| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Upstream supports zstd compression now as an alternative to zlib or xz.
zstd is only slightly larger than xz but is significantly faster. This
makes it a better candidate for the journal. Enable it in systemd
instead of xz.
Add an exemption for witherspoon to continue using xz instead. Having
both zstd and xz support built into the image puts us over the flash
limit and we currently need xz for phosphor-debug-collector.
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>
Change-Id: I6fb55b96776aeeea1c544743b08bdbbc2853e7b7
|
|
Testing:
Built image for p10bmc platform, installed on a test system,
and verified netipmid was started on both eth0 and eth1
interfaces. Also, verified that ipmitool commands work as
expected for IP addresses configured on eth0 and eth1 interfaces
Signed-off-by: Shantappa Teekappanavar <sbteeks@yahoo.com>
Change-Id: Ie8399ba1219dac2f9c4ac15b7dfcf7e746750f0d
|
|
We've typically kept these LAYERSERIES_COMPAT to 2 releases: the current
and the upcoming. Remove 'gatesgarth' is it is now 2 releases back.
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>
Change-Id: I5e812a94fed1738898af75c0fdee81996a5bbf20
|
|
Use nonarch_base_libdir instead of base_libdir, because the files are
always in /lib/udev/rules.d/. On ppc64le base_libdir is lib64, and the
files end up in the wrong spot.
Signed-off-by: Anton Blanchard <anton@ozlabs.org>
Change-Id: I3c24dae293f4166f1ed1826d09d4bde7cd7d0357
|
|
Configure the SOCSEC_SIGN_* variables to sign the SPL and exploit the
AST2600 hardware root-of-trust.
Note that this doesn't require that secure-boot is enabled on the
system, the SoC will bootstrap just fine with the signature in place
while secure-boot is disabled. Signing the SPL allows us to switch the
systems over to secure-boot at our leisure.
Change-Id: I07b5c4afb7bacc040cbdce6c82a0fb3a57d0f7f8
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
The hostboot dump collection initiated by watchdog_timeout
is disabled by default. This commit enables hostboot dump
collection on p10bmc platforms for watchdog timeout.
Build watchdog_timeout binary:
$ devtool modify -n openpower-debug-collector <local-source-path>
$ MACHINE=p10bmc bitbake openpower-debug-collector
Test:
Ran watchdog_timeout binary built by bitbake, and verified that
the binary takes the timeout value and times out after the specified
time interval
Signed-off-by: Shantappa Teekappanavar <sbteeks@yahoo.com>
Change-Id: I343aac95fd97feb226d8676c3c34c2902c1c6690
|
|
Running out of flash space on our witherspoon system so need to cut
back some function. Telemetry is fairly new and takes a sizeable chunk
of flash space (200KB)
Tested:
- Verified squashfs went from 19.20MB to 19.00MB with this change
Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: I1741649f573cd25363167d69b4a802f2f261d93a
|
|
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>
Change-Id: Iadecbc8418d901f82fcc741d3e88d2d202fe96fe
|
|
Update with the latest version from upstream.
Change-Id: I1a7da37b0457dab873afaf6445aca360d54b47ca
Signed-off-by: Adriana Kobylak <anoo@us.ibm.com>
|
|
Change-Id: I9116ed7260e369136acb39eec15075db2d4dbeba
Signed-off-by: Adriana Kobylak <anoo@us.ibm.com>
|
|
This commit is add D-bus monitor config.yaml
for errorlog entries for snmp eventing.
Tested by:
Configured SNMP and created errorlog
observed snmp traps recieved on SNMP server
Signed-off-by: Ravi Teja <raviteja28031990@gmail.com>
Change-Id: I59c1f8a3b03eb3aab4a55c3cfb0bfb3e9e600a4d
|
|
This commit would override response time out value for a request
to 4.8 seconds for all the ibm machines.
We have seen in our debugging that in rare cases host takes close
to 5 seconds to respond to the pldm message.This value (4.8 seconds)
is in sync with what phyp maintains for their messages(5 seconds).
