Mario Schuknecht [Tue, 30 Sep 2014 15:04:42 +0000 (17:04 +0200)]
mmc: Fix mmc bus width
After setting the bus width, the extended CSD register is read. Some selected
fields are compared with previously read extended CSD register fields. In this
comparison the EXT_CSD_ERASE_GROUP_DEF field is compared. But this field is
previously written under certain circumstances. And then the comparison fails.
Only compare read-only fields. Therefore compare field EXT_CSD_HC_WP_GRP_SIZE
instead of field EXT_CSD_ERASE_GROUP_DEF.
Signed-off-by: Mario Schuknecht <mario.schuknecht@dresearch-fe.de> Acked-by: Pantelis Antoniou <panto@antoniou-consulting.com>
Peter Bigot [Wed, 3 Sep 2014 17:22:43 +0000 (12:22 -0500)]
env_mmc: correct fini partition to match init partition
The code to set the MMC partition uses an weak function to obtain the
correct partition number. Use that instead of the compile-time default
when deciding whether it needs to switch back.
Fixes: 6e7b7df4df43574 ("env_mmc: support env partition setup in runtime") Signed-off-by: Peter A. Bigot <pab@pabigot.com> Acked-by: Dmitry Lifshitz <lifshitz@compulab.co.il> Acked-by: Pantelis Antoniou <panto@antoniou-consulting.com>
Peter Bigot [Tue, 2 Sep 2014 23:31:23 +0000 (18:31 -0500)]
mmc: restore capacity when switching to partition 0
The capacity and lba for an MMC device with part_num 0 reflects the
whole device. When mmc_switch_part() successfully switches to a
partition, the capacity is changed to that partition. As partition 0
does not physically exist, attempts to switch back to the whole device
will indicate an error, but the capacity setting for the whole device
must still be restored to match the partition.
Signed-off-by: Peter A. Bigot <pab@pabigot.com> Tested-by: Tom Rini <trini@ti.com> Acked-by: Pantelis Antoniou <panto@antoniou-consulting.com>
Simon Glass [Mon, 15 Sep 2014 12:33:20 +0000 (06:33 -0600)]
sandbox: config: Enable all SPI flash chips
Sandbox may as well support everything. This increases the amount of code
that is built/tested by sandbox, and also provides access to all the
supported SPI flash devices.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Jagannadha Sutradharudu Teki <jaganna@xilinx.com>
Simon Glass [Mon, 15 Sep 2014 12:33:19 +0000 (06:33 -0600)]
sandbox: Convert SPI flash emulation to use sf_params
At present sandbox has its own table of supported SPI flash chips. Now that
the SPI flash system is fully consolidated and has its own list, sandbox
should use that.
This enables us to expand the number of chips that sandbox supports.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Jagannadha Sutradharudu Teki <jaganna@xilinx.com>
To add the Denali NAND driver support into U-Boot.
This driver is leveraged from Linux with commit ID fdbad98dff8007f2b8bee6698b5d25ebba0471c9. For Denali
controller 64 variance, you need to declare macro
CONFIG_SYS_NAND_DENALI_64BIT.
Stefan Roese [Fri, 5 Sep 2014 07:57:01 +0000 (09:57 +0200)]
mtd: nand: Fix length bug in ioread16_rep() and iowrite16_rep()
The ioread16_rep() and iowrite16_rep() implementations are U-Boot specific
and have been introduced with the Linux MTD v3.14 sync. While introducing
these functions, the length for the loop has been miscalculated. The ">> 1"
is already present in the caller. So lets remove it in the function.
Tested on omap3_ha.
Signed-off-by: Stefan Roese <sr@denx.de> Cc: Heiko Schocher <hs@denx.de> Cc: Tom Rini <trini@ti.com> Cc: Scott Wood <scottwood@freescale.com> Acked-by: Heiko Schocher <hs@denx.de>
OMAP GPMC driver used with some NAND Flash devices
(e.g. Spansion S34ML08G1) causes that U-boot shows
hundreds of 'nand: bit-flip corrected' error messages.
