Re: [PATCH v5 04/17] odb: introduce parent pointers

From: Patrick Steinhardt <ps@pks.im>
Date: Jul 1, 2025, 12:17 UTC
Message-ID: <aGPR9zMyN1AZ4uUh@pks.im>
In-Reply-To: <xgbzzqyfnfmjmtucnq3qd3bfqdal6vjkaap3xfspeetbs6ontk@4vclbvlreere>

On Sun, Jun 29, 2025 at 09:34:12PM -0500, Justin Tobler wrote:

Show 16 quoted lines

> > @@ -167,22 +171,23 @@ static int link_alt_odb_entry(struct repository *r, const struct strbuf *entry,
> >  	while (pathbuf.len && pathbuf.buf[pathbuf.len - 1] == '/')
> >  		strbuf_setlen(&pathbuf, pathbuf.len - 1);
> >  
> > -	if (!alt_odb_usable(r->objects, &pathbuf, normalized_objdir, &pos))
> > +	if (!alt_odb_usable(odb, &pathbuf, normalized_objdir, &pos))
> >  		goto error;
> >  
> >  	CALLOC_ARRAY(alternate, 1);
> > -	/* pathbuf.buf is already in r->objects->source_by_path */
> > +	alternate->odb = odb;
> > +	/* pathbuf.buf is already in r->objects->alternate_by_path */
> 
> Should this comment instead say "odb->source_by_path"?
> 
> The remaining restructuring in this patch looks good.

Ah, good catch. I think I botched conflict resolution here.

Patrick