Age | Commit message (Collapse) | Author | Files | Lines |
|
IBMBASE is only used for pointing at licenses...point at the
licenses in oe-core in meta/files/common-licenses instead. to match the
defacto convention used in other oe layers like meta-openembedded.
(From meta-ibm rev: cbbb0e5b4e44d831fce5daafd13272d401440f40)
Change-Id: Ib214f92b2c384e3eb2a1f53ecf4b21034438d001
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
(From meta-ibm rev: effd83bdc65d9ae8f9d000205f8d55e0c29557e4)
Change-Id: I5d8fb1cd15e0b0bf2c2360631be8f259f42be98f
Signed-off-by: Alvin Wang <alvinwang@msn.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Now the mihawk fan event will build error on openbmc master.
So I made some changes in order to solve the problem.
Tested: Verify fan control functionality of mihawk image
(From meta-ibm rev: 3e3ec10428cabe6917ea982159565024fcd21f9b)
Signed-off-by: Ben Pai <Ben_Pai@wistron.com>
Change-Id: I84aa7e8fd85dba15a5a84bbf7944bb286e5ba9ae
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Add mihawk's ipmi sensor and fru yaml.
Tested: Use ipmi command to get the sensor list and fru data.
(From meta-ibm rev: 05c9a2734de75271837313f0e1794c58df50e990)
Signed-off-by: Ben Pai <Ben_Pai@wistron.com>
Change-Id: Ic85af9b0cc1b44a3377fa919e63ab04dbff7a31c
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Witherspoon has only one LAN interface which is channel number 1.
The json file provided by the phosphor layer has channel information
for two LAN interfaces. The phosphor layer config file is modified to
mark the channel number 2 as reserved.
(From meta-ibm rev: 48dc78b5e8b6deba59e4aa200b8d88f10e277b69)
Change-Id: I4af0b7e682d2ec266831c7053926d1515f62c726
Signed-off-by: Tom Joseph <tomjoseph@in.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Enable this config for mihawk so the fan sensors are working, otherwise
hwmon will fail and exit when fans are not working.
Tested: Verify that the fan_tach sensors return -ETIMEOUT when host is
powered off and do not exit with failure.
(From meta-ibm rev: 2e39633c12f7914b1a4a398bbff9b467930bb731)
Signed-off-by: Ben Pai <Ben_Pai@wistron.com>
Change-Id: I0a7e881de5d5baa8d4db3cac9ca3cb46ec08faf4
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Most of mihawk's sensors are different from witherspoon, so I temporarily
added the mihawk part and did not change the witherspoon.
Add mihawk version of fan control and d-bus monitor.
Mihawk does not have a water cooled mode, so another independent folder is added.
Tested: 1.Shut down the system if more than three cores
have a temperature greater than 100 degrees Celcius.
2.Fan control support cpu core, dimm, Onboard.
3.Confirm that sensor is able to return normally via REST.
(From meta-ibm rev: 949f91431bba2f0139cc823d674a741f9590c885)
Signed-off-by: Ben Pai <Ben_Pai@wistron.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
Change-Id: Icc401f32c67b622f0f1448e1166112e02c135a8f
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
This extension will (when it's complete) create IBM's Platform Event
Logs (PELs) for every OpenBMC event log, as well as provide features
to exchange them with the host, package them for Redfish, etc.
(From meta-ibm rev: 625fc419ae882dd572a1045bdca3286b896a69fe)
Change-Id: Ia215ce221ac58037c0f8029362181dde8a94d42e
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
The contents of this up-stream include the following items
1.Basic function
2.Led settings
3.GPIO monitor (id button and power supply)
4.Linux dts patch.
(From meta-ibm rev: c53430a2234917b71f29d7e61d6aea749c45a2fc)
Change-Id: Id0e47fd9197c8b2a70cbc94888bd572255664cdd
Signed-off-by: Ben Pai <Ben_Pai@wistron.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
A recent kernel update added a second input current/voltage/power
reading to the hwmon sysfs attributes for the ir35221 VRMs, and it threw
off the labels in the config files since they were hardcoded for the
specific previous numbers. For example, curr2_input previously
represented an output current but now represents an input current.