Possible cause was discussed in the mailinglist thread:
http://lists.denx.de/pipermail/u-boot/2014-April/177508.html
The issue was partially fixed with the cc81a5291910d7a.git
however this has to be done to fix the SPL.
The original author of the code is Belisko Marek
<marek.belisko@gmail.com>
Chris Packham [Wed, 10 Sep 2014 04:03:10 +0000 (16:03 +1200)]
powerpc: add --bss-plt to LDFLAGS
With some versions of gcc (that we know of 4.6.3 and 4.8.2 are affected)
it is necessary to specify --bss-plt to get the final blrl in the
_GOT2_TABLE_. Without this the last symbol does not get it's address
relocated. For the P2041RDB board this ended up being
NetArpWaitTimerStart which caused the ARP packets to timeout
immediately.
Signed-off-by: Joakim Tjernlund <joakim.tjernlund@transmode.se> Signed-off-by: Chris Packham <judge.packham@gmail.com> Acked-by: Joakim Tjernlund <joakim.tjernlund@transmode.se> Acked-by: Scott Wood <scottwood@freescale.com> Reviewed-by: York Sun <yorksun@freescale.com>
York Sun [Thu, 11 Sep 2014 20:32:07 +0000 (13:32 -0700)]
board/ls1021aqds: Add DDR4 support
LS1021AQDS has a variant with DDR4 slot. This patch adds a new defconfig
for this variant to enable DDR4 support. RAW timing parameters are not
added for DDR4. The board timing parameters are only tuned for single-
rank 1600 and 1800MT/s with Micron DIMM 9ASF51272AZ-2G1A1 due to DIMM
availability.
Signed-off-by: York Sun <yorksun@freescale.com> CC: Alison Wang <alison.wang@freescale.com>
York Sun [Thu, 11 Sep 2014 20:32:06 +0000 (13:32 -0700)]
driver/ddr/fsl: Fix DDR4 driver
When accumulated ECC is enabled, the DQ_MAP for ECC[4:7] needs to be set
to 0, i.e. 0->0, 1->1, etc., required by controller logic, even these pins
are not actually connected.
Also fix a bug when reading from DDR register to use proper accessor for
correct endianess.
York Sun [Mon, 8 Sep 2014 19:20:01 +0000 (12:20 -0700)]
ARMv8/ls2085a: Enable secondary cores
Spin table is at the very beginning of boot code. Each core has an individual
release address within the spin table, the ft_cpu_setup fn updates the
"cpu-release-addr" property of each cpu node with the corresponding release
address.
Also fix CPU_RELEASE_ADDR to point to secondary_boot_func.
Signed-off-by: York Sun <yorksun@freescale.com> Signed-off-by: Arnab Basu <arnab.basu@freescale.com>
York Sun [Mon, 8 Sep 2014 19:20:00 +0000 (12:20 -0700)]
armv8/fsl-lsch3: Release secondary cores from boot hold off with Boot Page
Secondary cores need to be released from holdoff by boot release
registers. With GPP bootrom, they can boot from main memory
directly. Individual spin table is used for each core. Spin table
and the boot page is reserved in device tree so OS won't overwrite.
Signed-off-by: York Sun <yorksun@freescale.com> Signed-off-by: Arnab Basu <arnab.basu@freescale.com>
fdt_support: Make of_bus_default_count_cells non static
of_bus_default_count_cells can be used to get the #address-cells
and #size-cells defined by the current node's parent node. This
is required when using of_read_number to read from FDT nodes that
can be 32 or 64 bytes depending on values defined by the parent.
Signed-off-by: Arnab Basu <arnab.basu@freescale.com> CC: Scott Wood <scottwood@freescale.com>
This is being done so that it can be used outside 'fdt_support.c'. Making
life more convenient when reading device node properties that can be 32
or 64 bits long.
