Currently the version defines reflect the digits in the version
number. MMC_VERSION_4_41 is defined as 0x441 and MMC_VERSION_4_5
is defined as 0x405. This results in MMC_VERSION_4_5 < MMC_VERSION_4_41
becoming true which was surely not intended. Fix this by
redefining the versions as 0x<major><minor><micro>. This makes the
string generation more complicated but makes versions comparable
again.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
At least NFS does not fulfill read requests > 1024 bytes at once.
Use read_full in read_file as read doesn't guarantee that the
whole request can be read at once.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We check whether the file passed as devicetree actually is a devicetree,
but we missed to return an error if it's not. Fix this.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This renames the Freescale and Phytec board directories and defconfig
files to a common naming scheme. The board directories are named
<vendor>-<board> and the defconfig files are named
<vendor>-<board>_defconfig. Also the DataModul realq7 is renamed to its
Marketing Name eDM-QMX6.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Some boards need gpio functions very early and also sometimes
is useful to toggle gpios during early code debug. This adds a header
file for setting i.MX gpios early.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Which bootmode is selected has no longer to be chosen by Kconfig. The
boards can decide themselves which bootmode they want to support. This
makes it unnecesary to ask the user which bootmode shall be supported,
so the "Select boot mode" becomes invisible and both support will be
compiled in as needed by the boards. NAND_IMX_BOOT goes away and the
already existing ARCH_IMX_EXTERNAL_BOOT_NAND can now be used for the
boards to depend on external nand boot.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Only one GPT will be used, but with devicetree support we can't predict
which one it is, so we need the clock lookup for all GPTs to ensure
that the timer gets its clock.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
If we are running from NFC SRAM and we are passed boarddata
containing a devicetree pointer then it point to an address relative
to the NFC SRAM start. First thing we do is to copy the initial
binary to SDRAM and jump there. The devicetree pointer has to be
adjusted by this offset.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
When compiling with multiimage support ld_var(_barebox_image_size) only
contains the length of the PBL image, but not including the appended
compressed data. With this patch the image size is read from the barebox
header instead which contains the correct size, either from the linker
or from the fix_size tool.
This makes the external_nand_boot compatible with multiimage support.
Tested on Phytec phyCARD-i.MX27 with and without PBL.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The pblx is a self extracting barebox binary. This doesn't have
the size of the image correctly set because the linker doesn't
generate it for relocatable binaries.
This currently only works on ARM, but this is the only architecture
supporting multi images anyway. TO make it work on other architectures
fix_size would have to be extended to recognize other images.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Instead of passing the offset to the fix_size tool check the image to
fixup for a valid header so that only recognized files are fixed up.
This makes the usage of this tool safer.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
on the edmqmx6 barebox is loaded into SRAM, so it must not get
too big. Disable some stuff to make the image fit into SRAM again.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
barebox only recognizes boarddata as dtb if it is inside SDRAM, so
copy the dtb there if necessary.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>