git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Taylor Blau <me@ttaylorr.com>
To: Jonathan Tan <jonathantanmy@google.com>
Cc: git@vger.kernel.org, peff@peff.net, dstolee@microsoft.com,
	avarab@gmail.com, gitster@pobox.com
Subject: Re: [PATCH v2 08/15] midx: allow marking a pack as preferred
Date: Tue, 2 Mar 2021 14:09:11 -0500	[thread overview]
Message-ID: <YD6NVxDib8ccf/6Z@nand.local> (raw)
In-Reply-To: <20210302041753.4037658-1-jonathantanmy@google.com>

On Mon, Mar 01, 2021 at 08:17:53PM -0800, Jonathan Tan wrote:
> I was initially confused that "preferred" was set twice, but this makes
> sense - the first one is when an existing midx is reused, and the second
> one is for objects in packs that the midx (if it exists) does not cover.

Yep. Those two paths permeate a lot of the MIDX writer code, since it
wants to reuse work from an existing MIDX if it can find one.

> > @@ -828,7 +869,19 @@ static int write_midx_internal(const char *object_dir, struct multi_pack_index *
> >  	if (ctx.m && ctx.nr == ctx.m->num_packs && !packs_to_drop)
> >  		goto cleanup;
> >
> > -	ctx.entries = get_sorted_entries(ctx.m, ctx.info, ctx.nr, &ctx.entries_nr);
> > +	if (preferred_pack_name) {
> > +		for (i = 0; i < ctx.nr; i++) {
> > +			if (!cmp_idx_or_pack_name(preferred_pack_name,
> > +						  ctx.info[i].pack_name)) {
> > +				ctx.preferred_pack_idx = i;
> > +				break;
> > +			}
> > +		}
> > +	} else
> > +		ctx.preferred_pack_idx = -1;
>
> Looks safer to put "ctx.preferred_pack_idx = -1" before the "if", just
> in case the given pack name does not exist?

Agreed.

> > @@ -889,6 +942,31 @@ static int write_midx_internal(const char *object_dir, struct multi_pack_index *
> >  			pack_name_concat_len += strlen(ctx.info[i].pack_name) + 1;
> >  	}
> >
> > +	/*
> > +	 * Recompute the preferred_pack_idx (if applicable) according to the
> > +	 * permuted pack order.
> > +	 */
> > +	ctx.preferred_pack_idx = -1;
> > +	if (preferred_pack_name) {
> > +		ctx.preferred_pack_idx = lookup_idx_or_pack_name(ctx.info,
> > +							     ctx.nr,
> > +							     preferred_pack_name);
> > +		if (ctx.preferred_pack_idx < 0)
> > +			warning(_("unknown preferred pack: '%s'"),
> > +				preferred_pack_name);
> > +		else {
> > +			uint32_t orig = ctx.info[ctx.preferred_pack_idx].orig_pack_int_id;
> > +			uint32_t perm = ctx.pack_perm[orig];
> > +
> > +			if (perm == PACK_EXPIRED) {
> > +				warning(_("preferred pack '%s' is expired"),
> > +					preferred_pack_name);
> > +				ctx.preferred_pack_idx = -1;
> > +			} else
> > +				ctx.preferred_pack_idx = perm;
> > +		}
> > +	}
>
> I couldn't figure out why the preferred pack index needs to be
> recalculated here, since the pack entries would have already been
> sorted. Also, the tests still pass when I comment this part out. A
> comment describing what's going on would be helpful.

Funny you mention that; I was wondering the same thing myself the other
day when reading these patches again before deploying them to a couple
of testing repositories at GitHub.

It is totally unnecessary: since we have already marked objects from the
preferred pack in get_sorted_entries(), the rest of the code doesn't
care if the preferred pack was permuted or not.

But we *do* care if the pack which was preferred expired. The 'git
repack --geometric --write-midx' caller (which will appear in a later
series) should never do that, so emitting a warning() is worthwhile. I
think ultimately you want something like this squashed in:

--- >8 ---

diff --git a/midx.c b/midx.c
index d2c56c4bc6..46f55ff6cf 100644
--- a/midx.c
+++ b/midx.c
@@ -582,7 +582,7 @@ static struct pack_midx_entry *get_sorted_entries(struct multi_pack_index *m,
 						  struct pack_info *info,
 						  uint32_t nr_packs,
 						  uint32_t *nr_objects,
-						  uint32_t preferred_pack)
+						  int preferred_pack)
 {
 	uint32_t cur_fanout, cur_pack, cur_object;
 	uint32_t alloc_fanout, alloc_objects, total_objects = 0;
@@ -869,6 +869,7 @@ static int write_midx_internal(const char *object_dir, struct multi_pack_index *
 	if (ctx.m && ctx.nr == ctx.m->num_packs && !packs_to_drop)
 		goto cleanup;

