ghw | Golang hardware discovery/inspection library | Security library

You can make ghw read from snapshots (created with ghw-snapshot) using environment variables or programmatically. Please check SNAPSHOT.md to learn more about how ghw creates and consumes snapshots. The environment variable GHW_SNAPSHOT_PATH let users specify a snapshot that ghw will automatically consume. All the needed chroot changes will be automatically performed. By default, the snapshot is unpacked on a temporary directory managed by ghw, and cleaned up when no longer needed, avoiding leftovers. The rest of the environment variables are relevant iff GHW_SNAPSHOT_PATH is given. GHW_SNAPSHOT_ROOT let users specify the directory on which the snapshot should be unpacked. This moves the ownership of that directory from ghw to users. For this reason, ghw will not clean up automatically the content unpacked in GHW_SNAPSHOT_ROOT. GHW_SNAPSHOT_EXCLUSIVE is relevant iff GHW_SNAPSHOT_ROOT is given. Set it to any value to toggle it on. This tells ghw that the directory is meant only to contain the given snapshot, thus ghw will not attempt to unpack it (and will go ahead silently!) unless the directory is empty. You can use both GHW_SNAPSHOT_ROOT and GHW_SNAPSHOT_EXCLUSIVE to make sure ghw unpacks the snapshot only once regardless of how many ghw packages (e.g. cpu, memory) access it. Set GHW_SNAPSHOT_PRESERVE to any value to enable it. If set, ghw will not clean up the unpacked snapshot once done, leaving it into the system.
You can create ghw snapshots in two ways. You can just consume the ghw-snapshot tool, or you can create them programmatically from your golang code. We explore now the latter case. Snapshotting takes two phases: 1. clone the relevant pseudofiles/pseudodirectories into a temporary tree This tree is usually deleted once the packing is successful. 2. pack the cloned tree into a tar.gz.