549: Error: operation combines symbols in different segments
{standard input}:2550: Error: operation combines symbols in different segments
make[3]: *** [drivers/block/systemace.o] Error 1
make[2]: *** [drivers/block] Error 2
make[1]: *** [drivers] Error 2
make: *** [__build_one_by_one] Error 2

It looks like the cause of this error message is the "#ident" directive.

Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Michal Simek <michal.simek@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
10 years agoMerge branch 'buildman' of git://git.denx.de/u-boot-x86
Tom Rini [Wed, 10 Sep 2014 00:02:43 +0000 (20:02 -0400)]
Merge branch 'buildman' of git://git.denx.de/u-boot-x86

10 years agoMerge branch 'master' of git://git.denx.de/u-boot-fsl-qoriq
Tom Rini [Wed, 10 Sep 2014 00:01:59 +0000 (20:01 -0400)]
Merge branch 'master' of git://git.denx.de/u-boot-fsl-qoriq

10 years agobuildman: fix typos of --dry-run help message
Masahiro Yamada [Thu, 4 Sep 2014 17:23:27 +0000 (02:23 +0900)]
buildman: fix typos of --dry-run help message

try run    => dry run
no nothing => do nothing
"..."      => '...'

The last one is for consistency with the other option helps.

Change-Id: I1d69047d1fae6ef095a18f69f44ee13c448db9b7
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Create parent directories as necessary
Thierry Reding [Tue, 19 Aug 2014 08:22:39 +0000 (10:22 +0200)]
buildman: Create parent directories as necessary

When creating build directories also create parents as necessary. This
fixes a failure when building a hierarchical branch (i.e. foo/bar).

Signed-off-by: Thierry Reding <treding@nvidia.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Tom Rini <trini@ti.com>
10 years agopatman: make run results better visible
Vadim Bendebury [Thu, 4 Sep 2014 17:45:13 +0000 (10:45 -0700)]
patman: make run results better visible

For an occasional user of patman some failures are not obvious: for
instance when checkpatch reports warnings, the dry run still reports
that the email would be sent. If it is not dry run, the warnings are
shown on the screen, but it is not clear that the email was not sent.

Add some code to report failure to send email explicitly.

Tested by running the script on a patch with style violations,
observed error messages in the script output.

Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-by: Doug Anderson <dianders@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Ignore conflicting tags
Simon Glass [Sat, 6 Sep 2014 01:00:23 +0000 (19:00 -0600)]
buildman: Ignore conflicting tags

Tags like Series-version are normally expected to appear once, and with a
unique value. But buildman doesn't actually look at these tags. So ignore
conflicts.

This allows bulidman to build a branch containing multiple patman series.

Reported-by: Steve Rae <srae@broadcom.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Permit branch names with an embedded '/'
Simon Glass [Sat, 6 Sep 2014 01:00:22 +0000 (19:00 -0600)]
buildman: Permit branch names with an embedded '/'

At present buildman naively uses the branch name as part of its directory
path, which causes problems if the name has an embedded '/'.

Replace these with '_' to fix the problem.

Reported-by: Steve Rae <srae@broadcom.com>
Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Expand output test to cover directory prefixes
Simon Glass [Sat, 6 Sep 2014 01:00:21 +0000 (19:00 -0600)]
buildman: Expand output test to cover directory prefixes

Now that buildman supports removing the build directory prefix from output,
add a test for it. Also ensure that output directories are removed when the
test completes.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Add additional functional tests
Simon Glass [Sat, 6 Sep 2014 01:00:20 +0000 (19:00 -0600)]
buildman: Add additional functional tests

This adds coverage of core features of the builder, including the
command-line options which affect building.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agopatman: Start with a clean series when needed
Simon Glass [Sat, 6 Sep 2014 01:00:19 +0000 (19:00 -0600)]
patman: Start with a clean series when needed

For reasons that are not well-understood, GetMetaDataForList() can end up
adding to an existing series even when it appears that it should be
starting a new one.

