Stefan Roese [Fri, 9 Jan 2015 13:54:55 +0000 (14:54 +0100)]
usb: gadget: composite: Fix NULL pointer crash in USB compliance test
On the DXR2 board (AM335x using MUSB) the USB compliance test suite
(USB 2.0 Command Verifier) will cause the board to crash and reset
upon the "BOS Descriptor Test - Addressed state". Here the output
from the DRX2 while running this test:
Simon Glass [Tue, 23 Dec 2014 19:04:52 +0000 (12:04 -0700)]
sunxi: Move SPL s_init() code to board_init_f()
The current sunxi implementation uses gdata, which is going away. It also
sets up DRAM before board_init_f() in SPL.
There is really no reason to do much in s_init() since board_init_f() is
called immediately afterwards. The only change is that we need our own
implementation of board_init_f() which sets up DRAM before the BSS (which
is in DRAM) is cleared.
The s_init() code runs once for SPL and again for U-Boot proper. We
shouldn't need to init the clock/timer/gpio/i2c init twice, so just have it
in SPL.
Simon Glass [Tue, 23 Dec 2014 19:04:51 +0000 (12:04 -0700)]
arm: Add warnings about using gdata
We need to get rid of this SPL-specific setting of the global_data pointer.
It is already set up in start.S immediately before board_init_f() is called,
and there may be information there that is needed (e.g. pre-reloc malloc
info).
Tom Rini [Fri, 19 Dec 2014 21:53:24 +0000 (16:53 -0500)]
TI ARMv7: Don't use GD before crt0.S has set it
Prior to this change we set the gd pointer early so that we can store
data in it. This becomes problematic for DM changes as well as being
odd in general. Re-work the code paths so that we don't need to set the
gd pointer so early and instead can rely upon the normal setting of it.
In order to do this we do need to move certain calls from s_init into
spl_board_init(), mainly preloader_console_init and
save_omap_boot_params.
Tested on: Beaglebone Black, AM43xx GP EVM, Beagleboard, Beagleboard xM,
OMAP5 uEVM, DRA7xx EVM Signed-off-by: Tom Rini <[email protected]> Tested-by: Simon Glass <[email protected]> Reviewed-by: Simon Glass <[email protected]>
net/vsc9953: Add driver for Vitesse VSC9953 L2 Switch IP
This patch adds a driver for VSC9953 L2 Switch. This Vitesse IP
is integrated in Freescale T1040 and T1020 SoCs.
The L2 switch has 10 Ethernet ports: 2 internal fixed-links
(ports 8 and 9) at 2.5 Gbps and and 8 external ports at 1 Gbps.
The external ports may be connected to PHYs over QSGMII and SGMII.
Commands have also been added to enable/disable a port and to
check a port's link speed, duplexity and status. The commands are:
ethsw port <port_nr> enable|disable - enable/disable an l2 switch port
ethsw port <port_nr> show - show an l2 switch port's configuration
port_nr=0..9; use "all" for all ports
For more detailse please see doc/README.t1040-l2switch
net/fm: Enable FMAN ports if l2switch ports are connected over SGMII
If SerDes is configured to connect L2 Switch ports from T1040
over SGMII or QSGMII, the two FMAN fixed ports (FM1@DTSEC1 and FM2@DTSEC2)
that are connected to two L2 swtch ports must be enabled. These
ports don't have PHYs and must be treated accordingly.
arch/powerpc: Fix mapping of Freescale SerDes protocols
The number of supported serdes protocols on Freescale SoCs
has increased over time. Until now, an u64 variable have been
initialized on boot with the configured protocols. However,
since this number has increased (enum srds_prtcl has more
than 64 values), 64 bits are no longer sufficient to hold track
of all the configured protocols.
This patch replaces the u64 map values with static arrays.
To keep track of the number of serdes protocols, the
SERDES_PRCTL_COUNT vale has been added at the end of
enum srds_prtcl. This value must always be the last one.
U-boot assumes that all FMAN ports have a PHY. Some SoCs (like T1040)
have fixed links. This means that the ports are connected MAC to MAc
and there is no Ethernet PHY attatched. This patch initializes a
FMAN MAC even if it doesn't have a PHY attached.
tang yuantian [Thu, 18 Dec 2014 02:26:34 +0000 (10:26 +0800)]
mpc85xx: clean up the old deep sleep framework
All the boards that support deep sleep feature are converted
to deep sleep generic board interface. The old interface which
support non-generic board is not used anymore. So clean it up.
