Unmount fixes #1147
Unmount fixes #1147
Conversation
- Fix typo against full dataset path argument - Make unmount with encryption keys idempotent
…sting non-empty path
|
I'll look into the config.fsType reference later this week (doc render issue) |
The value for device is copied into the SOURCE column of findmnt. By default this is "none". The mount and unmount scripts branch on SOURCE=fsType specifically, and the existence of "none" in the SOURCE column causes wrong branches to be executed. For a list of special filesystems, the default value "none" is replaced by the value for fsType.
Bert-Proesmans
force-pushed
the
unmount-fixes
branch
from
9c9042c to
b6a1b5e
Compare
|
How do these stacked PRs work? |
|
high level; the next PR incorporates all commits from the previous PR. the final PR runs tests on all code included in the entire stack why; i decided to split the commits made into chunks that (in my opinion) could be discussed together. stacking branches has better affordances for local development (and I wanted to try it out). what to do; approve and merge this (the first) PR, when the master REF updates approve and merge the next PR My goal is to activate unmount testing which requires various changes first before I can push the toggle. You can see that all tests (including the unmounting) are green through the entire stack of changes up to and including pushing the toggle to enable the tests. The requisite changes could be evaluated both standalone and in context of the stack, this approach provides the maximum flexibility for review (imo). |
github-merge-queue
bot
removed this pull request from the merge queue due to failed status checks
3 participants
Part of stacked pull requests to enable unmount testing. This is PR 1 out of 4.
These commits are grouped on the idea that they are non-breaking fixes to existing working config, and also unbreak non-working config.
Next PR #1148
See #1140 for ealier discussion.