scmi_voltage_descriptors_get() will incorrecly return an error code if
the last iteration of the for loop that retrieves the descriptors is
skipped due to an error. Skipping an iteration in the loop is not an
error, but the `ret` value from the last iteration will be propagated
when the function returns.
Fix by not saving return values that should not be propagated. This
solution also minimizes the risk of future patches accidentally
re-introducing this bug.
Link: https://lore.kernel.org/r/20220610140055.31491-1-ludvig.parsson@axis.com
Reviewed-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Ludvig Pärsson <ludvig.parsson@axis.com>
[sudeep.holla: Removed unneeded reset_rx_to_maxsz and check for return
value from scmi_voltage_levels_get as suggested by Cristian]
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Commit b260fccaeb ("firmware: arm_scmi: Add SCMI v3.1 protocol extended
names support") moved all the name string buffers to use the extended buffer
size of 64 instead of the required 16 bytes. While that should be fine if
the firmware terminates the string before 16 bytes, there is possibility
of copying random data if the name is not NULL terminated by the firmware.
SCMI base protocol agent_name/vendor_id/sub_vendor_id are defined by the
specification as NULL-terminated ASCII strings up to 16-bytes in length.
The underlying buffers and message descriptors are currently bigger than
needed; resize them to fit only the strictly needed 16 bytes to avoid
any possible leaks when reading data from the firmware.
Change the size argument of strlcpy to use SCMI_SHORT_NAME_MAX_SIZE always
when dealing with short domain names, so as to limit the possibility that
an ill-formed non-NULL terminated short reply from the SCMI platform
firmware can leak stale content laying in the underlying transport shared
memory area.
While at that, convert all strings handling routines to use the preferred
strscpy.
Link: https://lore.kernel.org/r/20220608095530.497879-1-cristian.marussi@arm.com
Fixes: b260fccaeb ("firmware: arm_scmi: Add SCMI v3.1 protocol extended names support")
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Avoid to invoke SENSOR_AXIS_NAME_GET on sensors that have not declared at
least one of their axes as supporting extended names.
Since the returned list of axes supporting extended names is not
necessarily comprising all the existing axes of the specified sensor,
take care also to properly pick the axis descriptor from the ID embedded
in the response.
Link: https://lore.kernel.org/r/20220608164051.2326087-1-cristian.marussi@arm.com
Fixes: 802b0bed01 ("firmware: arm_scmi: Add SCMI v3.1 SENSOR_AXIS_NAME_GET support")
Cc: Peter Hilber <peter.hilber@opensynergy.com>
Cc: Sudeep Holla <sudeep.holla@arm.com>
Reviewed-by: Peter Hilber <peter.hilber@opensynergy.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
While using SCMI iterators helpers a few local automatic variables are
defined but then used only as input for sizeof operators.
cppcheck is fooled to complain about this with:
| drivers/firmware/arm_scmi/sensors.c:341:48: warning: Variable 'msg' is
| not assigned a value. [unassignedVariable]
| struct scmi_msg_sensor_list_update_intervals *msg;
Even though this is an innocuos warning, since the uninitialized variable
is at the end never used in the reported cases, fix these occurences all
over SCMI stack to avoid keeping unneeded objects on the stack.
Link: https://lore.kernel.org/r/20220530115237.277077-1-cristian.marussi@arm.com
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Sudeep Holla <sudeep.holla@arm.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Even though malformed replies from firmware must be treated carefully to
avoid memory corruption in the kernel, some out-of-spec SCMI replies can
be tolerated to avoid breaking existing deployed system, as long as they
won't cause memory issues.
Relax the sanity checks on the recieved protocol list in the base protocol
to avoid breaking one of the deployed platform whose firmware is not easily
upgradable currently.