Signed-off-by: Arnab Basu <arnab.basu@freescale.com> Cc: Scott Wood <scottwood@freescale.com>
York Sun [Thu, 21 Aug 2014 23:13:22 +0000 (16:13 -0700)]
driver/ddr/fsl: Fix tXP and tCKE
The driver was written using old DDR3 spec which only covers low speeds.
The value would be suboptimal for higher speeds. Fix both timing according
to latest DDR3 spec, remove tCKE as an config option.
York Sun [Fri, 1 Aug 2014 22:51:00 +0000 (15:51 -0700)]
driver/ddr: Restruct driver to allow standalone memory space
U-boot has been initializing DDR for the main memory. The presumption
is the memory stays as a big continuous block, either linear or
interleaved. This change is to support putting some DDR controllers
to separated space without counting into main memory. The standalone
memory controller could use different number of DIMM slots.
driver/mtd: Use generic timer API for FSL IFC, eLBC
Freescale's flash control driver is using architecture specific timer API
i.e. usec2ticks
Replace usec2ticks with get_timer() (generic timer API)
Signed-off-by: Prabhakar Kushwaha <prabhakar@freescale.com> Acked-by: Scott Wood <scottwood@freescale.com> Reviewed-by: York Sun <yorksun@freescale.com>
Joe Perches [Tue, 23 Sep 2014 10:41:02 +0000 (12:41 +0200)]
checkpatch: remove unnecessary + after {8,8}
Pick the following commit from Linux kernel:
commit 66cb4ee0e52ca721f609fd5eec16187189ae5fda
Author: Joe Perches <joe@perches.com>
Date: Wed Sep 10 09:40:47 2014 +1000
checkpatch: remove unnecessary + after {8,8}
There's a useless "+" use that needs to be removed as perl 5.20 emits a
"Useless use of greediness modifier '+'" message each time it's hit.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Tom Rini <trini@ti.com>
kconfig: move CONFIG_DEFAULT_DEVICE_TREE to kconfig
This option specifies the default Device Tree used for the run-time
configuration of U-Boot.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Cc: Simon Glass <sjg@chromium.org> Cc: Stephen Warren <swarren@nvidia.com> Cc: Minkyu Kang <mk7.kang@samsung.com> Cc: Michal Simek <michal.simek@xilinx.com>
This commit moves:
CONFIG_OF_CONTROL
CONFIG_OF_SEPARATE
CONFIG_OF_EMBED
CONFIG_OF_HOSTFILE
Because these options are currently not supported for SPL,
the "Device Tree Control" menu does not appear in the SPL
configuration.
Note:
zynq-common.h should be adjusted so as not to change the
default value of CONFIG_SPL_FAT_LOAD_PAYLOAD_NAME.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Simon Glass <sjg@chromium.org> Cc: Stephen Warren <swarren@nvidia.com> Cc: Minkyu Kang <mk7.kang@samsung.com> Acked-by: Michal Simek <michal.simek@xilinx.com>
The mentioned binutils port got removed while the patch was
pending. As Ian pointed out there is another port providing
the binutils for arm now. Update the instructions accordingly.
Cc: ian@FreeBSD.org Cc: Tom Rini <trini@ti.com> Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
The libc headers on FreeBSD and likely related projects as well contain an
header file, cdefs.h which provides similiar functionality as linux/compiler.h.
It provides compiler independent defines like __weak __packed, to allow
compiling with multiple compilers which might have a different syntax for such
extension.
Since that header file is included in multiple standard headers, like stddef.h
and stdarg.h, multiple definitions of those defines will be present if both are
included. When compiling u-boot the compiler will warn about it hundreds of
times since e.g. common.h will include both files indirectly.
[1] Move driver/core/, driver/input/ and drivers/input/ entries
from the top Makefile to drivers/Makefile
[2] Remove the conditional by CONFIG_DM in drivers/core/Makefile
because the whole drivers/core directory is already selected
by CONFIG_DM in the upper level
[3] Likewise for CONFIG_DM_DEMO in drivers/demo/Makefile
[4] Simplify common/Makefile - both CONFIG_DDR_SPD and
CONFIG_SPD_EEPROM are boolean macros so they can directly
select objects
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Marek Vasut <marex@denx.de>
Since clang has a different definition for uninitialized_var
it will complain that it is redefined in include/compiler.h.