Ruchika Gupta [Wed, 10 Dec 2014 06:22:50 +0000 (11:52 +0530)]
powerpc: mpc85xx: Add dummy gpio.h to enable CONFIG_OF_CONTROL
If CONFIG_OF_CONTROL is enabled, lib/fdtdec.c is compiled.
It includes <asm/gpio.h> and then <asm/gpio.h> includes
<asm/arch/gpio.h>. As a result, all the SoCs that enable
CONFIG_OF_CONTROL must have <asm/arch/gpio.h>.
The right fix would be to split the lib/fdtdec.c to remove
dependency on GPIO.
This commit adds a dummy <asm/arch/gpio.h> to support OF_CONTROL
for mpc85xx platform. A file mpc85xx_gpio.h exists in
arch/powerpc/include/asm. The defintions in that file conflict
with the ones in asm-generic/gpio.h. Hence a dummy header file
has been added. This will be removed after FDT-GPIO stuff is
fixed correctly.
Alexey Brodkin [Thu, 25 Dec 2014 15:54:58 +0000 (18:54 +0300)]
arc: rename "arc700" in "arcv1"
As a preparation to ARCv2 port submission we rename "arc700" folder to
"arcv1" which stands for ARCv1 ISA also known as ARCompact.
This will allow us to add more flavours of binary-compatible ARCv1 CPUs
like ARC600 if needed later on and all required ARCv2 CPUs (which are
binary incompatible with ARCv1) in "arcv2" folder in subsequent commits.
Alexey Brodkin [Thu, 25 Dec 2014 15:47:45 +0000 (18:47 +0300)]
board/synopsys: remove selection of CPU from the board
Both ARCangel4 and AXS10x are FPGA-based boards so they may have
different CPUs. For now we have only 1 option (ARC700) and we define
this as default in arch Kconfig.
Alexey Brodkin [Sat, 27 Dec 2014 23:42:12 +0000 (02:42 +0300)]
arc: move common sources in library
"reset.c" and "cpu.c" have no architecture-specific code at all.
Others are applicable to either ARC CPU.
This change is a preparation to submission of ARCv2 architecture port.
Even though ARCv1 and ARCv2 ISAs are not binary compatible most of
built-in modules still have the same programming model - AUX registers
are mapped in the same addresses and hold the same data (new featues
extend existing ones).
So only low-level assembly code (start-up, interrupt handlers) is left
as CPU(actually ISA)-specific. This significantyl simplifies maintenance
of multiple CPUs/ISAs.
Alexey Brodkin [Fri, 26 Dec 2014 16:36:30 +0000 (19:36 +0300)]
arc: relocate - minor refactoring and clean-up
* use better symbols for relocatable region boundaries
("__image_copy_start" instead of "CONFIG_SYS_TEXT_BASE")
* remove useless debug messages because they will only show up in case
of both problem (when normal "if" branch won't be taken) and DEBUG take
place which is pretty rare situation.
Igor Guryanov [Wed, 24 Dec 2014 14:17:11 +0000 (17:17 +0300)]
arc: introduce separate section for interrupt vector table
Even though existing implementation works fine in preparation to
submission of ARCv2 architecture we need this change.
In case of ARCv2 interrupt vector table consists of just addresses
of corresponding handlers. And if those addresses will be in .text
section then assembler will encode them as everything in .text section
as middle-endian and then on real execution CPU will read swapped
addresses and will jump into the wild.
Once introduced new section is situated so .text section remains the
first which allows us to use common linker option for linking everything
to a specified CONFIG_SYS_TEXT_BASE.
Alexey Brodkin [Wed, 24 Dec 2014 14:00:29 +0000 (17:00 +0300)]
arc: add dependences on MMU presence
Depending on MMU presence in CPU there're differences in HW behavior.
For example address of instruction that caused exception is put in
ECR register if MMU exists and in ERET register otherwise.
Igor Guryanov [Wed, 24 Dec 2014 13:31:26 +0000 (16:31 +0300)]
arc: interrupts - fix mask setup
To disable interrupts we need to reset corresponding flags in STATUS32
register. For this we need to OR flags for interrupts level1 and level2
and then AND with current value in STATUS32.
Igor Guryanov [Wed, 24 Dec 2014 13:07:07 +0000 (16:07 +0300)]
arc: check caches existence before use
Some cache operations ({i|d}cache_{enable|disable|status} or
flush_dcache_all) are built and used even if CONFIG_SYS_{I|D}CACHE_OFF
is set.
This is required for force disable of caches on early boot.
What if something was executed before U-boot and enabled caches
(low-level bootloaders, previously run kernel etc.)?
But if CPU doesn't really have caches any attempt to access
cache-related AUX registers triggers instruction error exception.
