Re: [PATCH 2/5] apply: read in the index in --intent-to-add mode
- From
Raymond E. Pasco <ray@ameretat.dev>
- Date
- May 12, 2025, 02:13 UTC
- Message-ID
- <gz4sl3e3nqmkamkfmoktlqaefxteamj6v3fcs7dbavgys7ubng@utz2i5ahl77f>
- In-Reply-To
- <20250511003955.242889-3-ray@ameretat.dev>
On 25/05/10 08:36PM, Raymond E. Pasco wrote:
Show 17 quoted lines
> diff --git a/apply.c b/apply.c
> index 5e39cadde4..3bde54a04a 100644
> --- a/apply.c
> +++ b/apply.c
> @@ -4837,7 +4837,7 @@ static int apply_patch(struct apply_state *state,
> LOCK_DIE_ON_ERROR);
> }
>
> - if (state->check_index && read_apply_cache(state) < 0) {
> + if (state->update_index && read_apply_cache(state) < 0) {
> error(_("unable to read index file"));
> res = -128;
> goto end;
> --
> 2.49.0.1106.gc0efa3ba58
>
> For some reason, this causes checkout -p to break in a way which which I don't understand. Directly using (state->check_index || state->ita_only) doesn't break it, but because it doesn't seem logically possible to me for this to change any behavior besides behavior when ita_only is on (the update_index flag is in fact set just a few lines above!), I want to investigate a bit further to find out what I'm confused about.
Either way, reroll coming, though I'd still appreciate review on the series in general.