Since these are already defined in linux/compiler.h just remove
this instance.
cmd_nand: Update (nand_info_t*)nand after arg_off(_size) call
The arg_off() and arg_off_size() update the 'current NAND
device' variable (dev). This is then used when assigning the
(nand_info_t*)nand value. Place the assignment after the
arg_off(_size) calls to prevent using incorrect (nand_info_t*)
nand value.
Since CONFIG_CMD_IMPORTENV is defined in config_cmd_defaults.h,
it should be enabled for all the boards except bf506f-ezkit
that undefs it explicitely.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
Since CONFIG_CMD_GO is defined in config_cmd_defaults.h
(and no board undefs it its own header), it can be moved to
Kconfig with the default value "y".
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
Since CONFIG_CMD_EXPORTENV is defined in config_cmd_defaults.h,
it should be enabled for all the boards except bf506f-ezkit
that undefs it explicitely.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
Since CONFIG_CMD_CRC32 is defined in config_cmd_defaults.h,
it is enabled for all the boards except the ones undefining it
explicitly:
kwb
tseries_mmc
tseries_nand
tseries_spi
vct_platinum_onenand_small
vct_platinum_small
vct_platinumavc_onenand_small
vct_platinumavc_small
vct_premium_onenand_small
vct_premium_small
The default value of this config option should be "y" and
"# CONFIG_CMD_CRC32 is not set" should be added for those exceptions.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Simon Glass <sjg@chromium.org>
CONFIG_CMD_BOOTM is defined in config_cmd_defaults.h
which is forcebly included from each board.
So, the default value of "config CMD_BOOTM" should be "y".
For some boards undefining it (bf506f-ezkit, controlcenterd_TRAILBLA,
controlcenterd_TRAILBLAZER_DEVELOP, controlcenterd_TRAILBLAZER),
"# CONFIG_CMD_BOOTM is not set" should be added to their defconfig.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Simon Glass <sjg@chromium.org>
kconfig: move CONFIG_CC_OPTIMIZE_LIBS_FOR_SPEED to Kconfig
If this option is enabled, the objects under lib/ directory
are compiled with speed optimization, not size optimization.
(Currently, only used by some Blackfin boards.)
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Simon Glass <sjg@chromium.org> Tested-by: Simon Glass <sjg@chromium.org>
Since commit ddaf5c8f3030050fcd356a1e49e3ee8f8f52c6d4
(patman: RunPipe() should not pipe stdout/stderr unless asked),
Patman spits lots of "Invalid MAINTAINERS address: '-'"
error messages for patches with global changes.
It takes too long for Patman to process them.
Anyway, "M: -" does not carry any important information.
Rather, it is just like a place holder in case of assigning
a new board maintainer. Let's comment out.
This commit can be reproduced by the following command:
find . -name MAINTAINERS | xargs sed -i -e '/^M:[[:blank:]]*-$/s/^/#/'
The "S: Orphan" in MAINTAINERS means that the maintainer in the
"M:" field is unreachable (i.e. the email address is not working).
(Refer to the definition of "Orphan" adopted in U-Boot
in the log of commit 31f1b654b2f395b69faa5d0d3c1eb0803923bd3b,
"boards.cfg: move boards with invalid emails to Orphan")
For patch files adding global changes, scripts/get_maintainer.pl
adds bunch of such invalid email addresses, which results in
tons of annoying bounce emails.
This commit can be reproduced by the following command:
find . -name MAINTAINERS | xargs sed -i -e '
/^M:[[:blank:]]/ {
N
/S:[[:blank:]]Orphan/s/^/#/
}
'
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Acked-by: Simon Glass <sjg@chromium.org>
tools/genboardscfg.py: pick up also commented maitainers
We are still keeping invalid email addressed in MAINTAINERS
because they carry information.
