| Age | Commit message (Collapse) | Author |
|---|
|
When dm_remove_devices_active() removes devices using specialised flags
like DM_REMOVE_ACTIVE_ALL, a parent device may match (e.g. MMC has
DM_FLAG_OS_PREPARE) while its children do not. This deactivates the
parent but leaves children activated, an inconsistent state.
Later, when uclass_destroy() calls device_remove() with DM_REMOVE_NORMAL
on the already-deactivated parent, it returns early without touching the
children. The subsequent device_unbind() then fails because the children
are still activated.
Fix this by dropping only the DM_REMOVE_ACTIVE_ALL requirement for child
removal when the parent is being removed. This ensures children are
removed along with their parent, while still preserving other flags like
DM_REMOVE_NON_VITAL so that vital devices remain protected.
Signed-off-by: Simon Glass <[email protected]>
|
|
Move the flags_remove() call before device_chld_remove() and save the
result in a separate variable. This is just a refactoring with no
behaviour change, preparing for the next commit which needs to know
whether the parent will be removed before deciding how to remove its
children.
Signed-off-by: Simon Glass <[email protected]>
|
|
As part of bringing the master branch back in to next, we need to allow
for all of these changes to exist here.
Reported-by: Jonas Karlman <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
When bringing in the series 'arm: dts: am62-beagleplay: Fix Beagleplay
Ethernet"' I failed to notice that b4 noticed it was based on next and
so took that as the base commit and merged that part of next to master.
This reverts commit c8ffd1356d42223cbb8c86280a083cc3c93e6426, reversing
changes made to 2ee6f3a5f7550de3599faef9704e166e5dcace35.
Reported-by: Jonas Karlman <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
Remove <common.h> from this driver directory and when needed
add missing include files directly.
Signed-off-by: Tom Rini <[email protected]>
|
|
We have some nice macros for iterating over devices in device.h, but they
are not used by the driver core. Convert all the users I could find.
Signed-off-by: Sean Anderson <[email protected]>
|
|
Generate events when devices are probed or removed, allowing hooks
to be added for these situations.
This is controlled by the DM_EVENT config option.
Signed-off-by: Simon Glass <[email protected]>
|
|
Sphinx expects Return: and not @return to indicate a return value.
find . -name '*.c' -exec \
sed -i 's/^\(\s\)\*\(\s*\)@return\(\s\)/\1*\2Return:\3/' {} \;
find . -name '*.h' -exec \
sed -i 's/^\(\s\)\*\(\s*\)@return\(\s\)/\1*\2Return:\3/' {} \;
Signed-off-by: Heinrich Schuchardt <[email protected]>
|
|
This method is currently called after the platform data has been freed.
But the pre_unbind() method may wish to access this, e.g. to free some
data structures stored there.
Split the unbinding of devices into two pieces, as is done with removal.
This corrects the problem.
Also tidy a code-style issue in device_remove() while we are here.
Signed-off-by: Simon Glass <[email protected]>
|
|
- Merge the patch to take <asm/global_data.h> out of <common.h>
|
|
Add another flag to the DM core which could be assigned to drivers and
which makes those drivers call their remove callbacks last, just before
booting OS and after all the other drivers finished with their remove
callbacks. This is necessary for things like clock drivers, where the
other drivers might depend on the clock driver in their remove callbacks.
Prime example is the mmc subsystem, which can reconfigure a card from HS
mode to slower modes in the remove callback and for that it needs to
reconfigure the controller clock.
Signed-off-by: Marek Vasut <[email protected]>
Signed-off-by: Simon Glass <[email protected]>
|
|
At present if device_remove() decides that the device should not actually
be removed, it still calls the uclass pre_remove() method and powers the
device down.
Signed-off-by: Simon Glass <[email protected]>
|
|
At present the uclass pre-remove method is called before the children are
removed. But the children may refused to be removed, in whch case the
uclass is in a tricky situation. At present we handle this by calling
the uclass' post_probe() method. But it seems better to avoid doing
anything with the uclass in this case.