Link: https://lore.kernel.org/r/20220523171559.472112-1-cristian.marussi@arm.com
Cc: Etienne Carriere <etienne.carriere@linaro.org>
Cc: Sudeep Holla <sudeep.holla@arm.com>
Reported-by: Nicolas Frattaroli <frattaroli.nicolas@gmail.com>
Tested-By: Frank Wunderlich <frank-w@public-files.de>
Acked-by: Michael Riesch <michael.riesch@wolfvision.net>
Acked-by: Etienne Carriere <etienne.carriere@linaro.org>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
A few dereferences could happen before the iterator pointer argument was
checked for NULL, causing the following smatch warnings:
drivers/firmware/arm_scmi/driver.c:1214 scmi_iterator_run() warn: variable
dereferenced before check 'i' (see line 1210)
Fix by moving the checks early and dropping some unneeded local references.
No functional change.
Link: https://lore.kernel.org/r/20220503121047.3590340-1-cristian.marussi@arm.com
Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Add support for TEE shared memory in optee scmi transport. When using
tee shared memory, scmi optee transport manages SCMI messages using
msg protocol(from msg.c) in shared memory, whereas smt(from shmem.c)
protocol is used with static IOMEM based shared buffers.
Link: https://lore.kernel.org/r/20220425085127.2009-1-etienne.carriere@linaro.org
Reviewed-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Add SCMI v3.1 voltage protocol support for asynchronous VOLTAGE_LEVEL_SET
command.
Note that, if a voltage domain is advertised to support the asynchronous
version of VOLTAGE_LEVEL_SET, the command will be issued asynchronously
unless explicitly requested to use the synchronous version by setting the
mode to SCMI_VOLTAGE_LEVEL_SET_SYNC when calling voltage_ops->level_set.
The SCMI regulator driver level_set invocation has been left unchanged
so that it will transparently use the asynchronous version if available.
Link: https://lore.kernel.org/r/20220330150551.2573938-21-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Starting with SCMI v3.1, the PERFORMANCE_LIMITS_SET command allows a user
to request only one between max and min ranges to be changed, while leaving
the other untouched if set to zero in the request. Anyway SCMI v3.1 states
also explicitly that you cannot leave both of those unchanged (zeroed) when
issuing such command, so add a proper check for this condition.
Link: https://lore.kernel.org/r/20220330150551.2573938-23-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
[sudeep.holla: Dropped check for v3.0 and above to make the check
unconditional, updated the subject accordingly]
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Add SCMI v3.1 internal support for parsing message attributes reporting
the capability of a performance domain to report power-cost in microwatts.
Link: https://lore.kernel.org/r/20220330150551.2573938-22-cristian.marussi@arm.com
Cc: Lukasz Luba <lukasz.luba@arm.com>
Reviewed-by: Lukasz Luba <lukasz.luba@arm.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
SCMI specification defines some commands as optionally issued over multiple
messages in order to overcome possible limitations in payload size enforced
by the configured underlyinng transport.
Introduce some common protocol helpers to provide a unified solution for
issuing such SCMI multi-part commands.
Link: https://lore.kernel.org/r/20220330150551.2573938-14-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
The clock_enable_latency field in CLOCK_ATTRIBUTES response message has
been added only since SCMI v3.1. Use the advertised SCMI clock protocol
version as a proper condition check for parsing it, instead of the bare
message length lookup.
Link: https://lore.kernel.org/r/20220330150551.2573938-13-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
As per the spec, the clock_enable_delay is the worst case latency
incurred by the platform to enable the clock. The value of 0 indicates
that the platform doesn't support the same and must be considered as
maximum latency for practical purposes.
Currently the value of 0 is assigned as is and is propogated to the clock
framework which can assume that the clock can support atomic enable operation.
Link: https://lore.kernel.org/r/20220428122913.1654821-1-sudeep.holla@arm.com
Fixes: 18f295b758 ("firmware: arm_scmi: Add support for clock_enable_latency")
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Using the common protocol helper implementation add support for all new
SCMIv3.1 extended names commands related to all protocols with the
exception of SENSOR_AXIS_GET_NAME.
Link: https://lore.kernel.org/r/20220330150551.2573938-12-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Introduce a new set of common protocol operations bound to the protocol
handle structure so that can be invoked by the protocol implementation code
even when protocols are built as distinct loadable kernel module without
the need of exporting new symbols, like already done with scmi_xfer_ops.
Add at first, as new common protocol helper, an .extended_name_get helper
which will ease implementation and will avoid code duplication when adding
new SCMIv3.1 per-protocol _NAME_GET commands.