Fix this by using the LABEL mode in hwmon to find which sysfs file to
use for a particular sensor by the contents of its corresponding label
file. For example, the p0_vdd_current sensor should look at the sysfs
file that has a corresponding 'iout1' value in the currX_label file.
Tested: Test that the voltages are back within thresholds again and not
showing values around 12V.
(From meta-ibm rev: bca7ed0127f9e82895fb60b66b325347970f0860)
Change-Id: I9f28c9445daf9ce6206ff9e38e271ac805768fae
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Aatir Manzur (1):
Power supply input faults resolution
(From meta-ibm rev: bc9be1c977fcea7543aab6f2ffd51372bb3e805f)
Change-Id: I46d82875b474abe1af478e8bbfd7d1141d2d5c3d
Signed-off-by: Andrew Geissler <openbmcbump-github@yahoo.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
BMCWeb uses ObjectMapper associations to find sensor information for
Redfish. Due to BMCWeb enhancements for the Power and SensorCollection
schemas, the following Witherspoon association changes are required:
* Remove the "chassis" <-> "all_sensors" association for power supplies.
Power supplies were considered chassis by BMCWeb as a short-term
solution, but they have been moved to the Power schema.
* Add the power supply sensors to the "chassis" <-> "all_sensors"
association for the top level chassis.
* Add other missing power sensors to the "chassis" <->
"all_sensors" association for the top level chassis. These
associations were temporarily removed due to limitations in the BMCWeb
support for power sensors. These limitations have been addressed.
See https://github.com/openbmc/docs/blob/master/sensor-architecture.md
for more information on sensor associations used by BMCWeb.
(From meta-ibm rev: a1c35f612f58e5453fe988d59db21a8673fdcfa4)
Change-Id: Ibea5bdea58711e96faff4937ecccedcc695d9ba4
Signed-off-by: Shawn McCarney <shawnmm@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
As on Witherspoon, some VRMs on Swift need their AVS buses enabled
as part of the power on sequence. The AVS bus connects those VRMs
to the processors so the processors can set voltages.
On Witherspoon, the sequence was
1. Disable the AVS bus
2. Manually set VRM voltages
3. Enable AVS bus
Currently on Swift there is no need to manually set voltages, so
the AVS bus disable service and script isn't required.
Tested: Tested on swift, and ensured both witherspoon and
witherspoon-128 still have all avs scripts and
services installed.
(From meta-ibm rev: f1ceece0b75a4522ef3089319b12d22fd83ed9ef)
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: Ifd7ebdd5ec3dcdfef97c04e666d83fcb4991b452
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Matt Spinler (1):
Fix association path when resolving input faults
(From meta-ibm rev: aa31db48bd98f4d22b8b0b6a6316aefcd4bc5daa)
Change-Id: I3813ee4174d6aeebc6c7014d4b7d059e65b9ad99
Signed-off-by: Andrew Geissler <openbmcbump-github@yahoo.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
The Swift MRW has been updated and this commit utilizes some unused
sensors for Swift starting at 0xFE and counting backwards
(From meta-ibm rev: df212c93689bd73f0e760ed3cfffada6fa6b9ffa)
Change-Id: Ica981d322f51fec94c0db94884938a60b637eb73
Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Matt Spinler (2):
Update .gitignore
Don't try to read GPIOs when no device path
(From meta-ibm rev: adf123779de48ee17f2d8c53f4e3d865e575c2f9)
Change-Id: Ia97046f0d856266907831be9fc552d6d750d14f9
Signed-off-by: Andrew Geissler <openbmcbump-github@yahoo.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Use an override in bbappends in meta-witherspoon as described and
recommended in the "following best practices when creating layers"
section of the the Yocto development tasks manual.
ibm-ac-server represents common machine configuration shared by all of
IBMs "Accelerated Compute" class of servers. Apply it wherever any
variables are un-bounded by any override.
This enables more flexible use of meta-witherspoon in build
configurations, and easy addition of machines to the layer.
