Re: [PATCH v4 3/4] submodule: record superproject gitdir during absorbgitdirs

[Derrick Stolee <stolee@gmail.com>]

On 10/18/2021 7:18 PM, Jonathan Tan wrote:
>> Already during 'git submodule add' we record a pointer to the
>> superproject's gitdir. However, this doesn't help brand-new
>> submodules created with 'git init' and later absorbed with 'git
>> submodule absorbgitdir'. Let's start adding that pointer during 'git
>> submodule absorbgitdir' too.
> 
> s/absorbgitdir/absorbgitdirs/ (note the "s" at the end)
> 
>> @@ -2114,6 +2115,15 @@ static void relocate_single_git_dir_into_superproject(const char *path)
>>  
>>  	relocate_gitdir(path, real_old_git_dir, real_new_git_dir);
>>  
>> +	/* cache pointer to superproject's gitdir */
>> +	/* NEEDSWORK: this may differ if experimental.worktreeConfig is enabled */
>> +	strbuf_addf(&config_path, "%s/config", real_new_git_dir);
>> +	git_config_set_in_file(config_path.buf, "submodule.superprojectGitdir",
>> +			       relative_path(absolute_path(get_git_dir()),
>> +					     real_new_git_dir, &sb));
>> +
>> +	strbuf_release(&config_path);
>> +	strbuf_release(&sb);
>>  	free(old_git_dir);
>>  	free(real_old_git_dir);
>>  	free(real_new_git_dir);
> 
> Here [1] you mention that you'll delete the NEEDSWORK, but it's still
> there.
> 
> Having said that, it might be better to make a test in which we call
> this command while in a worktree of a superproject. The test might
> reveal that (as pointed out to me internally) you might need to use the
> common dir functions instead of the git dir functions to point to the
> directory that you want (git-worktree.txt distinguishes the 2 if you
> search for GIT_COMMON_DIR).

I came here to say the same thing. It's a bit too direct to compute
the location of a config file this way, so we should expose a method
that can create one for a given Git directory.

Since you're setting this config value inside the submodule's config,
what does it mean for a submodule to also be a worktree (and hence
require config.worktree)? What happens in this rough scenario?

  1. git init sub
  2. git init super
  3. git -C sub worktree add super/sub
  4. git -C super submodule absorbgitdir sub

I haven't actually tried running these things, but it seems unusual
and unexpected. This doesn't even account for cases where the repo
root and a worktree are both submodules within the superproject.

If we already have protections preventing these worktrees as
submodules, then perhaps there is no need for work here. I'm not
familiar enough with the area to make a claim one way or another.

Thanks,
-Stolee