- Apr 21, 2016
-
-
Evgeniy Borisov authored
The number of CSI TX lanes is hardcoded to 4. Removing this and start using value from configure stream response. NOTE: The patch depends on the CSI init change: "Use GB CSI params to init camera sub-devs" Signed-off-by:
Evgeniy Borisov <eborisov@mm-sol.com> Reviewed-by:
Gjorgji Rosikopulos <grosikopulos@mm-sol.com> Signed-off-by:
Greg Kroah-Hartman <gregkh@google.com>
-
David Lin authored
This change adds the AP Power Monitor functions to read out all the rails power information monitored by the SVC. Testing Done: - $ cat /d/greybus/1-svc/pwrmon/*/* and validate the output with the svc stub power monitor functions - $ tree /d/greybus/1-svc/pwrmon | | | |---pwrmon | | | | |---DUMMY_RAIL_1 | | | | | |---current_now | | | | | |---power_now | | | | | |---voltage_now | | | | |---DUMMY_RAIL_2 | | | | | |---current_now | | | | | |---power_now | | | | | |---voltage_now | | | | |---DUMMY_RAIL_3 | | | | | |---current_now | | | | | |---power_now | | | | | |---voltage_now | | | | |---DUMMY_RAIL_4 | | | | | |---current_now | | | | | |---power_now | | | | | |---voltage_now Signed-off-by:
David Lin <dtwlin@google.com> Reviewed-by:
Viresh Kumar <viresh.kumar@linaro.org> Signed-off-by:
-
David Lin authored
Read on apb_log causes null pointer exception due to the missing es2 device pointer passing to the debugfs. Testing done: - Enable apb_log and cat it Signed-off-by:
-
Axel Haslam authored
The check for line coding changed should use memcmp and not memcpy. Testing done: trivial Reviewed-by:
Johan Hovold <johan@hovoldconsulting.com> Signed-off-by:
Axel Haslam <ahaslam@baylibre.com> Signed-off-by:
-
Johan Hovold authored
Remove unnecessary interface-remove helper. Also add comment about why the disconnected flag is set. Signed-off-by:
-
Johan Hovold authored
Keep a detected interface registered until it is physically removed. Specifically, do not re-register an interface that is switching mode. Note that this also allows us to get rid of some nasty hacks from core. The Ara VID/PID bootrom hack for ES2 will continue to work, but is now mostly confined to the bootrom driver. Signed-off-by:
-
Johan Hovold authored
Make sure to deactivate all interfaces when the svc is going away. This is needed to eventually be able to do controlled teardown of the unipro network. Signed-off-by:
-
Johan Hovold authored
Make the control-device lifetime coincide with when the interface is enabled (enumerated). This is needed to be able register a new control device after a mode switch. Signed-off-by:
-
Johan Hovold authored
Return an error pointer when failing to create a control device. Signed-off-by:
-
Johan Hovold authored
The control device is an abstraction of the control connection over which a greybus manifest is retrieved. As interfaces switch modes (e.g. after boot-over-unipro) they expose new manifests, which can contain different vendor and product strings. Eventually control devices will be deregistered and recreated after an interface mode switch, while the interface itself remains registered. Note that only interfaces of type greybus will have control devices. Specifically, dummy interfaces will not. Signed-off-by:
-
Johan Hovold authored
Make sure to register also the control device along with any bundles when registering an interface. Note that we currently ignore failures to register the control device just as we do for bundle devices. Signed-off-by:
-
Johan Hovold authored
Make the control object be a greybus device. The control device will be used to expose attributes specific to greybus-type interfaces. Signed-off-by:
-
Johan Hovold authored
Move the timesync-operation functions above the control-object management functions, which is where all other operation implementations reside. Signed-off-by:
-
Johan Hovold authored
Disable and deactivate an interface immediately on registration failures. Signed-off-by:
-
Johan Hovold authored
