The post.c code is missing braces around the pass case, and as a
result, the diagnostic function will post both fail and pass for
a failed test. The reason for this bug is probably the incorrect
indentation used, so when reading the code it seems like there
are proper braces.
Indent the code to the correct depth and put proper braces around
the "else" branch of the "if" statement.
Signed-off-by: James Kosin <jkosin@intcomgrp.com>
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Now that the tools target requires the generated version header file, we
need to make sure that the directory it writes to exists. In a configured
tree, this is taken care of for us. But in an unconfigured one, the dir
does not yet exist causing a build error like so:
/bin/sh: line 5: ..../u-boot_build/include/version_autogenerated.h.tmp: No such file or directory
So create the dir for this file before we attempt to generate it.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Since panic() never returns, we should add an appropriate attribute to
let gcc improve optimization around it.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The targets/prerequisites are the same here; the rules only differ in
the recipes. So move the if logic protection to the recipe part so we
can keep the rest the same.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
This adds a simple flash test to automatically verify erasing,
writing, and reading of sectors. The code is based on existing
Blackfin tests but generalized for everyone to use.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Now that we have the generic GPIO layer, we can easily provide a common
implementation for the post_hotkeys_pressed() function based on it.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The current arch/driver specific UART posts basically boil down to setting
the UART to loop back mode, then reading and writing data. If we ignore
the loop back part, the rest can be built upon the existing common serial
API. So let's do just that.
First add a call back for serial drivers to implement loop back control.
Then write a post test that walks all of the serial drivers, puts them
into loop back mode, and verifies that reading/writing at all the diff
baud rates is OK.
If a serial driver doesn't support loop back mode (either it can't or
it hasn't done so yet), then skip it. This should allow for people to
easily migrate to the new post test with existing serial drivers.
I haven't touched the few already existing uart post tests as I don't
the hardware or knowledge of converting them over. So I've marked the
new test as weak which will allow the existing tests to override the
default until they are converted.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
The multi serial support has a "ctlr" field which almost no one uses,
but everyone is forced to set to useless strings. So punt it.
Funny enough, the only code that actually reads this field (the mpc8xx
driver) has a typo where it meant to look for the SCC driver. Fix it
while converting the check to use the name field.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
CC: Heiko Schocher <hs@denx.de>
CC: Anatolij Gustschin <agust@denx.de>
CC: Tom Rix <Tom.Rix@windriver.com>
CC: Minkyu Kang <mk7.kang@samsung.com>
CC: Craig Nauman <cnauman@diagraph.com>
CC: Marek Vasut <marek.vasut@gmail.com>
CC: Prafulla Wadaskar <prafulla@marvell.com>
CC: Mahavir Jain <mjain@marvell.com>
The serial_register function never fails (always return 0), so change it
to a void function to avoid wasting overhead on it.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Rather than sticking arch/board/driver specific logic in the common
serial code, push it all out to the respective drivers. The serial
drivers declare these funcs weak so that boards can still override
things with their own definition.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
CC: Heiko Schocher <hs@denx.de>
CC: Anatolij Gustschin <agust@denx.de>
CC: Tom Rix <Tom.Rix@windriver.com>
CC: Minkyu Kang <mk7.kang@samsung.com>
CC: Craig Nauman <cnauman@diagraph.com>
CC: Prafulla Wadaskar <prafulla@marvell.com>
CC: Mahavir Jain <mjain@marvell.com>
Tested-by: Minkyu Kang <mk7.kang@samsung.com>
Some toolchains enable security warning flags by default, but these don't
really make sense in the u-boot world. Such as forcing changes like:
-printf(foo);
+printf("%s", foo);
So disable the flags when the compiler supports them. Linux has already
merged a similar change in their build system.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
For people who want to manually extract the embedded environment so that
it can be manually packed into the final u-boot image, add a config opt
to force building of the envcrc tool.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Newer SST flashes have dropped the Auto Address Increment (AAI) word
programming (WP) modes in favor of the standard page programming mode
that most flashes now support. So add a flags field to the different
flashes to support both modes with new and old styles.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Every spi flash uses the same write disable command, so unify this in
the common code.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Fixed commit message.
