The mock-sensor-tool allows a user to overload sensor values retrieved from a phosphor-hwmon or dbus-sensors instance at the kernel interface level. The associated design document can be found here: https://gerrit.openbmc-project.xyz/c/openbmc/docs/+/37869 Usage: First, find the PID of the phosphor-hwmon or dbus-sensors instance to inject mock values in to. For example: systemctl status xyz.openbmc_project.Hwmon@-ahb-apb-pwm\\x2dtacho\\x2dcontroller\\x401e786000.service will return the PID of a phosphor-hwmon instance. Then, run the mock-sensor-tool and input the PID when prompted to. After a small delay, the tool will start and the user interface will instruct the user. Currently, all values that fit into a 32 bit long are supported, and all the standard linux errno codes are supported as well: https://www-numi.fnal.gov/offline_software/srt_public_context/WebDocs/Errors/unix_system_errors.html There is currently a temporarily "sleep" command that allows the user to put the tool in the background and run other commands in the BMC instance without the use of a multiplexer. While this solution isn't ideal, with the time constraints in mind this serves as a "band-aid" solution as putting the mock-sensor-tool in the background without sleeping the user interface thread will cause linux to stop the overloading thread as well (SIGTTIN) In the case that PTRACE_ATTACH is disabled on the machine, the parameter in the file /proc/sys/kernel/yama/ptrace_scope must be set to 0 (these are the same permissions tools like strace uses). More information can be found here: https://www.kernel.org/doc/Documentation/security/Yama.txt