Humility is the debugger for Hubris.
Hubris is the artifact that actually runs, and it must be allowed to be
optimized in terms of size, space, and run-time; in as much as contortions are
required for debuggability, they should be borne by Humility, not Hubris. As a
concrete example of this, Humility operates exclusively on release
builds of
Hubris, relying on (unloaded) DWARF information in the binaries to make sense of
the system.
Humility is Hubris-specific: it is not trying to be a generic in situ debugger, but rather specifically focused on debugging Hubris-based systems. Humility is therefore willing to encode Hubris-specific concepts like archives and tasks.
Debuggers must cut through abstractions, which often requires knowledge of underlying implementation detail. For Humility, this means being willing to take advantage of microcontroller-specific debug facilities where applicable. While Hubris may make decisions in the name of greater portability, Humility will generally make decisions to maximize debuggability, even where these facilities are highly specific to a particular MCU.
Humility is unafraid to offer device-specific functionality where that
functionality is useful for system debuggability. (For example,
humility i2c
functions as an I2C analyzer.) That said, all device
interaction uses Hubris as a proxy to actually communicate with the devices
themselves (e.g., HIF); Humility does not seek to create second I/O path.
Humility seeks to be useful, and therefore seeks to offer all manners of debugging: in situ, postmortem, dynamic instrumentation, static instrumentation, etc.
Humility operates by specifying a subcommand. There are options that are Humility-wide (that is, applying to every subcommand), as well as options that are specific to particular subcommands.
Humility needs to have some way of extracting information and/or controlling the microcontroller running Hubris. This is done through some variant of a debug probe, a separate microcontroller that speaks to debug-specific functionality on the target microcontroller.
For most evaluation boards, this debug probe is available on the board, and is connected to a host via USB, e.g. ST's STLink/V2 on the STM32F407 Discovery or the LPC-Link2 present on the LPCXpresso55S69. (On the Gemini board, there are two SWD headers, one for the LPC55S28 and the other for the STM32H753.)
While Humility allows for direct attachment to an on-chip debugger, doing so precludes other connections (from, for example, OpenOCD), making it too disruptive to development workflows. To allow for easier development workflows, Humility also has the option to attach via OpenOCD.
The debug probe to use is specified to Humility via the -p
option
(long form --probe
) or the HUMILITY_PROBE
environment variable,
which can have the following values:
-
auto
(default): Automatically determine how to attach to the microcontroller. If there is no probe found, this will default toarchive
. -
usb
: Attach directly via USB to a debug probe. When multiple probes are plugged in via USB, a probe index must be specified as a suffix (e.g.,usb-0
,usb-1
, etc.) To determine which probe is which, examine the serial number in the output ofhumility probe
. -
vid:pid[:serial]
: In some cases, the automatic algorithm may either find the wrong thing, or timeout attempting to search for non-existent probes. In these cases, it can help to explicitly set the vendor ID (VID) and product ID (PID) of the debug probe, which should be colon-delimited, e.g.,0483:374e
. (Determining the VID and PID of an attached probe is platform-specific; on Linux one can use thelsusb
command.) In cases where there are multiple probes with the same VID and PID, the serial number of the probe (as reported viahumility probe
or found in theiSerialNumber
field of the USB device descriptor) can be postpended, also delimited by a colon, e.g.0483:374e:004000343137510939383538
. -
ocd
: Attach via OpenOCD, which is presumed to have the TCL interface available on localhost on port 6666 (its default). -
jlink
: Attach via Segger JLink, which is presumed to have the GDB interface available on localhost on port 2331 (its default). Note that when semihosting is being used by Hubris, the Segger JLink GDB server will become confused when Humility attaches to it -- and subsequent calls to semihosting will cause a halt. A subsequent Humility invocation will resume the target (directing semihosting output correctly to the running GDB instance), but subsequent semihosting output will again cause a halt. To recover from this condition, send an explicit ^C to the running GDB and continue from the resulting stop. -
archive
: Do not attach to a probe at all, but rather use the specified Hubris archive as the target. Those commands that operate only by reading flash (e.g.,humility map
) can operate in this mode, but those that operate by reading RAM (e.g.,humility tasks
) will fail. Commands that operate on either (e.g.,humility readmem
orhumility readvar
) will succeed or fail depending on their input.
Many Humility commands require the complete Hubris archive. This is a ZIP
archive created by the build process, and includes all binaries as well as the
app.toml
file used to configure the Hubris archive. The archive can be
found in the target
for Hubris, and will end with (.zip
), e.g.:
/path/to/hurbis/target/demo-stm32h753-nucleo/dist/default/build-demo-stm32h753-nucleo.zip
.
The Hubris archive is specified via the -a
option or the HUMILITY_ARCHIVE
environment variable.
In the Humility examples in this documentation, unless otherwise specified,
the archive will be assumed to be set via -a
or HUMILITY_ARCHIVE
.
Many Humility commands are able to operate postmortem on a Hubris dump,
an ELF core file generated by the humility dump
command.
Dumps are offered in lieu of a probe and an archive and specified via
the -d
option (long form --dump
) or the HUMILITY_DUMP
environment
variable.
On Hubris systems that are so equipped, Humility can operate over a network
in lieu of a dump or probe. When operating over a network, the archive must
always be provided via -a
, as well as the IP address of the target via
-i
(long form --ip
) or the HUMILITY_IP
environment variable.
Note that not all commands can operate over the network, including (but not
limited to) any command that stops the target or reads memory directly.
On machines that have several different connected Hubris targets, Humility can
become thorny to manage. To aid this, Humility allows for environments:
JSON files that define how targets can be reached and auxiliary operations
that can be performed upon them. The environment file is specified via either
the --environment
argument or via the HUMILITY_ENVIRONMENT
environment
variable; a specific target within the environment is specified via the
--target
argument or via the HUMILITY_TARGET
environment variable.
The environment file contains a JSON object in which each key represents a
target. The members of each target object must include probe
and archive
,
which have the same meaning as the probe and archive as provided via the
command line. For example:
{
"lucky": {
"probe": "0483:374e:002A00174741500520383733",
"archive": "/gimlet/hubris/archives/lucky/build-gimlet.zip"
},
"sweaty": {
"probe": "0483:374e:000D00184741500520383733",
"archive": "/gimlet/hubris/archives/sweaty/build-gimlet.zip"
},
"grimey": {
"probe": "0483:374e:003400185553500820393256",
"archive": "/gimlet/hubris/archives/grimey/build-gimlet.zip"
}
}
Some targets may require multiple archives. These can be specified by
name. The archive to be used must be specified with the --archive-name
option to humility.
{
"lucky": {
"probe": "0483:374e:002A00174741500520383733",
"archive": {
imagea: "/gimlet/hubris/archives/lucky/build-a-gimlet.zip",
imageb: "/gimlet/hubris/archives/lucky/build-b-gimlet.zip"
}
},
}
The above definition -- when provided via --environment
or
HUMILITY_ENVIRONMENT
-- would allow one to (say) run humility --target grimey tasks
or humility --target lucky --image-name imagea tasks
.
In addition to probe
and archive
, one may also specify
associated commands in a cmds
object that contains a mapping of names to
commands to execute. For example:
{
"grimey": {
"probe": "0483:374e:003400185553500820393256",
"archive": "/gimlet/hubris/archives/grimey/build-gimlet.zip"
"cmds": {
"console": "/bin/sh -c \"grabserial.sh $(findtty.sh grimey)\"",
"power": {
"on": "power.sh --on grimey",
"off": "power.sh --off grimey",
"status": "power.sh --status grimey",
}
}
}
}
These commands can be in principle accessed by any debugging command, but the
humility exec
command in particular allows one to execute a command against
a specified target. (In the above example, one could execute humility --target grimey exec power.on
.)
- humility apptable: print Hubris apptable
- humility auxflash: manipulate auxiliary flash
- humility bankerase: Erase a bank
- humility console-proxy: SP/host console uart proxy
- humility counters: display event counters
- humility dashboard: dashboard for Hubris sensor data
- humility debugmailbox: interact with the debug mailbox on the LPC55
- humility diagnose: analyze a system to detect common problems
- humility doc: print command documentation
- humility dump: generate Hubris dump
- humility etm: commands for ARM's Embedded Trace Macrocell (ETM)
- humility exec: execute command within context of an environment
- humility extract: extract all or part of a Hubris archive
- humility flash: flash archive onto attached device
- humility gdb: Attach to a running system using GDB
- humility gimlet: Gimlet-specific diagnostic commands
- humility gpio: GPIO pin manipulation
- humility hash: Access to the HASH block
- humility hiffy: manipulate HIF execution
- humility host: Pretty-printing of host state
- humility hydrate: Rehydrate a bare memory dump
- humility i2c: scan for and read I2C devices
- humility ibc: interface to the BMR491 power regulator
- humility itm: commands for ARM's Instrumentation Trace Macrocell (ITM)
- humility jefe: influence jefe externally
- humility lpc55gpio: LPC55 GPIO pin manipulation
- humility lsusb: List all USB devices visible to Humility
- humility manifest: print archive manifest
- humility map: print memory map, with association of regions to tasks
- humility monorail: Management network control and debugging
- humility net: Management network device-side control and debugging
- humility openocd: Run OpenOCD for the given archive
- humility pmbus: scan for and read PMBus devices
- humility power: show power-related information
- humility powershelf: inspect powershelf over the management network
- humility probe: probe for any attached devices
- humility qspi: QSPI status, reading and writing
- humility readmem: read and display memory region
- humility readvar: read and display a specified Hubris variable
- humility rebootleby: Do something kind of like embootleby to recover a device with bad CFPA/CMPA
- humility registers: print Hubris registers
- humility rencm: query Renesas 8A3400X ClockMatrix parts
- humility rendmp: Renesas digital muliphase controller operations
- humility repl: read, eval, print, loop
- humility reset: Reset the chip using external pins
- humility ringbuf: read and display a specified ring buffer
- humility rpc: execute Idol calls over a network
- humility sbrmi: Sideband Remote Management Interface (SB-RMI) commands
- humility sensors: query sensors and sensor data
- humility spctrl: RoT -> SP control
- humility spd: scan for and read SPD devices
- humility spi: SPI reading and writing
- humility stackmargin: calculate and print stack margins by task
- humility stmsecure: change secure region settings on the stm32h7
- humility tasks: list Hubris tasks
- humility test: run Hubristest suite and parse results
- humility tofino-eeprom: read and write to the Tofino SPI EEPROM
- humility validate: validate presence and operation of devices
- humility vpd: read or write vital product data (VPD)
- humility writeword: writes one or more memory words
This is a deprecated command that allows for the display of the app table
found in old Hubris archives; see humility manifest
to understand
the contents of an archive.
Tools to interact with the auxiliary flash, described in RFD 311.
This subcommand should be rarely used; humility flash
will automatically
program auxiliary flash when needed.
Erase flash in "erase block" size chunks (32 KiB) on the RoT.
Erase len
is given in bytes and then rounded up to the nearest block
size so that all pages containing the range are erased. Then the erased
pages are overwritten with 0s, but only starting at the address given. In
other words, if the address is given as 0x40001 and len as 512, flash will
be erased from 0x40000 to 0x48000, but address 0x40000 will have value
0xff
, and every other byte will have value 0
.
Act as a proxy for the host serial console when it is jumpered to the SP.
humility counters
reads and displays any Hubris counters (as created via
the counters
crate and either the counters!
or counted_ringbuf!
macros).
For example:
$ humility -d ./hubris.core.0 counters
humility: attached to dump
gimlet_seq
|
+---> drv_gimlet_seq_server::__RINGBUF:
TOTAL VARIANT
1 Ice40Rails
1 IdentValid
1 ChecksumValid
1 Reprogram
1 Programmed
1 Programming
1 Ice40PowerGoodV1P2
1 Ice40PowerGoodV3P3
1 RailsOff
1 Ident
1 A2Status
1 A2
102 A1Status
1 CPUPresent
1 Coretype
292 A0Status
15 A0Power
...
If an argument is provided, only counters that have a name that
contains the argument as a substring, or are in a task that contains
the argument as a substring will be displayed. For example,
to display every counter that has thermal
in the name or the
containing task:
$ humility -d ./hubris.core.0 counters thermal
humility: attached to dump
thermal
|
+---> task_thermal::__RINGBUF:
TOTAL VARIANT
1 Start
1 ThermalMode(Auto)
3 AutoState(Boot)
2 AutoState(Running)
78 ControlPwm
2 PowerModeChanged
6 FanAdded
The ipc
subcommand shows IPC client counters generated automatically by
idol
, showing the total request count for a given IPC and per-client-task
breakdowns. For example:`
$ humility -d ./hubris.core.0 counters ipc`
humility: attached to dump
drv_gimlet_hf_api::__HOSTFLASH_CLIENT_COUNTERS
fn HostFlash::get_mux() .............................................. 6 calls
clients:
task host_sp_comms (0 restarts) .................... = 0 ........... = 6 ok
fn HostFlash::set_mux() .............................................. 2 calls
clients:
task gimlet_seq (0 restarts) ....................... + 0 ........... + 1 ok
task host_sp_comms (0 restarts) .................... + 0 ........... + 1 ok
--- ---
totals: = 0 err = 2 ok
fn HostFlash::get_dev() .............................................. 1 calls
clients:
task host_sp_comms (0 restarts) .................... = 0 ........... = 1 ok
drv_gimlet_seq_api::__SEQUENCER_CLIENT_COUNTERS
fn Sequencer::get_state() ......................................... 2017 calls
clients:
task thermal (0 restarts) .......................... + 0 ........ + 1386 ok
task power (0 restarts) ............................ + 0 ......... + 626 ok
task host_sp_comms (0 restarts) .................... + 0 ........... + 5 ok
--- ------
totals: = 0 err = 2017 ok
...
When displaying counters by IPC, substring filtering is performed on the counters variable, but not on the client task name. This allows filtering the output based on the IPC interface. For example:
$ humility -d ./hubris.core.0 counters ipc sensors
humility: attached to dump
task_sensor_api::__SENSOR_CLIENT_COUNTERS
fn Sensor::post() ................................................ 76717 calls
clients:
task power (0 restarts) ............................ + 0 ....... + 50300 ok
task thermal (0 restarts) .......................... + 0 ....... + 26417 ok
--- -------
totals: = 0 err = 76717 ok
fn Sensor::get_reading() ......................................... 19804 calls
clients:
task thermal (0 restarts) ...................................... = 18101 ok
- Err(NotPresent) ............................... + 1701 ..................
- Err(DeviceError) ................................. + 2 ..................
------ -------
totals: = 1703 err = 18101 ok
fn Sensor::nodata() ............................................... 6225 calls
clients:
task power (0 restarts) ............................ + 0 ........ + 3536 ok
task thermal (0 restarts) .......................... + 0 ........ + 2689 ok
--- ------
totals: = 0 err = 6225 ok
Instead, to show only the IPC counters recorded by specific client tasks,
use the --client
argument, which will filter the output counters to those
recorded in tasks whose names match the provided strings. For example, to
show only IPC counters recorded by the gimlet_seq
task, use:
$ humility -d ./hubris.core.0 counters ipc --client gimlet_seq
humility: attached to dump
drv_gimlet_hf_api::__HOSTFLASH_CLIENT_COUNTERS
fn HostFlash::set_mux() .............................................. 1 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ........... = 1 ok
drv_spi_api::__SPI_CLIENT_COUNTERS
fn Spi::exchange() ............................................... 67580 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ....... = 67580 ok
fn Spi::write() .................................................... 530 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ......... = 530 ok
fn Spi::lock() ....................................................... 4 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ........... = 4 ok
fn Spi::release() .................................................... 1 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ........... = 1 ok
drv_stm32xx_sys_api::__SYS_CLIENT_COUNTERS
fn Sys::gpio_read_input() ........................................ 16796 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ....... = 16796 ok
fn Sys::gpio_set_reset() ............................................ 15 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 .......... = 15 ok
fn Sys::gpio_configure_raw() ........................................ 14 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 .......... = 14 ok
task_jefe_api::__JEFE_CLIENT_COUNTERS
fn Jefe::set_state() ................................................. 5 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ........... = 5 ok
task_packrat_api::__PACKRAT_CLIENT_COUNTERS
fn Packrat::set_spd_eeprom() ........................................ 32 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 .......... = 32 ok
fn Packrat::set_mac_address_block() .................................. 1 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ........... = 1 ok
fn Packrat::set_identity() ........................................... 1 calls
clients:
task gimlet_seq (0 restarts) ....................... = 0 ........... = 1 ok
Multiple --client
arguments may be provided, to show IPCs from any of a
set of client tasks.