Link: https://lore.kernel.org/r/20220330150551.2573938-11-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Move some SCMI protocol specific definitions from common.h into a the new
dedicated protocols.h header so that SCMI protocols core code can include
only what it needs; this is going to be useful to avoid the risk of growing
indefinitely the dimension of common.h, especially when introducing some
common protocols helper functions.
Header common.h will continue to be included by SCMI core and transport
layers.
Link: https://lore.kernel.org/r/20220330150551.2573938-10-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
The string array 'name' inside struct scmi_clock_info holds the clock name
which was successfully retrieved by querying the SCMI platform, unless the
related underlying SCMI command failed.
Anyway, such scmi_clock_info structure is allocated using devm_kcalloc()
which in turn internally appends a __GFP_ZERO flag to its invocation:
as a consequence the string 'name' field does not need to be zeroed when
we fail to get the clock name via SCMI, it is already NULL terminated.
Remove unneeded explicit NULL termination.
Link: https://lore.kernel.org/r/20220330150551.2573938-9-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
When CLOCK_RATE_SET command is issued in asynchronous mode the delayed
response CLOCK_RATE_SET_COMPLETE comes back once the SCMI platform has
effectively operated the requested change: such delayed response carries
also the clock ID and the final clock rate that has been set.
As an aid to debug issues, check that the clock ID in the delayed
response matches the expected one and debug print the rate value.
Link: https://lore.kernel.org/r/20220330150551.2573938-8-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
A few protocol operations are available that returns a pointer to an
internal character array representing resource name. Make those functions
return a const pointer to such array.
Link: https://lore.kernel.org/r/20220330150551.2573938-7-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Move away from a statically allocated array for holding the current set of
protocols implemented by the platform in favour of allocating it
dynamically based on the number of protocols effectively advertised by the
platform via BASE protocol exchanges.
While at that, rectify the BASE_DISCOVER_LIST_PROTOCOLS loop iterations to
terminate only when a number of protocols equal to the advertised ones has
been received, instead of looping till the platform returns no more
protocols descriptors. This new behaviour is better compliant with the
specification and it has been tested to work equally well against an SCMI
stack running on top of an official SCP firmware on a JUNO board.
Link: https://lore.kernel.org/r/20220330150551.2573938-6-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Do not blindly trust SCMI platform response about list of implemented
protocols, instead validate the reported length of the list of protocols
against the real payload size of the message reply.
Link: https://lore.kernel.org/r/20220330150551.2573938-5-cristian.marussi@arm.com
Fixes: b6f20ff8bd ("firmware: arm_scmi: add common infrastructure and support for base protocol")
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
[sudeep.holla: Added early break if loop_num_ret = 0 and simplified calc_list_sz
calculation]
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
While enumerating protocols implemented by the SCMI platform using
BASE_DISCOVER_LIST_PROTOCOLS, the number of returned protocols is
currently validated in an improper way since the check employs a sum
between unsigned integers that could overflow and cause the check itself
to be silently bypassed if the returned value 'loop_num_ret' is big
enough.
Fix the validation avoiding the addition.
Link: https://lore.kernel.org/r/20220330150551.2573938-4-cristian.marussi@arm.com
Fixes: b6f20ff8bd ("firmware: arm_scmi: add common infrastructure and support for base protocol")
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Bail out of protocol initialisation routine early when basic information
about protocol version and attributes could not be retrieved. Failing to
act this way can lead to a successfully initialized SCMI protocol which
is in fact not fully functional.
Link: https://lore.kernel.org/r/20220330150551.2573938-3-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
The sparse checker complains about converting pointers between address
spaces. We correctly stored an __iomem pointer in struct scmi_optee_channel,
but discarded the __iomem when returning it from get_channel_shm, causing one
warning. Then we passed the non-__iomem pointer return from get_channel_shm
at two other places, where an __iomem pointer is expected, causing couple of
other warnings
Add the appropriate __iomem annotations at all places where it is missing.