As we do lot of fileIO operations, and the number of PDR's tend to grow,
picking 4.8 seconds seems to be less risky path & also larger timeout
would reduce the number of retries from BMC incase if the host does not
respond within 2-3 seconds.
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>
Change-Id: Ic0eac42b66dbc2c6d69bff440fe590775f35e390
|
|
Background:
OpenBMC provisions the BMC firmware image with the root account password
in a form which is no longer acceptable to Linux-PAM version 1.5.1.
Specifically, [phosphor-defaults.inc][] sets the password hash into
/etc/shadow as "\$1\$UGMqyqdG\$FZiylVFmRRfl9Z0Ue8G7e/", where $1
indicates the deprecated [MD5 hash algorithm][]. Ref: [wikipedia passwd
entry][]. Beginning around PAM version 1.5.1, when you log in, the
[pam_unix.so module][] authenticates okay but requires the password to
be changed. (For example, you'll get a message like "You are required
to change your password immediately (administrator enforced)." This
behavior is undesirable for OpenBMC project defaults, and is not
tolerated by the project's current continuous integration tools.)
This change is to replace the password hash to keep the same cleartext
password but hashed with an acceptable algorithm.
Specifically, the password hash supplied in phosphor-defaults.inc is
updated to use the same password as before but encoded
with the SHA-512 algorithm. The hash was generated by the
`openssl passwd -6 0penBmc` command. This change ought to be
transparent and forward and backward compatible.
Note various meta-layers use this same hash string in
conf/local.conf.sample files. They are changed to match.
References:
[phosphor-defaults.inc]: https://github.com/openbmc/openbmc/blob/1a977b269ed437bebb9ae7810e3157746ec9174d/meta-phosphor/conf/distro/include/phosphor-defa
ults.inc#L245
[wikipedia passwd entry]: https://en.wikipedia.org/wiki/Passwd
[pam_unix.so module]: https://github.com/linux-pam/linux-pam/tree/master/modules/pam_unix
[MD5 hash algorithm]: https://en.wikipedia.org/wiki/MD5
Tested:
Created image with new password hash and PAM 1.5.1 and checked that
login works okay and does not require the passwod to be changed.
Signed-off-by: Joseph Reynolds <joseph-reynolds@charter.net>
Change-Id: I5b189374f08ba506dbed7f8b9b991f2808cc3bc5
|
|
Background: The OpenBmc project default root account password is set
in meta-phosphor/conf/distro/include/phosphor-defaults.inc and can be
customized in each layer's local.conf file.
Many of these local.conf.sample files had redundant code to set the
password, which probably should not have been there. Removing them
allows the defaults in phosphor-defaults.inc to take effect.
Tested: No. Only meta-ibm was tested.
Signed-off-by: Joseph Reynolds <joseph-reynolds@charter.net>
Change-Id: I76dce00d269d7afa005d7bcfd63f846d3cf45596
|
|
Signed-off-by: Sumit Kumar <sumit_kumar@in.ibm.com>
Change-Id: I62027c739eb4cf95bb1bfc8ccddec30964bb1906
|
|
The phosphor-gpio-presence application is no longer needed for power
supply presence detection after the updates to using libgpiod in the
phosphor-psu-monitor application.
Remove the systemd parts of the bbappend, and remove the obmc conf
files.
Signed-off-by: B. J. Wyman <bjwyman@gmail.com>
Change-Id: I18a7f36e21d18a22f7625aadc3229ce5439c8d6a
|
|
Change Ie1a7c389edb6b7a048836a49283ceb62de51bba5 will be transitioning
the 'Type' property in Inventory.Item.Chassis to an enumeration.
In order to avoid crashing in PIM we need these default values in the
starting YAML to be fully-qualified enumeration names that will match
the values which will start in Ie1a7c.
This code is safe to merge as is without any PDI or PIM changes. Prior
to the PDI change, we must make a change to PIM that allows conversion
automatic from string->enum, leveraging library interfaces available in
sdbusplus. These will be submitted independently.
I checked the codebase for usage of this string. It appears that the
value is currently, effectively, write-only. There is code in bmcweb
that fills in the equivalent Redfish value but currently just hard-codes
the string 'RackMount'.