--client
may be combined with a filter matching counter variable names, to
show only the calls to specific IPC interfaces from specific tasks.
Provides a captive dashboard that graphs sensor values over time. (The
sensor
task is required for operation; see the documentation for
humility sensors
for more details.)
If -o
is provided, it specifies an output file for any raw sensor data
graphed by the dashboard.
The LPC55 includes an extra access port referred to as the Debug Mailbox. This allows for running a fixed set of commands to do useful things such as forcing SWD enablement and putting the chip into ISP mode without needing to touch an external pin
$ humility debugmailbox debug
Looks like a plausible debug mailbox
Reset chip successfully!
entering debug
$ humility debugmailbox isp
Looks like a plausible debug mailbox
Reset chip successfully!
entered ISP mode!
This subcommand provides a "firmware engineer in a can" for scanning and reporting issues in a running system. It's mostly concerned with
- Noticing things that seem fishy, and
- Taking action to get good snapshots of those things.
This is application-independent logic, so it doesn't have any visibility into things the application may think are fishy -- only general behaviors at the OS level, like faults.
Provides detailed documentation for Humility and its commands. To
get documentation on Humility, run humility doc
; to get documentation
for a specific command (like this one!) run humility doc
and specify
the command name -- and run humility --help
to list all commands.
humility dump
takes a dump of the attached system, writing out an ELF
core file:
$ humility dump
humility: attached via ST-Link
humility: core halted
humility: dumping to hubris.core.0
humility: dumped 1.12MB in 24 seconds
humility: core resumed
A dump file name may also be specified:
$ humility dump hubris.core.`date +%s`
humility: attached via ST-Link
humility: core halted
humility: dumping to hubris.core.1600718079
humility: dumped 1.12MB in 24 seconds
humility: core resumed
The resulting dump can be used with many commands (including manifest
,
map
, readvar
, and tasks
) -- and need not be run on the same machine
as the debugged MCU, e.g.:
$ humility -d hubris.core.0 tasks
humility: attached to dump
system time = 94529
ID TASK GEN PRI STATE
0 jefe 0 0 recv, notif: bit0 bit1(T+71)
1 net 1 5 recv, notif: bit0(irq61) bit2(T+213)
2 sys 0 1 recv
3 spi4_driver 0 3 recv
4 spi2_driver 0 3 recv
5 i2c_driver 0 3 recv
6 spd 0 2 notif: bit0(irq31/irq32)
7 thermal 0 5 recv, notif: bit0(T+673)
8 power 0 6 recv, notif: bit0(T+351)
9 hiffy 0 5 wait: reply from dump_agent/gen0
10 gimlet_seq 0 4 recv, notif: bit0
11 hash_driver 0 2 recv
12 hf 0 3 recv
13 update_server 0 3 recv
14 sensor 0 4 recv, notif: bit0(T+472)
15 host_sp_comms 0 7 recv, notif: bit0(irq82) bit1
16 udpecho 0 6 notif: bit0
17 udpbroadcast 0 6 notif: bit31(T+86)
18 udprpc 0 6 notif: bit0
19 control_plane_agent 0 6 recv, notif: bit0 bit1(irq37) bit2
20 sprot 0 4 notif: bit31(T+2)
21 validate 0 5 recv
22 vpd 0 4 recv
23 user_leds 0 2 recv
24 dump_agent 0 4 wait: reply from sprot/gen0
25 idle 0 8 RUNNING
Enables and operates upon the Embedded Trace Macrocell (ETM) found in some ARM Cortex-M parts.
humility exec
executes a command for a target within the specified
environment. The environment is specified to Humility via an argument
(--environment
) or an environment variable (HUMILITY_ENVIRONMENT
);
the target is similarly specified via an argument (--target
) or
an environment variable (HUMILITY_TARGET
). If specified via an
argument, note that both the environment and target must occur before
any subcommand, e.g.:
$ humility -e /path/to/env.json -t my-target exec power.on
To list available commands, use the --list
option.
For more details, see the Humliity documenation on environments.
humility extract
extracts a file from either a Hubris archive or a
dump. By default, the contents of the file are sent to standard output,
e.g.:
$ humility -a /path/to/my/hubris-archive.zip extract README.TXT
humility: extracting README.TXT to stdout
This is a build archive containing firmware build artifacts.
- app.toml is the config file used to build the firmware.
- git-rev is the commit it was built from, with optional dirty flag.
- info/ contains human-readable data like logs.
- elf/ contains ELF images for all firmware components.
- elf/tasks/ contains each task by name.
- elf/kernel is the kernel.
- img/ contains the final firmware images.
- debug/ contains OpenOCD and GDB scripts, if available.
To list all of the files in an archive, use the --list
(-l
) option:
$ humility -a /path/to/my/hubris-archive.zip extract --list
SIZE NAME
462 README.TXT
46 git-rev
5289 app.toml
1937 chip.toml
197456 elf/task/jefe
208836 elf/task/sys
288440 elf/task/i2c_driver
309100 elf/task/spi_driver
1059092 elf/task/net
65156 elf/task/user_leds
87284 elf/task/ping
31228 elf/task/pong
178600 elf/task/udpecho
370756 elf/task/hiffy
310184 elf/task/hf
286556 elf/task/hash_driver
17612 elf/task/idle
201964 elf/task/rng_driver
555820 elf/kernel
1550 info/allocations.txt
4195 info/map.txt
632426 img/combined.srec
307060 img/combined.elf
862933 img/combined.ihex
307200 img/combined.bin
632426 img/final.srec
307060 img/final.elf
862933 img/final.ihex
307200 img/final.bin
1790 img/flash.ron
1047 debug/script.gdb
1586 debug/openocd.cfg
235 debug/openocd.gdb
Note that the file specified is treated as a substring; if the substring matches exactly one file, it will be extracted:
$ humility -d ./hubris.core.79 extract map
humility: extracting info/map.txt to stdout
ADDRESS END SIZE FILE
08000000 08000298 298 target/demo-stm32h753-nucleo/dist/kernel
08000298 08004420 4188 target/demo-stm32h753-nucleo/dist/kernel
08004420 08005604 11e4 target/demo-stm32h753-nucleo/dist/kernel
08005a00 08005a58 58 target/demo-stm32h753-nucleo/dist/idle
08005c00 08005e10 210 target/demo-stm32h753-nucleo/dist/pong
08005e10 08005e88 78 target/demo-stm32h753-nucleo/dist/pong
08006000 080076f4 16f4 target/demo-stm32h753-nucleo/dist/jefe
080076f4 08007c08 514 target/demo-stm32h753-nucleo/dist/jefe
08007c08 08007c08 0 target/demo-stm32h753-nucleo/dist/jefe
08008000 08009798 1798 target/demo-stm32h753-nucleo/dist/udpecho
08009798 08009b90 3f8 target/demo-stm32h753-nucleo/dist/udpecho
08009b90 08009b90 0 target/demo-stm32h753-nucleo/dist/udpecho
08010000 08014020 4020 target/demo-stm32h753-nucleo/dist/hiffy
08014020 08015134 1114 target/demo-stm32h753-nucleo/dist/hiffy
08015134 0801574c 618 target/demo-stm32h753-nucleo/dist/hiffy
08018000 08019970 1970 target/demo-stm32h753-nucleo/dist/i2c_driver
08019970 08019ff4 684 target/demo-stm32h753-nucleo/dist/i2c_driver
08019ff4 0801a300 30c target/demo-stm32h753-nucleo/dist/i2c_driver
0801c000 0801de78 1e78 target/demo-stm32h753-nucleo/dist/spi_driver
0801de78 0801e898 a20 target/demo-stm32h753-nucleo/dist/spi_driver
0801e898 0801eca4 40c target/demo-stm32h753-nucleo/dist/spi_driver
08020000 0802a554 a554 target/demo-stm32h753-nucleo/dist/net
0802a554 0802d74c 31f8 target/demo-stm32h753-nucleo/dist/net
0802d74c 0802d7dc 90 target/demo-stm32h753-nucleo/dist/net
08040000 08041ea0 1ea0 target/demo-stm32h753-nucleo/dist/hf
08041ea0 080426a8 808 target/demo-stm32h753-nucleo/dist/hf
080426a8 080426a8 0 target/demo-stm32h753-nucleo/dist/hf
08044000 08044b8c b8c target/demo-stm32h753-nucleo/dist/ping
08044b8c 08044d34 1a8 target/demo-stm32h753-nucleo/dist/ping
08044d34 08044d34 0 target/demo-stm32h753-nucleo/dist/ping
08046000 080478b4 18b4 target/demo-stm32h753-nucleo/dist/hash_driver
080478b4 08047f68 6b4 target/demo-stm32h753-nucleo/dist/hash_driver
08047f68 08047f68 0 target/demo-stm32h753-nucleo/dist/hash_driver
08048000 08048fcc fcc target/demo-stm32h753-nucleo/dist/rng_driver
08048fcc 08049474 4a8 target/demo-stm32h753-nucleo/dist/rng_driver
08049474 08049474 0 target/demo-stm32h753-nucleo/dist/rng_driver
0804a000 0804a634 634 target/demo-stm32h753-nucleo/dist/sys
0804a634 0804a67c 48 target/demo-stm32h753-nucleo/dist/sys
0804a800 0804ac0c 40c target/demo-stm32h753-nucleo/dist/user_leds
0804ac0c 0804ac7c 70 target/demo-stm32h753-nucleo/dist/user_leds
20000400 20000400 0 target/demo-stm32h753-nucleo/dist/kernel
30000000 30000000 0 target/demo-stm32h753-nucleo/dist/net
If the substring matches more than one file, the command will fail and the matching files will be displayed:
$ humility -d ./hubris.core.79 extract toml
humility extract failed: "toml" matches multiple files: app.toml, stm32h7.toml
To redirect output to a particular file, use the --output
(-o
) option.
To dump the entire archive, leave the file unspecified. (Note that extracting the entire archive requires the specification of an output file to prevent accidental blasts of binary content to the console.)
Flashes the target with the image that is contained within the specified
archive (or dump). As a precautionary measure, if the specified archive
already appears to be on the target, humility flash
will fail unless the
-F
(--force
) flag is set. Because this will only check the image
ID (and not the entire image), humility flash
can be optionally told
to verify that all of the program text in the image is on the device
by specifying -V
(--verify
). Similarly, if one wishes to only
check the image against the archive (and not flash at all), specify
-C
(--check
).
This attempts to natively flash the part within Humility using probe-rs,
but for some parts or configurations, it may need to use OpenOCD as a
child process to flash it. If OpenOCD is required but not installed (or
isn't in the path), this will fail. If OpenOCD is used, temporary files
are created as part of this process; if they are to be retained, the -R
(--retain-temporaries
) flag should be set. To see what would be
executed without actually executing any commands, use the -n
(--dry-run
) flag. Should use of OpenOCD need to be forced (that is,
should probe-rs flashing fail), the -O
(--force-openocd
) flag can be
used. That said, OpenOCD should generally be discouraged; the disposition
is to extend probe-rs to support any parts that must be flashed via
OpenOCD.
If the specified archive includes auxiliary flash data and the new image
includes a task with the AuxFlash
API, two slots of auxiliary flash
will be programmed after the image is written. See RFD 311 for more
information about auxiliary flash management.
This command launches GDB and attaches to a running device.
By default, the user must be running openocd
or pyocd
in a separate
terminal.
The --run-openocd
option automatically launches openocd
based on the
openocd.cfg
file included in the build archive.
When using pyocd
, it must be launched with the --persist
option,
because humility gdb
connects to it multiple times (once to check the
app id, then again to run the console).
humility gimlet
contacts a (recent firmware version) Gimlet over the
management network and extracts certain live diagnostic information.
$ humility gimlet --ip fe80::0c1d:9aff:fe64:b8c2%en0 read-seq-regs
For a complete list of subcommands, use humility gimlet --help
.
humility gpio
allows for GPIO pins to be set, reset, queried or
configured on STM32 targets. Commands:
--set
(-s
): Sets a pin (sets it high)--reset
(-r
): Resets a pin (sets it low)--toggle
(-t
): Toggles a pin (sets it high if low, low if high)--input
(-i
): Queries the state of a pin (or all pins if no pin is specified)--configure
(-c
): Configures a pin
To change the state of a pin (or pins), specify the pin (or pins) and the desired command. For example, to toggle the state on pin 14 on port B:
$ humility gpio --toggle --pins B:14
humility: attached via ST-Link V3
[Ok([])]
To set pins B:0, B:14 and E:1:
$ humility gpio --set --pins B:0,B:14,E:1
humility: attached via ST-Link V3
[Ok([]), Ok([]), Ok([])]
To reset pin E:1:
$ humility gpio --reset --pins E:1
humility: attached via ST-Link V3
[Ok([])]
To get input values for a particular pin:
$ humility gpio --input --pins B:0,B:14,E:1
humility: attached via ST-Link V3
B:0 = 1
B:14 = 1
E:1 = 0
To get input values for all pins, leave the pin unspecified:
$ humility gpio --input
humility: attached via ST-Link V3
Pin 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
-----------------------------------------------------------------------
Port A 0 0 1 0 0 0 0 0 0 0 0 0 0 1 1 1
Port B 1 0 0 0 1 0 0 0 0 0 0 0 0 0 1 0
Port C 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Port D 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 0
Port E 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Port F 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Port G 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Port H 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Port I 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Port J 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Port K 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
To configure a pin, the configuration should be specified as a colon-delimited 5-tuple consisting of:
- Mode:
Input
,Output
,Alternate
, orAnalog
- Output type:
PushPull
orOpenDrain
- Speed:
Low
,Medium
,High
, orVeryHigh
- Pull direction:
None
,Up
, orDown
- Alternate function: one of
AF0
throughAF15
For example, to configure pin 5 on port A as a push-pull output:
$ humility gpio -c Output:PushPull:High:None:AF0 -p A:5
Uses the hash
task to hash sequences. The following are all equivalent:
$ humility hash --digest -s abc
$ humility hash --digest -x 61,62,63
$ echo -n abc > abc.txt ; humility hash --digest -f abc.txt
$ hash -i --update -s 'a' ; hash --update -s 'bc' ; hash --finalize
Note that --update
can also take a filename as a parameter.
humility hiffy
allows for querying and manipulation of hiffy
, the
HIF agent present in Hubris. To list all Idol interfaces present in
Hubris, use the -l
(--list
) option, optionally specifying a filter
for tasks or interface names if so desired:
$ humility hiffy -l user_leds
humility: attached via ST-Link
INTERFACE TASK
UserLeds user_leds
|
+--> UserLeds.led_on
| index usize
| <ok> ()
| <error> LedError
|
+--> UserLeds.led_off
| index usize
| <ok> ()
| <error> LedError
|
+--> UserLeds.led_toggle
index usize
<ok> ()
<error> LedError
To enlist the Hubris agent to call a particular interface and operation,
use -c
(--call
), using -a
(--arguments
) to indicate any arguments,
e.g.:
$ humility hiffy -c UserLeds.led_toggle -a index=0
humility: attached via ST-Link
UserLeds.led_toggle() = ()
To view the raw HIF functions provided to programmatic HIF consumers
within Humility, use -L
(--list-functions
).
humility host
pretty-prints host state, which is sent to the SP over IPCC.