So for convenience we'll try to avoid exceptions by checking if CPU
actually has caches (we check separately data and instruction cache
existence) at all.
Simon Glass [Tue, 2 Dec 2014 00:34:07 +0000 (17:34 -0700)]
buildman: Add an option to write the full build output
Normally buildman runs with 'make -s' meaning that only errors and warnings
appear in the log file. Add a -V option to run make in verbose mode, and
with V=1, causing a full build log to be created.
Simon Glass [Tue, 2 Dec 2014 00:34:06 +0000 (17:34 -0700)]
buildman: Add the option to download toolchains from kernel.org
The site at https://www.kernel.org/pub/tools/crosstool/ is a convenient
repository of toolchains which can be used for U-Boot. Add a feature to
download and install a toolchain for a selected architecture automatically.
It isn't clear how long this site will stay in the current place and
format, but we should be able to rely on bug reports if it changes.
Simon Glass [Tue, 2 Dec 2014 00:34:00 +0000 (17:34 -0700)]
buildman: Add an option to use the full tool chain path
In some cases there may be multiple toolchains with the same name in the
path. Provide an option to use the full path in the CROSS_COMPILE
environment variable.
Note: Wolfgang mentioned that this is dangerous since in some cases there
may be other tools on the path that are needed. So this is set up as an
option, not the default. I will need test confirmation (i.e. that this
commit fixes a real problem) before merging it.
Simon Glass [Tue, 2 Dec 2014 00:33:56 +0000 (17:33 -0700)]
buildman: Don't remove entire output directory when testing
When running tests the output directory is often wiped. This is only safe if
a branch is being built. The output directory may contain other things
besides the buildman test output.
Simon Glass [Tue, 2 Dec 2014 00:33:55 +0000 (17:33 -0700)]
buildman: Add an option to flatten output directory trees
When building current source for a single board, buildman puts the output
in <output_dir>/current/current/<board>. Add an option to make it use
<output_dir>/<board> instead. This removes the unnecessary directories
in that case, controlled by the --no-subdirs/-N option.
Simon Glass [Tue, 2 Dec 2014 00:33:54 +0000 (17:33 -0700)]
buildman: Try to guess the upstream commit
Buildman normally obtains the upstream commit by asking git. Provided that
the branch was created with 'git checkout -b <branch> <some_upstream>' then
this normally works.
When there is no upstream, we can try to guess one, by looking up through
the commits until we find a branch. Add a function to try this and print
a warning if buildman ends up relying on it.
Simon Glass [Tue, 2 Dec 2014 00:33:53 +0000 (17:33 -0700)]
buildman: Don't prune output space for 'current source' build
This is not needed since we always do a full (non-incremental) build. Also
it might be dangerous since it will try to delete everything below the
base directory.
Alexey Brodkin [Tue, 13 Jan 2015 15:49:01 +0000 (18:49 +0300)]
Kconfig: move CONFIG_SYS_CLK_FREQ to Kconfig
It makes sense to specify CONFIG_SYS_CLK_FREQ in "configs/xx_defconfig"
instead of "include/configs/xxx.h" because then header will be reusable
across boards with different CPU clocks.
Also this nice to have an ability for end user to tune this value
himself via "menuconfig".
For now I'm only applying this change to all ARC configs because
otherwise scope of change will be huge.
Masahiro Yamada [Thu, 8 Jan 2015 10:23:35 +0000 (19:23 +0900)]
spl: spl_nor: surround Linux-load code with #ifdef CONFIG_SPL_OS_BOOT
If CONFIG_SPL_NOR_SUPPORT is defined, spl_nor_load_image() requires
spl_start_uboot(), CONFIG_SYS_OS_BASE, CONFIG_SYS_SPL_ARGS_ADDR,
CONFIG_SYS_FDT_BASE to be defined even if users just want to run
U-Boot, not Linux. This is inconvenient.
This patch is following the codying style of common/spl/spl_nand.c.
Simon Glass [Tue, 2 Dec 2014 20:17:37 +0000 (13:17 -0700)]
bootm: Factor out common parts of image decompression code
Adjust the code so that the error reporting can all be done at the end,
and is the same for each decompression method. Try to detect when
decompression fails due to lack of space. Keep the behaviour of
resetting on failure even though there should be no memory corruption
now.
Simon Glass [Tue, 2 Dec 2014 20:17:31 +0000 (13:17 -0700)]
sandbox: Correct ordering of 'sb save' commands
Prior to commit d455d87 there was an inconsistency between the position of
the 'address' parameter in 'sb load' and 'sb save'. This was corrected but
it broke some tests. Fix the tests and also the help for 'sb save'.