(From meta-ibm rev: bb0e1d0699a4b9905c053c94af7d882cd71779fc)
Change-Id: I82f7338b0d7a64b5e7a67e19bcdff07f48a332d4
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
(From meta-ibm rev: 8f7ff601395419a933eb0506d5cc07fd805f7708)
Change-Id: Iffb2c550511b4024b721c2ba81aec1f909ce7703
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Using a distro feature here might be a case of too much abstraction.
Instead, just enable them because that is what is desired for those
systems. This is slightly easier to comprehend.
Revisit, if at a later time, several phosphor-software-manager users
enable signature verification and file syncing when using UBI (making
the abstraction useful). If/when that occurs, those configure options
and packages should be enabled via the distro feature in the base
recipe and not in a bbappend.
(From meta-ibm rev: 50eebe208babfe35ccd5a3ed8d560669381e82a5)
Change-Id: Ic4ac7f89eeb4575e6dabb68bb61aa55835316864
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
This commit adds the xyz.openbmc_project.Inventory.Item.Accelerator
interface to IPMI GPU functional sensors.
This was something done earlier by commit
f510f7b6b708b22db1f08faf369994f01edd0ddc.
The same is being ported now to the new sensor config YAML.
Tested:
Loaded this change up on a witherspoon. Verified that the
inventory objects created by host-ipmid now implement the
xyz.openbmc_project.Inventory.Item.Accelerator interface.
busctl call xyz.openbmc_project.ObjectMapper /xyz/openbmc_project/object_mapper
xyz.openbmc_project.ObjectMapper GetSubTree sias
"/xyz/openbmc_project/inventory" 0 1
"xyz.openbmc_project.Inventory.Item.Accelerator"
a{sa{sas}} 6
"/xyz/openbmc_project/inventory/system/chassis/motherboard/gv100card0"
....
....
(From meta-ibm rev: 2b99559bff775f719fb79a5a770b5267bbbd6dcc)
Signed-off-by: Santosh Puranik <santosh.puranik@in.ibm.com>
Change-Id: Ic132c417aca171e23f5d72705461fe3378e194b0
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
There are two directory names in the path that are different based
on the I2C bus.
(From meta-ibm rev: d4f13c672063f32ce6aafdaaf63ce34e79f7a4b5)
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I37288264ff5098f75780c243b47f46f312f98854
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Use typical bitbake metadata style and use double instead of single
quotes, and put the append operator prior to the override.
(From meta-ibm rev: 5670481726ee9e986090498e921f446ed61f1b84)
Change-Id: I744a54540b671b534f1eb1cac0a87e1891c782be
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Set the correct (noarch) package arch on a couple of
shell-script-providing packages.
(From meta-ibm rev: d2cb8b6c5c50794cdd97435d01fe9147010f89cd)
Change-Id: I5de4911460c141fef2200efa1f80c3f0b3a6c3dd
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Hard-coded values for fan tach sensor thresholds don't work because
these values vary. For accurate reporting, Redfish sensor output looks
at the functional status on the fan inventory.
Tested:
Same config files changed under /etc and change verified in the webUI
(From meta-ibm rev: 13ed7b8ed73d71206f6992a4c71294280c451446)
Signed-off-by: Aatir Manzur <aatrapps@gmail.com>
Change-Id: I80df620662d16cb677acbfe311bd2e709ae5976f
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Configure fan monitor to monitor the 4 single rotor chassis fans using
the same config parameters as witherspoon. These parameters will likely
change once complete hardware arrives and is thermally characterized.
Tested:
Built swift fan monitor with correctly generated config
Built witherspoon fan monitor resulting in no config change
(From meta-ibm rev: a7f9cd542edc7f09192ebe30ba6c286b1a01f578)
Change-Id: Iee6eccd75ead287db159b52c7f87a0d3a89e8c99
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Configure fan presence for the 4 single rotor chassis fans to use gpio
or tach feedback presence detection. For the internal boost fan,
currently only configure the use of gpio for presence detection. The
requirement of the boost fan is to assist in cooling specific card
locations when necessary and may be updated later to include presence
detection by tach feedback.
Tested:
Built swift fan presence with correctly generated config
Built witherspoon fan presence resulting in no config change
(From meta-ibm rev: 62395b7eb55fd36aa05bc8ed4593aea78c2e2dea)
Change-Id: I10d0bcf9311926cacb1c5310b80a3bc0f00ed782
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
first-boot-set-hostname service will set a unique hostname
to the machine on the very first boot. The idea of this
commit is to enable this service as a part of witherspoon
image.