It is only functional on a Gimlet SP image.
Pretty prints the value of LAST_HOST_PANIC
% humility: attached to dump
version: 2 (inferred)
cause: IPCC_PANIC_TRAP
error: 0
cpuid: 0
thread: 0xfffff78809c74c20
time: 74.639002555 (1970-01-01T00:01:14.639002555+00:00)
hrtime: 74639011110
addr: 0xfffffcf853a0f940
pc: 0xfffffffffbc5fb45
fp: 0xfffff78809c748f0
rp: 0xfffff78809c749e0
registers:
rdi 0xfffffffffc029d70 rsi 0xfffff78809c749d8 rdx 0xfffffffffc029d70
rcx 0x0 r8 0xfffffcf931453280 r9 0x6
rax 0x0 rbx 0xfffffcf931453280 rbp 0xfffff78809c74b60
r10 0x1 r11 0xfffffffff7c9c510 r12 0xfffffcf931da7210
r13 0xfffffcf931453370 r14 0x0 fsb 0xfffffffffc2604e8
gsb 0xfffff78809c74a90 es 0x0 fs 0x0
gs 0x0 tra 0x6 err 0x0
rip 0xfffffffffc029d70 cs 0x30 rfl 0x10246
rsp 0xfffff78809c74ad8 ss 0x38
message: BAD TRAP: type=6 (#ud Invalid opcode) rp=fffff78809c749e0 addr=fffffcf853a0f940
stack:
die+0x105 (0xfffffffffbc5fb45)
trap+0x855 (0xfffffffffbc606f5)
cmntrap+0xe9 (0xfffffffffbc49819)
clock+0x0 (0xfffffffffc029d70)
cbe_softclock+0x23 (0xfffffffffbc0c8b3)
av_dispatch_softvect+0x72 (0xfffffffffbccbf02)
apix_dispatch_softint+0x35 (0xfffffffff7c9c545)
switch_sp_and_call+0x15 (0xfffffffffbc834b5)
apix_do_softint+0x5a (0xfffffffff7c9c5ba)
apix_do_interrupt+0x2bf (0xfffffffff7c9d06f)
cmnint+0xc3 (0xfffffffffbc00233)
taskq_bucket_extend+0x15b (0xfffffffffc17a66b)
taskq_create_common+0x33a (0xfffffffffc17a19a)
system_taskq_init+0x4e (0xfffffffffc1788de)
main+0xbb (0xfffffffffc0c50bb)
_locore_start+0x88 (0xfffffffffbc49708)
Pretty-prints the contents of LAST_HOST_BOOT_FAIL
$ humility host boot-fail
humility: attached to dump
humility: reading LAST_HOST_BOOT_FAIL
[0; 4096]
humility hydrate
combines a raw memory dump (obtained from MGS) and a
Hubris archive, forming a proper Hubris dump.
$ humility -a build-grapefruit-image-default.zip hydrate hubris.dry.0
The archive is required, because the raw memory dump does not contain debug information.
By default, this writes to hubris.core.TASK_NAME.N
(where N
is the
lowest available value); use --out
to specify a different path name.
On platforms that have I2C support, humility i2c
can be used
to scan a bus, scan a device, read a register, or write a register. Its
usage will be specific to the board being examined; to specify a
controller use the -c
, to specify a port (if necessary), use -p
; to
specify a mux and segment (if necessary), use -m
.
For example, on a Gimletlet, here is a scan of controller I2C3, revealing
one device at address 0x48
:
$ humility i2c -s -c 3
humility: attached via ST-Link
Device scan on controller I2C3:
R = Reserved - = No device \o/ = Device found X = Timed out
ADDR 0x0 0x1 0x2 0x3 0x4 0x5 0x6 0x7 0x8 0x9 0xa 0xb 0xc 0xd 0xe 0xf
0x00 R R R R R R R R - - - - - - - -
0x10 - - - - - - - - - - - - - - - -
0x20 - - - - - - - - - - - - - - - -
0x30 - - - - - - - - - - - - - - - -
0x40 - - - - - - - - \o/ - - - - - - -
0x50 - - - - - - - - - - - - - - - -
0x60 - - - - - - - - - - - - - - - -
0x70 - - - - - - - - - - - - R R R R
To scan that device, specify its address via -d
:
$ humility i2c -s -c 3 -d 0x48
humility: attached via ST-Link
Register scan for device 0x48 on I2C3:
- = No register ! = No device X = Timed out
ADDR 0x0 0x1 0x2 0x3 0x4 0x5 0x6 0x7 0x8 0x9 0xa 0xb 0xc 0xd 0xe 0xf
0x00 0b c8 80 00 20 00 05 00 49 80 05 cb 00 00 00 00
0x10 00 00 00 00 00 00 00 00 6c 00 06 00 00 00 00 00
0x20 0b c8 80 00 20 00 05 00 49 80 05 cb 00 00 00 00
0x30 00 00 00 00 00 00 00 00 6c 00 06 00 00 00 00 00
0x40 00 00 80 00 20 00 05 00 49 80 05 cb 00 00 00 00
0x50 00 00 00 00 00 00 00 00 6c 00 06 00 00 00 00 00
0x60 00 00 80 00 20 00 05 00 49 80 05 cb 00 00 00 00
0x70 00 00 00 00 00 00 00 00 6c 00 06 00 00 00 00 00
0x80 00 00 80 00 20 00 05 00 49 80 05 cb 00 00 00 00
0x90 00 00 00 00 00 00 00 00 6c 00 06 00 00 00 00 00
0xa0 00 00 80 00 20 00 05 00 49 80 05 cb 00 00 00 00
0xb0 00 00 00 00 00 00 00 00 6c 00 06 00 00 00 00 00
0xc0 00 00 80 00 20 00 05 00 49 80 05 cb 00 00 00 00
0xd0 00 00 00 00 00 00 00 00 6c 00 06 00 00 00 00 00
0xe0 00 00 80 00 20 00 05 00 49 80 05 cb 00 00 00 00
0xf0 00 00 00 00 00 00 00 00 6c 00 06 00 00 00 00 00
(This device is an ADT7420 temp sensor.) To look at a particular
register, elide -s
and specify the register of interest via -r
:
$ humility i2c -c 3 -d 0x48 -r 0xb
humility: attached via ST-Link
Controller I2C3, device 0x48, register 0xb = 0xcb
To write a value to a register, specify the -w
flag, along with the
value to write, e.g. (for the ADT7420), the MSB of the THIGH
register:
$ humility i2c -c 3 -d 0x48 -r 0x4 -w 0x1f
humility: attached via ST-Link
Controller I2C3, device 0x48, register 0x4 = 0x1f
Note that if registers are not writable, the write will (generally) be silently discarded by the device; it can be useful to read the register after writing it to confirm that the value is as expected:
$ humility i2c -c 3 -d 0x48 -r 0x4
humility: attached via ST-Link
Controller I2C3, device 0x48, register 0x4 = 0x1f
To determine the last mux and segment to be enabled on a particular
controller/port, use --lastmux
(-l
):
$ humility i2c -b front --lastmux
humility: attached via ST-Link V3
last selected mux/segment for I2C2, port F: mux 3, segment 2
Interface to BMR491 power regulator
This regulator is present on Gimlet and Sidecar PCAs. Right now,
humility ibc
only exposes one subcommand: black-box
.
humility ibc black-box
allows you to read out the blackbox log from the
power converter. This can be used to debug previous faults. The log is
stored in non-volatile memory, so it should be persistent through power
loss.
Here's an example:
$ humility ibc black-box
humility: attached to 0483:374e:001B00083156501320323443 via ST-Link V3
FAULT EVENT
EVENT_INDEX: 0
TIMESTAMP 0x000a3747 = 18 hr, 35 min, 51.1 sec
EVENT_ID 0x0000
STATUS_WORD 0x0010
0x0010: Output overcurrent fault
STATUS_IOUT 0x80
0x80: Iout Overcurrent Fault
STATUS_MFR 0x00
V_IN 0xf84f = 39.500V
V_OUT 0x6000 = 12.000V
I_OUT 0x0075 = 117.000A
TEMPERATURE 0x0023 = 35.000°C
FAULT EVENT
EVENT_INDEX: 1
TIMESTAMP 0x000a3751 = 18 hr, 35 min, 52.1 sec
EVENT_ID 0x0001
STATUS_WORD 0x0001
0x0001: System event
STATUS_MFR 0x01
0x01: INPUT_LOW_EVENT
V_IN 0xf83c = 30.000V
V_OUT 0x0000 = 0.000V
I_OUT 0x0000 = 0.000A
TEMPERATURE 0x0000 = 0.000°C
FAULT EVENT
EVENT_INDEX: 2
TIMESTAMP 0x2809e751 = 777 day, 11 hr, 22 min, 48.1 sec
EVENT_ID 0x0002
STATUS_WORD 0x0001
0x0001: System event
STATUS_MFR 0x02
0x02: CANCEL_EVENT
V_IN 0xf86b = 53.500V
V_OUT 0x5f00 = 11.875V
I_OUT 0x0000 = 0.000A
TEMPERATURE 0x0012 = 18.000°C
FAULT EVENT
EVENT_INDEX: 3
TIMESTAMP 0x50099751 = 1554 day, 4 hr, 9 min, 44.1 sec
EVENT_ID 0x0003
STATUS_WORD 0x0001
0x0001: System event
STATUS_MFR 0x02
0x02: CANCEL_EVENT
V_IN 0xf86b = 53.500V
V_OUT 0x6000 = 12.000V
I_OUT 0x0000 = 0.000A
TEMPERATURE 0x0020 = 32.000°C
LIFECYCLE EVENT
EVENT_INDEX: 24
TIMESTAMP 0xffef8ad0 = 4969 day, 18 hr, 41 min, 12.0 sec
EVENT_ID 0x0680
STATUS_WORD 0x0001
0x0001: System event
STATUS_MFR 0x04
0x04: CLR_EVENT
V_IN 0xf86b = 53.500V
V_OUT 0x5f00 = 11.875V
I_OUT 0x0000 = 0.000A
TEMPERATURE 0x001a = 26.000°C
LIFECYCLE EVENT
EVENT_INDEX: 25
TIMESTAMP 0xffef8bd4 = 4969 day, 18 hr, 41 min, 38.0 sec
EVENT_ID 0x0681
STATUS_WORD 0x0001
0x0001: System event
STATUS_MFR 0x05
0x05: ERASE_OVFL_EVENT
V_IN 0xf860 = 48.000V
V_OUT 0x5f00 = 11.875V
I_OUT 0x0000 = 0.000A
TEMPERATURE 0x001b = 27.000°C
The log doesn't appear to be completely reliable, especially with respect to timestamps, so take it with a grain of salt and with the datasheet close at hand. For example, the machine in the example above had not be up for 777 days (or, for that matter, for 4,969).
humility itm
consumes data from the Instrumentation Trace Macrocell
(ITM) present in many ARM Cortex-M variants. ITM is problematic in many
dimensions: it is lossy; it requires knowledge of the target's clocking to
configure properly; it relies on functionality (SWO/SWV) that is often
buggy in chip debuggers; it isn't present everywhere (Cortex-M0+ in
particular doesn't have ITM). So in general, ITM isn't what Hubris
programmers should be looking for: those developing code and wishing to
see if and how that code is executed should prefer ring buffers to
ITM-based instrumentation. (See the documentation for humility ringbuf
for details.)
That said, ITM remains the best way to get certain messages from the
Hubris kernel (e.g., boot and panic messages); use humility itm -ea
to
enable ITM and attach to the connected device. For example, if running
with the ping
task, one will see messages from jefe
restarting it:
$ humility -a /path/to/my/hubris-archive.zip itm -ea
humility: attached via ST-Link
humility: core halted
humility: core resumed
humility: ITM synchronization packet found at offset 6
Task #7 Divide-by-zero
Task #7 Memory fault at address 0x0
Task #7 Divide-by-zero
Humility allows for some (well-defined) manipulation of tasks via jefe
,
the Hubris supervisor. By default, jefe
will carry out a configured set
of responses if tasks fault -- restarting them unless configured otherwise.
The policy that governs Jefe's response to task failure is called the
disposition. While debugging, it can be useful to change Jefe's
disposition on a particular task, so that the task will not automatically
restart on fault. This is done with the -H
/--hold
option:
$ humility jefe --hold ping
humility: attached via ST-Link
humility: successfully changed disposition for ping
This doesn't make any immediate changes to the task, but merely updates a
table inside jefe
. When the task next faults, it will not be restarted,
e.g.:
$ humility tasks ping
humility: attached via ST-Link
system time = 26597
ID TASK GEN PRI STATE
8 ping 121 4 FAULT: divide by zero (was: ready)
This can be particularly useful to couple with either humility dump
or
humility tasks -sl
:
$ humility tasks -sl ping
humility: attached via ST-Link
system time = 103879
ID TASK GEN PRI STATE
8 ping 121 4 FAULT: divide by zero (was: ready)
|
+---> 0x200065b0 0x0802a05e task_ping::divzero
@ /home/bmc/hubris/task/ping/src/main.rs:24
0x20006600 0x0802a0fe userlib::sys_panic
@ /home/bmc/hubris/sys/userlib/src/lib.rs:678
0x20006600 0x0802a0fe main
@ /home/bmc/hubris/task/ping/src/main.rs:35
To change the disposition of a task (back) to be restarted, use the
--release
/-r
flag:
$ humility jefe --release ping
humility: attached via ST-Link
humility: successfully changed disposition for ping
--release
undoes the effect of --hold
. If the task is parked at a fault,
it will be restarted immediately. Otherwise just the disposition will be
updated and the task will be left alone.
If you would like to cause a task to restart, use the --start
/-s
flag.
This will restart a running task immediately, and will re-run a held task
without releasing the hold. This means the task will still be caught at
the next fault, giving you a way of precisely trying an operation again.
To inject a fault into a task, use the --fault
/-f
flag. This will cause
jefe
to use Kernel Trickery(tm) to inject a fault into the target task,
and also switch its disposition to "held" so you can inspect the result.
The injected fault has a dedicated type to make clear that the fault
originated outside of the task:
$ humility jefe --fault pong
humility: attached via ST-Link
humility: successfully changed disposition for pong
$ humility tasks pong
humility: attached via ST-Link
system time = 191227
ID TASK GEN PRI STATE
7 pong 0 3 FAULT: killed by jefe/gen0 (was: recv, notif: bit0)
|
+---> 0x200063b8 0x08028c0a userlib::sys_recv_stub
@ /home/bmc/hubris/sys/userlib/src/lib.rs:288
0x20006400 0x080280ac userlib::sys_recv
@ /home/bmc/hubris/sys/userlib/src/lib.rs:236
0x20006400 0x080280ba main
@ /home/bmc/hubris/task/pong/src/main.rs:13
As with tasks held by --hold
, use --release
/-r
to set the task back to
normal, or --start
/-s
to run it once but catch the next fault.
