linux/poky.git/meta/classes/rootfs-postcommands.bbclass, branch uninative-2.9

rootfs-post: remove traling blanks from tasks

2020-07-27T09:37:57+00:00

remove the traling blanks before the ;-delimiter, so one could use "_remove" to avoid running tasks like 'rootfs_update_timestamp', which are currently hardcoded and not bound to any configurable feature flag (From OE-Core rev: 8c835f63c2bca0a66385eb7ea5ec021d86265777) Signed-off-by: Konrad Weihmann Signed-off-by: Richard Purdie

rootfs-postcommands: Improve/fix rootfs_check_host_user_contaminated

2020-07-03T19:17:29+00:00

Currently if the find command finds /home it stops checking the other files. Tweak the find expression to fix this. Also, from Alex Kiernan , generate output for each contaminated path so it's visible in the main log file. When matches are found, dump the data from the group and passwd files so useful debugging can be done based on the results. (From OE-Core rev: 504f1d158bf7c6ad08fbce755021db2b14565231) Signed-off-by: Richard Purdie

Revert "rootfs: Make BUILDNAME a weak default in reproducible_build_simple"

2020-02-21T17:48:09+00:00

This reverts commit 8380df6566db49ef184b837432558750f77f592f. Giving BUILDNAME a static default in reproducible_build_simple.bbclass to have /etc/version be generated with that name lead to other users of BUILDNAME no longer working as expected. E.g., buildstats.bcclass would now write its information in a single directory, overwriting the statistics for each build. (From OE-Core rev: dcd1fbe153369634c73370cbe22e1dbdac080146) Signed-off-by: Peter Kjellerstedt Signed-off-by: Richard Purdie

rootfs: Make BUILDNAME a weak default in reproducible_build_simple

2020-01-19T23:49:39+00:00

11e45082ad00 ("rootfs-postcommands.bbclass: improve binary reproducibility") fixed binary reproducibility of /etc/version, but with the move to reproducibilty in all builds, setting /etc/version to anything other than the default fixed timestamp is tricky because rootfs_reproducible() runs very late. rootfs.py uses BUILDNAME if set for /etc/version, so introduce a weak default for BUILDNAME of "REPRODUCIBLE_TIMESTAMP_ROOTFS", when enabling reproducible builds hence allowing BUILDNAME to be overridden elsewhere. (From OE-Core rev: bbf28ea9100a4f86b052c5cd53c77f9e1c03fb09) Signed-off-by: Alex Kiernan Signed-off-by: Richard Purdie

rootfs-postcommands: Avoid use of an hard-coded value

2019-10-02T19:22:25+00:00

(From OE-Core rev: 520c6f30cd571166a0d890d86b461f24f7a6998f) Signed-off-by: Alessio Igor Bogani Signed-off-by: Richard Purdie

rootfs-postcommands: check /etc/gconf exists before working on it

2019-09-19T09:54:32+00:00

Check that /etc/gconf exists before trying to find files under it, to avoid writing find error messages to the rootfs log. Also use ${sysconfdir}/gconf instead of hardcoding /etc/gconf. (From OE-Core rev: 70a77b91031388ea29eff6b8885fd25cc35b1dae) Signed-off-by: Ross Burton Signed-off-by: Richard Purdie

rootfs-postcommands.bbclass: add inittab tweak to read_only_rootfs_hook()

2019-08-28T10:31:22+00:00

It's not clear that we actually need to call mount -o remount,XX from the busybox-inittab inittab at all, but as we currently do, let's respect read-only-rootfs in IMAGE_FEATURES and not remount / as rw. (From OE-Core rev: 8840a831d47a94f36bb2e173164dbaf9b955e809) Signed-off-by: Andre McCurdy Signed-off-by: Richard Purdie

rootfs-postcommands: Cope with empty IMAGE_LINK_NAME in write_image_test_data

2019-07-16T12:53:17+00:00

Ensure that we don't create an image test data symlink named ".testdata.json" when IMAGE_LINK_NAME is empty. (From OE-Core rev: 97e1af51814c63963dc6eee003e0cf0e4dead024) Signed-off-by: Mike Crowe Signed-off-by: Richard Purdie

rootfs-postcommands: Cope with empty IMAGE_LINK_NAME in write_image_manifest

2019-07-16T12:53:17+00:00

Ensure that we don't create a symlink named ".manifest" if IMAGE_LINK_NAME is empty. (From OE-Core rev: 267697f5e1c931e39fd81dd1b14691e364be64f3) Signed-off-by: Mike Crowe Signed-off-by: Richard Purdie

systemd: do not create machine-id

2019-05-03T05:11:57+00:00

There is no reason to have an emtpy machine-id as part of the systemd package. Either: i) the filesystem is writable and the file will be created automatically; or ii) the filesystem is read-only, in which case the empty machine-id file should be created as part of the read-only-rootfs tweaks. (From OE-Core rev: 76444b63e614baea33c044851a5859f6d1e69729) Signed-off-by: Jonas Bonn Signed-off-by: Alex Kiernan Signed-off-by: Richard Purdie