diff options
| author | Jonathan Bauer | 2019-04-18 13:32:01 +0200 |
|---|---|---|
| committer | Jonathan Bauer | 2019-04-18 13:32:01 +0200 |
| commit | db9f15c4db32fabad70a43545324cb34d48ce68e (patch) | |
| tree | c192715f53e52a56a7e696e3fdca4cc2cd94a298 /core/modules/systemd/data/usr/lib | |
| parent | [rootfs-stage31] Scan pci device tree recursively (diff) | |
| download | mltk-db9f15c4db32fabad70a43545324cb34d48ce68e.tar.gz mltk-db9f15c4db32fabad70a43545324cb34d48ce68e.tar.xz mltk-db9f15c4db32fabad70a43545324cb34d48ce68e.zip | |
[slx-issue] use baseboard dmi info as fallback
some models like Intel's NUC5i5RYB do not provide
system-manufacturer/system-product-name in the DMI information -
that information is stored in baseboard-* instead. So we use that as a
fallback in case system-* was empty. Use '<unknown>' as ultimate
fallback.
This kinda works around an issue in the final loop filling in spaces: if
a line's value is long enough (e.g. 34 whitespaces like a untrimmed
'dmidecode -s' output...) the space variable would never be set. Thus,
${space:0:-2} would fail unless ${#space} is at least 2 and abort the
creation of /etc/issue altogether.
Diffstat (limited to 'core/modules/systemd/data/usr/lib')
0 files changed, 0 insertions, 0 deletions
