First steps

The structures of the system the tool runs on are shown with:

Reading the data of a platform usually requires privileges, because the table is taken from physical memory or from an interface the kernel restricts, so the tool is often run as the superuser.

Structures are selected by type, given by number or by code, or by handle:

The structures the table holds are listed, with their handles and types, with:

What a structure type is for, as the specification describes it, is told with:

The data is exported for a program to read, with every value addressed by a stable code, in JSON, XML or YAML:

The values are written by their codes, e.g. 4294967296 for a size, and are written for a person to read instead, e.g. 4 GiB, with --pretty.

The table is dumped into a file, which is read in place of the system with the -i option, on any system and without privileges:

Dump files are interchangeable with the ones dmidecode --dump-bin writes and dmidecode --from-dump reads.

The data is checked against the rules of the specification with:

Every issue found is reported with its offset within the data, the structure and the attribute it belongs to, and the rule which has found it:

Structures of the types the vendors define are decoded once the module of the vendor is enabled. The modules available are listed with opendmi modules, and one is enabled with the -m option: