Re: [PATCH 06/11] reset_head(): make default_reflog_action optional

From: Phillip Wood <phillip.wood123@gmail.com>
Date: Oct 4, 2021, 10:03 UTC
Message-ID: <0a580f49-4366-526b-d601-f8ce5f99711c@gmail.com>
In-Reply-To: <xmqqczoo5u7m.fsf@gitster.g>

Hi Junio

On 01/10/2021 22:08, Junio C Hamano wrote:

Show 30 quoted lines

> "Phillip Wood via GitGitGadget" <gitgitgadget@gmail.com> writes:
> 
>> From: Phillip Wood <phillip.wood@dunelm.org.uk>
>>
>> This parameter is only needed when a ref is going to be updated and
>> the caller does not pass an explicit reflog message. Callers that are
>> just discarding changes in the working tree like create_autostash() do
>> not update any refs so should not have to worry about passing this
>> parameter.
> 
> This only talks about internal implementation details of passing a
> parameter that may not be used, but ...
> 
>> -	ret = update_refs(oid, switch_to_branch, reflog_head, reflog_orig_head,
>> -			  default_reflog_action, flags);
>> +	if (oid != &head_oid || update_orig_head || switch_to_branch)
>> +		ret = update_refs(oid, switch_to_branch, reflog_head,
>> +				  reflog_orig_head, default_reflog_action,
>> +				  flags);
> 
> ... doesn't this have a more significant behaviour change?
> 
> I am not sure if comparison between oid and head_oid can safely
> cheat like the patch does, or if it is necessary to do a proper oid > comparison, but either way, this would stop calling update_refs(),
> which in turn means it would have an externally visible effect, like
> a hook no longer getting called, doesn't it?
> 
> It would be a change for the good---calling post-checkout hook when
> you did "git checkout (no other arguments)" feels wasteful, but it
> deserves to be told to end users, I would think.

That would be a change for the good, but it is not what this patch does. If oid == &head_oid then it means that the caller did not give us a tree to checkout so HEAD will not change unless we are switching branches. Using a pointer comparison rather than calling oidcmp() was a deliberate choice to avoid any user visible changes as there are no callers that try to run the checkout-hook without passing an oid or branch. The aim here is to avoid having to pass default_reflog when it is not needed without making any user visible changes.

> Again, a new test to protect the change would go well in the same
> patch.

I'm not sure what we could test

Best Wishes

Phillip

> Thanks.
> 
>