Add control devices to the example sysfs tree. Control devices are named <bus_id>-<module_id>.<intf_id>.ctrl and expose attributes that are specific to the greybus interface type. Specifically, dummy interfaces do not have a control device. Currently, only the vendor and product strings extracted from the manifest are exported. A subtree of the example tree now looks as follows: greybus1/ ├── 1-5 │ ├── 1-5.5 │ │ ├── 1-5.5.2 │ │ │ ├── bundle_class │ │ │ ├── bundle_id │ │ │ └── state │ │ ├── 1-5.5.ctrl │ │ │ ├── product_string │ │ │ └── vendor_string │ │ ├── ddbl1_manufacturer_id │ │ ├── ddbl1_product_id │ │ ├── interface_id │ │ ├── product_id │ │ ├── serial_number │ │ └── vendor_id │ ├── 1-5.6 │ │ └── interface_id │ ├── eject │ ├── module_id │ └── num_interfaces └── 1-svc Signed-off-by:
-
Johan Hovold authored
The svc eject attribute was added as an interim solution and is still used to implement a form of forced ejection. This will soon be superseded by the module eject attribute, which will provide an interface for clean eject. We may keep the forced-eject mechanism around indefinitely, albeit possibly with a different name (e.g. forced_intf_eject). Either way, update the example tree to reflect the actual name, intf_eject, which currently used for this svc attribute. Signed-off-by:
-
Johan Hovold authored
Remove the interface unique_id attribute, which there is currently no plan to ever implement. Note that the Ara serial numbers are already exposed through the serial_number attribute. Signed-off-by:
-
Johan Hovold authored
Move the interface power attributes after the other interface attributes to keep the attributes grouped by device type. Signed-off-by:
-
Johan Hovold authored
The manifest-parsing code could end up leaving the interface vendor_string set to an error pointer that we'd eventually try to free when destroying the interface. Signed-off-by:
-
Rui Miguel Silva authored
When transfer speed is too slow (less than 17Khz) the operation can take longer than the default greybus timeout. Because of this we need to use the request_send_sync_timeout and calculate the correct timeout for each operation. Signed-off-by:
Rui Miguel Silva <rui.silva@linaro.org> Tested-by:
Philip Yang <philipy@bsquare.com> Tested-by:
-
Rui Miguel Silva authored
The actual implementation of transfer_one_message have problems with some cases in the possible transfer options. We try to maximize the number of spi transfers in one greybus operation and need to save state until the full message is dispatch over greybus. Signed-off-by:
-
- Apr 16, 2016
-
-
David Lin authored
SVC watchdog should be disabled when device is entering suspend mode. Testing done: - Sanity tested on EVT1.5 - Check no SVC ping during the suspend process - Check SVC watchdog is back on pinging once device is resumed Signed-off-by:
Vaibhav Hiremath <vaibhav.hiremath@linaro.org> [vaibhav.hiremath@linaro.org: Removed unwanted check in notifier callback and Updated commit description] Tested-by:
Sandeep Patil <sspatil@google.com> Signed-off-by:
-
David Lin authored
Switch off APBs/SVC/Switch in the suspend call notifier. Note that this is an interim solution to enable fishfooding. Testing done: - Passed QA sanity test on EVT1.5 - Suspend current measured at ~70mW Signed-off-by:
-
- Apr 13, 2016
-
-
Bryan O'Donoghue authored
A number of data in TimeSync command structures are declared __u64/__u32 instead of __le64/__le32, I forgot to put this through an x86_64 compile before presentation for merge and as a result didn't catch this error. This patch fixes. Signed-off-by:
Bryan O'Donoghue <bryan.odonoghue@linaro.org> Signed-off-by:
-
Vaibhav Hiremath authored