The problem is that scripts/get_maintainer.pl adds emails in the
"M:" field including invalid ones.
We want to comment out invalid email addresses in MAINTAINERS
to prevent scripts/get_maintainer.pl from picking them up.
On the other hand, we want to collect them for boards.cfg
to know the last known maintainer of each board.
This commit adjusts tools/genboardscfg.py to parse also
the commented "M:" fields, which is useful for the next commit.
Simon Glass [Sun, 14 Sep 2014 18:40:15 +0000 (12:40 -0600)]
sandbox: Disable Ctrl-C
This is not supported properly on sandbox, and interferes with running
tests, since when a test script is piped in, some commands will call
ctrlc() which will drop characters from the test script.
Simon Glass [Sun, 14 Sep 2014 18:40:14 +0000 (12:40 -0600)]
Reactivate the tracing feature
This was lost sometime in the Kbuild conversion. Add it back.
Check that the trace test now passes:
$ ./test/trace/test-trace.sh
Simple trace test / sanity check using sandbox
/tmp/filemHKPGw
Build sandbox
O=sandbox FTRACE=1
GEN /home/sjg/c/src/third_party/u-boot/files/sandbox/Makefile
Configuring for sandbox board...
Check results
Test passed
mpc8xx: move common linker scripts into the CPU directory
Each CPU of PowerPC has its default linker script under the CPU
directory, except mpc8xx.
Every mpc8xx board has its own linker script under the board
directory, resulting in lots of duplication of linker scripts.
I notice eight mpc8xx boards have the same linker script.
We can decrease the number of linker scripts by putting a single
default linker script, arch/powerpc/cpu/mpc8xx/u-boot.lds.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Cc: Wolfgang Denk <wd@denx.de> Acked-by: Stefan Roese <sr@denx.de>
Fix the following build error in case CONFIG_E1000_NO_NVM is enabled:
CC drivers/net/e1000.o
drivers/net/e1000.c: In function ‘e1000_initialize’:
drivers/net/e1000.c:5365:5: error: ‘struct e1000_hw’ has no
member named ‘eeprom_semaphore_present’
make[1]: *** [drivers/net/e1000.o] Error 1
make: *** [drivers/net] Error 2 Acked-by: Marek Vasut <marex@denx.de>
Thomas Petazzoni [Wed, 27 Aug 2014 12:29:00 +0000 (14:29 +0200)]
tools/env: change stripping strategy to allow no-stripping
When building the U-Boot tools for non-ELF platforms (such as Blackfin
FLAT), since commit 79fc0c5f498c3982aa4740c273ab1a9255063d9c
("tools/env: cross-compile fw_printenv without setting HOSTCC"), the
build fails because it tries to strip a FLAT binary, which does not
make sense.
This commit solves this by changing the stripping logic in
tools/env/Makefile to be similar to the one in tools/Makefile. This
logic continues to apply strip to the final binary, but does not abort
the build if it fails, and does the stripping in place on the final
binary. This allows the logic to work fine if stripping doesn't work,
as it leaves the final binary untouched.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Masahiro Yamada <yamada.m@jp.panasonic.com> Cc: Sonic Zhang <sonic.zhang@analog.com> Reviewed-by: Masahiro Yamada <yamada.m@jp.panasonic.com> Reviewed-by: Sonic Zhang <sonic.zhang@analog.com>
board/t1040qds: Add sgmii ports support in 0xA7 protocol
T1042QDS (T1042 is T1040 Personality without L2 switch) supports following
sgmii interfaces with serdes protocol 0xA7
-SGMII-MAC3 on Lane B - slot 7
-SGMII-MAC5 on Lane H - slot 7
-SGMII2.5G-MAC1 on Lane C - slot 6
-SGMII2.5G-MAC2 on Lane D - slot 5
DDR-ODT require cfg_dram_type switch set properly as per DDR type.