Tested by:
1. Enable first-boot-set-hostname.service
- systemctl enable first-boot-set-hostname.service
2. Reboot the machine
- reboot
Post reboot, hostname will be changed to hostname-serialno
(From meta-ibm rev: 6c8fb8fd38fd7f627657268d64967474cdd27653)
Signed-off-by: asmithakarun <asmithkr@in.ibm.com>
Change-Id: I77ea7c8ddd8f0e0ce425f77262aea22712761769
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Refactor the config yaml to be machine specific
The only rule is the values in this file not conflict with the FRU_ID
fields in the Swift MRW. Swift MRU has FRU_ID values that go into the
50s so just add 10 to all values in this file.
(From meta-ibm rev: 89af26a2ca92f79493a652ae279a6c48a67d1f7f)
Change-Id: I4a2e3ca95791d0ec2201b027894833a7ef7b4cfb
Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
The Swift MRW does a few things different here:
- Renames these from gv100 to ga100
- Starts their FRU_ID at 15
- Only has 4 vs. 6 of them
(From meta-ibm rev: 91a87c8b0288f235452b3073e24a5519020553e8)
Change-Id: I0af55e2d0bf5e317dfae45b4e53810d3b6d8897c
Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
With updates from openbmc/openbmc#2911, event actions are now able to be
configured with groups subsets and event timers are now included in the
available event triggers. Actions with group subsets allow a single
event's set of triggers to run actions against a given group subset
instead of the entire event groups. Also, with timers being included as
a trigger, events can be configured with or without a timer instead of a
disabled timer being created for events that previously did not require
a timer.
Tested:
Generated code is functionally equivalent to previous yaml
Verified fan control functionality of wspoon image
(From meta-ibm rev: 1a43941fb100db7c46430f0dbbc82799276e9c13)
Change-Id: Ide4167ef1a202381b702b7f6fe58b7f30501ecb3
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
This allows upstream to drop some backward-incompatible changes while
the yaml in meta-ibm gets updated to the new format coming in from
upstream.
(From meta-ibm rev: 81ca4572471629b5114a52afb50a84d9e2cc92d1)
Change-Id: Ia554f14c624d236e02d7898d2f95e7a90384dadf
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
To better organize the policies(and the configurations) that are used by
phosphor-dbus-monitor, all of them are now located under the `dbus`
directory. This moves toward a more understandable and centralized
scheme for the recipe(s) used to configure and build
phosphor-dbus-monitor.
(From meta-ibm rev: 4b80c4f2b3a62d8084fc43f95177e299d53387ae)
Change-Id: Ie99c5dbe55c2295fc5cd7b5e7722cc10064e1b71
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Swift has two power supplies, same as Witherspoon, therefore will use
the same power supply policy configuration.
Swift and Witherspoon will also share the event policy configuration for
OCC throttling.
The thermal policy configuration will be shared between Swift and
Witherspoon, but may change for air cooled Swift to not also shutdown at
3 or more cores over 115C.
The fan policy configurations will be different for Swift, therefore
they will be separate. At this time, there are some areas unknown on how
fan presence and functional states will be handled for both air and
water cooled Swift machines, so these are configured to be the same.
Tested:
Built witherspoon phosphor-dbus-monitor resulting in no image change
Built swift phosphor-dbus-monitor resulting in correct policies
(From meta-ibm rev: fd06a2cbfd5f6e43aba4887ae1398984410d52cc)
Change-Id: I1125f00ecd51c23aac2da4cd2a47432ac3bc2de7
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Policy configurations will be loaded by machine within the system meta
layer (in this case meta-witherspoon). This allows different machines to
have different policies and/or different policy configurations.
The .bbappend can be used to alter which policies are used on a machine.
The thought here being that most machines within the same system meta
layer would use the same policies. However these policies would likely
have different configurations, this is where each machine would have its
own policy configuration which would be loaded per the policy recipe.