optee.c:414:20: warning: incorrect type in return expression (different address spaces)
optee.c:414:20: expected struct scmi_shared_mem *
optee.c:414:20: got struct scmi_shared_mem [noderef] __iomem *shmem
optee.c:426:26: warning: incorrect type in argument 1 (different address spaces)
optee.c:426:26: expected struct scmi_shared_mem [noderef] __iomem *shmem
optee.c:426:26: got struct scmi_shared_mem *shmem
optee.c:441:30: warning: incorrect type in argument 1 (different address spaces)
optee.c:441:30: expected struct scmi_shared_mem [noderef] __iomem *shmem
optee.c:441:30: got struct scmi_shared_mem *shmem
Link: https://lore.kernel.org/r/20220404102419.1159705-1-sudeep.holla@arm.com
Cc: Etienne Carriere <etienne.carriere@linaro.org>
Cc: Cristian Marussi <cristian.marussi@arm.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
During SCMI Clock protocol initialization, after having retrieved from the
SCMI platform all the available discrete rates for a specific clock, the
clock rates array is sorted, unfortunately using a pointer to its end as
a base instead of its start, so that sorting does not work.
Fix invocation of sort() passing as base a pointer to the start of the
retrieved clock rates array.
Link: https://lore.kernel.org/r/20220318092813.49283-1-cristian.marussi@arm.com
Fixes: dccec73de9 ("firmware: arm_scmi: Keep the discrete clock rates sorted")
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
On SCMI transports whose channels are based on a shared resource the TX
channel area has to be acquired by the agent before placing the desired
command into the channel and it will be then relinquished by the platform
once the related reply has been made available into the channel.
On an RX channel the logic is reversed with the platform acquiring the
channel area and the agent reliquishing it once done by calling the
scmi_clear_channel() helper.
As a consequence, even in case of error, the agent must never try to clear
a TX channel from its side: restrict the existing clear channel call on the
the reply path only to delayed responses since they are indeed coming from
the RX channel.
Link: https://lore.kernel.org/r/20220224152404.12877-1-cristian.marussi@arm.com
Fixes: e9b21c9618 ("firmware: arm_scmi: Make .clear_channel optional")
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
There are a few separately maintained driver subsystems that we merge through
the SoC tree, notable changes are:
- Memory controller updates, mainly for Tegra and Mediatek SoCs,
and clarifications for the memory controller DT bindings
- SCMI firmware interface updates, in particular a new transport based
on OPTEE and support for atomic operations.
- Cleanups to the TEE subsystem, refactoring its memory management
For SoC specific drivers without a separate subsystem, changes include
- Smaller updates and fixes for TI, AT91/SAMA5, Qualcomm and NXP
Layerscape SoCs.
- Driver support for Microchip SAMA5D29, Tesla FSD, Renesas RZ/G2L,
and Qualcomm SM8450.
- Better power management on Mediatek MT81xx, NXP i.MX8MQ
and older NVIDIA Tegra chips
-----BEGIN PGP SIGNATURE-----
iQIzBAABCgAdFiEEo6/YBQwIrVS28WGKmmx57+YAGNkFAmI4nOUACgkQmmx57+YA
GNlNNhAApPQw+FKQ6yVj2EZYcaAgik8PJAJoNQWYED52iQfm5uXgjt3aQewvrPNW
nkKx5Mx+fPUfaKx5mkVOFMhME5Bw9tYbXHm2/RpRp+n8jOdUlQpAhzIPOyWPHOJS
QX6qu4t+agrQzjbOCGouAJXgyxhTJFUMviM2EgVHbQHXPtdF8i2kyanfCP7Rw8cx
sVtLwpvhbLm849+deYRXuv2Xw9I3M1Np7018s5QciimI2eLLEb+lJ/C5XWz5pMYn
M1nZ7uwCLKPCewpMETTuhKOv0ioOXyY9C1ghyiGZFhHQfoCYTu94Hrx9t8x5gQmL
qWDinXWXVk8LBegyrs8Bp4wcjtmvMMLnfWtsGSfT5uq24JOGg22OmtUNhNJbS9+p
VjEvBgkXYD7UEl5npI9v9/KQWr3/UDir0zvkuV40gJyeBWNEZ/PB8olXAxgL7wZv
cXRYSaUYYt3DKQf1k5I4GUyQtkP/4RaBy6AqvH5Sx0lCwuY6G6ISK+kCPaaSRKnX
WR+nFw84dKCu7miehmW9qSzMQ4kiSCKIDqk7ilHcwv0J2oXDrlqVPKGGGTzZjUc8
+feqM/eSoYvDDEDemuXNSnl3hc1Zlvm7Apd5AN6kdTaNgoACDYdyvGuJ3CvzcA+K
1gBHUBvGS/ODA25KnYabr7wCMgxYqf7dXfkyKIBwFHwxOnRHtgs=
=Cfbk
-----END PGP SIGNATURE-----
Merge tag 'arm-drivers-5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc
Pull ARM driver updates from Arnd Bergmann:
"There are a few separately maintained driver subsystems that we merge
through the SoC tree, notable changes are:
- Memory controller updates, mainly for Tegra and Mediatek SoCs, and
clarifications for the memory controller DT bindings
- SCMI firmware interface updates, in particular a new transport
based on OPTEE and support for atomic operations.
