From: Elijah Newren <newren@gmail.com>
To: "Nguyễn Thái Ngọc" <pclouds@gmail.com>
Cc: Ben Peart <Ben.Peart@microsoft.com>,
Git Mailing List <git@vger.kernel.org>,
Junio C Hamano <gitster@pobox.com>,
Ben Peart <peartben@gmail.com>, Jeff King <peff@peff.net>
Subject: Re: [PATCH v3 2/4] unpack-trees: optimize walking same trees with cache-tree
Date: Fri, 10 Aug 2018 11:48:33 -0700 [thread overview]
Message-ID: <CABPp-BEOFSU2k+DKuTQZtz+c6eboiUo3RDzBZHkB6=V3SFAigQ@mail.gmail.com> (raw)
In-Reply-To: <CACsJy8BtgMSYqkD1EaFQ=S49BA-veyTO1qU0FaPMkHY-KeggfA@mail.gmail.com>
On Fri, Aug 10, 2018 at 9:29 AM Duy Nguyen <pclouds@gmail.com> wrote:
> On Wed, Aug 8, 2018 at 8:23 PM Elijah Newren <newren@gmail.com> wrote:
> > > + * cache-tree would be invalidated and we would never get here
> > > + * in the first place.
> > > + */
> > > + for (i = 0; i < nr_entries; i++) {
> > > + struct cache_entry *tree_ce;
> > > + int len, rc;
> > > +
> > > + src[0] = o->src_index->cache[pos + i];
> > > +
> > > + len = ce_namelen(src[0]);
> > > + tree_ce = xcalloc(1, cache_entry_size(len));
> > > +
> > > + tree_ce->ce_mode = src[0]->ce_mode;
> > > + tree_ce->ce_flags = create_ce_flags(0);
> > > + tree_ce->ce_namelen = len;
> > > + oidcpy(&tree_ce->oid, &src[0]->oid);
> > > + memcpy(tree_ce->name, src[0]->name, len + 1);
> >
> > We do a bunch of work to setup tree_ce...
> >
> > > + for (d = 1; d <= nr_names; d++)
> > > + src[d] = tree_ce;
> >
> > ...then we make nr_names copies of tree_ce (so that *way_merge or
> > bind_merge or oneway_diff or whatever will have the expected number of
> > entries).
> >
> > > + rc = call_unpack_fn((const struct cache_entry * const *)src, o);
> >
> > ...then we call o->fn (via call_unpack_fn) to do various complicated
> > logic to figure out which tree_ce to use?? Isn't that just an
> > expensive way to recompute that what we currently have in the index is
> > what we want to keep there?
> >
> > Granted, a caller of this may have set o->fn to something other than
> > {one,two,three}way_merge (or bind_merge), and that function might have
> > important side effects...but it just seems annoying to have to do so
> > much work when for most uses we already know the entry in the index is
> > the one we already want.
>
> I'm not so sure about that. Which is why I keep it generic.
>
> > In fact, the only other thing in the
> > codebase that o->fn is now set to is oneway_diff, which I think is a
> > no-op when the two trees match.
> >
> > Would be nice if we could avoid all this, at least in the common cases
> > where o->fn is a function known to not have side effects. Or did I
> > not read those functions closely enough and they do have important
> > side effects?
>
> In one of my earlier "how about this" attempts, I introduced fn_same
> [1] that can help achieve this without carving "known not to have side
> effects" in common code. Which I think is still a good direction to go
> if we want to optimize more aggressively. We could have something like
> this
>
> diff --git a/unpack-trees.c b/unpack-trees.c
> index 1f11991a51..01b80389e0 100644
> --- a/unpack-trees.c
> +++ b/unpack-trees.c
> @@ -699,6 +699,9 @@ static int traverse_by_cache_tree(int pos, int
> nr_entries, int nr_names,
> int ce_len = 0;
> int i, d;
>
> + if (o->fn_cache_tree)
> + return o->fn_cache_tree(pos, nr_entries, nr_names, names, info);
> +
> if (!o->merge)
> BUG("We need cache-tree to do this optimization");
>
> then you can add, say threeway_cache_tree_merge(), that does what
> traverse_by_cache_tree() does but more efficient. This involves a lot
> more work (mostly staring and those n-merge functions and making sure
> you don't set the right conditions before going the fast path).
>
> I didn't do it because.. well.. it's more work and also riskier. I
> think we can leave that for later, unless you think we should do it
> now.