The LPC55-equivalent of humility gpio
, allowing for GPIO pins to
be set, reset, queried or configured on LPC55 targets. Commands:
--set
(-s
): Sets a pin (sets it high)--reset
(-r
): Resets a pin (sets it low)--toggle
(-t
): Toggles a pin (sets it high if low, low if high)--input
(-i
): Queries the state of a pin (or all pins if no pin is specified)--configure
(-c
): Configures a pin--direction
(-d
): Configure the direction of a pin
To change the state of a pin (or pins), specify the pin (or pins) and the desired command. For example, to toggle the state on pin PIO0_17:
$ humility lpc55gpio --toggle --pins PIO0_17
humility: attached via CMSIS-DAP
[Ok([])]
To set pins PIO0_15, PIO0_16 and PIO0_17:
$ humility lpc55gpio --set --pins PIO0_15,PIO0_16,PIO0_17
humility: attached via CMSIS-DAP
[Ok([]), Ok([]), Ok([])]
To reset pin PIO0_17:
$ humility lpc55gpio --reset --pins PIO0_17
humility: attached via CMSIS-DAP
[Ok([])]
To get input values for a particular pin:
$ humility lpc55gpio --input --pins PIO0_10,PIO0_11,PIO1_0
humility: attached via CMSIS-DAP
PIO0_10 = 0
PIO0_11 = 1
PIO1_0 = 0
To get input values for all pins, leave the pin unspecified:
$ humility lpc55gpio --input
humility: attached via ST-Link V3
humility: attached to 1fc9:0143:12UNOSLDXOK51 via CMSIS-DAP
PIO0_0 = 0
PIO0_1 = 0
PIO0_2 = 0
PIO0_3 = 0
PIO0_4 = 0
PIO0_5 = 1
PIO0_6 = 0
PIO0_7 = 0
PIO0_8 = 0
PIO0_9 = 1
PIO0_10 = 0
PIO0_11 = 1
PIO0_13 = 0
...
To configure a pin, the configuration should be specified as a colon-delimited 6-tuple consisting of:
- Alternate function: one of
Alt0
throughAlt9
- Mode:
NoPull
,PullDown
,PullUp
, orRepeater
- Slew:
Standard
orFast
- Invert:
Disable
, orEnabled
- Digital mode:
Analog
orDigital
- Open drain:
Normal
orOpenDrain
Note that the direction of the pin should also likely be configured;
this is done via the --direction
option, specifying either Input
or
Output
. For example, to configure pin PIO0_17 to be an output:
$ humility lpc55gpio -c Alt0:NoPull:Standard:Disable:Digital:Normal -p PIO0_17
humility: attached via CMSIS-DAP
[Ok([])]
$ humility lpc55gpio -p PIO0_17 --direction Output
humility: attached via CMSIS-DAP
[Ok([])]
humility lsusb
will show you Humility's view of the USB devices available
on the system, to help you choose probes and/or diagnose permissions issues.
humility manifest
displays information about the Hubris archive. It
does not connect at all to a Hubris target to operate. In addition to
archive-wide attributes, humility manifest
displays the features enabled
for each task, as well as any device toplogy information present in
the archive, e.g.:
$ humility manifest
version => hubris build archive v1.0.0
git rev => 753a57169eba699e73ee59e0cf5345eb1d6e1ae2-dirty
board => nucleo-h743zi2
name => demo-stm32h753-nucleo
target => thumbv7em-none-eabihf
features => h743, itm
total size => 140K
kernel size => 30K
tasks => 12
ID TASK SIZE FEATURES
0 hiffy 42.9K h743, stm32h7, itm, i2c, gpio, qspi
1 jefe 6.7K itm
2 i2c_driver 9.8K h743
3 spi_driver 10.7K spi3, h743
4 hf 8.6K h743
5 rcc_driver 4.7K h743
6 gpio_driver 5.8K h743
7 usart_driver 5.9K h743
8 user_leds 5.5K stm32h7
9 pong 4.7K
10 ping 5.2K uart
11 idle 0.1K
i2c buses => 1 controller, 1 bus
C PORT MODE NAME DESCRIPTION
2 F init - -
humility manifest
can operate on either an archive or on a dump.
One common pathology in Hubris tasks is a fault induced when a task
attempts to access a memory address outside of its designated regions.
(This can happen, for example, because the task attempted to access device
memory that was not allocated to it in the build description, because it
exceeded the device memory allocated to it, or because it allocated the
memory allocated to it, e.g., because of a stack overflow.) humility tasks
can be useful to see this happening; a non-zero generation count
will indicate that a task has been restarted -- and the log message from
jefe
will indicate the specific address, e.g.:
$ humility itm -ea
humility: attached via OpenOCD
humility: core halted
humility: core resumed
humility: TPIU sync packet found at offset 1
humility: ITM synchronization packet found at offset 12
Task #7 Memory fault at address 0x200028fc
Task #7 Memory fault at address 0x200028fc
^C
To better understand the memory that a task is trying to access, one can
run the humility map
command, which shows the memory regions that have
been mapped into tasks, in address order:
$ humility -a ~/hubris/target/demo/dist/build-demo.zip map
humility: attached via OpenOCD
DESC LOW HIGH SIZE ATTR ID TASK
0x08004864 0x08010000 - 0x08017fff 32KiB r-x--- 0 jefe
0x08004884 0x08018000 - 0x08019fff 8KiB r-x--- 1 rcc_driver
0x080048a4 0x0801c000 - 0x0801ffff 16KiB r-x--- 2 usart_driver
0x080048c4 0x08020000 - 0x08023fff 16KiB r-x--- 3 user_leds
0x080048e4 0x08024000 - 0x08025fff 8KiB r-x--- 4 ping
0x08004904 0x08026000 - 0x08027fff 8KiB r-x--- 5 pong
0x08004924 0x08028000 - 0x080280ff 256 r-x--- 6 idle
0x08004944 0x0802a000 - 0x0802bfff 8KiB r-x--- 7 oh_no
0x08004964 0x0802c000 - 0x0802dfff 8KiB r-x--- 8 oh_no2
0x08004874 0x20001000 - 0x200013ff 1KiB rwx--- 0 jefe
0x08004894 0x20001400 - 0x200017ff 1KiB rwx--- 1 rcc_driver
0x080048b4 0x20001800 - 0x20001bff 1KiB rwx--- 2 usart_driver
0x080048d4 0x20001c00 - 0x20001fff 1KiB rwx--- 3 user_leds
0x080048f4 0x20002000 - 0x200021ff 512 rwx--- 4 ping
0x08004914 0x20002400 - 0x200027ff 1KiB rwx--- 5 pong
0x08004934 0x20002800 - 0x200028ff 256 rwx--- 6 idle
0x08004954 0x20002900 - 0x200029ff 256 rwx--- 7 oh_no
0x08004974 0x20002a00 - 0x20002aff 256 rwx--- 8 oh_no2
0x08004824 0x40004400 - 0x400047ff 1KiB rw-d-- 2 usart_driver
0x08004844 0x40020000 - 0x400203ff 1KiB rw-d-- 2 usart_driver
0x08004854 0x40020c00 - 0x40020fff 1KiB rw-d-- 3 user_leds
0x08004834 0x40023800 - 0x40023bff 1KiB rw-d-- 1 rcc_driver
(In this case, task 7, oh_no
, has overflowed its stack -- which
we can see from the map
output has been sized to only 256 bytes.)
humility monorail
exposes commands to interact with the management
network switch and PHYs. It is for management of the management network,
and can therefore only be run on two images:
app/sidecar/app.toml
running on Sidecar hardwareapp/gimletlet/app-vsc7448.toml
, running on a Gimletlet which is attached to a VSC7448 dev kit (VSC5627EV) via SPI. This setup is non-trivial; if you find yourself with a dev kit, talk to Matt about how to wire it up.
Use humility monorail -h
to see help, or humility monorail status
for
a bird's-eye view of the ports.
The read
subcommand allows you to read a register from the VSC7448 switch
IC. Registers can be specified in a few ways, but most of the time, you'll
want to execute a read by register name:
$ export HUMILITY_TARGET=sidecar
$ humility monorail read DEV1G[0]:DEV_RST_CTRL
humility: attached to 0483:374f:002A001C4D46500F20373033 via ST-Link V3
humility: Reading DEV1G[0]:DEV_CFG_STATUS:DEV_RST_CTRL from 0x71040000
DEV1G[0]:DEV_CFG_STATUS:DEV_RST_CTRL => 0x100000
bits | value | field
21:20 | 0x1 | SPEED_SEL
12 | 0x0 | PCS_TX_RST
8 | 0x0 | PCS_RX_RST
4 | 0x0 | MAC_TX_RST
0 | 0x0 | MAC_RX_RST
It's not necessary to use the fully qualified TARGET:GROUP:REGISTER
form;
any unambiguous subset will work (e.g. in the example above, we used
DEV1G[0]:DEV_RST_CTRL
instead of the full
DEV1G[0]:DEV_CFG_STATUS:DEV_RST_CTRL
). If your register name is not
unambiguous, an error will be printed.
Register names can be found in the
vsc7448-pac
crate,
which is automatically generated from Microchip's C SDK header files.
It's also possible to execute reads by raw address:
$ humility monorail read 0x71040000
humility: attached to 0483:374f:002A001C4D46500F20373033 via ST-Link V3
humility: Reading DEV1G[0]:DEV_CFG_STATUS:DEV_RST_CTRL from 0x71040000
DEV1G[0]:DEV_CFG_STATUS:DEV_RST_CTRL => 0x100000
bits | value | field
21:20 | 0x1 | SPEED_SEL
12 | 0x0 | PCS_TX_RST
8 | 0x0 | PCS_RX_RST
4 | 0x0 | MAC_TX_RST
0 | 0x0 | MAC_RX_RST
Modifies a register in the VSC7448 switch. Note that there is no checking of read/write vs read-only registers; good luck!
The info
subcommand looks up info on a register in the VSC7448 switch IC.
This command is offline, meaning it does not require an attached system.
$ humility monorail info HW_QSGMII_CFG
Register HSIO:HW_CFGSTAT:HW_QSGMII_CFG
Register address: 0x71460170
bits | field
13 | E_DET_ENA
11:0 | FLIP_LANES
14 | SHYST_DIS
12 | USE_I1_ENA
If you provide a value as the final argument, it will decode the register and pretty-print a table:
$ humility monorail info HW_QSGMII_CFG 0x2000
Register HSIO:HW_CFGSTAT:HW_QSGMII_CFG
Register address: 0x71460170
Register value: 0x2000
bits | value | field
14 | 0x0 | SHYST_DIS
13 | 0x1 | E_DET_ENA
12 | 0x0 | USE_I1_ENA
11:0 | 0x0 | FLIP_LANES
Prints a table showing the status of every port in the system, along with
their PHY (if present). The -p
argument allows you to specify a subset of
ports, e.g.
$ humility monorail status -p40,41,42,43,44,45
humility: attached to 0483:374f:002A001C4D46500F20373033 via ST-Link V3
PORT | MODE SPEED DEV SERDES LINK | PHY MAC LINK MEDIA LINK
-----|-------------------------------------|-------------------------------
40 | QSGMII 100M 1G_16 6G_14 up | VSC8504 down down
41 | QSGMII 100M 1G_17 6G_14 up | VSC8504 down up
42 | QSGMII 100M 1G_18 6G_14 up | VSC8504 down down
43 | QSGMII 100M 1G_19 6G_14 up | VSC8504 down down
44 | QSGMII 1G 1G_20 6G_15 up | VSC8562 up up
45 | QSGMII 1G 1G_21 6G_15 up | VSC8562 up up
Dumps an entire top-level target on the VSC7448, e.g. DEV1G[0]
$ humility monorail dump DEV1G[0]
humility: attached to 0483:374f:002A001C4D46500F20373033 via ST-Link V3
Dumping target DEV1G[0] (0x71040000 -> 0x710400a0)
DEV1G[0]:DEV_CFG_STATUS:DEV_RST_CTRL 0x00100000
DEV1G[0]:DEV_CFG_STATUS:DEV_STICKY 0x00004000
DEV1G[0]:DEV_CFG_STATUS:DEV_DBG_CFG 0x00000800
DEV1G[0]:DEV_CFG_STATUS:DEV_PORT_PROTECT 0x00000000
DEV1G[0]:DEV_CFG_STATUS:EEE_CFG 0x0011940a
DEV1G[0]:DEV_CFG_STATUS:PTP_CFG 0x00400000
DEV1G[0]:DEV_CFG_STATUS:PTP_EVENTS 0x00000000
...
This subcommand is rather fragile; large targets may overflow the HIF return stack, and it's possible to access invalid registers.
Prints the MAC table of the VSC7448 switch. This table shows which MAC addresses have be learned on each port.
$ humility monorail mac
humility: attached to 0483:374f:002A001C4D46500F20373033 via ST-Link V3
Reading 3 MAC addresses...
PORT | MAC
------|-------------------
18 | 0e:1d:23:88:1a:3b
41 | 0e:1d:7f:c3:07:31
48 | 0e:1d:15:70:3d:bb
Prints or resets (with -r
) counters for a port on the VSC7448.
$ humility monorail counters -p48
humility: attached to 0483:374f:002A001C4D46500F20373033 via ST-Link V3
Packet counters: (port 48)
Receive:
Unicast: 0
Multicast: 1049
Broadcast: 0
Transmit:
Unicast: 0
Multicast: 2099
Broadcast: 0
Subcommand to interact with PHYs on a per-port basis. Supports read
,
write
, info
, and dump
sub-subcommands, which behave similarly to the
commands to interact with VSC7448 registers.
PHY register names are also found in the
vsc7448-pac
crate.
humility net
exposes commands to interact with the management network from
the client's perspective.
(It is the counterpart to humility monorail
, which interacts with the
management network from the switch's perspective.)
It is fully functional on
- Gimlet
- Sidecar
- PSC
gimletlet-mgmt
These PCAs have the KSZ8463 switch + VSC85x2 PHY which is our standard management network interface.
The ip
subcommand works on any image with network functionality, and
prints the MAC and IPv6 address of the board. Note that on boards with
multiple ports, this is the lowest MAC / IPv6 address.
This subcommand prints the MAC table from the KSZ8463 switch. It is functional on the boards listed above and the Gimletlet (when the NIC daughterboard is installed)
This subcommand shows the status of the management network links. It is only functional on the fully supported boards listed above.
This subcommand shows the internal counters on the management network links.
It is only functional on the fully supported boards listed above. In
addition, the MAC-side counters are only supported on the VSC8562, not the
VSC8552; this is indicated with --
in the relevant table positions.
This command launches OpenOCD based on the config file in a build archive,
which then allows one to connect with either GDB or directly via telnet.
If the intention is to only run GDB, note that humility gdb --run-openocd
will both run OpenOCD and run a foreground GDB that is connected to it.
Operates on PMBus devices in the system. To list all PMBus devices, use
--list
(-l
):
$ humility pmbus --list
C P MUX ADDR DEVICE RAILS
3 H - 0x24 tps546b24a V3P3_SP_A2
3 H - 0x26 tps546b24a V3P3_SYS_A0
3 H - 0x27 tps546b24a V5_SYS_A2
3 H - 0x29 tps546b24a V1P8_SYS_A2
3 H - 0x5a raa229618 VDD_VCORE, VDD_MEM_ABCD
3 H - 0x5b raa229618 VDDCR_SOC, VDD_MEM_EFGH
3 H - 0x5c isl68224 VPP_ABCD, VPP_EFGH, V1P8_SP3
4 F - 0x10 adm1272 V54_FAN
4 F - 0x14 adm1272 V54_HS_OUTPUT
4 F - 0x25 tps546b24a V0P96_NIC_VDD_A0HP
4 F - 0x67 bmr491 V12_SYS_A2
To query a particular device, specify the device by its I2C identity (controller/port, segment, address):
$ humility pmbus -d 0x67 -c 4 -p f
0x01 OPERATION 0x84
0x02 ON_OFF_CONFIG 0x18
0x10 WRITE_PROTECT 0x00
0x19 CAPABILITY 0xb0
0x20 VOUT_MODE 0x15
0x21 VOUT_COMMAND 0x6000 = 12.000V
0x22 VOUT_TRIM 0x0000 = 0.000V
0x23 VOUT_CAL_OFFSET 0xffb4 = 31.963V
0x24 VOUT_MAX 0x7333 = 14.400V
0x25 VOUT_MARGIN_HIGH 0x699a = 13.200V
0x26 VOUT_MARGIN_LOW 0x5666 = 10.800V
0x27 VOUT_TRANSITION_RATE 0x9b02 = 0.094V/ms
0x28 VOUT_DROOP 0xe800 = 0.000mV/A
...
