Re: [Outreachy PATCH v6 1/3] environment: stop storing `core.attributesFile` globally

From: Junio C Hamano <gitster@pobox.com>
Date: Feb 11, 2026, 16:46 UTC
Message-ID: <xmqq5x83w7m1.fsf@gitster.g>
In-Reply-To: <37abfe28-ff89-460a-962d-4503ec3f7975@gmail.com>

Phillip Wood <phillip.wood123@gmail.com> writes:

Show 21 quoted lines

> Thanks for the backtrace which helped me spot the problem though I 
> should have spotted this yesterday. The problem is in 
> initialize_repository()
>
>>   void initialize_repository(struct repository *repo)
>>   {
>> +	if (repo->initialized)
>> +		BUG("repository initialized already");
>> +	repo->initialized = true;
>> +
>>   	repo->remote_state = remote_state_new();
>>   	repo->parsed_objects = parsed_object_pool_new(repo);
>>   	ALLOC_ARRAY(repo->index, 1);
>>   	index_state_init(repo->index, repo);
>>   	repo->check_deprecated_config = true;
>> +	repo_config_values_init(repo_config_values(repo));
>
> Here you need to use repo->config_values_private_ instead of using the 
> accessor as it is fine to initialize the config values to their defaults 
> in any instance, it is only when we read them that we want to assert 
> that we're reading from "the_repository".

Sorry, but ...

At the beginning of repo_config_values() in the patch, there is a check to ensure that repo->initialized is true and otherwise you get an error. But the initialization is already done in the early part of initialize_repository() as quoted above. So I do not see what difference it would make if we rewrote the last line as

	repo_config_values_init(&repo->config_values_private_);

I am confused.