T1040RDB, T1042RDB boards have DDR3L type DDR, so cfg_dram_type
should be set to OFF for DDR3L
Update t104xrdb/README for switch setting
Signed-off-by: Priyanka Jain <Priyanka.Jain@freescale.com> Reviewed-by: York Sun <yorksun@freescale.com>
ramneek mehresh [Fri, 22 Aug 2014 05:26:05 +0000 (10:56 +0530)]
powerpc/8xxx: Fix in USB device-tree fixup
Fix following issues in USB device-tree fixup:
- returns when either dr_mode or phy_type not defined.
This was terminating fix-up when only either property
was defined in hwconfig string
- updates dr_mode_type or dr_phy_type with junk value when
their index is -1. Now these are updated only when their
respective index is pointing to relevant types
in modes[] and phys[] array
- dr_mode_type and dr_phy_type were not NULL for
each controller
Signed-off-by: Ramneek Mehresh <ramneek.mehresh@freescale.com> Reviewed-by: York Sun <yorksun@freescale.com>
vijay rai [Tue, 19 Aug 2014 07:16:53 +0000 (12:46 +0530)]
powerpc/t104xrdb: Add T1042RDB board support
T1042RDB is a Freescale reference board that hosts the T1042 SoC
(and variants). The board is similar to T1040RDB, T1042 is a reduced
personality of T1040 SoC without Integrated 8-port Gigabit(L2 Switch).
T1042RDB is configured with serdes protocol 0x86 which can support
following interfaces
- 2 RGMII's on DTSEC4, DTSEC5
- 1 SGMII on DTSEC3
DTSEC1, DTSEC2 are not connected on board.
This Patch
- add T1042RDB support
- updates README file for T1042RDB details and update commands for switching
to alternate banks from vBank0 to vBank4 and vice versa
This patch also does minor clean ups for fdt defines for T1042RDB and
T1042RDB_PI board
Signed-off-by: Vijay Rai <vijay.rai@freescale.com> Signed-off-by: Priyanka Jain <Priyanka.Jain@freescale.com> Reviewed-by: York Sun <yorksun@freescale.com>
vijay rai [Wed, 23 Jul 2014 12:55:47 +0000 (18:25 +0530)]
powerpc/t104xrdb: Add Support of rcw for T1042RDB in u-boot
This patch adds support of rcw for T1042RDB, it makes following changes :
- Adds t1042_rcw.cfg file for serdes protocol 0x86 for T1042RDB
- Renames t1042_pi_rcw.cfg file from t1042_rcw.cfg and also updates
comments for valid serdes protocol which is 0x06
- Also updates CONFIG_SYS_FSL_PBL_RCW for T1042RDB
Signed-off-by: Vijay Rai <vijay.rai@freescale.com> Signed-off-by: Priyanka Jain <Priyanka.Jain@freescale.com> Reviewed-by: York Sun <yorksun@freescale.com>
Here the bit KWSPI_CSN_ACT (0x1) should have been cleared. Instead
0xfffffffe is written into this control register. This is the main
reason to use the clrsetbits() functions now. As they make clearing
bits much less error prone.
Additionally KWSPI_IRQUNMASK is not used in spi_cs_activate() and
spi_cs_deactivate() any more. Its the wrong macro but has the same
value as the correct one (KWSPI_CSN_ACT).
This is in preparation for use of this driver on the Marvell Armada XP
platform as well.
Signed-off-by: Stefan Roese <sr@denx.de> Acked-by: Prafulla Wadaskar <prafulla@marvell.com> Tested-by: Luka Perkov <luka@openwrt.org> Reviewed-by: Jagannadha Sutradharudu Teki <jaganna@xilinx.com>
Nikita Kiryanov [Wed, 20 Aug 2014 12:08:48 +0000 (15:08 +0300)]
spl: replace CONFIG_SPL_SPI_* with CONFIG_SF_DEFAULT_*
Currently, CONFIG_SPL_SPI_* #defines are used for controlling SPI boot in
SPL. These #defines do not allow the user to select SPI mode for the SPI flash
(there's no CONFIG_SPL_SPI_MODE, so the SPI mode is hardcoded in
spi_spl_load.c), and duplicate information already provided by
CONFIG_SF_DEFAULT_* #defines.