In the unusual case that a device is unknown to the system (that is, it does
not appear in humility manifest
), you can force a particular PMBus driver
by using --driver
(-D
).
For the common case of devices known to the system, you can specify a device
by name if it matches a single device in the system (e.g., humility pmbus -d bmr491
). In lieu of specifying a device, you can specify a rail via
--rail
(-r
), e.g.:
$ humility pmbus --rail VDD_MEM_EFGH
humility: attached via ST-Link V3
0x00 PAGE 0x01 = 1
0x01 OPERATION 0x48
0x02 ON_OFF_CONFIG 0x16
0x04 PHASE 0x00 = 0
0x07 ZONE_CONFIG 0xfefe
0x08 ZONE_ACTIVE 0x0000
0x10 WRITE_PROTECT 0x00
0x19 CAPABILITY 0xd4
0x20 VOUT_MODE 0x40
0x21 VOUT_COMMAND 0x04ce = 1.230V
0x22 VOUT_TRIM 0x0000 = 0.000V
0x23 VOUT_CAL_OFFSET 0x0000 = 0.000V
0x24 VOUT_MAX 0x05dc = 1.500V
0x25 VOUT_MARGIN_HIGH 0x03b1 = 0.945V
0x26 VOUT_MARGIN_LOW 0x0357 = 0.855V
0x27 VOUT_TRANSITION_RATE 0x09c4 = 0.025V/ÎĽs
0x28 VOUT_DROOP 0x0014 = 0.200mV/A
...
You can specify a command (or commands) to run with --command
(-C
);
this can be useful when paired with the --verbose
(-v
) flag to
deconstruct a particular command result:
$ humility pmbus -r VDD_MEM_ABCD --command STATUS_WORD --verbose
humility: attached via ST-Link V3
0x79 STATUS_WORD 0x0000
|
| b15 0b0 = no fault <= OutputVoltageFault
| b14 0b0 = no fault <= OutputCurrentFault
| b13 0b0 = no fault <= InputFault
| b12 0b0 = no fault <= ManufacturerFault
| b11 0b0 = POWER_GOOD set <= PowerGoodStatus
| b10 0b0 = no fault <= FanFault
| b9 0b0 = no fault <= OtherFault
| b8 0b0 = no fault <= UnknownFault
| b7 0b0 = no fault <= Busy
| b6 0b0 = power is not off <= Off
| b5 0b0 = no fault <= OutputOvervoltageFault
| b4 0b0 = no fault <= OutputOvercurrentFault
| b3 0b0 = no fault <= InputUndervoltageFault
| b2 0b0 = no fault <= TemperatureFault
| b1 0b0 = no fault <= CMLFault
| b0 0b0 = no fault <= NoneOfTheAbove
+-----------------------------------------------------------------------
You can also write a PMBus command with --write
(-w
), which allows for
for particular fields to be written, e.g.:
$ humility pmbus -r VDD_VCORE -w OPERATION.MarginFaultResponse=ActUpon
humility: attached via ST-Link V3
humility: I2C3, port H, dev 0x5a, rail 0: successfully wrote OPERATION
It should go without saying that this should be done carefully!
To get specific help on what fields may be written and the legal values for
those fields, use --commandhelp
(-H
):
$ humility pmbus -r VDD_VCORE --commandhelp OPERATION
0x01 OPERATION
| b7 OnOffState <= On/off state
| 0b0 = Off <- output off
| 0b1 = On <- output on
| b6 TurnOffBehavior <= Power down behavior
| 0b0 = Clear <- powers down immediately
| 0b1 = Set <- powers down based on TOFF_DELAY
| b5:4 VoltageCommandSource <= Source of output voltage
| 0b00 = VOUT_COMMAND <- set by VOUT_COMMAND
| 0b01 = VOUT_MARGIN_LOW <- set by VOUT_MARGIN_LOW
| 0b10 = VOUT_MARGIN_HIGH <- set by VOUT_MARGIN_HIGH
| 0b11 = AVS_VOUT_COMMAND <- set by AVSBus
| b3:2 MarginFaultResponse <= Margin fault response
| 0b01 = Ignore <- ignore margin faults
| 0b10 = ActUpon <- act upon margin faults
| b1 TransitionControl <= Transition control
| 0b0 = Clear <- not set
| 0b1 = Set <- set
+-----------------------------------------------------------------------
To get a summary of all PMBus rails in the system, use --summarize
(-s
):
$ humility pmbus --summarize
humility: attached via ST-Link V3
DEVICE RAIL PG? #FLT VIN VOUT IOUT TEMP_1
tps546b24a V3P3_SP_A2 Y 0 11.969V 3.311V 0.404A 33.500°C
tps546b24a V3P3_SYS_A0 Y 0 11.969V 3.309V 1.457A 36.250°C
tps546b24a V5_SYS_A2 Y 0 11.969V 4.982V 0.518A 36.000°C
tps546b24a V1P8_SYS_A2 Y 0 11.984V 1.797V 3.316A 34.750°C
raa229618 VDD_VCORE Y 0 11.990V 1.181V 53.200A 40.000°C
raa229618 VDD_MEM_ABCD Y 0 11.990V 1.224V 28.600A 41.000°C
raa229618 VDDCR_SOC Y 0 11.950V 0.890V 18.200A 42.000°C
raa229618 VDD_MEM_EFGH Y 0 11.950V 1.223V 27.400A 43.000°C
isl68224 VPP_ABCD Y 0 11.950V 2.500V 0.400A 40.000°C
isl68224 VPP_EFGH Y 0 11.950V 2.499V 0.500A 30.000°C
isl68224 V1P8_SP3 Y 0 11.950V 1.796V 1.600A 0.000°C
adm1272 V54_FAN Y 4 0x088c 0x088c 0x0810 30.452°C
adm1272 V54_HS_OUTPUT Y 4 0x088c 0x088b 0x092e 30.690°C
tps546b24a V0P96_NIC_VDD_A0HP Y 0 11.984V 0.955V 3.074A 37.500°C
bmr491 V12_SYS_A2 Y 1 53.625V 11.995V 19.250A 35.750°C
Note that for some devices, it is not possible to get accurate voltage and
current readings from pmbus
alone, as knowledge of how the device is
integrated into a larger system is required to interpret raw values. For
these devices, the raw value is provided (as in the adm1272
output,
above); to get the interpreted value, use humility power
instead (which
has the added advantage of displaying all power rails in the systemn, not
just PMBus devices.)
humility pmbus
can use two different mechanisms to perform PMBus actions,
selected by the --agent
command-line argument.
--agent=i2c
uses direct construction and execution of raw I2C commands. This only works when connected to the target via a debugger.--agent=idol
uses Idol operations, which can be executed over the network. This could also be used (in theory) when connected with a debugger; in practice, the Idol operations have a larger encoding compared to raw I2C operations, so the HIF program may not fit.--agent=auto
(the default) selectsi2c
if we're connected with a debugger oridol
if we're connected over the network.
In practice, allowing humility pmbus
to select the agent is almost always
what you want.
humility power
displays the values associated with power rails,
displaying output voltage, output current, input voltate, input current,
and temperature. Not all measurements are available for all rails; if a
measurement is not provided for a given rail, "-" is printed. To specify
which rails are displayed, use the --rails
option. If a rail can
provide a given measurement, but that measurement is unavailable (say, due
to being in a power state whereby the rail is not powered), "x" is
displayed. Some rails can determine current by output phase; to display
these, use the --phase-current
option.
humility powershelf
allows for remotely dumping the state of the PSC
power shelves.
This command is currently hard-coded to support only the MWOCP68, and it dumps 50+ properties described in the ACAN-114 application note. It will only dump the properties from a single shelf+rail combination, so seeing properties of all 6 shelves requires calling this command 6 time (with indices 0 through 5).
humility probe
attempts to infer as much about the hardware state as it
can, e.g.:
$ humility probe
humility: attached via ST-Link
humility: probe => STLink V3, VID 0483, PID 374e
humility: probe serial => 003700303137511139383538
humility: core => Cortex-M7
humility: manufacturer => STMicroelectronics
humility: chip => STM32H7, revision 0x2003
humility: status => executing
humility: debug units => CSTF(x2) CTI(x2) DWT ETM FPB ITM SCS SWO TMC TPIU
humility: CSTF => 0x5c004000, 0x5c013000
humility: CTI => 0x5c011000, 0xe0043000
humility: DWT => 0xe0001000
humility: ETM => 0xe0041000
humility: FPB => 0xe0002000
humility: ITM => 0xe0000000
humility: SCS => 0xe000e000
humility: SWO => 0x5c003000
humility: TMC => 0x5c014000
humility: TPIU => 0x5c015000
humility: ITM status => TRCENA enabled, TCR disabled, TER=0x0
humility: R0 => 0x20006000
humility: R1 => 0x20006000
humility: R2 => 0x0
humility: R3 => 0x0
humility: R4 => 0x0
humility: R5 => 0x0
humility: R6 => 0x0
humility: R7 => 0x0
humility: R8 => 0x0
humility: R9 => 0x0
humility: R10 => 0x0
humility: R11 => 0x0
humility: R12 => 0x0
humility: SP => 0x20006100
humility: LR => 0x802404f
humility: PC => 0x8024052
humility: xPSR => 0x61000000
humility: MSP => 0x20000f48
humility: PSP => 0x20006100
humility: SPR => 0x7000000
If provided a Hubris archive, humility probe
will display any register
contents symbolically, e.g.:
$ humility -a ~/hubris/target/demo/dist/build-demo.zip probe
humility: attached via ST-Link
humility: probe => STLink V2-1, VID 0483, PID 374b
humility: probe serial => 066DFF383032534E43132614
humility: core => Cortex-M4
humility: manufacturer => STMicroelectronics
humility: chip => STM32F40x/STM32F41x, revision 0x1007
humility: debug units => DWT ETM FPB ITM SCS TPIU
humility: status => executing
humility: ITM => TRCENA enabled, TCR enabled, TER=0x3
humility: R0 => 0x0
humility: R1 => 0x0
humility: R2 => 0x1
humility: R3 => 0x20001bd4
humility: R4 => 0x20001bd4
humility: R5 => 0x801d988
humility: R6 => 0xb004
humility: R7 => 0x20001bf0
humility: R8 => 0x40004400
humility: R9 => 0x1
humility: R10 => 0x0
humility: R11 => 0xffff
humility: R12 => 0x0
humility: SP => 0x20001ba8
humility: LR => 0x801c12b <- main+0xef
humility: PC => 0x801d290 <- sys_recv_stub+0x1e
humility: xPSR => 0x61000000
humility: MSP => 0x20000f48
humility: PSP => 0x20001ba8
humility: SPR => 0x7000000
humility qspi
manipulates (and importantly, writes to) QSPI-attached
flash in Hubris. To read the device identifier, use the --id
(-i
)
option:
$ humility qspi -i
humility: attached via ST-Link V3
DeviceIdData {
manufacturer_id: Micron(0x20)
memory_type: 3V(186)
memory_capacity: 33554432
uid_n: 16
# If a Micron device:
ext_device_id: 0b1000100
2nd device generation,
Standard BP scheme,
HOLD#/RESET#=HOLD
Additional HW RESET# is available,
Sector size is Uniform 64KB,
device_configuration_info: 0
uid: [9a, ec, 0b, 00, 19, f9, ff$, 39, 00, be, 69, 97, f4, a2]
}
[Ok([20, ba, 19, 10, 44, 0, 9a, ec, b, 0, 19, f9, ff, 39, 0, be, 69, 97, f4, a2])]
To write an image from a file, use the --writefile
(-W
) option:
$ humility -W ./milan-spew-115k2-2dpc-0.4.1-dataeye.bin
humility: attached via ST-Link V3
humility: erasing 16777216 bytes...
humility: ... done
humility: flashed 16.00MB in 5 minutes
If writing similar images, it is much faster to write only those blocks
that differ. To perform a differential write, use the --diffwrite
(-D
)
option:
$ humility qspi -D ./milan-spew-115k2-2dpc-0.4.1.bin
humility: attached via ST-Link V3
humility: erasing 65536 bytes...
humility: ... done
humility: hashed 16.00MB, wrote 64.00KB in 16 seconds
To read, write or hash a particular region, use the --read
(-r
),
--write
(-w
), or --hash
(-H
) respectively -- giving the address
via --address
(-a
) and the number of bytes via --nbytes
(-n
).
For example, to read 128 bytes from address 0x120000:
$ humility qspi -r -a 0x120000 -n 128
\/ 1 2 3 4 5 6 7 8 9 a b c d e f
0x00120000 | 24 50 53 50 ee 7a 31 28 10 00 00 00 20 cd 48 20 | $PSP.z1(.... .H
0x00120010 | 00 00 00 00 40 04 00 00 00 30 12 00 00 00 00 00 | ....@....0......
0x00120020 | 01 00 00 00 00 54 01 00 00 40 12 00 00 00 00 00 | .....T...@......
0x00120030 | 03 00 00 00 00 54 01 00 00 a0 13 00 00 00 00 00 | .....T..........
0x00120040 | 08 00 00 00 40 ea 01 00 00 00 15 00 00 00 00 00 | ....@...........
0x00120050 | 09 00 00 00 40 06 00 00 00 f0 16 00 00 00 00 00 | ....@...........
0x00120060 | 0a 00 00 00 40 06 00 00 00 00 17 00 00 00 00 00 | ....@...........
0x00120070 | 0b 00 00 00 ff ff ff ff 01 00 00 00 00 00 00 00 | ................
To get the SHA256 hash for that same region:
$ humility qspi -H -a 0x120000 -n 128
humility: attached via ST-Link V3
120000..000080: 4d07112733efe240f990fad785726c52de4335d6c5c30a33e60096d4c2576742
By default, Hubris reserved sector 0 (the lowest 64 KiB) of QSPI for its own
internal bookkeeping. To override this, use the --write-sector0
flag;
this is required for erases and writes that would otherwise modify sector 0,
as well as bulk erase.
humility readmem
allows one to read a specified range of memory:
$ humility readmem 0x00011b00
humility: attached via DAPLink
humility: reading at 0x11b00 for 256 bytes
\/ 1 2 3 4 5 6 7 8 9 a b c d e f
0x00011b00 | 00 0f 1a bf 81 54 bc f1 01 0f d0 bd 02 44 bc f1 | .....T.......D..
0x00011b10 | 02 0f 51 70 00 d1 d0 bd 91 70 d0 bd 69 00 01 00 | ..Qp.....p..i...
0x00011b20 | 04 00 00 00 04 00 00 00 17 01 01 00 6b 00 01 00 | ............k...
0x00011b30 | f1 00 01 00 65 78 70 6c 69 63 69 74 20 70 61 6e | ....explicit pan
0x00011b40 | 69 63 00 00 3c 1f 01 00 49 00 00 00 0a 00 00 00 | ic..<...I.......
0x00011b50 | 09 00 00 00 76 69 76 61 20 65 6c 20 6a 65 66 65 | ....viva el jefe
0x00011b60 | 0a 54 61 73 6b 20 23 20 50 61 6e 69 63 21 0a 00 | .Task # Panic!..
0x00011b70 | 61 1b 01 00 06 00 00 00 67 1b 01 00 08 00 00 00 | a.......g.......