Simon Glass [Tue, 2 Dec 2014 20:17:30 +0000 (13:17 -0700)]
bootm: Move compression progress/error messages into a function
This code is repeated in several places, and does not detect a common
fault where the image is too large. Move it into its own function and
provide a more helpful messages in this case, for compression schemes
which support this.
Simon Glass [Tue, 2 Dec 2014 20:17:29 +0000 (13:17 -0700)]
lzma: fix buffer bound check error further
Commit 4d3b8a0d fixed a problem with lzma decompress where it would
run out of bytes to decompress. The algorithm needs to know how many
uncompressed bytes it is expected to produce.
However, the fix introduced a potential buffer overrun, and causes
the compression test to fail (test_compression command in sandbox).
The correct fix seems to be to use the minimum of the expected number
of uncompressed bytes and the amount of output space available. That
way things work normally when there is enough space, and return an
error (without overrunning available space) when there is not.
Masahiro Yamada [Wed, 14 Jan 2015 03:35:24 +0000 (12:35 +0900)]
README.scrapyard: fill commit and date fields
This commit was generated by the following command:
scripts/fill_scrapyard.py
The commit-ID of CPCIISER4 removal has been fixed by hand because
the board was removed by commit 370572601027 (ppc4xx: remove CPCIISER4
board), but it was added to README.scrapyard by commit 9a4018e09a2f
(ppc4xx: remove DP405 board).
Hans de Goede [Sun, 11 Jan 2015 18:33:35 +0000 (19:33 +0100)]
sunxi: usbc: Add support for usb-vbus0 controller by axp drivebus pin
The axp221 / axp223's N_VBUSEN pin can be configured as an output rather
then an input, and this is used on some boards to control usb-vbus0, add
support for this.
Hans de Goede [Sun, 11 Jan 2015 18:58:03 +0000 (19:58 +0100)]
sunxi: axp221: Add support for controlling the drivebus pin
The axp221 / axp223's N_VBUSEN pin can be configured as an output rather
then an input, add axp_drivebus_enable() and _disable() functions to set
the pin in output mode and control it.
Hans de Goede [Sun, 11 Jan 2015 18:43:56 +0000 (19:43 +0100)]
sunxi: axp221: Protect axp221_init against multiple calls
The voltage setting code knows it needs to call axp221_init before calling
the various voltage setting functions.
But users of axp utility functions like axp221_get_sid() do not know this,
so the utility functions always call axp221_init() to ensure that the
p2wi / rsb setup magic has been done.
Since doing this repeatedly is quite expensive, add a check to axp221_init
so that it only does the initialization once.
Chen-Yu Tsai [Mon, 12 Jan 2015 10:02:12 +0000 (18:02 +0800)]
sunxi: Hummingbird_A31_defconfig: Enable VGA output through external DAC
The Hummingbird A31 uses an external DAC connected to the LCD0 outputs
for the on board VGA output. The DAC has a power control that's toggled
by GPIO.
Chen-Yu Tsai [Mon, 12 Jan 2015 10:02:10 +0000 (18:02 +0800)]
sunxi: video: Allow external DACs for VGA on platforms without TV encoders
Using an external DAC for VGA output was available on sun5i. Since
some other SoCs don't have a builtin TV encoder, but might have
use for a VGA output, enable the option for the platforms that
don't have TV encoders.
console: Use pre-console buffer to get complete log on all consoles
Currently the pre-console buffer can accumulate early log messages
and flush them to the serial console as soon as it becomes available.
This patch just adds one more pre-console buffer flushing point and
does all the same for the other consoles too. This is particularly
useful for the vga/hdmi/lcd console, where we can see all the older
messages now (except for the log messages from SPL).
Naturally, we don't want to get an extra copy of the log messages
on the serial console again at the second flushing point, so the
serial console has to be explicitly filtered out.
Jan Kiszka [Wed, 31 Dec 2014 12:46:42 +0000 (13:46 +0100)]
sun7i: Move psci_arch_init close to text_end
"adr rX, text_end" only works if the label is close. Adding further code
to the other functions will prevent this. So move the containing
function close to label. No functional change.
Jan Kiszka [Wed, 31 Dec 2014 12:46:40 +0000 (13:46 +0100)]
sun7i: Add support for taking CPUs offline via PSCI
Based on the original version by Marc Zyngier. It adds a psci_cpu_off
implementation for the A20 SoC. The mechanism works by first preparing
the calling CPU to go offline (disable and flush cache, disable SMP),
then requesting CPU 0 to pull the plug. The request is sent as FIQ on
SGI15.
projects / u-boot.git / log