Tested: Booted Witherspoon in a QEMU model with this change and proposed
changes to sdbusplus + PIM. PIM no longer coredumps with the PDI change
and yields an expected persistence file:
```
$ pwd
/var/lib/phosphor-inventory-manager/xyz/openbmc_project/inventory/system/chassis
$ cat xyz.openbmc_project.Inventory.Item.Chassis
{
"value0": {
"cereal_class_version": 2,
"Type": 3
}
}
```
Signed-off-by: Patrick Williams <patrick@stwcx.xyz>
Change-Id: Icaf9447f31ccdd945cdf74b3e017682e4aed686f
|
|
Change-Id: Iac6362f744113d9a405e4a4ec30077e513dda2d3
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
As for Rainier and Everest, use KCS2 as the debug-trigger interface.
Change-Id: I2614ffc3d97164658f2ad4fbb916cbf44be152e1
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
Currently required for the debug-trigger application.
Change-Id: I585ccc5591b3f66db41bfd9799df05cec902403d
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
Replaces the obsolete MCTP LPC driver.
Change-Id: Ic5f7f6ec796bff8f0628d53e09a7437a74399ef4
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
This sensor on the Op_Panel, one of them being a SI7020 type sensor
which needed to be enabled.
Signed-off-by: Bruce Mitchell <bruce.mitchell@linux.vnet.ibm.com>
Change-Id: Ib6a86a6fae52414ad9c6901e5f6f8aa8c8a383c3
|
|
Witherspoon uses a specialized phosphor-fan-control@.service file and
uses the phosphor-fan-control-init@.service file still to just set the
fans ready target. To correctly be included in the FILES directive,
these services need to be explicitly added on wiherspoon machines.
Change-Id: I76129a7833337aaf74fd83b0e28f3083f1a41b0a
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
|
|
- ddimm
- capi connector
- nvme
- pcie slot
- pcie cable card top and bottom enclosures
- base op panel
Signed-off-by: Lakshminarayana R. Kammath <lkammath@in.ibm.com>
Change-Id: I7de498f83ab3bace4ce39deb98c31c71c40f2dd4
|
|
MaxBootCycleCount is a dbus property that is used to indicate
the maximum number of boot cycles that the post-code-manager
daemon can store the progress codes for, before it starts
wrapping.
Unlike Intel postcodes, IBM progress codes are huge and it is
possible to get close to 230+ progress codes per boot cycle, so
tweaking this value to 5 boot cycles for IBM systems so that
we can get less number of log entries on a redfish GET request.
Signed-off-by: Manojkiran Eda <manojkiran.eda@gmail.com>
Change-Id: Ida10e73992b01535d57c844d83366d786cd448e2
|
|
The Altitude measurement is from the computation of
the DPS310 pressure sensor with this
"Expression": "44330.0 * (1.0 - ((P0/101325.0) ^ (1/5.255)))"
Tested: busctl introspect xyz.openbmc_project.VirtualSensor \
/xyz/openbmc_project/sensors/altitude/Altitude returned
results consistent with location of the Rainier system.
Signed-off-by: Bruce Mitchell <bruce.mitchell@linux.vnet.ibm.com>
Change-Id: I07b83dd6b8f4ccf0755ccbdc81c40976a3e571cd
|
|
Phosphor logging PEL infrastructure provide supports to
process SBE chip-op failures and creating PEL.
PEL creation required to process SBE provided FFDC packets.
This FFDC is initially created by SBE using FAPI based
infrastructure for the hardware procedure execution failures.
To extract this data to convert into PEL required format
requires additional processing. This processing required
PHAL(Power Hardware Abstraction layer) based back-end
support.
Added required packages support for PHAL feature enabled
systems.
Test: verified build.
Signed-off-by: Jayanth Othayoth <ojayanth@in.ibm.com>
Change-Id: I5cf1a30514aa0a5bfdf2fe3a288dac9d5a0b5b47
|
|
OpenBMC requires additional capabilities for the
OpenPOWER systems such as
- A common mechanism to access and control the host
hardware from the BMC.