0x00011b80 | 20 42 61 64 20 53 79 73 63 61 6c 6c 20 55 73 61 | Bad Syscall Usa
0x00011b90 | 67 65 20 0a 61 1b 01 00 06 00 00 00 80 1b 01 00 | ge .a...........
0x00011ba0 | 13 00 00 00 93 1b 01 00 01 00 00 00 20 53 74 61 | ............ Sta
0x00011bb0 | 63 6b 20 6f 76 65 72 66 6c 6f 77 20 61 74 20 61 | ck overflow at a
0x00011bc0 | 64 64 72 65 73 73 20 30 78 00 00 00 61 1b 01 00 | ddress 0x...a...
0x00011bd0 | 06 00 00 00 ac 1b 01 00 1d 00 00 00 93 1b 01 00 | ................
0x00011be0 | 01 00 00 00 20 4d 65 6d 6f 72 79 20 66 61 75 6c | .... Memory faul
0x00011bf0 | 74 20 61 74 20 61 64 64 72 65 73 73 20 30 78 00 | t at address 0x.
If an argument is present, it is the number of bytes to read:
$ humility readmem 0x00011d00 100
humility: attached via DAPLink
humility: reading at 0x11d00 for 100 bytes
\/ 1 2 3 4 5 6 7 8 9 a b c d e f
0x00011d00 | 20 62 6f 75 6e 64 73 3a 20 74 68 65 20 6c 65 6e | bounds: the len
0x00011d10 | 20 69 73 20 20 62 75 74 20 74 68 65 20 69 6e 64 | is but the ind
0x00011d20 | 65 78 20 69 73 20 30 30 30 31 30 32 30 33 30 34 | ex is 0001020304
0x00011d30 | 30 35 30 36 30 37 30 38 30 39 31 30 31 31 31 32 | 0506070809101112
0x00011d40 | 31 33 31 34 31 35 31 36 31 37 31 38 31 39 32 30 | 1314151617181920
0x00011d50 | 32 31 32 32 32 33 32 34 32 35 32 36 32 37 32 38 | 2122232425262728
0x00011d60 | 32 39 33 30 | 2930
Both arguments can be in either hex, decimal, octal or binary (addresses and contents will always be printed in hex):
humility readmem 0o216401 0b110
humility: attached via DAPLink
humility: reading at 0x11d01 for 6 bytes
0 \/ 2 3 4 5 6 7 8 9 a b c d e f
0x00011d00 | 62 6f 75 6e 64 73 | bounds
The length argument can have an optional size suffix. Note that "k" is used to to denote the SI kilobytes (that is, 1000 bytes); if one wishes to have a multiples of 1024 bytes (a kibibyte), "KiB" should be used instead.
To display as half-words (16-bits) use -h
; to display as words (32-bits)
use -w
. (The addresses must be 2-byte and 4-byte aligned, respectively.)
$ humility readmem -w 0x20000000 0x40
humility: attached via DAPLink
humility: reading at 0x20000000 for 64 bytes
\/ 4 8 c
0x20000000 | 00000001 20000180 0000000b 00005020 | ....... .... P..
0x20000010 | 00000002 200001f0 00838042 00000000 | ....... B.......
0x20000020 | 00004db8 00004dc8 00004d28 00004d28 | .M...M..(M..(M..
0x20000030 | 00004d28 00004d28 00004d28 00004d28 | (M..(M..(M..(M..
A frequent use of readmem
is to read peripheral memory; as a
convenience, a peripheral name can be used in lieu of an address, provided
that an archive or dump is also specified:
$ humility -a ~/hubris/target/gemini-bu/dist/build-gemini-bu.zip readmem -w i2c4
0x30
humility: attached via ST-Link
\/ 4 8 c
0x58001c00 | 000000f7 00020490 00008000 00008000 | ................
0x58001c10 | 10c0ecff 00000000 00000021 00000000 | ........!.......
0x58001c20 | 00000000 00000080 00000000 00000000 | ................
Note that reading some peripheral memory may have side effects!
It can also be useful to interpret memory contents symbolically; to do this,
provide a dump or achive and specify the -s
option, e.g.:
$ humility -a ~/hubris/target/gemini-bu-rot/dist/build-gemini-bu-rot.zip readmem -s 0x20004b30 0x40
humility: attached via DAPLink
0x20004b30 | 0x20004bdc
0x20004b34 | 0x0000000a <- __EXCEPTIONS+0x2
0x20004b38 | 0x80000000
0x20004b3c | 0x00002e2e <- syscall_entry+0xcf6
0x20004b40 | 0x00000000
0x20004b44 | 0x0003c2e3 <- spi:main+0x5b
0x20004b48 | 0x0003ceda <- spi:sys_send_stub+0xe
0x20004b4c | 0x01000000
0x20004b50 | 0x00000000
0x20004b54 | 0x00000000
0x20004b58 | 0x00000000
0x20004b5c | 0x00000000
0x20004b60 | 0x00000000
0x20004b64 | 0x00000000
0x20004b68 | 0x00000000
0x20004b6c | 0x00000000
humility readvar
allows one to read a global static variable.
To list all such variables, use the -l
option:
$ humility readvar -l
humility: MODULE VARIABLE ADDR SIZE
humility: kernel CORE_PERIPHERALS 0x20000000 1
humility: kernel CURRENT_TASK_PTR 0x20000018 4
humility: kernel DEVICE_PERIPHERALS 0x20000001 1
humility: kernel FAULT_NOTIFICATION 0x20000004 4
humility: kernel IRQ_TABLE_BASE 0x20000010 4
humility: kernel IRQ_TABLE_SIZE 0x20000014 4
humility: kernel TASK_TABLE_BASE 0x20000008 4
humility: kernel TASK_TABLE_SIZE 0x2000000c 4
humility: kernel __EXCEPTIONS 0x08000008 56
humility: kernel __INTERRUPTS 0x08000040 620
humility: kernel __RESET_VECTOR 0x08000004 4
humility: adt7420 TEMPS_BYMINUTE 0x2000b848 17288
humility: adt7420 TEMPS_BYSECOND 0x20008000 14408
To read a variable, specify it:
$ humility readvar CURRENT_TASK_PTR
humility: attached via ST-Link
CURRENT_TASK_PTR (0x20000018) = Some(NonNull<kern::task::Task> {
pointer: 0x20000558 (*const kern::task::Task)
})
Recovers an LPC55 target by reflashing the CFPA, CMPA, and bootleby image. This will not work if the target has a locked CMPA or SWD access has been disabled!
!!! Using this can be dangerous and should be undertaken with caution
humility registers
displays the registers from either a live system
or a dump, e.g.:
$ humility registers
humility: attached via ST-Link V3
R0 = 0x00000000
R1 = 0x0000000a
R2 = 0x80000000
R3 = 0x00000000
R4 = 0x00000000
R5 = 0x0000f406
R6 = 0x00002711
R7 = 0x20000310
R8 = 0x00000000
R9 = 0x00000000
R10 = 0x20000f68
R11 = 0x00000001
R12 = 0x200002b4
SP = 0x200002e8
LR = 0x0800414f
PC = 0x08004236
PSR = 0x4100000f <- 0100_0001_0000_0000_0000_0000_0000_1111
|||| | || | | |
|||| | || | | + Exception = 0xf
|||| | || | +------------ IC/IT = 0x0
|||| | || +-------------------- GE = 0x0
|||| | |+------------------------------ T = 1
|||| | +------------------------------- IC/IT = 0x0
|||| +--------------------------------- Q = 0
|||+----------------------------------- V = 0
||+------------------------------------ C = 0
|+------------------------------------- Z = 1
+-------------------------------------- N = 0
MSP = 0x200002e8
PSP = 0x20011ab0
SPR = 0x01000001 <- 0000_0001_0000_0000_0000_0000_0000_0001
||| | | |
||| | | + PRIMASK = 1
||| | +---------- BASEPRI = 0x0
||| +-------------------- FAULTMASK = 0
||+------------------------------ CONTROL.nPRIV = 1
|+------------------------------- CONTROL.SPSEL = 0
+-------------------------------- CONTROL.FPCA = 0
FPSCR = 0x00000000
If an archive is provided or if displaying registers from a dump, the symbol that corresponds to register's value (if any) is displayed, e.g.:
$ humility -d ./hubris.core.81 registers
humility: attached to dump
R0 = 0x00000000
R1 = 0x0000000a
R2 = 0x80000000
R3 = 0x00000000
R4 = 0x00000000
R5 = 0x0000f406
R6 = 0x00002711
R7 = 0x20000310 <- kernel: 0x20000000+0x310
R8 = 0x00000000
R9 = 0x00000000
R10 = 0x20000f68 <- kernel: DEVICE_PERIPHERALS+0x0
R11 = 0x00000001
R12 = 0x200002b4 <- kernel: 0x20000000+0x2b4
SP = 0x200002e8 <- kernel: 0x20000000+0x2e8
LR = 0x0800414f <- kernel: write_str<cortex_m::itm::Port>+0xd
PC = 0x08004236 <- kernel: panic+0x36
PSR = 0x4100000f <- 0100_0001_0000_0000_0000_0000_0000_1111
|||| | || | | |
|||| | || | | + Exception = 0xf
|||| | || | +------------ IC/IT = 0x0
|||| | || +-------------------- GE = 0x0
|||| | |+------------------------------ T = 1
|||| | +------------------------------- IC/IT = 0x0
|||| +--------------------------------- Q = 0
|||+----------------------------------- V = 0
||+------------------------------------ C = 0
|+------------------------------------- Z = 1
+-------------------------------------- N = 0
MSP = 0x200002e8 <- kernel: 0x20000000+0x2e8
PSP = 0x20011ab0 <- pong: 0x20011800+0x2b0
SPR = 0x01000001 <- 0000_0001_0000_0000_0000_0000_0000_0001
||| | | |
||| | | + PRIMASK = 1
||| | +---------- BASEPRI = 0x0
||| +-------------------- FAULTMASK = 0
||+------------------------------ CONTROL.nPRIV = 1
|+------------------------------- CONTROL.SPSEL = 0
+-------------------------------- CONTROL.FPCA = 0
To display a stack backtrace, use the --stack
(-s
) option:
$ humility -d ./hubris.core.81 registers --stack
humility: attached to dump
R0 = 0x00000000
...
R10 = 0x20000f68 <- kernel: DEVICE_PERIPHERALS+0x0
R11 = 0x00000001
R12 = 0x200002b4 <- kernel: 0x20000000+0x2b4
SP = 0x200002e8 <- kernel: 0x20000000+0x2e8
|
+---> 0x20000318 0x08004236 rust_begin_unwind
0x20000330 0x08000558 core::panicking::panic_fmt
0x20000358 0x08000ad8 core::panicking::panic
0x20000390 0x08003ba6 kern::arch::arm_m::safe_sys_tick_handler
0x20000390 0x08003ba6 kern::arch::arm_m::SysTick::{{closure}}
0x20000390 0x08003ba6 kern::arch::arm_m::with_task_table
0x20000390 0x08003bb0 SysTick
LR = 0x0800414f <- kernel: write_str<cortex_m::itm::Port>+0xd
...
To additionally display floating point registers on platforms that support
floating point, use the --floating-point
(-f
) option.
Query the Renesas 8A3400X ClockMatrix part -- or process a trace from Renesas configuration software.
humility rendmp
allows for querying, dumping and flashing the Renesas
Digital Multiphase family of voltage regulators.
Like humility pmbus
, a device can be specified in terms of an
address (which requires the further specification of a controller and
port) or a PMbus rail.
To view the number of NVM OTP slots that remain, use the --slots
option:
$ humility rendmp -b mid -d 0x5a --slots
humility: attached via ST-Link V3
humility: RAA229618 at I2C3, port H, dev 0x5a has 28 slots available
To determine the OTP CRC, use the --crc
option. Note that an
entirely unprogrammed part generally has a CRC of 0:
$ humility rendmp -b mid -d 0x5a --crc
humility: attached via ST-Link V3
humility: RAA229618 at I2C3, port H, dev 0x5a has CRC 0x00000000
To flash a part (that is, to program its one-time programmable non-volatile memory), specify the HEX file as generated by the Renesas PowerNavigator. Note that this file specifies the address of the device; mismatches are not permitted, e.g.:
$ humility rendmp -b mid -d 0x5b --flash ./raa229618-0x5a.hex
humility: attached via ST-Link V3
humility rendmp failed: image specifies address to be 0x5a; can't flash 0x5b
Specify the proper device to flash:
$ humility rendmp -b mid -d 0x5a --flash ./raa229618-0x5a.hex
humility: attached via ST-Link V3
humility: 28 NVM slots remain
humility: flashing 2871 bytes
humiility: flashed 2.80KB in 4 seconds
humility: bank 0: bank written successfully
humility: flashed successfully after 246 ms; power cycle to load new configuration
To check a configuration, specify the image and the --check
option:
$ humility rendmp -b mid -d 0x5c -f ./isl68224-0x5c.hex --check
humility: attached via ST-Link V3
humility: 27 NVM slots remain
humility: image CRC (0x841f35a5) matches OTP CRC
The Renesas voltage regulators include a black box which stores fault
information. This can be queried using the --blackbox
subcommand,
specifying a device (I2C) address to pick a specific power converter:
$ humility rendmp --blackbox --device=0x5b
humility: attached to 0483:374f:000C001F4D46500F20373033 via ST-Link V3
rail0 uptime: 0 sec
rail1 uptime: 0 sec
controller fault: 0
rail0 fault: 00000000000000000000000000000000 ()
rail1 fault: 00000000000000000000000000000000 ()
phase fault uc: 00000000000000000000000000000000 ()
phase fault oc: 00000000000000000000000000000000 ()
adc fault uc: 00000000000000000000000000000000 ()
adc fault oc: 00000000000000000000000000000000 ()
rail0 status: 0000000000000000 ()
rail1 status: 0000000000000000 ()
status cml: 00000000 ()
status mfr: 00000000 ()
rail1 status vout: 00000000 ()
rail0 status vout: 00000000 ()
rail1 status iout: 00000000 ()
rail0 status iout: 00000000 ()
rail1 status temperature: 00000000 ()
rail0 status temperature: 00000000 ()
rail1 status input: 00000000 ()
rail0 status input: 00000000 ()
| RAIL 0 | RAIL 1
-----|---------|-----------
VIN | 0.00 V | 0.00 V
VOUT | 0.000 V | 0.000 V
IIN | 0.00 A | 0.00 A
IOUT | 0.0 A | 0.0 A
TEMP | 0°C | 0°C
controller read temperature: 0°C
PHASE | TEMPERATURE | CURRENT
-------|-------------|----------
0 | 0°C | 0.0 A
1 | 0°C | 0.0 A
2 | 0°C | 0.0 A
3 | 0°C | 0.0 A
4 | 0°C | 0.0 A
5 | 0°C | 0.0 A
6 | 0°C | 0.0 A
7 | 0°C | 0.0 A
8 | 0°C | 0.0 A
9 | 0°C | 0.0 A
10 | 0°C | 0.0 A
11 | 0°C | 0.0 A
12 | 0°C | 0.0 A
13 | 0°C | 0.0 A
14 | 0°C | 0.0 A
15 | 0°C | 0.0 A
16 | 0°C | 0.0 A
17 | 0°C | 0.0 A
18 | 0°C | 0.0 A
19 | 0°C | 0.0 A
(In the example above, there have been no faults so the blackbox is empty)
To check individual phases for errors, use the --phase-check
subcommand:
$ humility rendmp --phase-check --device=0x5c
humility: attached to 0483:3754:000D00344741500820383733 via ST-Link V3
Phase check for ISL68221 at 0x5c
PHASE | VOUT | IOUT | TEMP | RAIL
------|------------|------------|------------|------------
0 | 0.441V | -56.100A | 25.000°C | VPP_ABCD
1 | 0.448V | -55.800A | 26.000°C | VPP_EFGH
2 | 0.453V | 0.900A | 0.000°C | V1P8_SP3
This must be run with the system in the A2 power state, and on a machine
with no DIMMs. This latter constraint can be overridden with
--force-phase-check
, but this should be used carefully: if a checked
phase powers a DIMM, an I2C hang can result.
humility repl
is an interactive prompt that you can use with humility.