Earlier during boot sequence implementation, we had seen race between USb3613 and APB boot, and since APB boot time is ~2sec, we delayed USb3613 configuration for 2sec after APB deassertion of reset. This obviously won't work in the case of suspend/resume, where we would like to put APB into OFF state and coldboot in resume. With the latest FW changes, we do not see any race issue. I have done regression testing (> 50 iteration of reboot + unipro link up and down) without any issues. So lets get rid of the 2sec delay with this patch. Testing Done: Tested on EVT 1.5 platform. Signed-off-by:
-
- Apr 12, 2016
-
-
Evgeniy Borisov authored
Extending the configure streams interface with CSI params. Getting CSI frequency data form configure streams response. * num_lanes - Number of CSI data lanes * clk_freq - CSI clock frequency in Hz * lines_per_second - Total number of lines in a second of transmission (blanking included) From the AP side we need to know for the CSI speed configuration. This information is needed for dynamically bandwidth calculations. NOTE: Change should be along merged with corresponding interface change in kernel: "camera: Extend the configure streams interface with CSI params" Signed-off-by:
-
- Apr 08, 2016
-
-
David Lin authored
This change implements the AP Power Monitor functions for obtaining current/voltage/power on a specific rail of an Interface. Testing Done: $ cat /sys/bus/greybus/devices/1-3/current_now 103 $ cat /sys/bus/greybus/devices/1-3/power_now 303 $ cat /sys/bus/greybus/devices/1-3/voltage_now 203 Signed-off-by:
Bartosz Golaszewski <bgolaszewski@baylibre.com> Signed-off-by:
-
- Apr 05, 2016
-
-
Vaibhav Agarwal authored
Device type info shared to above HAL is currently hard coded to SPK only. Actual device type is identifed while parsing widget types from topology FW shared by codec module. Signed-off-by:
Vaibhav Agarwal <vaibhav.agarwal@linaro.org> Reviewed-by:
Mark Greer <mgreer@animalcreek.com> Signed-off-by:
-
Vaibhav Agarwal authored
Currently, single field is used to report device type say SPK, MIC, HS, HP, etc. However above HAL expects separate fields for input & ouput device types. Signed-off-by:
-
Gjorgji Rosikopulos authored
Add support for greybus metadata format. Greybus metadata format id is 0x41. Signed-off-by:
Laurent Pinchart <laurent.pinchart@linaro.org> Signed-off-by:
-
Viresh Kumar authored
Align with Greybus specifications and rename Firmware Protocol driver as Bootrom Protocol driver. Signed-off-by:
-
- Apr 03, 2016
-
-
Viresh Kumar authored
Don't know why, but checkpatch checks if we are running it from top of a kernel tree or not, but then it also provides an option to suppress the warning using --no-tree. Instead of forcing everyone to use this every time, lets make this behavior default. Signed-off-by:
-
- Apr 02, 2016
-
-
Akash Choudhari authored
This reverts commit a1d8f2c3856804ed26157104bb203edf4c882a6c.
-
Georgi Dobrev authored
Added a sysfs entry called pwr_off. When a "1" is passed to it, it sends a GB_SVC_TYPE_PWR_DOWN command to the SVC, powering it down along with the switch and INA231 chips. Testing Done: Tested on EVT1_5, works. Signed-off-by:Georgi Dobrev <dobrev_georgi@projectara.com>
-
- Apr 01, 2016
-
-
Greg Kroah-Hartman authored
They belong in a subdir. Signed-off-by: -
Greg Kroah-Hartman authored
For those who are stuck using old kernel trees, let's include the latest version of checkpatch.pl into our tree to help prevent coding style mistakes from creeping in. Also add spelling.txt to catch spelling errors in comments. Signed-off-by: -
Vaibhav Agarwal authored
Capture path related settings during startup, perpare & hwparams were earlier missing. Signed-off-by:
-
Vaibhav Agarwal authored
Register jack with ASoC sound card in case audio module populates it via codec FW. Currently, only a single jack with 4 buttons can be registered for each module. Signed-off-by:
-
Vaibhav Agarwal authored
widget pointer was incorrectly modfied while parsing kcontrol Signed-off-by:
-
Gjorgji Rosikopulos authored
No need to duplicate module ops on every registration. NOTE: Change should be along merged with: "msm: camera: Change gb_camera_module ops to pointer" Signed-off-by:
-