>
> [1] https://public-inbox.org/git/20180726163049.GA15572@duynguyen.home/
Yeah, from your other thread, I think I was missing some of the
intracacies of how the cache-tree works and the extra work that'd be
needed to bring it along. Deferring until later makes sense.
next prev parent reply other threads:[~2018-08-10 18:48 UTC|newest]
Thread overview: 121+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-07-18 20:45 [PATCH v1 0/3] [RFC] Speeding up checkout (and merge, rebase, etc) Ben Peart
2018-07-18 20:45 ` [PATCH v1 1/3] add unbounded Multi-Producer-Multi-Consumer queue Ben Peart
2018-07-18 20:57 ` Stefan Beller
2018-07-19 19:11 ` Junio C Hamano
2018-07-18 20:45 ` [PATCH v1 2/3] add performance tracing around traverse_trees() in unpack_trees() Ben Peart
2018-07-18 20:45 ` [PATCH v1 3/3] Add initial parallel version of unpack_trees() Ben Peart
2018-07-18 22:56 ` Junio C Hamano
2018-07-18 21:02 ` [PATCH v1 0/3] [RFC] Speeding up checkout (and merge, rebase, etc) Stefan Beller
2018-07-18 21:34 ` Jeff King
2018-07-23 15:48 ` Ben Peart
2018-07-23 17:03 ` Duy Nguyen
2018-07-23 20:51 ` Ben Peart
2018-07-24 4:20 ` Jeff King
2018-07-24 15:33 ` Duy Nguyen
2018-07-25 20:56 ` Ben Peart
2018-07-26 5:30 ` Duy Nguyen
2018-07-26 16:30 ` Duy Nguyen
2018-07-26 19:40 ` Junio C Hamano
2018-07-27 15:42 ` Duy Nguyen
2018-07-27 16:22 ` Ben Peart
2018-07-27 18:00 ` Duy Nguyen
2018-07-27 17:14 ` Junio C Hamano
2018-07-27 17:52 ` Duy Nguyen
2018-07-29 6:24 ` Duy Nguyen
2018-07-29 10:33 ` [PATCH v2 0/4] Speed up unpack_trees() Nguyễn Thái Ngọc Duy
2018-07-29 10:33 ` [PATCH v2 1/4] unpack-trees.c: add performance tracing Nguyễn Thái Ngọc Duy
2018-07-30 20:16 ` Ben Peart
2018-07-29 10:33 ` [PATCH v2 2/4] unpack-trees: optimize walking same trees with cache-tree Nguyễn Thái Ngọc Duy
2018-07-30 20:52 ` Ben Peart
2018-07-29 10:33 ` [PATCH v2 3/4] unpack-trees: reduce malloc in cache-tree walk Nguyễn Thái Ngọc Duy
2018-07-30 20:58 ` Ben Peart
2018-07-29 10:33 ` [PATCH v2 4/4] unpack-trees: cheaper index update when walking by cache-tree Nguyễn Thái Ngọc Duy
2018-08-08 18:46 ` Elijah Newren
2018-08-10 16:39 ` Duy Nguyen
2018-08-10 18:39 ` Elijah Newren
2018-08-10 19:30 ` Duy Nguyen
2018-08-10 19:40 ` Elijah Newren
2018-08-10 19:48 ` Duy Nguyen
2018-07-30 18:10 ` [PATCH v2 0/4] Speed up unpack_trees() Ben Peart
2018-07-31 15:31 ` Duy Nguyen
2018-07-31 16:50 ` Ben Peart
2018-07-31 17:31 ` Ben Peart
2018-08-01 16:38 ` Duy Nguyen
2018-08-08 20:53 ` Ben Peart
2018-08-09 8:16 ` Ben Peart
2018-08-10 16:08 ` Duy Nguyen
2018-08-10 15:51 ` Duy Nguyen
2018-07-30 21:04 ` Ben Peart
2018-08-04 5:37 ` [PATCH v3 " Nguyễn Thái Ngọc Duy
2018-08-04 5:37 ` [PATCH v3 1/4] unpack-trees: add performance tracing Nguyễn Thái Ngọc Duy
2018-08-04 5:37 ` [PATCH v3 2/4] unpack-trees: optimize walking same trees with cache-tree Nguyễn Thái Ngọc Duy
2018-08-08 18:23 ` Elijah Newren
2018-08-10 16:29 ` Duy Nguyen
2018-08-10 18:48 ` Elijah Newren [this message]
2018-08-04 5:37 ` [PATCH v3 3/4] unpack-trees: reduce malloc in cache-tree walk Nguyễn Thái Ngọc Duy
2018-08-08 18:30 ` Elijah Newren
2018-08-04 5:37 ` [PATCH v3 4/4] unpack-trees: cheaper index update when walking by cache-tree Nguyễn Thái Ngọc Duy
2018-08-06 15:48 ` [PATCH v3 0/4] Speed up unpack_trees() Junio C Hamano
2018-08-06 15:59 ` Duy Nguyen
2018-08-06 18:59 ` Junio C Hamano
2018-08-08 17:00 ` Ben Peart
2018-08-08 17:46 ` Junio C Hamano