+	ctx.preferred_pack_idx = -1;
 	if (preferred_pack_name) {
 		for (i = 0; i < ctx.nr; i++) {
 			if (!cmp_idx_or_pack_name(preferred_pack_name,
@@ -877,8 +878,7 @@ static int write_midx_internal(const char *object_dir, struct multi_pack_index *
 				break;
 			}
 		}
-	} else
-		ctx.preferred_pack_idx = -1;
+	}

 	ctx.entries = get_sorted_entries(ctx.m, ctx.info, ctx.nr, &ctx.entries_nr,
 					 ctx.preferred_pack_idx);
@@ -942,28 +942,21 @@ static int write_midx_internal(const char *object_dir, struct multi_pack_index *
 			pack_name_concat_len += strlen(ctx.info[i].pack_name) + 1;
 	}

-	/*
-	 * Recompute the preferred_pack_idx (if applicable) according to the
-	 * permuted pack order.
-	 */
-	ctx.preferred_pack_idx = -1;
+	/* Check that the preferred pack wasn't expired (if given). */
 	if (preferred_pack_name) {
-		ctx.preferred_pack_idx = lookup_idx_or_pack_name(ctx.info,
-							     ctx.nr,
-							     preferred_pack_name);
-		if (ctx.preferred_pack_idx < 0)
+		int preferred_idx = lookup_idx_or_pack_name(ctx.info,
+							    ctx.nr,
+							    preferred_pack_name);
+		if (preferred_idx < 0)
 			warning(_("unknown preferred pack: '%s'"),
 				preferred_pack_name);
 		else {
-			uint32_t orig = ctx.info[ctx.preferred_pack_idx].orig_pack_int_id;
+			uint32_t orig = ctx.info[preferred_idx].orig_pack_int_id;
 			uint32_t perm = ctx.pack_perm[orig];

-			if (perm == PACK_EXPIRED) {
+			if (perm == PACK_EXPIRED)
 				warning(_("preferred pack '%s' is expired"),
 					preferred_pack_name);
-				ctx.preferred_pack_idx = -1;
-			} else
-				ctx.preferred_pack_idx = perm;
 		}
 	}


  reply	other threads:[~2021-03-03  6:41 UTC|newest]