- Cleanups to the TEE subsystem, refactoring its memory management
For SoC specific drivers without a separate subsystem, changes include
- Smaller updates and fixes for TI, AT91/SAMA5, Qualcomm and NXP
Layerscape SoCs.
- Driver support for Microchip SAMA5D29, Tesla FSD, Renesas RZ/G2L,
and Qualcomm SM8450.
- Better power management on Mediatek MT81xx, NXP i.MX8MQ and older
NVIDIA Tegra chips"
* tag 'arm-drivers-5.18' of git://git.kernel.org/pub/scm/linux/kernel/git/soc/soc: (154 commits)
ARM: spear: fix typos in comments
soc/microchip: fix invalid free in mpfs_sys_controller_delete
soc: s4: Add support for power domains controller
dt-bindings: power: add Amlogic s4 power domains bindings
ARM: at91: add support in soc driver for new SAMA5D29
soc: mediatek: mmsys: add sw0_rst_offset in mmsys driver data
dt-bindings: memory: renesas,rpc-if: Document RZ/V2L SoC
memory: emif: check the pointer temp in get_device_details()
memory: emif: Add check for setup_interrupts
dt-bindings: arm: mediatek: mmsys: add support for MT8186
dt-bindings: mediatek: add compatible for MT8186 pwrap
soc: mediatek: pwrap: add pwrap driver for MT8186 SoC
soc: mediatek: mmsys: add mmsys reset control for MT8186
soc: mediatek: mtk-infracfg: Disable ACP on MT8192
soc: ti: k3-socinfo: Add AM62x JTAG ID
soc: mediatek: add MTK mutex support for MT8186
soc: mediatek: mmsys: add mt8186 mmsys routing table
soc: mediatek: pm-domains: Add support for mt8186
dt-bindings: power: Add MT8186 power domains
soc: mediatek: pm-domains: Add support for mt8195
...
An SCMI platform can optionally advertise an enable latency typically
associated with a specific clock resource: add support for parsing such
optional message field and export such information in the usual publicly
accessible clock descriptor.
Link: https://lore.kernel.org/r/20220217131234.50328-8-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Introduce new _atomic variant for SCMI clock protocol operations related
to enable disable operations: when an atomic operation is required the xfer
poll_completion flag is set for that transaction.
Link: https://lore.kernel.org/r/20220217131234.50328-7-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
An SCMI agent can be configured system-wide with a well-defined atomic
threshold: only SCMI synchronous command whose latency has been advertised
by the SCMI platform to be lower or equal to this configured threshold will
be considered for atomic operations, when requested and if supported by the
underlying transport at all.
Link: https://lore.kernel.org/r/20220217131234.50328-6-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Add support for .mark_txdone and .poll_done transport operations to SCMI
VirtIO transport as pre-requisites to enable atomic operations.
Add a Kernel configuration option to enable SCMI VirtIO transport polling
and atomic mode for selected SCMI transactions while leaving it default
disabled.
