build: update clang-format

Update to the latest OpenBMC clang-format.

Change-Id: I19424fe32f32900d21272e343a5bda8675edbe60
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

meson: add checks, defaults

Default to debugoptimized builds instead of the meson default of debug
since that is the autotools default.

Anything other than c++17 is not supported, so error out if s

meson: add checks, defaults

Default to debugoptimized builds instead of the meson default of debug
since that is the autotools default.

Anything other than c++17 is not supported, so error out if something
else is selected by the user.

Make tiny builds even smaller by removing asserts and remove asserts for
release builds.

Change-Id: I034438263b1125c476dccb377dc52142415dddd2
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

show more ...

meson: fix systemd service bugs

The logic for ENABLE_PHAL_TRUE was reversed under meson.

prefix was omitted when computing the path for the bindir replacement
token in the system unit files.

Fixes

meson: fix systemd service bugs

The logic for ENABLE_PHAL_TRUE was reversed under meson.

prefix was omitted when computing the path for the bindir replacement
token in the system unit files.

Fixes: 919e408aaad453f94e1ff3d35440b8fc8a2f3e13
Change-Id: I096d1f69554b23c20bd4836b36ffa5656a219333
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

show more ...

meson: install executables

This was missed in the initial meson port.

Change-Id: I71a27cfe16bfff38c74bb82169e57f6e729bc827
Fixes: 919e408aaad453f94e1ff3d35440b8fc8a2f3e13
Signed-off-by: Brad Bishop

meson: install executables

This was missed in the initial meson port.

Change-Id: I71a27cfe16bfff38c74bb82169e57f6e729bc827
Fixes: 919e408aaad453f94e1ff3d35440b8fc8a2f3e13
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

show more ...

test: remove gtest subproject fallback

We don't have a gtest subproject, so remove the logic to fallback on
one. This is a copy/paste mistake.

Change-Id: I06b7361f0b0f93f9358d4f0615bdb9d2130fa92d

test: remove gtest subproject fallback

We don't have a gtest subproject, so remove the logic to fallback on
one. This is a copy/paste mistake.

Change-Id: I06b7361f0b0f93f9358d4f0615bdb9d2130fa92d
Fixes: 919e408aaad453f94e1ff3d35440b8fc8a2f3e13
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

show more ...

test: Makefile.am: Remove -lstdc++fs

56d14d04 dropped experimental::filesystem and removed one
-lstdc++fs, this removes the other.

Tested: Built the repo.
Change-Id: I085ad483b317b0018e45001d6d1274

test: Makefile.am: Remove -lstdc++fs

56d14d04 dropped experimental::filesystem and removed one
-lstdc++fs, this removes the other.

Tested: Built the repo.
Change-Id: I085ad483b317b0018e45001d6d12740ec662d811
Signed-off-by: Gunnar Mills <gmills@us.ibm.com>

show more ...

host-stop: ensure stop-instructions run in correct seq

The flow of synchronization targets during a obmc-host-stop@.target is
the following:
obmc-host-stop-pre -> obmc-host-stopping -> obmc-host-sto

host-stop: ensure stop-instructions run in correct seq

The flow of synchronization targets during a obmc-host-stop@.target is
the following:
obmc-host-stop-pre -> obmc-host-stopping -> obmc-host-stopped

The obmc-host-stop-pre is where we notify the host that a shutdown is
occurring and allow it to gracefully shut itself down. Once that is
complete we run obmc-host-stopping services (i.e. things the BMC does to
actually stop the host like stop-instructions).

Move the op-stop-instructions@.service to the correct synchronization
targets and remove the soft off service dependency since that will now
be enforced via the correct synchronization targets.

Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: I9b4c36b44eef2dab38ad1ab26eb10ee7e1b9a6dc

show more ...

build: add support for meson

Change-Id: I5a76281cc6060afffc1d77e52f03acb34669856c
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

fix compiler warnings

Fix a couple pedantic, format=, and unused-parameter warnings so we can
turn the warnings all the way up.

Change-Id: I7c6309e30147f54fc64ace99dc784fbd96c77c70
Signed-off-by: B

fix compiler warnings

Fix a couple pedantic, format=, and unused-parameter warnings so we can
turn the warnings all the way up.

Change-Id: I7c6309e30147f54fc64ace99dc784fbd96c77c70
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

show more ...

c++17: drop experimental::filesystem

Use std::filesystem, and drop support for building with experimental
under c++14.