2018-08-08 18:12 ` Junio C Hamano
2018-08-08 18:39 ` Junio C Hamano
2018-08-10 16:53 ` Duy Nguyen
2018-08-12 8:15 ` [PATCH v4 0/5] " Nguyễn Thái Ngọc Duy
2018-08-12 8:15 ` [PATCH v4 1/5] trace.h: support nested performance tracing Nguyễn Thái Ngọc Duy
2018-08-13 18:39 ` Ben Peart
2018-08-12 8:15 ` [PATCH v4 2/5] unpack-trees: add " Nguyễn Thái Ngọc Duy
2018-08-12 10:05 ` Thomas Adam
2018-08-13 18:50 ` Junio C Hamano
2018-08-13 18:44 ` Ben Peart
2018-08-13 19:25 ` Jeff King
2018-08-13 19:36 ` Stefan Beller
2018-08-13 20:11 ` Ben Peart
2018-08-13 19:52 ` Duy Nguyen
2018-08-13 21:47 ` Jeff King
2018-08-13 22:41 ` Junio C Hamano
2018-08-14 18:19 ` Jeff Hostetler
2018-08-14 18:32 ` Duy Nguyen
2018-08-14 18:44 ` Stefan Beller
2018-08-14 18:51 ` Duy Nguyen
2018-08-14 19:54 ` Jeff King
2018-08-14 20:52 ` Junio C Hamano
2018-08-15 16:32 ` Duy Nguyen
2018-08-15 18:28 ` Junio C Hamano
2018-08-14 20:14 ` Jeff Hostetler
2018-08-12 8:15 ` [PATCH v4 3/5] unpack-trees: optimize walking same trees with cache-tree Nguyễn Thái Ngọc Duy
2018-08-13 18:58 ` Ben Peart
2018-08-15 16:38 ` Duy Nguyen
2018-08-12 8:15 ` [PATCH v4 4/5] unpack-trees: reduce malloc in cache-tree walk Nguyễn Thái Ngọc Duy
2018-08-12 8:15 ` [PATCH v4 5/5] unpack-trees: reuse (still valid) cache-tree from src_index Nguyễn Thái Ngọc Duy
2018-08-13 15:48 ` Elijah Newren
2018-08-13 15:57 ` Duy Nguyen
2018-08-13 16:05 ` Ben Peart
2018-08-13 16:25 ` Duy Nguyen
2018-08-13 17:15 ` Ben Peart
2018-08-13 19:01 ` [PATCH v4 0/5] Speed up unpack_trees() Junio C Hamano
2018-08-14 19:19 ` Ben Peart
2018-08-18 14:41 ` [PATCH v5 0/7] " Nguyễn Thái Ngọc Duy
2018-08-18 14:41 ` [PATCH v5 1/7] trace.h: support nested performance tracing Nguyễn Thái Ngọc Duy
2018-08-18 14:41 ` [PATCH v5 2/7] unpack-trees: add " Nguyễn Thái Ngọc Duy
2018-08-18 14:41 ` [PATCH v5 3/7] unpack-trees: optimize walking same trees with cache-tree Nguyễn Thái Ngọc Duy
2018-08-20 12:43 ` Ben Peart
2018-08-18 14:41 ` [PATCH v5 4/7] unpack-trees: reduce malloc in cache-tree walk Nguyễn Thái Ngọc Duy
2018-08-18 14:41 ` [PATCH v5 5/7] unpack-trees: reuse (still valid) cache-tree from src_index Nguyễn Thái Ngọc Duy
2018-08-18 14:41 ` [PATCH v5 6/7] unpack-trees: add missing cache invalidation Nguyễn Thái Ngọc Duy
2018-08-18 14:41 ` [PATCH v5 7/7] cache-tree: verify valid cache-tree in the test suite Nguyễn Thái Ngọc Duy
2018-08-18 21:45 ` Elijah Newren
2018-08-18 22:01 ` [PATCH v5 0/7] Speed up unpack_trees() Elijah Newren
2018-08-19 5:09 ` Duy Nguyen
2018-08-25 12:18 ` [PATCH] Document update for nd/unpack-trees-with-cache-tree Nguyễn Thái Ngọc Duy
2018-08-25 12:31 ` Martin Ågren
2018-08-25 13:02 ` [PATCH v2] " Nguyễn Thái Ngọc Duy
2018-07-27 15:50 ` [PATCH v1 0/3] [RFC] Speeding up checkout (and merge, rebase, etc) Ben Peart
2018-07-26 16:35 ` Duy Nguyen
2018-07-24 5:54 ` Junio C Hamano
2018-07-24 15:13 ` Duy Nguyen
2018-07-24 21:21 ` Jeff King
2018-07-25 16:09 ` Ben Peart
2018-07-24 4:27 ` Jeff King
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: http://vger.kernel.org/majordomo-info.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CABPp-BEOFSU2k+DKuTQZtz+c6eboiUo3RDzBZHkB6=V3SFAigQ@mail.gmail.com' \
--to=newren@gmail.com \
--cc=Ben.Peart@microsoft.com \
--cc=git@vger.kernel.org \
--cc=gitster@pobox.com \
--cc=pclouds@gmail.com \
--cc=peartben@gmail.com \
--cc=peff@peff.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).