Link: https://lore.kernel.org/r/20220217131234.50328-4-cristian.marussi@arm.com
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Igor Skalkin <igor.skalkin@opensynergy.com>
Cc: Peter Hilber <peter.hilber@opensynergy.com>
Cc: virtualization@lists.linux-foundation.org
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Add a new spinlock dedicated to the access of the TX free list and a couple
of helpers to get and put messages back and forth from the free_list.
Link: https://lore.kernel.org/r/20220217131234.50328-3-cristian.marussi@arm.com
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Igor Skalkin <igor.skalkin@opensynergy.com>
Cc: Peter Hilber <peter.hilber@opensynergy.com>
Cc: virtualization@lists.linux-foundation.org
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Currently SCMI VirtIO channels are marked with a ready flag and related
lock to track channel lifetime and support proper synchronization at
shutdown when virtqueues have to be stopped.
This leads to some extended spinlocked sections with IRQs off on the RX
path to keep hold of the ready flag and does not scale well especially when
SCMI VirtIO polling mode will be introduced.
Add an SCMI VirtIO channel dedicated refcount to track active users on both
the TX and the RX path and properly enforce synchronization and cleanup at
shutdown, inhibiting further usage of the channel once freed.
Link: https://lore.kernel.org/r/20220217131234.50328-2-cristian.marussi@arm.com
Cc: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Igor Skalkin <igor.skalkin@opensynergy.com>
Cc: Peter Hilber <peter.hilber@opensynergy.com>
Cc: virtualization@lists.linux-foundation.org
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
modprobe can't handle spaces in aliases. Get rid of it to fix the issue.
Link: https://lore.kernel.org/r/20220211102704.128354-1-sudeep.holla@arm.com
Fixes: aa4f886f38 ("firmware: arm_scmi: add basic driver infrastructure for SCMI")
Reviewed-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Alyssa Ross <hi@alyssa.is>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
The SMC calling convention designates R0-R7 as input registers in
AArch32 mode, and this conflicts with the compiler's use of R7 as a
frame pointer when building in Thumb2 mode. Generally, we don't enable
the frame pointer, and GCC happily enables the -pg profiling hooks
without them. However, Clang refuses, and errors out with the message
below:
drivers/firmware/arm_scmi/smc.c:152:2: error: write to reserved register 'R7'
arm_smccc_1_1_invoke(scmi_info->func_id, 0, 0, 0, 0, 0, 0, 0, &res);
^
include/linux/arm-smccc.h:550:4: note: expanded from macro 'arm_smccc_1_1_invoke'
arm_smccc_1_1_smc(__VA_ARGS__); \
^
Let's just disable ftrace for the compilation unit when building this
configuration.
Link: https://lore.kernel.org/r/20220203082204.1176734-11-ardb@kernel.org
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Couple of main additions:
- Support for OPTEE based SCMI transport to enable using SCMI service
provided by OPTEE on some platforms
- Support for atomic SCMI transports which enables few SCMI transactions
to be completed in atomic context. This involves other refactoring work
associated with it. It also marks SMC and OPTEE as atomic transport as
the commands are completed once the return
Other changes involves some trace and log enhancements and a miscellaneous
bug fix.
-----BEGIN PGP SIGNATURE-----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=12H2
-----END PGP SIGNATURE-----
Merge tag 'scmi-updates-5.17' of git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux into for-next/scmi
Arm SCMI firmware interface updates for v5.17
Couple of main additions:
- Support for OPTEE based SCMI transport to enable using SCMI service
provided by OPTEE on some platforms
- Support for atomic SCMI transports which enables few SCMI transactions
to be completed in atomic context. This involves other refactoring work
associated with it. It also marks SMC and OPTEE as atomic transport as
the commands are completed once the return
Other changes involves some trace and log enhancements and a miscellaneous
bug fix.