Change from using a default constructor parameter to an explicit one, to
work around this problem.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Provide an internal option to clean the outpur dir
Simon Glass [Sat, 6 Sep 2014 01:00:18 +0000 (19:00 -0600)]
buildman: Provide an internal option to clean the outpur dir

For testing it is useful to clean the output directory before running a
test. This avoids a test interfering with the results of a subsequent
test by leaving data around.

Add this feature as an optional parameter to the control logic.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Correct counting of build failures on retry
Simon Glass [Sat, 6 Sep 2014 01:00:17 +0000 (19:00 -0600)]
buildman: Correct counting of build failures on retry

When a build is to be performed, buildman checks to see if it has already
been done. In most cases it will not bother trying again. However, it was
not reading the return code from the 'done' file, so if the result was a
failure, it would not be counted. This depresses the 'failure' count stats
that buildman prints in this case.

Fix this bug by always reading the return code.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Allow tests to have their own boards
Simon Glass [Sat, 6 Sep 2014 01:00:16 +0000 (19:00 -0600)]
buildman: Allow tests to have their own boards

Rather than reading boards.cfg, which may take time to generate and is not
necessarily suitable for running tests, create our own list of boards.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Avoid looking at config file or toolchains in tests
Simon Glass [Sat, 6 Sep 2014 01:00:15 +0000 (19:00 -0600)]
buildman: Avoid looking at config file or toolchains in tests

These files may not exist in the environment, or may not be suitable for
testing. Provide our own config file and our own toolchains when running
tests.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Set up bsettings outside the control module
Simon Glass [Sat, 6 Sep 2014 01:00:14 +0000 (19:00 -0600)]
buildman: Set up bsettings outside the control module

Move the bsettings code back to the main buildman.py file, so we can do
something different when testing.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Add a functional test
Simon Glass [Sat, 6 Sep 2014 01:00:13 +0000 (19:00 -0600)]
buildman: Add a functional test

Buildman currently lacks testing in many areas, including its use of git,
make and many command-line flags.

Add a functional test which covers some of these areas. So far it does
a fake 'build' of all boards for the current source tree.

This version reads the real ~/.buildman and boards.cfg files. Future work
will improve this.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agopatman: Provide a way to intercept commands for testing
Simon Glass [Sat, 6 Sep 2014 01:00:12 +0000 (19:00 -0600)]
patman: Provide a way to intercept commands for testing

Add a test point for the command module. This allows tests to emulate
the execution of commands. This provides more control (since we can make
the fake 'commands' do whatever we like), makes it faster to write tests
since we don't need to set up as much environment, and speeds up test
execution.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Move full help code into the control module
Simon Glass [Sat, 6 Sep 2014 01:00:11 +0000 (19:00 -0600)]
buildman: Move full help code into the control module

There is no good reason to keep this code separate. Move it into control.py
so it is easier to test.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Move the command line code into its own file
Simon Glass [Sat, 6 Sep 2014 01:00:10 +0000 (19:00 -0600)]
buildman: Move the command line code into its own file

We want to be able to issue parser commands from within buildman for test
purposes. Move the parser code into its own file so we don't end up needing
the buildman and test modules to reference each other.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agopatman: RunPipe() should not pipe stdout/stderr unless asked
Simon Glass [Sat, 6 Sep 2014 01:00:09 +0000 (19:00 -0600)]
patman: RunPipe() should not pipe stdout/stderr unless asked

RunPipe() currently pipes the output of stdout and stderr to a pty, but
this is not the intended behaviour. Fix it.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Enhance basic test to check summary output
Simon Glass [Sat, 6 Sep 2014 01:00:08 +0000 (19:00 -0600)]
buildman: Enhance basic test to check summary output

Adjust the basic test so that it checks all console output. This will help
to ensure that the builder is behaving correctly with printing summary
information.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agobuildman: Send builder output through a function for testing
Simon Glass [Sat, 6 Sep 2014 01:00:07 +0000 (19:00 -0600)]
buildman: Send builder output through a function for testing

To allow us to verify the builder's console output, send it through a
function which can collect it when running in test mode.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agopatman: Add a way of recording terminal output for testing
Simon Glass [Sat, 6 Sep 2014 01:00:06 +0000 (19:00 -0600)]
patman: Add a way of recording terminal output for testing

When running unit tests we don't want output to go to the terminal.
Provide a way of collecting it so that it can be examined by test code
later.