Thread overview: 171+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-02-10 23:02 [PATCH 0/9] midx: implement a multi-pack reverse index Taylor Blau
2021-02-10 23:02 ` [PATCH 1/9] t/helper/test-read-midx.c: add '--show-objects' Taylor Blau
2021-02-11  2:27   ` Derrick Stolee
2021-02-11  2:34     ` Taylor Blau
2021-02-10 23:02 ` [PATCH 2/9] midx: allow marking a pack as preferred Taylor Blau
2021-02-11 19:33   ` SZEDER Gábor
2021-02-15 15:49     ` Taylor Blau
2021-02-15 17:01       ` Ævar Arnfjörð Bjarmason
2021-02-15 18:41         ` [PATCH 0/5] commit-graph: parse_options() cleanup Ævar Arnfjörð Bjarmason
2021-02-15 18:41         ` [PATCH 1/5] commit-graph: define common usage with a macro Ævar Arnfjörð Bjarmason
2021-02-16 11:33           ` Derrick Stolee
2021-02-15 18:41         ` [PATCH 2/5] commit-graph: remove redundant handling of -h Ævar Arnfjörð Bjarmason
2021-02-16 11:35           ` Derrick Stolee
2021-02-15 18:41         ` [PATCH 3/5] commit-graph: use parse_options_concat() Ævar Arnfjörð Bjarmason
2021-02-15 18:51           ` Taylor Blau
2021-02-15 19:53             ` Taylor Blau
2021-02-15 20:39             ` Ævar Arnfjörð Bjarmason
2021-09-17 21:13               ` SZEDER Gábor
2021-09-17 22:03                 ` Jeff King
2021-09-18  4:30                   ` Taylor Blau
2021-09-18  7:20                     ` Ævar Arnfjörð Bjarmason
2021-09-18 15:56                       ` Taylor Blau
2021-09-18 15:58                         ` Taylor Blau
2021-09-18  0:58                 ` Ævar Arnfjörð Bjarmason
2021-02-15 18:41         ` [PATCH 4/5] commit-graph: refactor dispatch loop for style Ævar Arnfjörð Bjarmason
2021-02-15 18:53           ` Taylor Blau
2021-02-16 11:40             ` Derrick Stolee
2021-02-16 12:02               ` Ævar Arnfjörð Bjarmason
2021-02-16 18:28                 ` Derrick Stolee
2021-02-15 18:41         ` [PATCH 5/5] commit-graph: show usage on "commit-graph [write|verify] garbage" Ævar Arnfjörð Bjarmason
2021-02-15 19:06           ` Taylor Blau
2021-02-16 11:43           ` Derrick Stolee
2021-02-15 21:01         ` [PATCH v2 0/4] midx: split out sub-commands Taylor Blau
2021-02-15 21:01           ` [PATCH v2 1/4] builtin/multi-pack-index.c: inline 'flags' with options Taylor Blau
2021-02-15 21:01           ` [PATCH v2 2/4] builtin/multi-pack-index.c: don't handle 'progress' separately Taylor Blau
2021-02-15 21:39             ` Ævar Arnfjörð Bjarmason
2021-02-15 21:45               ` Taylor Blau
2021-02-16 11:47                 ` Derrick Stolee
2021-02-15 21:01           ` [PATCH v2 3/4] builtin/multi-pack-index.c: define common usage with a macro Taylor Blau
2021-02-15 21:01           ` [PATCH v2 4/4] builtin/multi-pack-index.c: split sub-commands Taylor Blau
2021-02-15 21:54             ` Ævar Arnfjörð Bjarmason
2021-02-15 22:34               ` Taylor Blau
2021-02-15 23:11                 ` Ævar Arnfjörð Bjarmason
2021-02-15 23:49                   ` Taylor Blau
2021-02-16 11:50           ` [PATCH v2 0/4] midx: split out sub-commands Derrick Stolee
2021-02-16 14:28             ` Taylor Blau
2021-02-10 23:02 ` [PATCH 3/9] midx: don't free midx_name early Taylor Blau
2021-02-10 23:02 ` [PATCH 4/9] midx: keep track of the checksum Taylor Blau
2021-02-11  2:33   ` Derrick Stolee
2021-02-11  2:35     ` Taylor Blau
2021-02-10 23:03 ` [PATCH 5/9] midx: make some functions non-static Taylor Blau
2021-02-10 23:03 ` [PATCH 6/9] Documentation/technical: describe multi-pack reverse indexes Taylor Blau
2021-02-11  2:48   ` Derrick Stolee
2021-02-11  3:03     ` Taylor Blau
2021-02-10 23:03 ` [PATCH 7/9] pack-revindex: read " Taylor Blau
2021-02-11  2:53   ` Derrick Stolee
2021-02-11  3:04     ` Taylor Blau
2021-02-11  7:54   ` Junio C Hamano
2021-02-11 14:54     ` Taylor Blau
2021-02-10 23:03 ` [PATCH 8/9] pack-write.c: extract 'write_rev_file_order' Taylor Blau