Switch the ordering so that we make sure the children can be removed
before advising the uclass.
Signed-off-by: Simon Glass <[email protected]>
|
|
This flag has the word 'REMOVE' in it which means it conflicts with
the DM_REMOVE flags. Rename it to DM_FLAG_LEAVE_PD_ON which seems to
indicate its purpose well enough.
Signed-off-by: Simon Glass <[email protected]>
|
|
Move this out of the common header and include it only where needed. In
a number of cases this requires adding "struct udevice;" to avoid adding
another large header or in other cases replacing / adding missing header
files that had been pulled in, very indirectly. Finally, we have a few
cases where we did not need to include <asm/global_data.h> at all, so
remove that include.
Signed-off-by: Simon Glass <[email protected]>
Signed-off-by: Tom Rini <[email protected]>
|
|
Driver model: make some udevice fields private
Driver model: Rename U_BOOT_DEVICE et al.
dtoc: Tidy up and add more tests
ns16550 code clean-up
x86 and sandbox minor fixes for of-platdata
dtoc prepration for adding build-time instantiation
|
|
Prepare v2021.01-rc5
Signed-off-by: Tom Rini <[email protected]>
|
|
At present flags are stored as part of the device. In preparation for
storing them separately, change the access to go through inline functions.
Signed-off-by: Simon Glass <[email protected]>
|
|
Use these functions in the core code as much as possible. With this, there
are only two places where each priv/plat pointer is accessed, one for read
and one for write.
Signed-off-by: Simon Glass <[email protected]>
|
|
Most drivers use these access methods but a few do not. Update them.
In some cases the access is not permitted, so mark those with a FIXME tag
for the maintainer to check.
Signed-off-by: Simon Glass <[email protected]>
Acked-by: Andy Shevchenko <[email protected]>
Acked-by: Pratyush Yadav <[email protected]>
|
|
Now that migration to the new sequence numbers is complete, drop the old
fields. Add a test that covers the new behaviour.
Also drop the check for OF_PRIOR_STAGE since we always assign sequence
numbers now.
Signed-off-by: Simon Glass <[email protected]>
|
|
We use 'priv' for private data but often use 'platdata' for platform data.
We can't really use 'pdata' since that is ambiguous (it could mean private
or platform data).
Rename some of the latter variables to end with 'plat' for consistency.
Signed-off-by: Simon Glass <[email protected]>
|
|
This construct is quite long-winded. In earlier days it made some sense
since auto-allocation was a strange concept. But with driver model now
used pretty universally, we can shorten this to 'auto'. This reduces
verbosity and makes it easier to read.
Coincidentally it also ensures that every declaration is on one line,
thus making dtoc's job easier.
Signed-off-by: Simon Glass <[email protected]>
|
|
The test should be checking whether $flags are non-zero and $drv_flags
contain specific flags, however these two sets of flags are separate,
and the two tests should be logically ANDed, not bitwise ANDed.
Signed-off-by: Marek Vasut <[email protected]>
Cc: Simon Glass <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
When removing a device the power domains it uses are generally powered
off. But when we are trying to unbind all devices (e.g. for running tests)
we don't want to probe a device in the 'remove' path.
Add a new flag to skip this power-down step.
Signed-off-by: Simon Glass <[email protected]>
|
|
This failure path is tricky to debug since it continues after failure and
there are a lot of error paths. Add logging to help.
Signed-off-by: Simon Glass <[email protected]>
|
|
In various cases a power domain must stay enabled after device
removal when booting OS (i.e. serial debug console or display).
Add a flag to selectively skip switching off a power domain.
Fixes: 52edfed65de9 ("dm: core: device: switch off power domain after device removal")
Signed-off-by: Anatolij Gustschin <[email protected]>
Acked-by: Neil Armstrong <[email protected]>
Tested-by: Guillaume La Roque <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
We want to avoid allocating platform data twice. This could happen if
device_probe() is called after device_ofdata_to_platdata() for the same
device.