Change-Id: I32755f0d1b119c041978c2f8b988a16a14ba88d4
Signed-off-by: Brad Bisho

c++17: drop experimental::filesystem

Use std::filesystem, and drop support for building with experimental
under c++14.

Change-Id: I32755f0d1b119c041978c2f8b988a16a14ba88d4
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

show more ...

phal: use unix line breaks

OpenBMC is a Linux distribution, so use unix newlines instead of dos.

Change-Id: Ic26fa702baa8c80b901011e0606c3e7ee1409f61
Signed-off-by: Brad Bishop <bradleyb@fuzziesqui

phal: use unix line breaks

OpenBMC is a Linux distribution, so use unix newlines instead of dos.

Change-Id: Ic26fa702baa8c80b901011e0606c3e7ee1409f61
Signed-off-by: Brad Bishop <bradleyb@fuzziesquirrel.com>

show more ...

phal: Added HW callout info into PEL

- If any hardware procedure (hwp) failed during execution of ipling then
that particular hwp will request the required hardware's (which are defined
error xm

phal: Added HW callout info into PEL

- If any hardware procedure (hwp) failed during execution of ipling then
that particular hwp will request the required hardware's (which are defined
error xml of respective hwp) need to be callout. So, this patch will create
PEL records based on given FFDC (First Failure Data Collection) by phal
during failure.

- Also, this patch will deconfigure (marking hardware as non-functional
i.e update HWAS_STATE attribute with respective bitfield value in POWER CEC
device tree) required hardwares which are given as part of phal ffdc data.

- Used "libfmt" to format trace with field value instead adding as entry
in journal so, while debugging it will be easy instead looking each entry.

TestedBy:

"User Data 1": {
"Section Version": "1",
"Sub-section type": "1",
"Created by": "0x2000",
"HWP_CDG_TGT_01_CO_PRIORITY": "MEDIUM",
"HWP_CDG_TGT_01_CO_REQ": "true",
"HWP_CDG_TGT_01_DECONF_REQ": "false",
"HWP_CDG_TGT_01_GUARD_REQ": "false",
"HWP_CDG_TGT_01_GUARD_TYPE": "GARD_Fatal",
"HWP_CDG_TGT_01_LOC_CODE": "Ufcs-P0-C15",
"HWP_CDG_TGT_01_PHYS_PATH": "physical:sys-0/node-0/proc-0",
"HWP_FFDC_ATTR_CP_REFCLOCK_SELECT_VALUE": "00",
"HWP_FFDC_RCS_CLOCK_TEST_IN": "01",
"HWP_FFDC_READ_SNS1LTH": "0c000000",
"HWP_HW_CO_01_CLK_POS": "0",
"HWP_HW_CO_01_HW_ID": "PROC_REF_CLOCK",
"HWP_HW_CO_01_LOC_CODE": "Ufcs-P0-C15",
"HWP_HW_CO_01_PHYS_PATH": "physical:sys-0/node-0/proc-0",
"HWP_HW_CO_01_PRIORITY": "HIGH",
"HWP_RC": "RC_CLOCK_TEST_OUT_RCS_ERR",
"HWP_RC_DESC": "One or more processor reference clocks not toggling",
"LOG000 2020-09-22 04:12:58": "Running p10_clock_test HWP on processor 0\n",
"LOG001 2020-09-22 04:12:58": "inf: p10_clock_test:60
p10_clock_test: Entering ...\n",
"LOG002 2020-09-22 04:12:58": "inf: p10_clock_test_latches:106
p10_clock_test_latches: Entering
...\n",
"LOG003 2020-09-22 04:12:58": "err: CLOCK_TEST_OUT_RCS_ERR:4036
One or more processor reference clocks not toggling\n",
"LOG004 2020-09-22 04:12:58": "err: _setHwpError:120 _setHwpError:
Creating HWP error 0x927e1c\n",
"LOG005 2020-09-22 04:12:58": "inf: addErrorInfo:344
addErrorInfo: Adding cdg (1:0:0), pri: 1, gtyp: 227\n",
"LOG006 2020-09-22 04:12:58": "err: p10_clock_test_latches:130 Clock A is bad\n",
"LOG007 2020-09-22 04:12:58": "err: p10_clock_test:76
p10_clock_test_latches(i_target_chip, true) failed.\n",
"LOG008 2020-09-22 04:12:58": "HWP clock_test failed on proc 0, rc=2124995320\n",
"_PID": "1209"
},
"User Data 2": {
"Section Version": "1",
"Sub-section type": "1",
"Created by": "0x2000",
"Data": [
{
"LocationCode": "Ufcs-P0-C15",
"MRUs": [
{
"ID": 65536,
"Priority": "H"
}
],
"Priority": "H"
},
{
"Deconfigured": false,
"Guarded": false,
"LocationCode": "Ufcs-P0-C15",
"MRUs": [
{
"ID": 65536,
"Priority": "M"
}
],
"Priority": "M"
}
]
}

Signed-off-by: Ramesh Iyyar <rameshi1@in.ibm.com>
Change-Id: I4462c46747598166b085990e1d82cb2998b9e06b

show more ...

mpreboot: do not run stop instructions

The MPIPL path will stop instructions by default so there is no need to
run the service.