- A storage mechanism to store the data that can be
used for initialize, control and access the CEC
hardware.
- An infrastructure to initialize the hardware using
the hardware procedures provided by the hardware team.
- Common mechanism to handle the hardware failures for
the host hardware.
These above features are essential for the BMC to provide
better RAS capability and to support lab debug for the
enterprise class systems.
The phal(POWER Hardware Abstraction Layer) aims to provide
a common layout such that the OpenPOWER application can
use this infrastructure to boot, enable RAS features like dump,
diagnostics, and lab debug tooling purpose.
The phal is a combination of open source packages that are
tailored/customised to use in OpenBMC. For instance POWER
hardware access and control is managed through pdbg and
host hardware data modelling is done via devicetree.
phal is group of libraries running in BMC. These libraries
are used by OpenPOWER specific application for host hardware
interactions, Hostboot and SBE initialization, diagnostics
and debugging.
Signed-off-by: Jayanth Othayoth <ojayanth@in.ibm.com>
Change-Id: I187d9582598743871279a1e407451f4cf4f72668
|
|
Install the udev files to cover the transition between drivers and
devices.
Change-Id: I87ffe3ae82bb7a073a8140adcb308b09fc43c9b9
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
libmctp provides the MCTP implementation for communication between the
host and the BMC.
Change-Id: I6ae5a23fbb34c27db16a152afefb79ca6a649759
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
These were separated in a strange fashion. Perhaps we should reorganise
the whole file in platform-specific sections.
Change-Id: I51b8bb20060447c6e88557824e9465b1713f082f
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
Currently /proc/sysrq-trigger is the only supported backend for
debug-trigger. Once we have kdump support, we'll add a kexec backend and
drop the magic sysrq.
Change-Id: I07dfa50474d9f6333428b6fb38020c7b7fd0adec
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
Change-Id: Ic1cde1cd67fa6f8f68df80a1f327ace64a39d72b
Signed-off-by: Andrew Jeffery <andrew@aj.id.au>
|
|
Change-Id: I4453f9b01f986a529c76710d665ab57639232b2d
Signed-off-by: Anton D. Kachalov <gmouse@google.com>
|
|
It's been found that the processors are generating more preheat than
anticipated and the rainier 4u systems currently being built may not
have the baffles available to be installed. Increasing the poweron fan
speed to overcome this preheat affect to the PCIE area so PCIE cards do
not overtemp.
Change-Id: I03414fc4cc9cd3aa8d248c4b1408bab658978f48
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
|
|
p10bmc systems have their own checkstop monitor, provided by the
openpower-hw-diags package. Having both installed on p10bmc systems is
wasteful and has shown some intermittent issues (checkstop-monitor
service core dumping when checkstops occur).
Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: I118b359bf551e6502086800a9081ca8856ecf4ae
|
|
This is apparently not actually working anymore and is removed in the
next poky update.
Change-Id: Ia1c6a258d124a4a30a14fc42e8e0bba95e64faeb
Signed-off-by: William A. Kennington III <wak@google.com>
|
|
The IBM host firmware team has requested an initial timeout of 10
minutes for the host watchdog. The goal is to bring this down once
bringup is complete.
When inband IPMI was used, there was a mechanism for the host to adjust
this timeout so the BMC firmware would set the initial 30 second timeout
and the host firmware would change it to 10 minutes once they started.
With PLDM there is no way for the host to adjust this timeout so we need
to start with the larger value.
Tested:
- Verified a p10bmc image had the new poweron file installed with 10
minute timeout.
Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: If94766d5f2d62271c4adf366631f0b31a75ef8d0
|
|
Signed-off-by: Chris Cain <cjcain@us.ibm.com>
Change-Id: I2ce6a862e0d16fe8a67c9dc12ceacbc01d3a0b1b
|
|
This commit adds associations for all the inventory that
gets represented as Assemblies in Redfish.
Change-Id: Id4e9e4ac870e5bed665462d161abde00c4b85554
Signed-off-by: Vishwanatha Subbanna <vishwa@linux.vnet.ibm.com>
|
|
The rainier machine has been replaced with the generic p10bmc.