Add a flag to track whether device_ofdata_to_platdata() has been called on
a device. Check the flag to make sure it doesn't happen twice, and clear
the flag when the data is freed.
Signed-off-by: Simon Glass <[email protected]>
|
|
The power domain associated with a device is enabled when probing,
but currently the domain remains enabled when the device is removed.
Some boards started to disable power domains for selected devices
via custom board_quiesce_devices(), but it doesn't work in many
cases, i. e. because devices still can be accessed later in
.remove() callback on behalf of dm_remove_devices_flags().
Utilize the DM core to power off the device power domain, but add a
device flag to be able to selectively let the power domain enabled
after device removal. This might be required for devices that must
remain enabled when booting OS, i. e. serial console for debug
output, etc.
Signed-off-by: Anatolij Gustschin <[email protected]>
Signed-off-by: Lokesh Vutla <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
Also add a 'drv' parameter to filter the children to remove/unbind.
Exporting those functions is a preparatory work for the addition of the
bind/unbind commands.
Signed-off-by: Jean-Jacques Hiblot <[email protected]>
|
|
When U-Boot started using SPDX tags we were among the early adopters and
there weren't a lot of other examples to borrow from. So we picked the
area of the file that usually had a full license text and replaced it
with an appropriate SPDX-License-Identifier: entry. Since then, the
Linux Kernel has adopted SPDX tags and they place it as the very first
line in a file (except where shebangs are used, then it's second line)
and with slightly different comment styles than us.
In part due to community overlap, in part due to better tag visibility
and in part for other minor reasons, switch over to that style.
This commit changes all instances where we have a single declared
license in the tag as both the before and after are identical in tag
contents. There's also a few places where I found we did not have a tag
and have introduced one.
Signed-off-by: Tom Rini <[email protected]>
|
|
This new flag can be added to DM device drivers, which need to do some
final configuration before U-Boot exits and the OS (e.g. Linux) is
started. The remove functions of those drivers will get called at
this stage to do these last-stage configuration steps.
Signed-off-by: Stefan Roese <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
Cc: Bin Meng <[email protected]>
|
|
The new function dm_remove_devices_flags() is intented for driver specific
last-stage cleanup operations before the OS is started. This patch adds
this functionality and hooks it into the common device_remove()
function.
Drivers wanting to use this feature for some last-stage removal calls,
need to add one of the DM_REMOVE_xx flags to their driver .flags.
Signed-off-by: Stefan Roese <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
|
|
This patch adds the flags parameter to device_remove() and changes all
calls to this function to provide the default value of DM_REMOVE_NORMAL
for "normal" device removal.
This is in preparation for the driver specific pre-OS (e.g. DMA
cancelling) remove support.
Signed-off-by: Stefan Roese <[email protected]>
Cc: Simon Glass <[email protected]>
Acked-by: Simon Glass <[email protected]>
|
|
This is a flag. Adjust the name to be consistent with the other flags.
Signed-off-by: Simon Glass <[email protected]>
|
|
Some devices have a name that is stored in allocated memory. At present
there is no mechanism to free this memory when the device is unbound.
Add a device flag to track whether a name is allocated and a function to
add the flag. Free the memory when the device is unbound.
Signed-off-by: Simon Glass <[email protected]>
|
|
This reverts commit bb52b367f6ca4a3a918e77737f4ff6a1089912d9.
Signed-off-by: Simon Glass <[email protected]>
|
|
In U-Boot's driver model, memory is basically allocated and freed
in the core framework. So, low level drivers generally only have
to specify the size of needed memory with .priv_auto_alloc_size,
.platdata_auto_alloc_size, etc. Nevertheless, some drivers still
need to allocate/free memory on their own in case they cannot
statically know the necessary memory size. So, I believe it is
reasonable enough to port Devres into U-boot.
