t2021: move directory-to-file checkout test out of t0050 - #2210
Conversation
The test 'checkout with no pathspec and a case insensitive fs' was added to t0050 as a minimal reproduction of a segfault. Reviewing it later, its author noted it had ended up in a script it does not belong in: https://lore.kernel.org/git/20191007180409.GD11529@szeder.dev/ Three things were wrong with it: - It carried the CASE_INSENSITIVE_FS prerequisite. The segfault could only be triggered on a case insensitive filesystem, but the sequence of commands itself succeeds anywhere, so the prerequisite only prevented the test from running on most systems. - t0050 collects filesystem capability tests. What this exercises is checkout replacing a tracked directory with a tracked file when switching branches, which is what t2021 already covers. - It asserted nothing beyond 'git checkout main' succeeding, so it would not have noticed checkout leaving the wrong contents behind. Move it to t2021, drop the prerequisite, and check the resulting worktree: Gitweb is restored as a file and the gitweb directory is gone. The repository is created with an explicit initial branch, since t2021 does not set GIT_TEST_DEFAULT_INITIAL_BRANCH_NAME the way t0050 does. Signed-off-by: skdas20 <skdas5405@gmail.com>
Welcome to GitGitGadgetHi @skdas20, and welcome to GitGitGadget, the GitHub App to send patch series to the Git mailing list from GitHub Pull Requests. Please make sure that either:
You can CC potential reviewers by adding a footer to the PR description with the following syntax: NOTE: DO NOT copy/paste your CC list from a previous GGG PR's description, Also, it is a good idea to review the commit messages one last time, as the Git project expects them in a quite specific form:
It is in general a good idea to await the automated test ("Checks") in this Pull Request before contributing the patches, e.g. to avoid trivial issues such as unportable code. Contributing the patchesBefore you can contribute the patches, your GitHub username needs to be added to the list of permitted users. Any already-permitted user can do that, by adding a comment to your PR of the form Both the person who commented An alternative is the channel Once on the list of permitted usernames, you can contribute the patches to the Git mailing list by adding a PR comment If you want to see what email(s) would be sent for a After you submit, GitGitGadget will respond with another comment that contains the link to the cover letter mail in the Git mailing list archive. Please make sure to monitor the discussion in that thread and to address comments and suggestions (while the comments and suggestions will be mirrored into the PR by GitGitGadget, you will still want to reply via mail). If you do not want to subscribe to the Git mailing list just to be able to respond to a mail, you can download the mbox from the Git mailing list archive (click the curl -g --user "<EMailAddress>:<Password>" \
--url "imaps://imap.gmail.com/INBOX" -T /path/to/raw.txtTo iterate on your change, i.e. send a revised patch or patch series, you will first want to (force-)push to the same branch. You probably also want to modify your Pull Request description (or title). It is a good idea to summarize the revision by adding something like this to the cover letter (read: by editing the first comment on the PR, i.e. the PR description): To send a new iteration, just add another PR comment with the contents: Need help?New contributors who want advice are encouraged to join git-mentoring@googlegroups.com, where volunteers who regularly contribute to Git are willing to answer newbie questions, give advice, or otherwise provide mentoring to interested contributors. You must join in order to post or view messages, but anyone can join. You may also be able to find help in real time in the developer IRC channel, |
This addresses #429.
The test
'checkout with no pathspec and a case insensitive fs'in t0050 was added as a minimal reproduction of a segfault. Its author reviewed it afterwards and listed what was wrong with it:https://lore.kernel.org/git/20191007180409.GD11529@szeder.dev/
This patch acts on three of those points.
Dropping
CASE_INSENSITIVE_FS. The segfault needed a case insensitive filesystem, but the command sequence does not. I checked this rather than assuming it: with the prerequisite removed the test passes on ext4, so as it stood the prerequisite was only stopping it from running on most systems.Moving it to t2021. t0050 collects filesystem capability tests. What this actually exercises is checkout replacing a tracked directory with a tracked file across a branch switch, and t2021 (
checkout must not overwrite an untracked objects) is where that behaviour is already covered.Checking the result. The test previously ended at
git checkout main, so it passed as long as checkout exited successfully and would not have noticed the wrong contents being left behind. It now asserts the end state I observed by running the sequence:Gitwebpresent as a file, and thegitwebdirectory gone.One detail worth flagging for review: t2021 does not set
GIT_TEST_DEFAULT_INITIAL_BRANCH_NAMEthe way t0050 does, so the repository is created with an explicitgit init -b main. Adding the variable to t2021 would have changed the environment for the tests already in the script, which seemed worse than being explicit in the one test that needs it.I have left the fourth point from that message alone — why the extra subdirectory was needed to trigger the segfault. I could not determine it, and guessing at it in the commit message seemed worse than not raising it.
Both scripts pass: t0050 12/12, t2021 10/10.