There are actually some issues within the SBE if instructions are
sto

mpreboot: do not run stop instructions

The MPIPL path will stop instructions by default so there is no need to
run the service.

There are actually some issues within the SBE if instructions are
stopped before it is told to enter MPIPL.

Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: If4aeaa7c4aac6a74ad5d3d8aa90d693a17c6c1f9

show more ...

mpreboot: ensure started after host stopped

Some race conditions have been hit where the enterMpReboot command is
processed before the host is fully stopped.

Change the service dependencies to ensu

mpreboot: ensure started after host stopped

Some race conditions have been hit where the enterMpReboot command is
processed before the host is fully stopped.

Change the service dependencies to ensure it is not started until the
host is fully stopped while also ensuring it is executed prior to the
host reboot service starting.

Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: I7bdc2ce407f2212a7e856e6dce275b553c8ec0ab

show more ...

cfam-reset: do not run if host is up

The cfam reset service should not be run in situations where the host is
already running. For example, if the system is booted and the BMC is
rebooted, this serv

cfam-reset: do not run if host is up

The cfam reset service should not be run in situations where the host is
already running. For example, if the system is booted and the BMC is
rebooted, this service should not run. Utilize the
/run/openbmc/host@%i-on feature to know if the host is running.

Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: I0f7660fd07bdc5c9d4989c612d8660d82982c7a6

show more ...

phal: Added service to reinit CEC device tree

The phal will initialize the required targets (hardware) as functional
in POWER CEC system device tree during ipling but, later timeframe of
ipling thos

phal: Added service to reinit CEC device tree

The phal will initialize the required targets (hardware) as functional
in POWER CEC system device tree during ipling but, later timeframe of
ipling those targets may be marked as non-functional (for example,
based on available guard records or hwp's failure case) so, when
doing normal ipl next time then ipling will behave differently due to
those marked targets as non-functional. So, before doing normal ipl
need to reinitialize the POWER CEC system device tree.

Hence, the added service will try phal to make reinitialize the device
tree.

Signed-off-by: Ramesh Iyyar <rameshi1@in.ibm.com>
Change-Id: Iaf04af076b61f3c4aa9cee641d951e2d843e996b

show more ...

Add cfam reset and FSI scans to the SPI mux service

Setting the SPI mux first requires FSI to be scanned, which requires
the CFAMs to be reset. Then, once the SPI mux is set, in order to probe
the S

Add cfam reset and FSI scans to the SPI mux service

Setting the SPI mux first requires FSI to be scanned, which requires
the CFAMs to be reset. Then, once the SPI mux is set, in order to probe
the SPI controllers, FSI must be scanned once again. Add all of this to
the service which sets the SPI mux.

Signed-off-by: Eddie James <eajames@linux.ibm.com>
Change-Id: I2d26b9810f82f256996c10ee72e52674c749163e

show more ...

mpreboot: Adding PDBG device tree environment variable

Change:
- This commit sets PDBG device tree environment variable in
mpreboot services, for custom device tree usage in libpdbg
and is a

mpreboot: Adding PDBG device tree environment variable

Change:
- This commit sets PDBG device tree environment variable in
mpreboot services, for custom device tree usage in libpdbg
and is a pre requisite for pHAL feature usage.

For phal feature disabled build the pdbg environment variable
won't be set.

Signed-off-by: Chirag Sharma <chirshar@in.ibm.com>
Change-Id: I7ba7823099f5e9c26103a3dc89e865512dc4ec7a

show more ...

cfam-reset: utilize new sysfs interface if avaialble

The kernel introduced support on some systems for a sysfs based
mechanism to issue cfam resets. See the following commit for more
information:
ht

cfam-reset: utilize new sysfs interface if avaialble

The kernel introduced support on some systems for a sysfs based
mechanism to issue cfam resets. See the following commit for more
information:
https://github.com/openbmc/linux/commit/0298c9857e284aae7841a31c2a6cd3954afec6c9

Userspace software should first look to utilize the new sysfs based
mechanism and if not available, fall back to the libgpiod mechanism.