Change-Id: Ia7b6e8b6ad569d6f607d4a38318704fb402f2a1f
Signed-off-by: Adriana Kobylak <anoo@us.ibm.com>
|
|
Use the 'insecure/imprint' development key to sign the p10bmc SPL. The
key can be overriden for a production key if necessary.
Signed-off-by: Klaus Heinrich Kiwi <klaus@linux.vnet.ibm.com>
Change-Id: I6e4abecb5859fb59c6185a097cf88bdcb958e207
|
|
Use the insecure key provided by commit '748d586bc1 meta-aspeed: Add
development key for Kernel sign' to sign both the Kernel as well as
U-Boot fitImages. This is used for U-Boot FIT Signature Verification
using a known key, fit for development purposes. For production
purposes, a secure private key must be used.
Signed-off-by: Klaus Heinrich Kiwi <klaus@linux.vnet.ibm.com>
Change-Id: If0c39f4aa17e6eaa5f6952a90283457f252a64d3
|
|
To support the new rainier 2U fan hardware, the fan monitor config needs
to be updated to reflect the factor/offset between the target and
feedback speeds of the new hardware.
Change-Id: Ic63d678e89b210da77d4681ad44a209dd2807836
Signed-off-by: Matthew Barth <msbarth@us.ibm.com>
|
|
The p10bmc system uses a MMC flash chip with much larger storage than
a NOR chip. Increase the current journal size allocation from 2MB to 16MB
so that p1bmc saves more data for debug.
Tested: Verified that a witherspoon and p10bmc image had the expected
journal size file (2MB for witherspoon and 16MB for p10bmc) under
rootfs/lib/systemd/journald.conf.d/
Change-Id: Idcdb01bbf017c411c7b97406191e90774e8a8753
Signed-off-by: Adriana Kobylak <anoo@us.ibm.com>
|
|
We want to use packageconfig options instead of build system specific
ones.
Change-Id: I587d33e3dae82c4c9d85f59f4ee34618443f1dcb
Signed-off-by: William A. Kennington III <wak@google.com>
|
|
Switch over to the 'count' method of determining fan faults. As a
starting point, use a count interval of 1 and a threshold of 30. This
is the preferred method as it is better at catching flakey fans.
See docs/monitor/method.md in the phosphor-fan-presence repo for more
info on how this works.
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I153d05779a5a7af854f351c1be3a3448b85fc461
|
|
The fan presence detect application uses I2C device paths in its config
files to address the IO expanders that IBM systems use to monitor fan
presence detects.
Use the /sys/bus/i2c/devices/... paths in these config files instead of
the absolute paths. These new paths are links to the actual paths, and
shouldn't change even if some internal kernel change drives an actual
device path change. There are other config files today that do this the
same way.
Signed-off-by: Matt Spinler <spinler@us.ibm.com>
Change-Id: I39ed9d703cf6bdc821a54c6d5d087f7fc080705b
|
|
This creates a new DISTRO_FEATURE "ibm-service-account-policy" which
- Adds an admin account which cannot SSH to the BMC's command shell.
- Adds a service account which can SSH and has passwordless sudo access.
This feature is applied to witherspoon-tacoma and p10bmc (rainier).
Tested:
The image behaves as before when the distro feature is not configured.
When the distro feature is configured:
The root user has the same access as before.
The admin user:
- Is not allowed to access the BMC's command shell.
- Console login gets: This account is currently not available.
- SSH login gets: Permission denied, please try again.
- Redfish and REST API access works with role=Administrator.
The service user:
- Console login to the BMC's command shell works. The home
directory is /. Passwordless sudo works.
- SSH login works and using sudo from a SSH session works.
- Redfish and REST API access works with role=Administrator.
Change-Id: Icac5ba7f4fa663047709ab55007bbcfec8158f5e
Signed-off-by: Joseph Reynolds <joseph-reynolds@charter.net>
|
|
Change-Id: Ib7a9802582e146b5e2e5edb7b2df1800dfe36b22
Signed-off-by: Vishwanatha Subbanna <vishwa@linux.vnet.ibm.com>
|