Signed-off-by: Simon Glass <sjg@chromium.org>
10 years agoclang: workaround for generated constants
Jeroen Hofstee [Wed, 30 Jul 2014 19:54:54 +0000 (21:54 +0200)]
clang: workaround for generated constants

KBuild abuses the asm statement to write to a file and
clang chokes about these invalid asm statements. Hack it
even more by fooling this is actual valid asm code.

cc: Masahiro Yamada <yamada.m@jp.panasonic.com>
cc: Tom Rini <trini@ti.com>
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agoeabi_compat: add __aeabi_memcpy __aeabi_memset
Jeroen Hofstee [Wed, 30 Jul 2014 19:54:53 +0000 (21:54 +0200)]
eabi_compat: add __aeabi_memcpy __aeabi_memset

cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agoARM: make gd a function for clang
Jeroen Hofstee [Wed, 30 Jul 2014 19:54:52 +0000 (21:54 +0200)]
ARM: make gd a function for clang

"clang does not support global register variables; this is
unlikely to be implemented soon because it requires additional
LLVM backend support" [1]

Workaround it by obtaining the value of gd/r9 by an inline
asm routine. Note there is no set routine added for ARM at the
moment, since most if not all updates of gd from c are actually
not needed for ARM.

[1] http://clang.llvm.org/docs/UsersManual.html

cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agocc-option: also detect unsupported warnings options
Jeroen Hofstee [Wed, 30 Jul 2014 19:54:51 +0000 (21:54 +0200)]
cc-option: also detect unsupported warnings options

By default clang will echo a warning if a warning option is
unknown. Turning warnings into errors when polling for options
also catches such cases and prevents passing arguments to the
compiler which cause warnings.

cc: Masahiro Yamada <yamada.m@jp.panasonic.com>
cc: Tom Rini <trini@ti.com>
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agoARM: SPL: do not set gd again
Jeroen Hofstee [Wed, 30 Jul 2014 19:54:50 +0000 (21:54 +0200)]
ARM: SPL: do not set gd again

Just before calling board_init_f, crt0.S has already
reserved space for the initial gd on the stack. There
should be no need to allocate it again.

cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agoboard_r: ARM[64] do not set gd again
Jeroen Hofstee [Wed, 30 Jul 2014 19:54:49 +0000 (21:54 +0200)]
board_r: ARM[64] do not set gd again

For ARM / ARM64 the relocation routines already updated
gd to the new value. Don't set it again. This allows
compilation with clang as it cannot update gd directly.

cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
10 years agomicroblaze: Fix printf size_t format related warnings (again...)
Vasili Galka [Tue, 26 Aug 2014 10:45:34 +0000 (13:45 +0300)]
microblaze: Fix printf size_t format related warnings (again...)

The basic idea: Define size_t using the __SIZE_TYPE__ compiler-defined
type.

For detailed explanation see similar patch for the nios2 arch:
"nios2: Fix printf size_t format related warnings (again...)"
(sha1: 00a2517fcb5159ed016b25130184638b1dbf2f02)

Signed-off-by: Vasili Galka <vvv444@gmail.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
10 years agoMerge branch 'u-boot-sunxi/master' into 'u-boot-arm/master'
Albert ARIBAUD [Tue, 9 Sep 2014 07:19:10 +0000 (09:19 +0200)]
Merge branch 'u-boot-sunxi/master' into 'u-boot-arm/master'

10 years agoMerge branch 'u-boot-samsung/master' into 'u-boot-arm/master'
Albert ARIBAUD [Mon, 8 Sep 2014 22:21:24 +0000 (00:21 +0200)]
Merge branch 'u-boot-samsung/master' into 'u-boot-arm/master'

10 years agols102xa: dcu: Add platform support for DCU on LS1021ATWR board
Wang Huan [Fri, 5 Sep 2014 05:52:50 +0000 (13:52 +0800)]
ls102xa: dcu: Add platform support for DCU on LS1021ATWR board

This patch adds the TWR_LCD_RGB card/HDMI options and the common
configuration for DCU on LS1021ATWR board.