Kill CONFIG_SPL_SPI_*, and use CONFIG_SF_DEFAULT_* instead.
Cc: Tom Rini <trini@ti.com> Cc: Marek Vasut <marex@denx.de> Cc: Sudhakar Rajashekhara <sudhakar.raj@ti.com> Cc: Lokesh Vutla <lokeshvutla@ti.com> Cc: Vitaly Andrianov <vitalya@ti.com> Cc: Lars Poeschel <poeschel@lemonage.de> Cc: Bo Shen <voice.shen@atmel.com> Cc: Hannes Petermaier <hannes.petermaier@br-automation.com> Cc: Michal Simek <monstr@monstr.eu> Acked-by: Marek Vasut <marex@denx.de> Signed-off-by: Nikita Kiryanov <nikita@compulab.co.il> Reviewed-by: Jagannadha Sutradharudu Teki <jaganna@xilinx.com>
Nikita Kiryanov [Wed, 20 Aug 2014 12:08:50 +0000 (15:08 +0300)]
spi: mxc: fix sf probe when using mxc_spi
MXC SPI driver has a feature whereas a GPIO line can be used to force CS high
across multiple transactions. This is set up by embedding the GPIO information
in the CS value:
cs = (cs | gpio << 8)
This merge of cs and gpio data into one value breaks the sf probe command:
if the use of gpio is required, invoking "sf probe <cs>" will not work, because
the CS argument doesn't have the GPIO information in it. Instead, the user must
use "sf probe <cs | gpio << 8>". For example, if bank 2 gpio 30 is used to force
cs high on cs 0, bus 0, then instead of typing "sf probe 0" the user now must
type "sf probe 15872".
This is inconsistent with the description of the sf probe command, and forces
the user to be aware of implementaiton details.
Fix this by introducing a new board function: board_spi_cs_gpio(), which will
accept a naked CS value, and provide the driver with the relevant GPIO, if one
is necessary.
Cc: Eric Nelson <eric.nelson@boundarydevices.com> Cc: Eric Benard <eric@eukrea.com> Cc: Fabio Estevam <fabio.estevam@freescale.com> Cc: Tim Harvey <tharvey@gateworks.com> Cc: Stefano Babic <sbabic@denx.de> Cc: Tom Rini <trini@ti.com> Cc: Marek Vasut <marex@denx.de> Reviewed-by: Marek Vasut <marex@denx.de> Signed-off-by: Nikita Kiryanov <nikita@compulab.co.il> Reviewed-by: Jagannadha Sutradharudu Teki <jaganna@xilinx.com>
Robert Baldyga [Thu, 18 Sep 2014 15:13:07 +0000 (17:13 +0200)]
dm: avoid dev->req_seq overflow
Since dev->req_seq value is initialized from "reg" property of fdt node,
there is posibility, that address value contained in fdt is greater than
INT_MAX, and then value in dev->req_seq is negative which led to probe()
fail.
This patch fix this problem by ensuring that req_seq is positive, unless
it's one of errno codes.
Signed-off-by: Robert Baldyga <r.baldyga@samsung.com> Acked-by: Simon Glass <sjg@chromium.org>
Simon Glass [Wed, 17 Sep 2014 15:02:38 +0000 (09:02 -0600)]
dm: core: Allow device_bind() to used without CONFIG_OF_CONTROL
The sequence number support in driver model requires device tree control.
It should be skipped if CONFIG_OF_CONTROL is not defined, and should not
require functions from fdtdec.
Simon Glass [Mon, 15 Sep 2014 12:33:36 +0000 (06:33 -0600)]
sf: Add an empty entry to the parameter list
The list is supposed to be terminated with a NULL name, but is not. If a
board probes a chip which does not appear in the table, U-Boot will crash
(at least on sandbox).
projects / karo-tx-uboot.git / log