On Mon, Dec 02, 2019 at 02:42:31PM -0500, Jeff King wrote:
> On Mon, Dec 02, 2019 at 06:40:35PM +0100, SZEDER Gábor wrote:
>
> > > When loading packfiles on start-up, we traverse the internal packfile
> > > list once per file to avoid reloading packfiles that have already
> > > been loaded. This check runs in quadratic time, so for poorly
> > > maintained repos with a large number of packfiles, it can be pretty
> > > slow.
> > >
> > > Add a hashmap containing the packfile names as we load them so that
> > > the average runtime cost of checking for already-loaded packs becomes
> > > constant.
> > [...]
> > This patch break test 'gc --keep-largest-pack' in 't6500-gc.sh' when
> > run with GIT_TEST_MULTI_PACK_INDEX=1, because there is a duplicate
> > entry in '.git/objects/info/packs':
>
> Good catch. The issue is that we only add entries to the hashmap in
> prepare_packed_git(), but they may be added to the pack list by other
> callers of install_packed_git(). It probably makes sense to just push
> the hashmap maintenance down into that function, like below. That
> requires an extra strhash() when inserting a new pack, but I don't think
> that's a big deal.
Ah, great catch, and thanks for pointing it out. We have been running
this patch in production at GitHub for a few weeks now, but didn't
notice this because we never run tests with
'GIT_TEST_MULTI_PACK_INDEX=1' in the environment.
Perhaps in the future that might change, but I think that for now that
can explain why the failure wasn't noticed earlier.
> diff --git a/packfile.c b/packfile.c
> index 253559fa87..f0dc63e92f 100644
> --- a/packfile.c
> +++ b/packfile.c
> @@ -757,6 +757,9 @@ void install_packed_git(struct repository *r, struct packed_git *pack)
>
> pack->next = r->objects->packed_git;
> r->objects->packed_git = pack;
> +
> + hashmap_entry_init(&pack->packmap_ent, strhash(pack->pack_name));
> + hashmap_add(&r->objects->pack_map, &pack->packmap_ent);
> }
>
> void (*report_garbage)(unsigned seen_bits, const char *path);
> @@ -864,11 +867,8 @@ static void prepare_pack(const char *full_name, size_t full_name_len,
> /* Don't reopen a pack we already have. */
> if (!hashmap_get(&data->r->objects->pack_map, &hent, pack_name)) {
> p = add_packed_git(full_name, full_name_len, data->local);
> - if (p) {
> - hashmap_entry_init(&p->packmap_ent, hash);
> - hashmap_add(&data->r->objects->pack_map, &p->packmap_ent);
> + if (p)
> install_packed_git(data->r, p);
> - }
> }
> free(pack_name);
> }
>
> -Peff
Thanks,
Taylor