This allows you to run several commands in succession without needing to type in some core settings over and over again.
humility repl
takes the same top level arguments as any other subcommand,
and will remember them inside of the prompt. For example:
$ humility -a ../path/to/hubris/archive.zip repl
humility: attached via ST-Link V2-1
Welcome to the humility REPL! Try out some subcommands, or 'quit' to quit!
humility> tasks
system time = 7209837
ID TASK GEN PRI STATE
0 jefe 0 0 recv, notif: bit0 bit1(T+63)
1 rcc_driver 0 1 recv
2 usart_driver 0 2 RUNNING
3 user_leds 0 2 recv
4 ping 47524 4 wait: reply from usart_driver/gen0
5 pong 0 3 recv, notif: bit0(T+163)
6 hiffy 0 3 notif: bit31(T+121)
7 idle 0 5 ready
humility> tasks
system time = 7212972
ID TASK GEN PRI STATE
0 jefe 0 0 recv, notif: bit0 bit1(T+28)
1 rcc_driver 0 1 recv
2 usart_driver 0 2 recv, notif: bit0(irq38)
3 user_leds 0 2 recv
4 ping 47544 4 RUNNING
5 pong 0 3 recv, notif: bit0(T+28)
6 hiffy 0 3 notif: bit31(T+252)
7 idle 0 5 ready
humility> quit
Quitting!
As you can see, we can run the tasks
subcommand twice, without passing our
archive each time. In the output above, you can see the ping task faulting
in the background; your code is still running in the background while you
use the repl!
Finally, as you can see, quit
will quit the repl. There is also a
history
command, which will show you recent commands you've put into the
prompt.
The repl is still very early days! We'll be adding more features in the future.
humility reset
will hard reset the system using the debug pin
or using software reset with the appropriate flag
humility ringbuf
reads and displays any Hubris ring buffers (as created
via the ringbuf!
macro in the Hubris ringbuf
crate). e.g.:
$ humility -d ./hubris.core.5 ringbuf
ADDR NDX LINE GEN COUNT PAYLOAD
0x2000a288 552 92 1 5 (21.5, 70.69999694824219)
0x2000a298 553 92 1 1 (21.4375, 70.58749389648438)
0x2000a2a8 554 92 1 1 (21.5, 70.69999694824219)
0x2000a2b8 555 92 1 1 (21.4375, 70.58749389648438)
0x2000a2c8 556 92 1 5 (21.5, 70.69999694824219)
0x2000a2d8 557 92 1 1 (21.5625, 70.8125)
0x2000a2e8 558 92 1 15 (21.5, 70.69999694824219)
0x2000a2f8 559 92 1 1 (21.4375, 70.58749389648438)
0x2000a308 560 92 1 10 (21.5, 70.69999694824219)
0x2000a318 561 92 1 2 (21.4375, 70.58749389648438)
0x2000a328 562 92 1 2 (21.5, 70.69999694824219)
0x2000a338 563 92 1 1 (21.4375, 70.58749389648438)
0x2000a348 564 92 1 9 (21.5, 70.69999694824219)
0x2000a358 565 92 1 3 (21.4375, 70.58749389648438)
0x2000a368 566 92 1 4 (21.5, 70.69999694824219)
0x2000a378 567 92 1 1 (21.4375, 70.58749389648438)
...
If an argument is provided, only ring buffers that have a name that
contains the argument as a substring, or are in a task that contains
the argument as a substring will be displayed. For example,
to display every ring buffer that has i2c
in the name or the
containing task:
$ humility -d ./hubris.core.76 ringbuf ksz
humility: attached to dump
humility: ring buffer ksz8463::__RINGBUF in net:
NDX LINE GEN COUNT PAYLOAD
2 134 89 1 Read(IADR5, 0x4000)
3 134 89 1 Read(IADR4, 0x0)
4 134 89 1 Read(P1MBSR, 0x780c)
5 148 89 1 Write(IACR, 0x1c00)
6 134 89 1 Read(IADR5, 0x4000)
7 134 89 1 Read(IADR4, 0x0)
8 148 89 1 Write(IACR, 0x1c14)
...
By default, all integer values in a ring buffer are displayed in hex;
this can be overridden with the --decimal
option.
Unless explicitly disabled, ring buffer entries are de-deduplicated (with
the count of a specific entry being indicated by the COUNT
column). In
some cases, it can be useful to expand these de-duplicated entries (for
example, if each entry represents a measurement, and one wishes to see the
entire series); this can be effected with the --expand
option.
See the ringbuf
documentation for more details.
humility rpc
allows for execution of Idol commands over a network, rather
than through a debugger.
It requires the Hubris udprpc
task to be listening on port 8. This task
decodes bytes from a UDP packet, and shoves them directly into sys_send
to
a target task.
An archive is required so that humility
knows what functions are available
and how to call them. The archive ID is checked against the image ID on the
target; udprpc
will refuse to execute commands when the ID does not match.
Function calls are handled identically to the humility hiffy
subcommand,
except that an --ip
address is required:
$ humility rpc --ip fe80::0c1d:9aff:fe64:b8c2%en0 -c UserLeds.led_on -aindex=0
UserLeds.led_on() = ()
Alternatively, you can set the HUMILITY_RPC_IP
environmental variable.
You may need to configure an IPv6 network for humility rpc
to work. On
illumos, it looks like this:
$ pfexec ipadm create-addr -t -T addrconf e1000g0/addrconf
To listen for compatible devices on your network, run humility rpc --listen
$ humility rpc --listen
humility: listening... (ctrl-C to stop, or timeout in 5s)
MAC IPv6 COMPAT PART REV SERIAL
a8:40:25:04:02:81 fe80::aa40:25ff:fe04:281 Yes 913-0000019 6 BRM42220066
a8:40:25:05:05:00 fe80::aa40:25ff:fe05:500 No (legacy) 0 (legacy)
a8:40:25:05:05:00 fe80::aa40:25ff:fe05:501 No (legacy) 0 (legacy)
Under the hood, this listens for packets from the Hubris udpbroadcast
task, which includes MAC address and image ID (checked for compatibility).
When listening, it is mandatory to specify the interface (e.g. humility rpc --listen -i en0
on MacOS). If the Part
/ Serial
columns are marked as
(legacy)
, the SP is running an older version of udpbroadcast
that did
not include identity information. If they are marked as (vpdfail)
, they
are running a new-enough udpbroadcast
, but the SP was unable to read its
identity from its VPD.
To call all targets that match an archive, --listen
can be combined with
--call
Print out information retrieved via AMD's sideband remote management interface (SB-RMI). This interface is somewhat limited in its utility, but can sometimes yield information not available via other means.
When run without arguments, this will indicate the presence of each thread; if all are present and not in alert status, one will see output like:
THR 0x0 0x1 0x2 0x3 0x4 0x5 0x6 0x7 0x8 0x9 0xa 0xb 0xc 0xd 0xe 0xf
0x00 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x10 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x20 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x30 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x40 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x50 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x60 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x70 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
If any CPU has its alert bit set, it will be marked as MCE
, and the
machine check exception information will be displayed, e.g.:
THR 0x0 0x1 0x2 0x3 0x4 0x5 0x6 0x7 0x8 0x9 0xa 0xb 0xc 0xd 0xe 0xf
0x00 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x10 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x20 ok ok ok ok ok MCE ok ok ok ok ok ok ok ok ok ok
0x30 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x40 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x50 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x60 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
0x70 ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok ok
=== MCE on thread 0x25 (37), bank 0 ===
MCA_CTL(0) 0x0000000000ffffff
MCA_STATUS(0) 0xb420000001130839
|
+---> Valid (VAL) = true
Overflow (OVER) = false
Uncorrected error (UC) = true
Error condition enabled (EN) = true
Misc error register valid (MISCV) = false
Address register valid (ADDRV) = true
Processor context corrupt (PCC) = false
Task content corrupt (TCC) = false
Syndrome register valid (SYNDV) = true
Deferred error (Deferred) = false
Poisoned data consumed (Poison) = false
Extended Error Code = 0x113
MCA Error Code = 0x839
MCA_ADDR(0) 0x000ffffe14700008
MCA_MISC(0) 0xd010000000000000
MCA_CONFIG(0) 0x00000007000001fd
|
+---> Interrupt Enable (IntEn) = false
Deferred error type = 0x0
MCAX enable (McaxEn) = true
MCA FRU text supported = false
Address LSB in MCA_STATUS = true
Deferred error status supported = true
System fatal error event supported = true
Deferred interrupt type supported = true
Deferred error logging supported = true
MCAX capable = true
MCA_SYND(0) 0x000000005c000002
MCA_IPID(0) 0x001000b000000000
|
+---> MCA bank type = 0x10
Instance ID (high) = 0x0
Hardware ID = 0xb0
Instance ID (low) = 0x0
MCA_DESTAT(0) 0x0000000000000000
MCA_DEADDR(0) 0x0000000000000000
CPU identification information as provided by the cpuid
instruction
can be retrieved by using the --cpuid
option and specifying a
desired target thread; full MCA information can similarly be retrieved
using the --mca
option and specifyin a desired thread.
humility sensors
communicates with the sensor
Hubris task via its
Sensor
Idol interface to get sensor data. If there is no sensor
task
or if there are no sensors defined in the in Hubris application
description, this command will not provide any meaningful output. To list
all available sensors, use -l
(--list
). To constrain sensors by type,
use the -t
(--types
) option; to constrain sensors by device, use the
-d
(--devices
) option; to constrain sensors by name, use the -n
(--named
) option. Within each option, multiple specifications serve as
a logical OR (that is, (-d raa229618,tmp117
would yield all sensors from
either device), but if multiple kinds of specifications are present, they
serve as a logical AND (e.g., -t thermal -d raa229618,tmp117
would yield
all thermal sensors from either device). Alternatively, sensors can be
listed or queried by specifying the ID (or IDs) via -i
(--id
).
By default, humility sensors
displays the value of each specified sensor
and exits; to read values once per second, use the -s
(--sleep
)
option. To print values as a table with individual sensors as columns,
--tabular
. In its default output (with one sensor per row), error
counts are also displayed.
humility spctrl
runs commands on the RoT to control the SP.
You must run humility spctrl init
before any other commands
$ humility spctrl init
[Ok([])]
You can read/write memory on the SP via the RoT
$ humility spctrl -W read 0x08000000 64
humility: attached via CMSIS-DAP
\/ 4 8 c
0x08000000 | 20000400 08000299 08003b6d 08004271 | ... ....m;..qB..
0x08000010 | 08003c8d 08003ccd 08003cd3 00000000 | .<...<...<......
0x08000020 | 00000000 00000000 00000000 0800398b | .............9..
0x08000030 | 08003b6d 00000000 08003aa9 080039dd | m;.......:...9..
$ humility spctrl -W read 0x00000000 64
humility: attached via CMSIS-DAP
\/ 4 8 c
0x00000000 | 3d0fbf49 991373d9 9107611c f6d84242 | I..=.s...a..BB..
0x00000010 | 742397db c7c60242 decc7515 ce719848 | ..#tB....u..H.q.
0x00000020 | b2d2639e faf8049b e202de8c 2ae12025 | .c..........% .*
0x00000030 | f739ba6f 20067a60 310c4e08 e42eca28 | o.9.`z. .N.1(...
$ humility spctrl -W write 0x00000000 0x11,0x22,0x33,0x44,0x55,0x66,0x77,0x88
humility: attached via CMSIS-DAP
[Ok([])]
$ humility spctrl -W read 0x00000000 64
humility: attached via CMSIS-DAP
\/ 4 8 c
0x00000000 | 44332211 88776655 9107611c f6d84242 | ."3DUfw..a..BB..
0x00000010 | 742397db c7c60242 decc7515 ce719848 | ..#tB....u..H.q.
0x00000020 | b2d2639e faf8049b e202de8c 2ae12025 | .c..........% .*
0x00000030 | f739ba6f 20067a60 310c4e08 e42eca28 | o.9.`z. .N.1(...
Scan for and read devices implementing Serial Presence Detect (SPD).
When run without arguments, humility spd
will display the SPD data
as gathered and cached by the system:
$ humility spd
humility: attached via ST-Link V3
ADDR MANUFACTURER PART WEEK YEAR
0 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
1 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
2 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
3 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
4 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
5 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
6 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
7 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
8 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
9 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
10 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
11 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
12 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
13 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
14 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
15 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
This performs no I2C reads, and because it operates on cached data, can be run postmortem.
To force an I2C read of a given SPD device, specify the desired bus in terms of either a named bus or controller/port/mux:
% humility spd -b mid
humility: attached via ST-Link V3
ADDR MANUFACTURER PART WEEK YEAR
0 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
1 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
2 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
3 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
4 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
5 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
6 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
7 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
Note that a given bus can have up to 8 DIMMs on it.
To dump the entire contents of one more SPDs, use the --verbose
(-v
)
option:
% humility spd --bus mid --address 5 --verbose
humility: attached via ST-Link V3
ADDR MANUFACTURER PART WEEK YEAR
5 Micron Technology 36ASF8G72PZ-3G2E1 44 2021
|
+----> 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f
0x000 | 23 12 0c 01 86 31 00 08 00 60 00 03 08 0b 80 00 | #....1...`......
0x010 | 00 00 05 0d f8 ff 02 00 6e 6e 6e 11 00 6e f0 0a | ........nnn..n..
0x020 | 20 08 00 05 00 50 14 28 28 00 78 00 14 3c 00 00 | ....P.((.x..<..
0x030 | 00 00 00 00 00 00 00 00 00 00 00 00 16 16 15 16 | ................
0x040 | 03 16 03 16 03 16 03 16 0d 16 16 16 16 16 00 00 | ................
0x050 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x060 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x070 | 00 00 00 00 00 00 9c 00 00 00 00 00 e7 00 fd a3 | ................
0x080 | 31 11 61 19 00 86 9d 22 01 65 45 00 00 00 00 00 | 1.a....".eE.....
0x090 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x0a0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x0b0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x0c0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x0d0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x0e0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x0f0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 b8 ce | ................
0x100 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x110 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x120 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x130 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x140 | 80 2c 06 21 44 32 52 c5 7e 33 36 41 53 46 38 47 | .,.!D2R.~36ASF8G
0x150 | 37 32 50 5a 2d 33 47 32 45 31 20 20 20 31 80 2c | 72PZ-3G2E1 1.,
0x160 | 45 4a 41 41 42 4a 35 50 30 30 31 00 00 00 00 00 | EJAABJ5P001.....
0x170 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x180 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x190 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x1a0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x1b0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x1c0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x1d0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x1e0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
0x1f0 | 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 | ................
To dump a given SPD to a file, additionally provide the --output
(-o
)
option and specify a desired output file:
% humility spd --bus mid --address 5 --output spd.5.out
humility: attached via ST-Link V3
humility: wrote SPD data for address 5 as binary to spd.5.out
humility spi
can be used to read or write to attached SPI devices,
(as defined in the application TOML).
The SPI peripheral (block) to be used can be specified with the
--peripheral
(-p
) option. This should be a number that matches SPI
peripheral number; if it is not specified (and there is only one
SPI-controlling task found), the peripheral associated with that task will
be assumed.
