O_CREAT and O_RWSIZE_8 are both defined as 0100. Fix this by moving the
O_RWSIZE_* flags to unused bits.
This bug leads to incomplete writes when the destination file is created
and mem_write is involved, for example with the memcpy command:
memcpy -s /some/file -d /dev/ram0 0 0 10
In this case only 8 bytes will be copied and it will be done using 8
byte accesses which may not work properly if the destination is not
sufficiently aligned, i.e.:
memcpy -s /some/file -d /dev/ram0 0 1 8
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Reported-by: Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>
Often enough the biggest free region spans most free RAM, so
it doesn't add much value to test the remaining free regions. This
patch changes the default behaviour to only test the biggest free
region and adds the -t option to test all regions.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The memtest does a bus integrity check and a moving inversions test.
Split the tests into two separate functions so that the can be called
separately.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
remap_range is for remapping regions with different cache attributes.
It is implemented for ARM and PowerPC only, the other architectures only
provide stubs.
Currently the new cache attributes are passed in an architecture specific
way and the attributes have to be retrieved by calls to
mmu_get_pte_cached_flags() and mmu_get_pte_uncached_flags().
Make this simpler by providing architecture independent flags which can
be directly passed to remap_range()
Also provide a MAP_ARCH_DEFAULT flag and a arch_can_remap() function.
The MAP_ARCH_DEFAULT defaults to whatever caching type the architecture
has as default. the arch_can_remap() function returns true if the
architecture can change the cache attributes, false otherwise. This
allows the memtest code to better find out what it has to do.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Normally code providing a feature should be implemented in common/ which
is then called from the command code. Follow this rule and move some more
of the memtest code to common/.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
detect_fs would usually mount a device on a directory,
so, use a device-specific type detection.
Signed-off-by: Vicente Bergas <vicencb@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
it's currently only handled through DT and for i.MX25 OTG port (at least),
we need to configure portsc to get host mode working but it actually fails
as ci->phymode is 0 which is USBPHY_INTERFACE_MODE_UNKNOWN
Signed-off-by: Eric Bénard <eric@eukrea.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
putting an error pointer into strerror can be a bit confusing since
strerror takes a positive error code but PTR_ERR returns a negative
number, so we have to do strerror(-PTR_ERR(errp)). Some places got
this wrong already, so introduce a strerrorp function which directly
takes an error pointer.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
ndelay is declared in include/clock.h, udelay in include/common.h and
mdelay in include/common.h and include/clock.h. Move them all to
include/clock.h and remove duplicates.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This one is available in the kernel and used by the denali driver
Signed-off-by: Enrico Jorns <ejo@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The host may assume that the usb device is still up and running after
booting if we do not deregister the udc here. I observed issues when the
linux kernel was using a usb gadget directly where the complete USB Hub
got disconnected through this.
This patch adds a proper USB disconnect for gadget devices.
Signed-off-by: Markus Pargmann <mpa@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Get a 'struct mci' by search after the device name.
Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Added missing defines for the extended CSD register until standard 5.1.
Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
There is no possibility to read/write to the extended CSD register of
MMC devices from a command. To avoid duplicated driver code, two driver
functions have to be public.
Signed-off-by: Daniel Schultz <d.schultz@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Allow drivers to build even if reset source support isn't built in.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Allow drivers to build even if WD core isn't enabled.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This patch is based on
commit 3d6f4a20cc287a8980c6186624834cf10a70752b
Author: David Miller <davem@davemloft.net>
Date: Thu Jul 24 23:38:31 2008 -0700
endian: Always evaluate arguments.
In barebox.git/drivers/usb/core/usb.c,
function usb_parse_config() we have had
le16_to_cpus(&(dev->config.wTotalLength));
which evaluates to "do { } while (0)" on little endian,
so struct usb_configuration field misuse will never
be discovered on little endian.
Therefore, always evaluate the arguments to nop endian
transformation operations.
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds a device_node member to struct cdev and a function
to find a cdev by device_node.
This also removes the setting of cdev->dev->device_node in
the of partition parser. We must not set the device since it
may not refer to a partition but to a whole device with partitions.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Instead of writing one word to the txfifo and then wait until
one is received in the rxfifo we can write until the txfifos
are not full and read as long the rxfifos contain data. This
makes transfers for the m25p80 driver around 7 times faster
here.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The i.MX SPI controller in version 2.3 can immediately start a transfer
when the txfifo is written to. In this mode we no longer have to trigger
the transfer with the xch bit which makes the code a bit simpler.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
- make some locally used functions static
- remove read_passwd which is unused
- some refactoring to make code pathes clearer
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We used to have the login functionality in the /env/bin/init script.
This is hard to review and it's too easy to break the login functionality
with changes to this script. Move the places to ask for a password to
C code where we have only a few places where we have to ask for a password.
Mainly these are run_shell() and the menutree command.
This patch introduces a login() function which will only return if the correct
password has been entered. Following calls will return immediately without
asking for a password again.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Different devices may report the reset source with different levels
of certainty. For example a SoC may see a power-on reset, but this
may only be because an external PMIC has power cycled the SoC. This
means the PMIC knows the real reason better and thus the reset reason
from the PMIC should be preferred.
This patch introduces priorities for the reset_source to handle the
above scenario. Also add a of_get_reset_source_priority() function
to retrieve the desired priority from the device tree.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds a dev and name member to struct watchdog which helps distinguishing
between different watchdogs. Also add some debugging aids.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Put watchdogs on a list to allow multiple watchdogs. Add a priority
field to be able to pick the highest priority watchdog.
This patch also provides a of_get_watchdog_priority() function to
allow configuring the watchdog priority from the device tree.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This replaces the reset_cpu() function which every SoC or board must
provide with registered handlers. This makes it possible to have multiple
reset functions for boards which have multiple ways to reset the machine.
Also boards which have no way at all to reset the machine no longer
have to provide a dummy reset_cpu() function.
The problem this solves is that some machines have external PMICs or
similar to reset the system which have to be preferred over the
internal SoC reset, because the PMIC can reset not only the SoC but also
the external devices.
To pick the right way to reset a machine each handler has a priority. The
default priority is 100 and all currently existing restart handlers are
registered with this priority. of_get_restart_priority() allows to retrieve
the priority from the device tree which makes it possible for boards to
give certain restart handlers a higher priority in order to use this one
instead of the default one.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Some SPI NOR flashes support 4K erase blocks. 4K erase blocks do not
work with UBIFS which needs a minimum erase block size of 15360 bytes.
Also bigger sectors are faster to erase. This patch adds a device tree
option to use the bigger blocks instead of the default 4K blocks.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: Philipp Zabel <p.zabel@pengutronix.de>
Sascha Hauer
Antony Pavlov
Peter Mamonov
Vicente Bergas
Eric Bénard
Enrico Jorns
Markus Pargmann
Daniel Schultz
Lucas Stach
David S. Miller