Signed-off-by: Alison Wang <alison.wang@freescale.com>
10 years agovideo: dcu: Add Sii9022A HDMI Transmitter support
Wang Huan [Fri, 5 Sep 2014 05:52:49 +0000 (13:52 +0800)]
video: dcu: Add Sii9022A HDMI Transmitter support

On LS1021ATWR, Silicon's Sii9022A HDMI Transmitter
is used. This patch adds the common setting for this
chip.

Signed-off-by: Alison Wang <alison.wang@freescale.com>
10 years agovideo: dcu: Add DCU driver support
Wang Huan [Fri, 5 Sep 2014 05:52:48 +0000 (13:52 +0800)]
video: dcu: Add DCU driver support

This patch is to add DCU driver support. DCU also named
2D-ACE(Two Dimensional Animation and Compositing Engine)
is a system master that fetches graphics stored in internal
or external memory and displays them on a TFT LCD panel.

Signed-off-by: Alison Wang <alison.wang@freescale.com>
10 years agoserial: lpuart: add 32-bit registers lpuart support
Jingchang Lu [Fri, 5 Sep 2014 05:52:47 +0000 (13:52 +0800)]
serial: lpuart: add 32-bit registers lpuart support

On vybrid, lpuart's registers are 8-bit. On LS102xA, lpuart's registers
are 32-bit. This patch adds the support for 32-bit registers on
LS102xA.

Signed-off-by: Jingchang Lu <jingchang.lu@freescale.com>
Signed-off-by: Yuan Yao <yao.yuan@freescale.com>
10 years agonet: tsec: Remove tx snooping support from LS1
Claudiu Manoil [Fri, 5 Sep 2014 05:52:46 +0000 (13:52 +0800)]
net: tsec: Remove tx snooping support from LS1

Remove the DMCTRL Tx snooping bits (TDSEN and TBDSEN) as a
workaround for LS1.  It has been observed that currently
the Tx stops functioning after a fair amount of Tx traffic
with these settings on.  These bits are sticky and once set
they cannot be reset from Linux, for instance.

Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
10 years agoarm: ls102xa: Add basic support for LS1021ATWR board
Wang Huan [Fri, 5 Sep 2014 05:52:45 +0000 (13:52 +0800)]
arm: ls102xa: Add basic support for LS1021ATWR board

LS102xA is an ARMv7 implementation. This patch is to add
basic support for LS1021ATWR board.
 One DDR controller
 DUART1 is used as the console

For the detail board information, please refer to README.

Signed-off-by: Chen Lu <chen.lu@freescale.com>
Signed-off-by: Yuan Yao <yao.yuan@freescale.com>
Signed-off-by: Alison Wang <alison.wang@freescale.com>
10 years agoarm: ls102xa: Add basic support for LS1021AQDS board
Wang Huan [Fri, 5 Sep 2014 05:52:44 +0000 (13:52 +0800)]
arm: ls102xa: Add basic support for LS1021AQDS board

LS102xA is an ARMv7 implementation. This patch is to add
basic support for LS1021AQDS board.
 One DDR controller
 DUART1 is used as the console

For the detail board information, please refer to README.

Signed-off-by: Alison Wang <alison.wang@freescale.com>
Signed-off-by: Jason Jin <jason.jin@freescale.com>
Signed-off-by: York Sun <yorksun@freescale.com>
Signed-off-by: Yuan Yao <yao.yuan@freescale.com>
Signed-off-by: Prabhakar Kushwaha <prabhakar@freescale.com>
10 years agodriver/ddr/fsl: Add support of overriding chip select write leveling
York Sun [Fri, 5 Sep 2014 05:52:43 +0000 (13:52 +0800)]
driver/ddr/fsl: Add support of overriding chip select write leveling

JEDEC spec allows DRAM vendors to use prime DQ for write leveling. This
is not an issue unless some DQ pins are not connected. If a platform uses
regular DIMMs but with reduced DDR ECC pins, the prime DQ may end up on
those floating pins for the second rank. The workaround is to use a known
good chip select for this purpose.

Signed-off-by: York Sun <yorksun@freescale.com>