Devres, which originates in Linux, manages device resources for each
device and automatically releases them on driver detach. With devres,
device resources are guaranteed to be freed whether initialization
fails half-way or the device gets detached.
The basic idea is totally the same to that of Linux, but I tweaked
it a bit so that it fits in U-Boot's driver model.
In U-Boot, drivers are activated in two steps: binding and probing.
Binding puts a driver and a device together. It is just data
manipulation on the system memory, so nothing has happened on the
hardware device at this moment. When the device is really used, it
is probed. Probing initializes the real hardware device to make it
really ready for use.
So, the resources acquired during the probing process must be freed
when the device is removed. Likewise, what has been allocated in
binding should be released when the device is unbound. The struct
devres has a member "probe" to remember when the resource was
allocated.
CONFIG_DEBUG_DEVRES is also supported for easier debugging.
If enabled, debug messages are printed each time a resource is
allocated/freed.
Signed-off-by: Masahiro Yamada <[email protected]>
Acked-by: Simon Glass <[email protected]>
|
|
Currently, we only have DM_FLAG_ACTIVATED to indicate the device
status, but we still cannot know in which stage is in progress,
binding or probing.
This commit introduces a new flag, DM_FLAG_BOUND, which is set when
the device is really bound, and cleared when it is unbound.
Signed-off-by: Masahiro Yamada <[email protected]>
Acked-by: Simon Glass <[email protected]>
|
|
These functions are useful to remove all children from an usb bus before
rescanning the bus. Give them a better name and export them.
Signed-off-by: Hans de Goede <[email protected]>
Acked-by: Simon Glass <[email protected]>
|
|
This commit adds 'uclass_platdata' field to 'struct udevice', which
can be automatically allocated at bind. The allocation size is defined
in 'struct uclass_driver' as 'per_device_platdata_auto_alloc_size'.
New device's flag is added: DM_FLAG_ALLOC_UCLASS_PDATA, which is used
for memory freeing at device unbind method.
As for other udevice's fields, a complementary function is added:
- dev_get_uclass_platdata()
Signed-off-by: Przemyslaw Marczak <[email protected]>
Cc: Simon Glass <[email protected]>
Acked-by: Simon Glass <[email protected]>
|
|
The driver is not modified by driver model, so update driver_bind() to
recognise that.
Signed-off-by: Simon Glass <[email protected]>
Reviewed-by: Marek Vasut <[email protected]>
|
|
In many cases the per-child private data for a device's children is defined
by the uclass rather than the individual driver. For example, a SPI bus
needs to store information about each of its children, but all SPI drivers
store the same information. It makes sense to allow the uclass to define
this data.
If the driver provides a size value for its per-child private data, then use
it. Failng that, fall back to that provided by the uclass.
Signed-off-by: Simon Glass <[email protected]>
Reviewed-by: Masahiro Yamada <[email protected]>
|
|
For buses it is common for parents to need to know the address of the child
on the bus, the bus speed to use for that child, and other information. This
can be provided in platform data attached to each child.
Add driver model support for this, including auto-allocation which can be
requested using a new property to specify the size of the data.
Signed-off-by: Simon Glass <[email protected]>
Reviewed-by: Masahiro Yamada <[email protected]>
|
|
When using allocated platform data, allocate it when we bind the device.
This makes it possible to fill in this information before the device is
probed.
This fits with the platform data model (when not using device tree),
since platform data exists at bind-time.
Signed-off-by: Simon Glass <[email protected]>
Reviewed-by: Masahiro Yamada <[email protected]>
|
|
For SPL we don't expect to need to remove a device. Save some code space
by dropping this feature. The board config can define
CONFIG_DM_DEVICE_REMOVE if this is in fact needed.
Signed-off-by: Simon Glass <[email protected]>
Reviewed-by: Tom Rini <[email protected]>
|