* tag 'scmi-updates-5.17' of git://git.kernel.org/pub/scm/linux/kernel/git/sudeep.holla/linux:
firmware: arm_scmi: Add new parameter to mark_txdone
firmware: arm_scmi: Add atomic mode support to smc transport
firmware: arm_scmi: Add support for atomic transports
firmware: arm_scmi: Make optee support sync_cmds_completed_on_ret
firmware: arm_scmi: Make smc support sync_cmds_completed_on_ret
firmware: arm_scmi: Add sync_cmds_completed_on_ret transport flag
firmware: arm_scmi: Make smc transport use common completions
firmware: arm_scmi: Add configurable polling mode for transports
firmware: arm_scmi: Use new trace event scmi_xfer_response_wait
include: trace: Add new scmi_xfer_response_wait event
firmware: arm_scmi: Refactor message response path
firmware: arm_scmi: Set polling timeout to max_rx_timeout_ms
firmware: arm_scmi: Perform earlier cinfo lookup call in do_xfer
firmware: arm_scmi: optee: Drop the support for the OPTEE shared dynamic buffer
firmware: arm_scmi: optee: Fix missing mutex_init()
firmware: arm_scmi: Make virtio Version_1 compliance optional
firmware: arm_scmi: Add optee transport
dt-bindings: arm: Add OP-TEE transport for SCMI
firmware: arm_scmi: Review some virtio log messages
This will enable cleanups down the road.
The idea is to disable cbs, then add "flush_queued_cbs" callback
as a parameter, this way drivers can flush any work
queued after callbacks have been disabled.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Link: https://lore.kernel.org/r/20211013105226.20225-1-mst@redhat.com
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Add a new xfer parameter to mark_txdone transport operation which enables
the SCMI core to optionally pass back into the transport layer a reference
to the xfer descriptor that is being handled.
Link: https://lore.kernel.org/r/20211220195646.44498-9-cristian.marussi@arm.com
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Add a Kernel configuration option to enable SCMI SMC transport atomic
mode operation for selected SCMI transactions and leave it as default
disabled.
Substitute mutex usages with busy-waiting and declare smc transport as
.atomic_enabled if such Kernel configuration option is enabled.
Link: https://lore.kernel.org/r/20211220195646.44498-8-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
An SCMI transport can be configured as .atomic_enabled in order to signal
to the SCMI core that all its TX path is executed in atomic context and
that, when requested, polling mode should be used while waiting for command
responses.
When a specific platform configuration had properly configured such a
transport as .atomic_enabled, the SCMI core will also take care not to
sleep in the corresponding RX path while waiting for a response if that
specific command transaction was requested as atomic using polling mode.
Asynchronous commands should not be used in an atomic context and so a
warning is emitted if polling was requested for an asynchronous command.
Add also a method to check, from the SCMI drivers, if the underlying SCMI
transport is currently configured to support atomic transactions: this will
be used by upper layers to determine if atomic requests can be supported at
all on this SCMI instance.
Link: https://lore.kernel.org/r/20211220195646.44498-7-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Declare each OPTEE SCMI channel as not having a completion_irq so as to
enable polling mode and then enable also .sync_cmds_completed_on_ret flag
in the OPTEE transport descriptor so that real polling is itself
effectively bypassed on the rx path: once the optee command invocation has
successfully returned the core will directly fetch the response from the
shared memory area.
Remove OPTEE SCMI transport specific .poll_done callback support since
real polling is effectively bypassed when .sync_cmds_completed_on_ret is
set.
Add OPTEE SCMI transport specific .mark_txdone callback support in order to
properly handle channel locking along the tx path.
Link: https://lore.kernel.org/r/20211220195646.44498-6-cristian.marussi@arm.com
Cc: Etienne Carriere <etienne.carriere@linaro.org>
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Enable sync_cmds_completed_on_ret in the SMC transport descriptor and
remove SMC specific .poll_done callback support since polling is bypassed
when sync_cmds_completed_on_ret is set.
Link: https://lore.kernel.org/r/20211220195646.44498-5-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Add a flag to let the transport signal to the core if its handling of sync
command implies that, after .send_message has returned successfully, the
requested command can be assumed to be fully and completely executed on
SCMI platform side so that any possible response value is already
immediately available to be retrieved by a .fetch_response: in other words
the polling phase can be skipped in such a case and the response values
accessed straight away.
Note that all of the above applies only when polling mode of operation was
selected by the core: if instead a completion IRQ was found to be available
the normal response processing path based on completions will still be
followed.
Link: https://lore.kernel.org/r/20211220195646.44498-4-cristian.marussi@arm.com
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Ludvig Pärsson