Tested:
Built witherspoon phosphor-dbus-monitor resulting in no image change
Built swift phosphor-dbus-monitor where no polices exist now
(From meta-ibm rev: 6275fda44b3fd1a039f6aef55e920b542b8135ef)
Change-Id: I9c86e9eb15756236e1e802f18345fb8c7ab5400a
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Remove `witherspoon` from recipe file names, which includes the
directory containing the policy configuration files.
Tested:
Built witherspoon phosphor-dbus-monitor resulting in no image change
(From meta-ibm rev: c8562bc0066f8657d2fbb5c1962826e686a7f652)
Change-Id: I7d7d247865bc2f8432e01a46caa43cc8770ae100
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
The native recipes are no longer used after the switch to non-native
recipes in the meta-phosphor layer.
Tested:
Built witherspoon phosphor-dbus-monitor resulting in no image change
Built swift phosphor-dbus-monitor resulting in no image change
(From meta-ibm rev: c50b717ce219342c4730d3f029712d5a7d10da07)
Change-Id: I0026c06592921d0f1633bd3a24e6cd0530d3cc19
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Create the phosphor-dbus-monitor non-native recipes to allow the switch
from native to non-native within the meta-ibm layer. Upon switching to
the non-native recipe in meta-phosphor, these non-native recipes will
then replace the native recipes in providing the policy configuration
files(yaml).
Tested:
Built witherspoon phosphor-dbus-monitor resulting in no image change
(From meta-ibm rev: 0ca32c6b6d0566a29756ecd05ffb748764296a27)
Change-Id: Idf800c0d537b6051c3bc263908d9bbb2bc6ab2de
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
This sensor is needed under the chassis association so that bmcweb can report
the PowerControl/PowerConsumedWatts Redfish property.
(From meta-ibm rev: bc0b221dda4f9aad8589469066d8f075b69789f0)
Signed-off-by: Eddie James <eajames@linux.ibm.com>
Change-Id: I9a5daf64e9888a5b7c817fab59658a054de25a83
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
BMCWeb uses ObjectMapper associations to find sensor information for
Redfish. The names of these associations have changed. Updating the
Witherspoon association names to match the new naming scheme.
See https://gerrit.openbmc-project.xyz/c/openbmc/docs/+/22976 for more
information on the new sensor association names.
(From meta-ibm rev: 0c2b3acd905de8d8547139075ecc034f213636d9)
Change-Id: Iccb0f3d4f75d8716447cc4bb488cf268374309c1
Signed-off-by: Shawn McCarney <shawnmm@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
avsbus-control provides scripts, so set the package arch appropriately.
(From meta-ibm rev: 3168dfa433d1ecda4fb41612942951d06d075157)
Change-Id: Ic136d2cdf03019b69da45062057866bad75c1085
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
A prior patch added a new recipe acx22-yaml-config to provide all the
YAML configuration for ACx22 systems. This patch switches the layer to
that recipe by setting EXTRA_OECONF and DEPENDS as appropriate in the
required bbappends.
With this switch, a number of recipes and their provided YAML are now
dead metadata, so remove those.
(From meta-ibm rev: 3828b0070fd01e948fea02545da2bd91f96add36)
Change-Id: I723320ed88cf2fb7cbc063e4f41637fc98a5787e
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Swift doesn't need to run the vrm-control.sh script to set VRM
voltages, which also means it doesn't need to disable/enable their
AVS buses around it, nor bind/unbind the VRM device drivers.
Remove all of these services, and also create a new recipe for
power-workarounds.sh that was in the avsbus-control recipe so that
it can remain.
(From meta-ibm rev: 65ced2221d38ce08dd7dcc46ec254b1bdd4a714f)
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: Ib64cc45499292ae84689fb11400ff9ff0f5f5942
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Since the power supply I2C addresses are swapped between the 2 systems,
there needs to be separate config files.
(From meta-ibm rev: 7b91758259c860052d5f292a86a56c0d18884616)
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I3eb8ec881a7e4d7887c79a53ef142ee2f948b53f
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Fan presence config files reside under the machine name directory within
`phosphor-fan-presence-config`.