Because of the full duplex nature of SPI, bytes will always be written
and read; to actually write specific bytes, the bytes to be written
should be specified via --write
(-w
). To report bytes read back,
--read
(-r
) should be specified, along with the number of bytes via
--nbytes
(-n
).
For example, to write the byte sequence 0x1
, 0x0
, 0x0
and then read
32 bytes (discarding the first three) from device 0 on SPI2:
$ humility spi -p 2 --nbytes 32 --write 0x1,0x0,0x0 --read --discard 3
humility: attached to 0483:374e:003C00174741500520383733 via ST-Link V3
humility: SPI master is spi2_driver
\/ 1 2 3 4 5 6 7 8 9 a b c d e f
0x00000000 | 01 de aa 55 00 00 00 ff 06 00 00 00 1c 01 0b 00 | ...U............
0x00000010 | 00 00 00 00 ff ff ff 06 12 00 00 00 06 | .............
humility stackmargin
calculates and print stack margins by task. The
margin is determined by walking up each stack, looking for the first
word that does not contain the uninitialized pattern (0xbaddcafe
),
from which it infers maximum depth, and therefore margin:
$ humility -d ./hubris.core.10 stackmargin
humility: attached to dump
ID TASK STACKBASE STACKSIZE MAXDEPTH MARGIN
0 jefe 0x20001000 1024 768 256
1 rcc_driver 0x20001400 1024 176 848
2 usart_driver 0x20001800 1024 216 808
3 user_leds 0x20001c00 1024 208 816
4 ping 0x20002000 512 224 288
5 pong 0x20002400 1024 208 816
6 idle 0x20002800 256 104 152
Note that the margin is only valid for the task's lifetime -- and in particular, will not be correct if the task has restarted due to a stack overflow!
Humility has support to manage the Root Security Services (RSS) and various flash options bits
A typical sequence to set the secure region at 0x08000000
humility stmsecure set-secure-bit
humility stmsecure set-secure-region 0x08000000 0xa000
To undo the secure region:
humility stmsecure unset-secure-region
humility stmsecure unset-secure-bit
The STM32 has support for flash bank swapping as well
humility stmsecure bank-swap
humility tasks
offers a ps-like view of a system, e.g.:
$ humility tasks
humility: attached via ST-Link
system time = 1764993
ID TASK GEN PRI STATE
0 jefe 0 0 recv, notif: bit0 bit1(T+7)
1 rcc_driver 0 1 recv
2 gpio_driver 0 2 recv
3 usart_driver 0 2 recv, notif: bit0(irq39)
4 i2c_driver 0 2 recv
5 spi_driver 0 2 recv
6 user_leds 0 2 recv
7 pong 0 3 FAULT: killed by jefe/gen0 (was: recv, notif: bit0)
8 ping 14190 4 wait: send to pong/gen0
9 hiffy 0 3 notif: bit0(T+7)
10 hf 0 3 notif: bit0(T+18)
11 idle 0 5 RUNNING
To see every field in each task, you can use the -v
flag:
$ humility -d hubris.core.4 tasks -v
humility: attached to dump
system time = 1791860
ID TASK GEN PRI STATE
...
7 pong 0 3 FAULT: killed by jefe/gen0 (was: recv, notif: bit0)
|
+-----------> Task {
save: SavedState {
r4: 0x200063c4,
r5: 0x10,
r6: 0x1,
r7: 0x0,
r8: 0x60003,
r9: 0x4,
r10: 0x200063d4,
r11: 0x1,
psp: 0x20006330,
exc_return: 0xffffffed,
...
},
priority: Priority(0x3),
state: Faulted {
fault: Injected(TaskId(0x0)),
original_state: InRecv(None)
},
...
...
To see a task's registers, use the -r
flag:
$ humility tasks -r user_leds
humility: attached via ST-Link
system time = 1990498
ID TASK GEN PRI STATE
6 user_leds 0 2 recv
|
+---> R0 = 0x20005fc8 R1 = 0x0000000c R2 = 0x00000000 R3 = 0x20005fd8
R4 = 0x20005fc8 R5 = 0x0000000c R6 = 0x00000000 R7 = 0x00000000
R8 = 0x08027154 R9 = 0x00000000 R10 = 0xfffffe00 R11 = 0x00000001
R12 = 0x00000000 SP = 0x20005fa0 LR = 0x08026137 PC = 0x08026e42
To see a task's stack backtrace, use the -s
flag:
$ humility tasks -s user_leds
humility: attached via ST-Link
system time = 2021382
ID TASK GEN PRI STATE
6 user_leds 0 2 recv
|
+---> 0x20005fc0 0x08026e42 userlib::sys_recv_stub
0x20006000 0x08026128 userlib::sys_recv
0x20006000 0x08026128 idol_runtime::dispatch
0x20006000 0x08026136 main
To additionally see line number information on a stack backtrace, also provide
-l
flag:
$ humility tasks -sl user_leds
humility: attached via ST-Link
system time = 2049587
ID TASK GEN PRI STATE
6 user_leds 0 2 recv
|
+---> 0x20005fc0 0x08026e42 userlib::sys_recv_stub
@ /home/bmc/hubris/sys/userlib/src/lib.rs:288
0x20006000 0x08026128 userlib::sys_recv
@ /home/bmc/hubris/sys/userlib/src/lib.rs:236
0x20006000 0x08026128 idol_runtime::dispatch
@ /home/bmc/.cargo/git/checkouts/idolatry-1ebf1c2fd2f30300/6d18e14/runtime/src/lib.rs:137
0x20006000 0x08026136 main
@ /home/bmc/hubris/drv/user-leds/src/main.rs:110
These options can naturally be combined, e.g. humility tasks -slvr
.
When run against a test archive, humility test
kicks off the test suite.
Humility is responsible for getting the number of tests and then running
each one in succession.
humility: attached via CMSIS-DAP
Total test cases: 50
humility: running test_send ...ok
humility: running test_recv_reply ...ok
humility: running test_recv_reply_fault ...ok
humility: running test_floating_point_lowregs ...ok
humility: running test_floating_point_highregs ...ok
humility: running test_floating_point_fault ...ok
humility: running test_fault_badmem ...ok
humility: running test_fault_stackoverflow ...ok
humility: running test_fault_execdata ...ok
humility: running test_fault_illop ...ok
humility: running test_fault_nullexec ...ok
humility: running test_fault_textoob ...ok
humility: running test_fault_stackoob ...ok
humility: running test_fault_buserror ...ok
humility: running test_fault_illinst ...ok
humility: running test_fault_divzero ...ok
humility: running test_fault_maxstatus ...ok
humility: running test_fault_badstatus ...ok
humility: running test_fault_maxrestart ...ok
humility: running test_fault_badrestart ...ok
humility: running test_fault_maxinjection ...ok
humility: running test_fault_badinjection ...ok
humility: running test_fault_superinjection ...ok
humility: running test_fault_selfinjection ...ok
humility: running test_panic ...ok
humility: running test_restart ...ok
humility: running test_restart_taskgen ...ok
humility: running test_borrow_info ...ok
humility: running test_borrow_read ...ok
humility: running test_borrow_write ...ok
humility: running test_borrow_without_peer_waiting ...ok
humility: running test_supervisor_fault_notification ...ok
humility: running test_timer_advance ...ok
humility: running test_timer_notify ...ok
humility: running test_timer_notify_past ...ok
humility: running test_task_config ...ok
humility: running test_task_status ...ok
humility: running test_task_fault_injection ...ok
humility: running test_refresh_task_id_basic ...ok
humility: running test_refresh_task_id_off_by_one ...ok
humility: running test_refresh_task_id_off_by_many ...ok
humility: running test_post ...ok
humility: running test_idol_basic ...ok
humility: running test_idol_bool_arg ...ok
humility: running test_idol_bool_ret ...ok
humility: running test_idol_bool_xor ...ok
humility: running test_idol_err_ret ...ok
humility: running test_idol_ssmarshal ...ok
humility: running test_idol_ssmarshal_multiarg ...ok
humility: running test_idol_ssmarshal_multiarg_enum ...ok
Ran a total of 50 cases
All tests will produce an output file. This contains information about the hubris archive as well as task state after each test run. Search for "fail" to see any failed tests.
$ cat hubris.testout.15
...
==== Test results
...
==== Test test_task_status result: "ok"
==== Task state
system time = 27941
ID TASK GEN PRI STATE
0 runner 0 0 recv, notif: bit0 bit1 bit2 bit3 bit4 bit5 bit6 bit7 bit8 bit9 bit10 bit11 bit12 bit13 bit14 bit15 bit16 bit17 bit18 bit19 bit20 bit21 bit22 bit23 bit24 bit25 bit26 bit27 bit28 bit29 bit30 bit31
1 suite 33 2 recv
2 assist 37 1 FAULT: in syscall: used bogus task index (was: ready)
3 idol 0 1 recv
4 hiffy 0 3 notif: bit31(T+8)
5 idle 0 4 RUNNING
Older versions of the test suite gave streaming output. This is no longer available. If the information in the output is not enough to debug, the recommendation is to extend the state that is captured.
Tools to interact with the Tofino EEPROM
humility validate
uses the Hubris validate
task to validate the
correct presence of devices as described by the application TOML. To
view all devices, use the --list
option; to validate them, run
without any additional arguments:
$ humility validate
humility: attached via ST-Link V3
ID VALIDATION C P MUX ADDR DEVICE DESCRIPTION
0 removed 2 F - 0x48 tmp117 Southwest temperature sensor
1 removed 2 F - 0x49 tmp117 South temperature sensor
2 removed 2 F - 0x4a tmp117 Southeast temperature sensor
3 present 2 F - 0x70 pca9545 U.2 ABCD mux
4 present 2 F - 0x71 pca9545 U.2 EFGH mux
5 present 2 F - 0x72 pca9545 U.2 IJ/FRUID mux
6 timeout 2 B - 0x73 pca9545 M.2 mux
7 timeout 2 B 1:4 0x4c tmp451 T6 temperature sensor
8 validated 3 H - 0x24 tps546b24a A2 3.3V rail
9 validated 3 H - 0x26 tps546b24a A0 3.3V rail
10 validated 3 H - 0x27 tps546b24a A2 5V rail
11 validated 3 H - 0x29 tps546b24a A2 1.8V rail
12 present 3 H - 0x3a max5970 M.2 hot plug controller
13 absent 3 H - 0x4c sbtsi CPU temperature sensor
14 present 3 H - 0x58 idt8a34003 Clock generator
15 validated 3 H - 0x5a raa229618 CPU power controller
16 validated 3 H - 0x5b raa229618 SoC power controller
17 validated 3 H - 0x5c isl68224 DIMM/SP3 1.8V A0 power controller
18 validated 4 F - 0x10 adm1272 Fan hot swap controller
19 validated 4 F - 0x14 adm1272 Sled hot swap controller
20 validated 4 F - 0x20 max31790 Fan controller
21 validated 4 F - 0x25 tps546b24a T6 power controller
22 removed 4 F - 0x48 tmp117 Northeast temperature sensor
23 removed 4 F - 0x49 tmp117 North temperature sensor
24 validated 4 F - 0x4a tmp117 Northwest temperature sensor
25 validated 4 F - 0x67 bmr491 Intermediate bus converter
Reads from (or writes to) EEPROMs that contain vital product data (VPD).
To list all eligible devices, use --list
:
$ humility vpd --list
humility: attached via ST-Link V3
ID C P MUX ADDR DEVICE DESCRIPTION LOCKED
0 1 B 1:1 0x50 at24csw080 Sharkfin VPD locked
1 1 B 1:2 0x50 at24csw080 Gimlet Fan VPD unlocked
2 1 B 1:3 0x50 at24csw080 Sidecar Fan VPD unlocked
To read from all devices, combine --list
with --read
. To read from a
particular device, use --read
alone, and specify the device by either id
(--id
) or by some (case-insensitive) substring of its description
(--device
):
$ humility vpd --read --id 0
humility: attached via ST-Link V3
[
("FRU0", [
("BARC", [
"OXC11R00241",
]),
]),
]
In this example, this could also be phrased as:
$ humility vpd --read --device sharkfin
humility: attached via ST-Link V3
[
("FRU0", [
("BARC", [
"OXC11R00241",
]),
]),
]
You can also use the --raw
flag to --read
to see the raw bytes:
$ humility vpd --read -i 10 --raw
humility: attached via ST-Link V3
\/ 1 2 3 4 5 6 7 8 9 a b c d e f
0x00000000 | 46 52 55 30 4c 00 00 00 fd c6 3b db 42 41 52 43 | FRU0L.....;.BARC
0x00000010 | 1f 00 00 00 ce 3d d7 f7 30 58 56 31 3a 39 31 33 | .....=..0XV1:913
0x00000020 | 30 30 30 30 30 31 39 3a 30 30 36 3a 42 52 4d 34 | 0000019:006:BRM4
0x00000030 | 32 32 32 30 30 32 33 00 b9 1f 7f e3 4d 41 43 30 | 2220023.....MAC0
0x00000040 | 09 00 00 00 61 64 d1 7e a8 40 25 04 01 00 08 00 | ....ad.~.@%.....
0x00000050 | 08 00 00 00 26 27 3d 9d ee a4 f9 bb ff ff ff ff | ....&'=.........
Note that this will fail if the description matches more than one device, e.g.:
$ humility vpd --read --device fan
humility: attached via ST-Link V3
humility vpd failed: multiple devices match description "fan"
To write VPD data, pass a filename of a file that contains a RON description of a valid TLV-C payload, e.g.:
$ cat vpd.in
[("BARC", [
("FOOB", [ [8, 6, 7, 5, 3, 0, 9] ]),
("QUUX", []),
])]
$ humility vpd --write ./vpd.in --device sharkfin
humility: attached via ST-Link V3
humility: successfully wrote 56 bytes of VPD
You can also use a file as a loopback device via --loopback
, allowing
you to, e.g., read binary data and format it (i.e., via --read
).
To lock a VPD device, use the --lock
command. This will lock the VPD
permanently and cannot be undone; subsequent attempts to write to (or
lock) a locked VPD device will result in an error. The lock status of
each device is shown in --list
.
To lock all VPD devices, use the --lock-all
command. This will exit
successfully if all devices were successfully locked or are already
locked; if a device is missing or otherwise cannot be locked, all other
devices will be locked, but the command will exit with a non-zero exit
status.
Given a word-aligned address, writes the specified 32-bit value. If multiple values are specified, writes each in turn, incrementing the address by the word size after each write.
It should go without saying that this should be only used with care.
In particular, the target is not halted before any writes (but that
functionality can be affected by using writeword
).
For example, to write the value 0x11223344 to 0x24018900:
$ humility writeword 0x24018900 0x11223344
humility: attached via ST-Link V3
humility: writing 0x11223344 to 0x24018900
Note that the word is written as single, 32-bit value -- and will therefore be little-endian if reading memory:
$ humility readmem 0x24018900 16
humility: attached via ST-Link V3
\/ 1 2 3 4 5 6 7 8 9 a b c d e f
0x24018900 | 44 33 22 11 00 00 00 00 00 00 00 00 00 00 00 00 | D3".............
To write multiple values to multiple words starting at the specified address, specify them as additional arguments, e.g.:
$ humility writeword 0x24047800 0xaa 0xbb 0xcc 0xdd
humility: attached via ST-Link V3
humility: writing 0xaa to 0x24047800
humility: writing 0xbb to 0x24047804
humility: writing 0xcc to 0x24047808
humility: writing 0xdd to 0x2404780c
$ humility readmem 0x24047800 16
humility: attached via ST-Link V3
\/ 1 2 3 4 5 6 7 8 9 a b c d e f
0x24047800 | aa 00 00 00 bb 00 00 00 cc 00 00 00 dd 00 00 00 | ................