Currently only witherspoon-tacoma and rainier support this new sysfs
based cfam reset mechanism.

Tested:
- Built a witherspoon image and verified libgpiod path was used
- Built a witherspoon-tacoma and verified new sysfs path was used

Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: I3a99899cb7d46a878a62c93954c59b7853c2b759

show more ...

pHAL: Enabled environment variable based pHAL logging

pHAL repos logging level is initialised based on repository
specific environment variables. If the environment variable
is not set by the user,

pHAL: Enabled environment variable based pHAL logging

pHAL repos logging level is initialised based on repository
specific environment variables. If the environment variable
is not set by the user, default logging level ( INFO only)
will be getting initialised.

Environment variable names:
- PDBG_LOG pdbg library
- IPL_LOG ipl library
- LIBEKB_LOG libekb library

Refer individual repository for logging level details.

Example: To set pdbg log level to PDBG_DEBUG
export PDBG_LOG=4
systemctl import-environment PDBG_LOG

Signed-off-by: Jayanth Othayoth <ojayanth@in.ibm.com>
Change-Id: I3f9f35c0b5d23c5635001379bd330d56c2a468b7

show more ...

Fix for missing phal traces in journal log

Added support to include traces provided by phal
callback functions to journal log.

Removed additional check to skip unwanted data from
phal call back , w

Fix for missing phal traces in journal log

Added support to include traces provided by phal
callback functions to journal log.

Removed additional check to skip unwanted data from
phal call back , which is fixed in the latest libekb
logging.

Signed-off-by: Jayanth Othayoth <ojayanth@in.ibm.com>
Change-Id: I43533f18f6efaded31bf57e8c10834def75f59ae

show more ...

Skip changing boot seeprom in memory preserving reboot

In memory preserving reboot path skip updating the attributes
to switch the seeprom. The boot seeprom should be not changed
during a memory pre

Skip changing boot seeprom in memory preserving reboot

In memory preserving reboot path skip updating the attributes
to switch the seeprom. The boot seeprom should be not changed
during a memory preserving reboot

Signed-off-by: Dhruvaraj Subhashchandran <dhruvaraj@in.ibm.com>
Change-Id: I16a9a79df69d66c4cf03517f1a072f2e2c0e534e

show more ...

Add startHostMpReboot support to PHAL based booting

Add support for setting the boot type as MPIPL to skip the
normal reboot and execute memory preserving reboot specific
steps.

Signed-off-by: Dhru

Add startHostMpReboot support to PHAL based booting

Add support for setting the boot type as MPIPL to skip the
normal reboot and execute memory preserving reboot specific
steps.

Signed-off-by: Dhruvaraj Subhashchandran <dhruvaraj@in.ibm.com>
Change-Id: I477a01f2d0d3c3574a3e1335be666703b71713ab

show more ...

Remove multithreading while calling enter_mpipl

libpdbg which implements the enter_mpipl is not thread-safe
so removing multithreading and creating processes.

Signed-off-by: Dhruvaraj Subhashchandr

Remove multithreading while calling enter_mpipl

libpdbg which implements the enter_mpipl is not thread-safe
so removing multithreading and creating processes.

Signed-off-by: Dhruvaraj Subhashchandran <dhruvaraj@in.ibm.com>
Change-Id: I5ca1ff2f203203b387486d9ed410eb4517443474

show more ...

clear host-on file when host stopped

This file is currently removed in the chassis poweroff target within the
op-wait-power-off service. This was fine until our recent introduction of
warm resets.

clear host-on file when host stopped

This file is currently removed in the chassis poweroff target within the
op-wait-power-off service. This was fine until our recent introduction of
warm resets.

Here's the scenario that causes trouble:

- Reboot BMC with host up, /run/openbmc/host@0-on is created to indicate
the host is already on and the services involved in booting the host
should do nothing.
- Now issue a warm reboot of the host. Because the chassis power is not
cycled on a warm reboot, the service is not run which clears the
host@0-on file. This then leads to the service responsible for starting
the host not running.

The solution is to clear this file when the host is stopped.

The removal of the file should stay within chassis poweroff target to
handle corner cases where only a hard power off is done to the system.

Tested:
Built image with fix and verified that test scenario now works and that
file was cleared as expected as a part of the warm reboot.

Signed-off-by: Andrew Geissler <geissonator@yahoo.com>
Change-Id: I6383d96f7a6be32ca19b998d9cd1cce81a307412

show more ...