Tested:
Verified config.yaml picked up for witherspoon machine
Verified config.yaml picked up for swift machine
(From meta-ibm rev: 86142b4ce89084bf3310e3c9cbebc32a9c17afd2)
Change-Id: I5b379d593394218c9149cf1648f3288b37f6f96a
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Fan monitor config files reside under the machine name directory within
`phosphor-fan-monitor-config`.
Tested:
Verified monitor.yaml picked up for witherspoon machine
Verified monitor.yaml picked up for swift machine
(From meta-ibm rev: 470fe1f5de23711c3348d17638e65c106f54304f)
Change-Id: I87065dc923243e05b4ddbc54b0c1290617cb16f0
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
YAML configuration files exist scattered throughout the OpenBMC tree and
how they are used is controlled with layers dependencies of virtuals and
preferred providers.
Most of the time the above scheme is very difficult to comprehend. This
patch continues a re-thinking of that approach towards a more
centralized scheme.
Specifically this patch implements a single YAML config recipe for the
ACx22 systems. The logic contained in the recipe was pulled from
all over the OpenBMC tree - the ability to comprehend how the different
YAML files are generated and consumed should be greatly eased.
A couple notable details:
- Specialized tasks for doing deep YAML dictionary merges existed in
the upstream recipes (-hardcoded.yaml, -config.yaml), but it was
observed that simply concatenating the YAML files produced the same
results, so the specialized deep merge logic was not copied here and can
be deprecated upstream.
- Unlike the upstream recipes, acx22-yaml-config is a target recipe and
as such enables MACHINE based overrides.
YAML files were copied from different locations in the tree, and run
through a styling application (pyyaml dump(load(yaml))):
phosphor-ipmi-fru-read-bmc-inventory-native:bmc-fru-config.yaml ->
acx22-yaml-config:acx22-ipmi-fru-bmc.yaml
phosphor-ipmi-fru-read-not-sent-by-host-inventory-native:fru-config-not-sent-by-host.yaml ->
acx22-yaml-config:acx22-ipmi-fru-not-sent-by-host.yaml
phosphor-ipmi-host:occ_sensors.hardcoded.yaml ->
acx22-yaml-config:acx22-ipmi-occ-sensors.yaml
phosphor-ipmi-host:hwmon_sensors.hardcoded.yaml ->
acx22-yaml-config:acx22-ipmi-hwmon-sensors.yaml
phosphor-ipmi-host:witherspoon_hwmon_sensors.hardcoded.yaml ->
acx22-yaml-config:ac922-ipmi-extra-hwmon-sensors.yaml
phosphor-ipmi-sensor-inventory-mrw-config-native:config.yaml ->
acx22-yaml-config:acx22-ipmi-sensors-mrw.yaml
(From meta-ibm rev: 3cb4a62140dca941e801026f447f13374cf68968)
Change-Id: Id0584812bbf40c0118c8c4383d9dd67c38a45dec
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Initially configure Swift's power supply sensors as they were on
Witherspoon. Keep the config files separate between systems as they will
be different due to differences between the power supplies used.
Tested:
Built Swift image that contained the config files
Built Witherspoon image that still contained the config files
(From meta-ibm rev: d8d27673f1b277db83859408842619ae24606235)
Change-Id: I64d893c6c5dbea0e4d4a592042b17d6bad5b4596
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
Create Swift's OCC sensors initially using the same thresholds as they
were on Witherspoon. Swift has 12 dimms/proc with 2 temp sensors per
dimm. Also there are 2 GPUs/proc.
Tested:
Built image that contained the config files
(From meta-ibm rev: 2bd366ab4bd4b9e085b70d46f07f5f1acdd35790)
Change-Id: Ia381c561f283c4d542e0035534375fa14c493278
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
The top supply, PS0, is 0x68, and PS1 is 0x69.
(From meta-ibm rev: c4df414e0284f1859ff1c709eb84eace6d848e87)
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I4a9f38830f86b18932a918119bafaf1f5bd5d72a
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|
|
More than 1 directory in the path is dependent on the I2C bus.
(From meta-ibm rev: dc4b6022d75a5ed50588c71f1890f5cb7df0dcd4)
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: Iec2026ab941e735e684b939bca40482a7c603150
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>
|