In certain circumstances rpmsg devices needs to acquire a handle to the
ancestor remoteproc instance, e.g. to invoke rproc_report_crash() when a
fatal error is detected. Introduce an interface that walks the device
tree in search for a remoteproc instance and return this.
Tested-by: NSuman Anna <s-anna@ti.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Subdevices might depend on earlier registered subdevices for
communication purposes, as such they should be stopped in reverse order
so that said communication channel is removed after the dependent
subdevice is stopped.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Make this const as it is only stored in the type field of a device
structure, which is const.
Done using Coccinelle.
Signed-off-by: NBhumika Goyal <bhumirks@gmail.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The additional arguments in the internal __rproc_boot() function
were dropped in commit 2bfc311a ("remoteproc: Drop wait in
__rproc_boot()"). The exported rproc_boot() is now just a wrapper
around this internal function, so merge them together.

While at this, also remove the declaration for the previously
cleaned up rproc_boot_nowait() function.
Signed-off-by: NSuman Anna <s-anna@ti.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Trivial fix to spelling mistake in dev_err message
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Replace rproc_shutdown() by rproc_stop() and rproc_boot() by
rproc_start() in the recovery path, in order to avoid remoteproc
resources re-allocation overhead and to assist with extracting the
coredumps after stopping the remote processor.
Signed-off-by: NSarangdhar Joshi <spjoshi@codeaurora.org>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

In the context of recovering from crash, rproc_trigger_recovery() does
rproc_shutdown() followed by rproc_boot(). The remoteproc resources are
cleaned up in rproc_shutdown() and immediately reallocated in
rproc_boot() which is an unnecessary overhead. Furthermore, we want the
memory regions to be accessible after stopping the remote processor, to
be able to extract the memory content for a coredump.

This patch factors out the code in rproc_boot() and rproc_shutdown()
path and introduces rproc_{start,stop}() in order to avoid resource
allocation overhead.
Signed-off-by: NSarangdhar Joshi <spjoshi@codeaurora.org>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The functions rproc_add_virtio_devices() and rproc_fw_config_virtio()
are reduced to trigger auto-boot only. Modify these function names and
related comments to reflect their current state.

This patch does not add any functional change.
Signed-off-by: NSarangdhar Joshi <spjoshi@codeaurora.org>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The rproc_add_virtio_devices() requests firmware asynchronously and
triggers boot if the auto_boot flag is set. However, this
asynchronous call seems to be redundant for non auto-boot scenario
since the rproc_boot() would call request_firmware() anyways. Move
the auto_boot check to rproc_add() so that a redundant call to
_request_firmware can be avoided for non auto-boot case.
Signed-off-by: NSarangdhar Joshi <spjoshi@codeaurora.org>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

firmware_loading_complete is used to synchronize operations
on rproc while asynchronous firmware loading is in progress.
However, rproc_boot() no longer waits on
firmware_loading_complete. Hence drop this completion
variable altogether and handle the race between rproc_del()
and rproc_boot() using new state RPROC_DELETED.

The request_firmware_nowait() will hold the reference to
rproc device by using a get_device()/put_device(), so the
rproc struct will remain valid even when we return from
rproc_del() before the asynchronous call to
rproc_fw_config_virtio() completes.

CC: Loic Pallardy <loic.pallardy@st.com>
CC: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: NSarangdhar Joshi <spjoshi@codeaurora.org>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The "remoteproc{0,1...}" sysfs entries are added in
rproc_add() and deleted in rproc_type_release() instead of
in rproc_del(). That leaves these lingering entries sticking
around after we return from rproc_del(). Move the
rproc_delete_debug_dir() to rproc_del() to fix this.
Signed-off-by: NSarangdhar Joshi <spjoshi@codeaurora.org>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Following any fw_rsc_vdev entries in the resource table are two variable
length arrays, the first one reference vring resources and the second
one is the virtio config space.  The virtio config space is used by
virtio to communicate status and configuration changes and must as such
be shared with the remote.

The reverted commit incorrectly made any changes to the virtio config
space only affect the local copy, in an attempt to allowing memory
protection of the shared resource table.

This reverts commit cda85293.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Commit 2b45cef5 ("remoteproc: Further extend the vdev life cycle")
extends kref support for vdev management.
It introduces a regression when following sequence is executed:
rproc_boot --> rproc_shutdown --> rproc_boot
Second rproc_boot call crashes on register_virtio_device as device
is already existing.
Issue is previous vdev is never released when rproc is stop because
associated refcount is too high.

kref_get introduces is not needed as kref_init already initializes
krefcount to 1 because it considers associated variable as used.
This introduces a misalignment between kref_get and kref_put calls.

Fixes: 2b45cef5 ("remoteproc: Further extend the vdev life cycle")
Signed-off-by: NLoic Pallardy <loic.pallardy@st.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

In the event that rproc_boot() is called before the firmware loaded
completion has been flagged it will wait with the mutex held,
obstructing the request_firmware_nowait() callback from completing the
wait.

As rproc_fw_config_virtio() has been reduced to only triggering
auto-boot there is no longer a reason for waiting in rproc_boot(), so
drop this.

Cc: Sarangdhar Joshi <spjoshi@codeaurora.org>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Warning users that remoteproc and it's binary format are under
development doesn't serve much of a purpose. Different drivers support
different image formats and the resource table has a version field that
would need to be bumped when incompatible changes are introduced.

So lets drop this warning to clean up the kernel log.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

As all vdev resources are allocated before we boot the remote processor
we no longer need to support modifying the resource table while the
remote is running.

