A few scripts were added without set -o pipefail since the original
change to add that everywhere. This will get the dib-lint check
passing again.
Change-Id: I96bef45cc10ff9bbcf2c4f1b796b8cd188e10485
Add the ability to supply a gerrit ref in DIB_REPOREF_* and
have that reference fetched from the remote and checked out for
building within the image.
Closes-Bug: #1314021
Change-Id: I5e5742c4cbff98e14121c50dde5e512f192b5415
Co-Authored-By: Robert Collins <rbtcollins@hp.com>
Using -n for bash arrays was a little bit bad, and also resulted in
errors when running under set -u. Wean the yum element off the array
and switch to a straight string.
Change-Id: Ifd61fce7dadd5820022e9b2ab42020ba0cb71ce6
When using custom-built kernels, their filenames may not necessarily
follow the standard naming conventions for the stock systems in question.
This patch permits the specification of an alternative pattern to use
when looking for the correct kernel to construct the PXE images from.
Example:
export DIB_BAREMETAL_KERNEL_PATTERN="kernel*jang"
export DIB_BAREMETAL_INITRD_PATTERN="initrd*jang"
Change-Id: I86e5254051addd5e89dd4e363e9bba117e65353d
Add support for supplying a custom keyring and debootstrap script
using two new environment variables:
* DIB_DEBIAN_KEYRING - Path to a gpg keyring.
* DIB_DEBIAN_DEBOOTSTRAP_SCRIPT - Path to a alternative debootstrap
script.
Change-Id: I42e8832b435b8a77b0681b5bc9d9332bbe860cd5
Although set -u allows empty WHITELIST=() list to be set, you can't test
-n "$WHITELIST". Looks like the intention of the commit that caused
this issue (Change-ID Iad2d490c86dceab148ea9ab08f457c49a5d5352e) is to
set a policy to prevent future use of bugs due to unbound variables.
Thanks StevenK for pairing with me and helping me hone my bash skills.
Change-Id: I2349c990ba53b0aaecd5119812582c7c848f1e4e
Closes-Bug: #1319987
If the kernel is restricted read (as a security measure) we may not be
able to read it without chmodding it.
Change-Id: Ib577705c27d8c351288334de03a7848cf49f2de9
If 'DIB_RHSM_USER' isn't set, the image build with rhel
element fails because of check 'set -u'. So, we should export
'DIB_RHSM_USER' in script to avoid error during build
Change-Id: Ic63a20e3d91a25bbe804440a5919b035489c60a3
Closes-bug: #1318717
With this change, dib-lint raises an error if when finding an element
using space indentation that is not multiple of 4.
Co-Authored-By: Jon-Paul Sullivan <jonpaul.sullivan@hp.com>
Change-Id: I470e1fdfc38a3f3c7ba5644c5103f2a9ef073005
When the cache tarball gets fail the output doesn't indicate
what url is failing. Adding '-x' displays the urls attempted.
Change-Id: I71b2dbfd712e72c591410ca9b68edc76d493ad6e
This will hopefully catch bugs where they happen rather than
allowing scripts to continue on and fail later.
Change-Id: Idacd9274415b21db285a198dafff19b1d19a4a68
As with the previous similar changes, this is intended to catch
problems as they happen instead of ignoring them and continuing on
to potentially fail later. Setting this on all existing scripts
will allow us to enforce use via Jenkins.
Change-Id: Iad2d490c86dceab148ea9ab08f457c49a5d5352e
Updates the fedora element so that DIB_DISTRIBUTION_MIRROR can
be used to configure a custom Yum mirror on Fedora. This
implementation replaces the baseurl in each of the default
Fedora Yum .repo files so that a custom mirror URL/path
can be supplied. The URL should point to the directory
where the updates and release directories exist. Example:
http://download.fedoraproject.org/pub/fedora/linux
As part of this change we also remove any existing
metalink settings in the .repo files to cover the case
where an end user runs their own createrepo command
to create the mirror... which means your metadata won't match that
of the normal Fedora mirrors and your baseurl won't get used.
Change-Id: I1a49a2941b1cc489e91a72b87430393cd32e865d
With this change, dib-lint ensure elements do not use tab
indentation. The following files are checked:
- executable file
- .md file
Change-Id: I071262ff9f6599548f869f5439ee127f64eeb46f
DIB's map-services is currently used in TIE's os-svc-config
binaries to support service name lookups. Many of the systemd
specific functions require bin/map-services to exist.
This change adds bin/map-services files that are noops
to both the debian and ubuntu elements so that as those
distros move towards using systemd they can make use
of the functions we already have in place.
Change-Id: If8d61f3858b11de86c3292e840d033e5e3cecedb
If scripts are not set -e then errors can be ignored, causing more
confusing failures later.
Also adds an exclusion comment to the ramdisk init script since we
don't want that to exit on failure.
Change-Id: Idf43993bd10b1ef16c1d3b0d9df8d0ad94c46458
There are certain scripts (such as the ramdisk init script) that
need to ignore linting rules for one reason or another. This adds
support for exclusions via a comment in the file like:
"# dib-lint: disable=executable sete"
There should be no ", but I need those to prevent git from treating
that as a comment. This syntax is similar to the pylint exclusion
mechanism.
Exclusion support is added to the executable check, but not
the alphabetical ordering one because I can't imagine a reason we
would need to disable that, and I don't know that comments are
supported in those files anyway.
Change-Id: I9ecfb47269841dc75a005855455ac26ad2cbc642
Adds a new element, package-installs, that provides an interface for
declarative package installs and uninstalls. Packages to install can be
added to an install.d/package-installs-<element-name> file. The set of
packages listed across such files are installed in a single transaction
at the beginning of install.d.
Prefacing the package name with a "-" indicates that the package should
be uninstalled at the end of the install.d phase. Again, the full set of
uninstalls are done in a single transaction.
An element providing a package-installs file should add package-installs
to its element-deps file.
Change-Id: I5b540388eff1079c8dee933b869463371481152b
If the tmpfs being used to hold the image filesystem is close to being
full then housing the raw disk image in the same place can be
problematic as it tends to grow faster then the filesystem shrinks when
data is being moved into it.
Putting them both into separate tmpfs's will allow there to be an
overlap where they jointly use more then the size limit for a singe tmpfs.
Change-Id: Ia17ca357d3b865d70a2d3e13e0479d008ca5f924
Closes-Bug: #1289582
