Git submodule shows new commits, submodule status says nothing to commit

It's because Git records which commit (not a branch or a tag, exactly one commit represented in SHA-1 hash) should be checked out for each submodule. If you change something in submodule dir, Git will detect it and urge you to commit those changes in the top-level repoisitory.

Run git diff in the top-level repository to show what has actually changed Git thinks. If you've already made some commits in your submodule (thus "clean" in submodule), it reports submodule's hash change.

$ git diff
diff --git a/src/repo b/src/repo
index b0c86e2..a893d84 160000
--- a/src/repo
+++ b/src/repo
@@ -1 +1 @@
-Subproject commit b0c86e28675c9591df51eedc928f991ca42f5fea
+Subproject commit a893d84d323cf411eadf19569d90779610b10280

Otherwise it shows -dirty hash change which you cannot stage or commit in the top-level repository. git status also claims submodule has untracked/modified content.

$ git diff
diff --git a/src/repo b/src/repo
--- a/src/repo
+++ b/src/repo
@@ -1 +1 @@
-Subproject commit b0c86e28675c9591df51eedc928f991ca42f5fea
+Subproject commit b0c86e28675c9591df51eedc928f991ca42f5fea-dirty

$ git status
On branch master
Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git checkout -- <file>..." to discard changes in working directory)
  (commit or discard the untracked or modified content in submodules)

    modified:   src/repo (untracked content)

no changes added to commit (use "git add" and/or "git commit -a")

To update which commit records should be checked out for the submodule, you need to git commit the submodule in addition to committing the changes in the submodule:

git add src/repo

I just ran into this same class of problem, and I was able to use the solution offered by @AugustinAmenabar in the comments section of the accepted answer. My setup was a bit more complex, so I added the --recursive flag to bring all dependencies up to date.

git submodule update --recursive src/repo

None of the answers here solve my problem.

I am documenting/sharing here what worked for me. Hoping it helps someone else.

Initially my submodule was at commit A (at the time of adding submodule to main repo), then I checked out a branch (let us call it new-submodule-branch) and made commits B and C to it and pushed it to remote (github.com)

Post this, my main repo started showing

Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git restore <file>..." to discard changes in working directory)
    modified:   <submodule_name> (new commits)

If, from the main repo root, I ran git submodule update --remote --init --recursive, it kept reverting my submodule's HEAD in detached state to commit A

So then I set the branch value as new-submodule-branch in <MainRepo>/.gitmodules as follows

[submodule "<submodule_name>"]
    path = <submodule_name>
    url = [email protected]:ProProgrammer/<submodule_name>.git
    branch = new-submodule-branch

post this, when I ran git submodule update --remote --init --recursive, it would no longer revert my submodule's HEAD in detached state to commit A however it still kept showing the annoying

    modified:   <submodule_name> (new commits)

So far, I was following the official git reference for submodules, now I decided to do some more Googling, and I stumbled an article titled Getting git submodule to track a branch, this clearly said

You have to go and update that submodule commit reference to the latest code in the remote branch to avoid this

So finally, I did what I was trying to avoid:

git add <submodule_name>
git commit --amend --no-edit # I combined this with the previous commit where I added the 'branch' value in .gitmodules

If you want to see what this looks like once pushed to remote (github.com in my case), you can see this exact commit here