2021-02-10 23:03 ` [PATCH 9/9] pack-revindex: write multi-pack reverse indexes Taylor Blau
2021-02-11  2:58 ` [PATCH 0/9] midx: implement a multi-pack reverse index Derrick Stolee
2021-02-11  3:06   ` Taylor Blau
2021-02-11  8:13 ` Junio C Hamano
2021-02-11 18:37   ` Derrick Stolee
2021-02-11 18:55     ` Junio C Hamano
2021-02-24 19:09 ` [PATCH v2 00/15] " Taylor Blau
2021-02-24 19:09   ` [PATCH v2 01/15] builtin/multi-pack-index.c: inline 'flags' with options Taylor Blau
2021-02-24 19:09   ` [PATCH v2 02/15] builtin/multi-pack-index.c: don't handle 'progress' separately Taylor Blau
2021-02-24 19:09   ` [PATCH v2 03/15] builtin/multi-pack-index.c: define common usage with a macro Taylor Blau
2021-02-24 19:09   ` [PATCH v2 04/15] builtin/multi-pack-index.c: split sub-commands Taylor Blau
2021-03-02  4:06     ` Jonathan Tan
2021-03-02 19:02       ` Taylor Blau
2021-03-04  1:54         ` Jonathan Tan
2021-03-04  3:02           ` Taylor Blau
2021-02-24 19:09   ` [PATCH v2 05/15] builtin/multi-pack-index.c: don't enter bogus cmd_mode Taylor Blau
2021-02-24 19:09   ` [PATCH v2 06/15] builtin/multi-pack-index.c: display usage on unrecognized command Taylor Blau
2021-02-24 19:09   ` [PATCH v2 07/15] t/helper/test-read-midx.c: add '--show-objects' Taylor Blau
2021-02-24 19:09   ` [PATCH v2 08/15] midx: allow marking a pack as preferred Taylor Blau
2021-03-02  4:17     ` Jonathan Tan
2021-03-02 19:09       ` Taylor Blau [this message]
2021-03-04  2:00         ` Jonathan Tan
2021-03-04  3:04           ` Taylor Blau
2021-02-24 19:09   ` [PATCH v2 09/15] midx: don't free midx_name early Taylor Blau
2021-02-24 19:10   ` [PATCH v2 10/15] midx: keep track of the checksum Taylor Blau
2021-02-24 19:10   ` [PATCH v2 11/15] midx: make some functions non-static Taylor Blau
2021-02-24 19:10   ` [PATCH v2 12/15] Documentation/technical: describe multi-pack reverse indexes Taylor Blau
2021-03-02  4:21     ` Jonathan Tan
2021-03-02  4:36       ` Taylor Blau
2021-03-02 19:15       ` Taylor Blau
2021-03-04  2:03         ` Jonathan Tan
2021-02-24 19:10   ` [PATCH v2 13/15] pack-revindex: read " Taylor Blau
2021-03-02 18:36     ` Jonathan Tan
2021-03-03 15:27       ` Taylor Blau
2021-02-24 19:10   ` [PATCH v2 14/15] pack-write.c: extract 'write_rev_file_order' Taylor Blau
2021-02-24 19:10   ` [PATCH v2 15/15] pack-revindex: write multi-pack reverse indexes Taylor Blau
2021-03-02 18:40     ` Jonathan Tan
2021-03-03 15:30       ` Taylor Blau
2021-03-04  2:04         ` Jonathan Tan
2021-03-04  3:06           ` Taylor Blau
2021-03-11 17:04 ` [PATCH v3 00/16] midx: implement a multi-pack reverse index Taylor Blau
2021-03-11 17:04   ` [PATCH v3 01/16] builtin/multi-pack-index.c: inline 'flags' with options Taylor Blau
2021-03-29 11:20     ` Jeff King
2021-03-11 17:04   ` [PATCH v3 02/16] builtin/multi-pack-index.c: don't handle 'progress' separately Taylor Blau
2021-03-29 11:22     ` Jeff King
2021-03-11 17:04   ` [PATCH v3 03/16] builtin/multi-pack-index.c: define common usage with a macro Taylor Blau
2021-03-11 17:04   ` [PATCH v3 04/16] builtin/multi-pack-index.c: split sub-commands Taylor Blau
2021-03-29 11:36     ` Jeff King
2021-03-29 20:38       ` Taylor Blau
2021-03-30  7:04         ` Jeff King
2021-03-11 17:04   ` [PATCH v3 05/16] builtin/multi-pack-index.c: don't enter bogus cmd_mode Taylor Blau
2021-03-11 17:04   ` [PATCH v3 06/16] builtin/multi-pack-index.c: display usage on unrecognized command Taylor Blau
2021-03-29 11:42     ` Jeff King
2021-03-29 20:41       ` Taylor Blau
2021-03-11 17:05   ` [PATCH v3 07/16] t/helper/test-read-midx.c: add '--show-objects' Taylor Blau
2021-03-11 17:05   ` [PATCH v3 08/16] midx: allow marking a pack as preferred Taylor Blau
2021-03-29 12:00     ` Jeff King
2021-03-29 21:15       ` Taylor Blau
2021-03-30  7:11         ` Jeff King
2021-03-11 17:05   ` [PATCH v3 09/16] midx: don't free midx_name early Taylor Blau
