Re: [PATCH 1/3] wt-status: implement opportunisitc index update correctly

From: Stefan Beller <sbeller@google.com>
Date: Dec 7, 2016, 20:56 UTC
Message-ID: <CAGZ79kbZBzF20=cfF85_fSuFdpYZuNyhDNCSKCb9HBpLOSV0hA@mail.gmail.com>
In-Reply-To: <CAPc5daX2CZ0=UBMuz70KwFPBDTUgAdi8WoVUJ7gNTq+QEXKxbg@mail.gmail.com>

On Wed, Dec 7, 2016 at 12:53 PM, Junio C Hamano <gitster@pobox.com> wrote:

Show 7 quoted lines

> On Wed, Dec 7, 2016 at 12:48 PM, Stefan Beller <sbeller@google.com> wrote:
>>
>> So I would expect that we'd rather fix the update_index_if_able instead by
>> checking for the lockfile to be in the correct state?
>
> I actually don't expect that, after looking at other call sites of
> that function.

Yes I checked the other callers as well right now and you seem to be correct. My initial response was based on the name of the function, specifically the _if_able part as that hinted to me that I can call the function with no precondition and the _if_able will figure out when to do the actual update_index.

The first part of the condition of the function
    (istate->cache_changed || has_racy_timestamp(istate)
reads rather as a _if_needed instead of an _if_able to me.