This saves us from the table_ptr dance, but more importantly allow the
remote processor to enable security lock down of the loaded table memory
region.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The vdev handler is now just another resource allocator, so handle all
resource types in a single pass.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Vrings are now allocated as we parse the resource table, before we
boot the rproc or register any virtio devices, so it's safe to bump
max_notifyid as part of this process.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Represent the virtio device part of the vdev resources as remoteproc
subdevices to finalize the decoupling of the virtio resource and device
handling.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Tie the vdev (and hence vring) life cycle to the resource parsing and
resource cleanup operations, allowing us to safely register and
unregister virtio devices on the go.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Instead of having the vrings being allocated and freed as they are
requested by the virtio device tie their life cycle to the vdev
resource. This allows us to decouple the vdev resource management from
the virtio device management.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

No functional change
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

A subdevice is an abstract entity that can be used to tie actions to the
booting and shutting down of a remote processor. The subdevice object is
expected to be embedded in concrete implementations, allowing for a
variety of use cases to be implemented.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

This patch adds a sysfs interface to rproc allowing the firmware name
and processor state to be changed dynamically.

State was previously available in debugfs, and is replicated here. The
firmware file allows retrieval of the running firmware name, and a new
one to be specified at run time, so long as the remote processor has
been stopped.
Signed-off-by: NMatt Redfearn <matt.redfearn@imgtec.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Storage of the firmware name was inconsistent, either storing a pointer
to a name stored with unknown ownership, or a variable length tacked
onto the end of the struct proc allocated in rproc_alloc.

In preparation for allowing the firmware of an already allocated struct
rproc to be changed, instead always keep a locally maintained copy of
the firmware name.
Signed-off-by: NMatt Redfearn <matt.redfearn@imgtec.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Lock the implementation as we hand out references to client drivers
rather than when they try to boot the remote processor. This allows
auto-booting remote processors to be shut down by unloading their
module, in addition to first unbinding them.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

In order to be able to lock a rproc driver implementations only when
used by a client, we must differ between the dereference operation of a
client and the implementation itself.

This patch brings no functional change.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The freeing of resources will attempt to clear values previously set in
the cached resource table, so make sure to free the table after we have
cleaned up the resources.

Fixes: 988d204c ("remoteproc: Move handling of cached table to boot/shutdown")
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

rproc_alloc() will make sure that the "firmware" pointer is either a
driver supplied value or pointing to a generated firmware filename, it
can't be NULL. So drop the extra check in the rproc_boot() path.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

In current implementation, struct fw_rsc_vdev_vring which describes
vring resource in firmware resource table owns only device address,
because it assumes that host is responsible of vring allocation and
only device address is needed by coprocessor.
But if vrings need to be fixed in system memory map for any reasons
(security, SoC charactieristics...), physical address is needed exatly
identified the memory chunck by host.

For that let's transform reserved field of struct fw_rsc_vdev_vring
to pa (physical address).
Signed-off-by: NLoic Pallardy <loic.pallardy@st.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

In the case that we have a resource table, but not a loaded one we
should leave the table_ptr intact, as subsequent resource handling could
otherwise dereference the NULL pointer.
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

As we moved the vdev handling to the main boot/shutdown code path we can
further simplify the resource table handling by moving the parsing spet
to boot as well. The lifespan of the resource table is changed to live
from rproc_boot() to rproc_shutdown().

Cc: Lee Jones <lee.jones@linaro.org>
Cc: Loic Pallardy <loic.pallardy@st.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The newly introduced "always-on" flag allows us to stop giving the vdevs
special treatment. The ordering of resource allocation and life cycle of
the remote processor is kept intact.

This allows us to mark a remote processor with vdevs to not boot unless
explicitly requested to do so by a client driver.

Cc: Lee Jones <lee.jones@linaro.org>
Cc: Loic Pallardy <loic.pallardy@st.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The calculation of max_notifyid must only be done before we call start()
on the remoteproc drivers, so move the calculation to be part of the
loading steps.

Cc: Lee Jones <lee.jones@linaro.org>
Cc: Loic Pallardy <loic.pallardy@st.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Introduce an "auto-boot" flag on rprocs to make it possible to flag
remote processors without vdevs to automatically boot once the firmware
is found.

Preserve previous behavior of the wkup_m3 processor being explicitly
booted by a consumer.

Cc: Lee Jones <lee.jones@linaro.org>
Cc: Loic Pallardy <loic.pallardy@st.com>
Cc: Suman Anna <s-anna@ti.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

There are couple of debug statements that are printing hexadecimal
numbers without the leading 0x. Fix these and use the standard 0x%x
format specifier so that there is no confusion when looking at the
traces.
Signed-off-by: NSuman Anna <s-anna@ti.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

This patch fixes the existing alignment checkpatch check
warnings of the type "Alignment should match open parenthesis"
in the remoteproc core source files.
Signed-off-by: NSuman Anna <s-anna@ti.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Fix the code formatting to use the kernel preferred style
of using the actual variables to determize the size using
the sizeof() operator.
Signed-off-by: NSuman Anna <s-anna@ti.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

Fix couple of minor mis-spelled words in all the remoteproc
source files.
Signed-off-by: NSuman Anna <s-anna@ti.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>

The dma_addr_t types can be printed properly using the %pad
printk format-specifier, there is no need to resort to the
unsigned long long type-casting to deal with different possible
type sizes.
Signed-off-by: NSuman Anna <s-anna@ti.com>
Signed-off-by: NBjorn Andersson <bjorn.andersson@linaro.org>