Signed-off-by: Wolfgang Denk <wd@denx.de>
Now that the common spi_flash structure tracks all the info that these
drivers need, kill off their local state indirection and use just what
the common code provides.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Once we add a new page_size field for write lengths, we can unify the
write methods for most of the spi flash drivers.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Convert all the comments at the top of the file into help text for people
to easily get at with standard -h/--help options.
Signed-off-by: Mike Frysinger <vapier@gentoo.org>
Acked-by: Detlev Zundel <dzu@denx.de>
In some circumstances, reset_timer_masked() was called be timer_init() in
order to perform architecture specific timer initialisation. In such
cases, the required code in reset_timer_masked() has been moved into
timer_init()
There is no need to use set_timer(). Replace with appropriate use of
get_timer()
Signed-off-by: Graeme Russ <graeme.russ@gmail.com>
Acked-by: Wolfgang Denk <wd@denx.de>
* Fix: if using md5 command watchdog timed out
* change function call md5(..) to the watchdog-safe variant
md5_wd(..) to support watchdog reset
Signed-off-by: Jens Scharsig <esw@bus-elektronik.de>
* Fix: if using sha1 command watchdog timed out
* change function call sha1_csum(..) to the watchdog-safe variant
sha1_csum_wd(..) to support watchdog reset
Signed-off-by: Jens Scharsig <esw@bus-elektronik.de>
It might be desirable to have the ability to flush icache/dcache
within u-boot, this patch gives each arch the ability to provide
a flush_dcache/flush_icache function to let u-boot flush caches
from the prompt
Signed-off-by: Matthew McClintock <msm@freescale.com>
Print a message if we do not have the ability to uncompress a gzip
image. Before, u-boot would just assume the routines were available
Signed-off-by: Matthew McClintock <msm@freescale.com>
If we don't want to build support for any partition types we can now
add #undef CONFIG_PARTITIONS in a board config file to keep this from
being compiled in. Otherwise boards assume this is compiled in by
default
Signed-off-by: Matthew McClintock <msm@freescale.com>
In miiphy_register() the new device's name was initialised by passing a
string parameter as the format string to sprintf(). As this would cause
problems if it ever contained a '%' symbol, switch to using strncpy()
instead.
Signed-off-by: Laurence Withers <lwithers@guralp.com>
Cc: Andy Fleming <afleming@freescale.com>
When booting with a ramdisk we bump the amount of memory reserved for
the device tree by FDT_RAMDISK_OVERHEAD. However we did not increase
the actual size in the device tree blob to match.
Its possible on boundary cases that we dont have enough memory according
to the device tree blob and get errors like:
WARNING: could not set linux,initrd-end FDT_ERR_NOSPACE
We can easily fix this by setting the device tree size at the same time
we bump the amount of memory reserved for the device tree.
Signed-off-by: Kumar Gala <galak@kernel.crashing.org>
Acked-by: Gerald Van Baren <vanbaren@cideas.com>
* Fix: if using crc32 command watchdog timed out
* change function call crc32(..) to the watchdog-safe variant
crc_32_wd(..) to support watchdog reset
Signed-off-by: Jens Scharsig <esw@bus-elektronik.de>
In some cases (e.g. bootm with a elf payload which is already at the right
position) there is a in place copy of data to the same address. Catching this
saves some ms while booting.
Signed-off-by: Matthias Weisser <weisserm@arcor.de>
This include is needed, if this memory test is used "outside"
from post code, for example booting with nand_spl, and using
this memory test before copying u-boot code to RAM and jumping
to it.
Signed-off-by: Heiko Schocher <hs@denx.de>
James Kosin
Mike Frysinger
Graeme Russ
Daniel Schwierzeck
Aneesh V
Jens Scharsig
Matthew McClintock
Laurence Withers
Kumar Gala
Wolfgang Denk
Matthias Weisser
Rob Herring
Heiko Schocher