2021-03-11 17:05   ` [PATCH v3 10/16] midx: keep track of the checksum Taylor Blau
2021-03-11 17:05   ` [PATCH v3 11/16] midx: make some functions non-static Taylor Blau
2021-03-11 17:05   ` [PATCH v3 12/16] Documentation/technical: describe multi-pack reverse indexes Taylor Blau
2021-03-29 12:12     ` Jeff King
2021-03-29 21:22       ` Taylor Blau
2021-03-11 17:05   ` [PATCH v3 13/16] pack-revindex: read " Taylor Blau
2021-03-29 12:43     ` Jeff King
2021-03-29 21:27       ` Taylor Blau
2021-03-11 17:05   ` [PATCH v3 14/16] pack-write.c: extract 'write_rev_file_order' Taylor Blau
2021-03-11 17:05   ` [PATCH v3 15/16] pack-revindex: write multi-pack reverse indexes Taylor Blau
2021-03-29 12:53     ` Jeff King
2021-03-29 21:30       ` Taylor Blau
2021-03-11 17:05   ` [PATCH v3 16/16] midx.c: improve cache locality in midx_pack_order_cmp() Taylor Blau
2021-03-29 12:59     ` Jeff King
2021-03-29 21:34       ` Taylor Blau
2021-03-30  7:15         ` Jeff King
2021-03-12 15:16   ` [PATCH v3 00/16] midx: implement a multi-pack reverse index Derrick Stolee
2021-03-29 13:05   ` Jeff King
2021-03-29 21:30     ` Junio C Hamano
2021-03-29 21:37     ` Taylor Blau
2021-03-30  7:15       ` Jeff King
2021-03-30 13:37         ` Taylor Blau
2021-03-30 15:03 ` [PATCH v4 " Taylor Blau
2021-03-30 15:03   ` [PATCH v4 01/16] builtin/multi-pack-index.c: inline 'flags' with options Taylor Blau
2021-03-30 15:03   ` [PATCH v4 02/16] builtin/multi-pack-index.c: don't handle 'progress' separately Taylor Blau
2021-03-30 15:03   ` [PATCH v4 03/16] builtin/multi-pack-index.c: define common usage with a macro Taylor Blau
2021-03-30 15:03   ` [PATCH v4 04/16] builtin/multi-pack-index.c: split sub-commands Taylor Blau
2021-03-30 15:04   ` [PATCH v4 05/16] builtin/multi-pack-index.c: don't enter bogus cmd_mode Taylor Blau
2021-03-30 15:04   ` [PATCH v4 06/16] builtin/multi-pack-index.c: display usage on unrecognized command Taylor Blau
2021-03-30 15:04   ` [PATCH v4 07/16] t/helper/test-read-midx.c: add '--show-objects' Taylor Blau
2021-03-30 15:04   ` [PATCH v4 08/16] midx: allow marking a pack as preferred Taylor Blau
2021-04-01  0:32     ` Taylor Blau
2021-03-30 15:04   ` [PATCH v4 09/16] midx: don't free midx_name early Taylor Blau
2021-03-30 15:04   ` [PATCH v4 10/16] midx: keep track of the checksum Taylor Blau
2021-03-30 15:04   ` [PATCH v4 11/16] midx: make some functions non-static Taylor Blau
2021-03-30 15:04   ` [PATCH v4 12/16] Documentation/technical: describe multi-pack reverse indexes Taylor Blau
2021-03-30 15:04   ` [PATCH v4 13/16] pack-revindex: read " Taylor Blau
2021-03-30 15:04   ` [PATCH v4 14/16] pack-write.c: extract 'write_rev_file_order' Taylor Blau
2021-09-08  1:08     ` [PATCH] pack-write: skip *.rev work when not writing *.rev Ævar Arnfjörð Bjarmason
2021-09-08  1:35       ` Carlo Arenas
2021-09-08  2:42         ` Taylor Blau
2021-09-08 15:47           ` Junio C Hamano
2021-09-08  2:50       ` Taylor Blau
2021-09-08  3:50         ` Taylor Blau
2021-09-08 10:18           ` Ævar Arnfjörð Bjarmason
2021-09-08 16:32             ` Taylor Blau
2021-03-30 15:04   ` [PATCH v4 15/16] pack-revindex: write multi-pack reverse indexes Taylor Blau
2021-03-30 15:04   ` [PATCH v4 16/16] midx.c: improve cache locality in midx_pack_order_cmp() Taylor Blau
2021-03-30 15:45   ` [PATCH v4 00/16] midx: implement a multi-pack reverse index Jeff King
2021-03-30 15:49     ` Taylor Blau
2021-03-30 16:01       ` 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

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=YD6NVxDib8ccf/6Z@nand.local \
    --to=me@ttaylorr.com \
    --cc=avarab@gmail.com \
    --cc=dstolee@microsoft.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=jonathantanmy@google.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.
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 NNTP newsgroup(s).