[PATCH] gc: add `--expire-to` option

[PATCH] gc: add `--expire-to` option

[ZheNing Hu via GitGitGadget]

From: ZheNing Hu <adlternative@gmail.com>

This commit extends the functionality of `git gc`
by adding a new option, `--expire-to=<dir>`. Previously,
this feature was implemented in `git repack` (see 91badeb),
allowing users to specify a directory where unreachable and
expired cruft packs are stored during garbage collection.
However, users had to run `git repack --cruft --expire-to=<dir>`
followed by `git prune` to achieve similar results within `git gc`.

By introducing `--expire-to=<dir>` directly into `git gc`,
we simplify the process for users who wish to manage their
repository's cleanup more efficiently. This change involves
passing the `--expire-to=<dir>` parameter through to `git repack`,
making it easier for users to set up a backup location for cruft
packs that will be pruned.

Signed-off-by: ZheNing Hu <adlternative@gmail.com>
---
    gc: add --expire-to option
    
    I want to perform a "safe" garbage collection for the Git repository on
    the server, which avoids data corruption issues caused by concurrent
    pushes during git GC. To achieve this, I currently need to use git
    repack --cruft --expire-to=<dir> and git prune in combination. However,
    it would be simpler if we could directly use --expire-to=<dir> with the
    git-gc command.

Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-1843%2Fadlternative%2Fzh%2Fgc-expire-to-v1
Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-1843/adlternative/zh/gc-expire-to-v1
Pull-Request: https://github.com/gitgitgadget/git/pull/1843

 Documentation/git-gc.txt | 6 ++++++
 builtin/gc.c             | 6 +++++-
 t/t6500-gc.sh            | 6 ++++++
 3 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
index 370e22faaeb..b4c0cf02972 100644
--- a/Documentation/git-gc.txt
+++ b/Documentation/git-gc.txt
@@ -69,6 +69,12 @@ be performed as well.
 	the `--max-cruft-size` option of linkgit:git-repack[1] for
 	more.
 
+--expire-to=<dir>::
+	When packing unreachable objects into a cruft pack, write a cruft
+	pack containing pruned objects (if any) to the directory `<dir>`.
+	See the `--expire-to` option of linkgit:git-repack[1] for
+	more.
+
 --prune=<date>::
 	Prune loose objects older than date (default is 2 weeks ago,
 	overridable by the config variable `gc.pruneExpire`).
diff --git a/builtin/gc.c b/builtin/gc.c
index d52735354c9..77904694c9f 100644
--- a/builtin/gc.c
+++ b/builtin/gc.c
@@ -136,6 +136,7 @@ struct gc_config {
 	char *prune_worktrees_expire;
 	char *repack_filter;
 	char *repack_filter_to;
+	char *repack_expire_to;
 	unsigned long big_pack_threshold;
 	unsigned long max_delta_cache_size;
 };
@@ -441,6 +442,8 @@ static void add_repack_all_option(struct gc_config *cfg,
 		if (cfg->max_cruft_size)
 			strvec_pushf(&repack, "--max-cruft-size=%lu",
 				     cfg->max_cruft_size);
+		if (cfg->repack_expire_to)
+			strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);
 	} else {
 		strvec_push(&repack, "-A");
 		if (cfg->prune_expire)
@@ -675,7 +678,6 @@ struct repository *repo UNUSED)
 	const char *prune_expire_sentinel = "sentinel";
 	const char *prune_expire_arg = prune_expire_sentinel;
 	int ret;
-
 	struct option builtin_gc_options[] = {
 		OPT__QUIET(&quiet, N_("suppress progress reporting")),
 		{ OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),
@@ -694,6 +696,8 @@ struct repository *repo UNUSED)
 			   PARSE_OPT_NOCOMPLETE),
 		OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
 			 N_("repack all other packs except the largest pack")),
+		OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
+			   N_("pack prefix to store a pack containing pruned objects")),
 		OPT_END()
 	};
 
diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
index ee074b99b70..d4b0653a9b7 100755
--- a/t/t6500-gc.sh
+++ b/t/t6500-gc.sh
@@ -339,6 +339,12 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
 	test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
 '
 
+test_expect_success '--expire-to sets appropriate repack options' '
+	mkdir expired &&
+	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to=./expired/pack &&
+	test_subcommand $cruft_max_size_opts --expire-to=./expired/pack <trace2.txt
+'
+
 run_and_wait_for_gc () {
 	# We read stdout from gc for the side effect of waiting until the
 	# background gc process exits, closing its fd 9.  Furthermore, the

base-commit: 92999a42db1c5f43f330e4f2bca4026b5b81576f
-- 
gitgitgadget

[PATCH v2 0/2] gc: add --expire-to option

[ZheNing Hu via GitGitGadget]

I want to perform a "safe" garbage collection for the Git repository on the
server, which avoids data corruption issues caused by concurrent pushes
during git GC. To achieve this, I currently need to use git repack --cruft
--expire-to=<dir> and git prune in combination. However, it would be simpler
if we could directly use --expire-to=<dir> with the git-gc command.

ZheNing Hu (2):
  gc: add `--expire-to` option
  fix(gc): make --prune=now compatible with --expire-to

 Documentation/git-gc.txt | 6 ++++++
 builtin/gc.c             | 9 +++++++--
 t/t6500-gc.sh            | 6 ++++++
 3 files changed, 19 insertions(+), 2 deletions(-)


base-commit: 92999a42db1c5f43f330e4f2bca4026b5b81576f
Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-1843%2Fadlternative%2Fzh%2Fgc-expire-to-v2
Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-1843/adlternative/zh/gc-expire-to-v2
Pull-Request: https://github.com/gitgitgadget/git/pull/1843

Range-diff vs v1:

 1:  14e94bf04e5 = 1:  14e94bf04e5 gc: add `--expire-to` option
 -:  ----------- > 2:  579757957d2 fix(gc): make --prune=now compatible with --expire-to

-- 
gitgitgadget

[PATCH v2 1/2] gc: add `--expire-to` option

[ZheNing Hu via GitGitGadget]

From: ZheNing Hu <adlternative@gmail.com>

This commit extends the functionality of `git gc`
by adding a new option, `--expire-to=<dir>`. Previously,
this feature was implemented in `git repack` (see 91badeb),
allowing users to specify a directory where unreachable and
expired cruft packs are stored during garbage collection.
However, users had to run `git repack --cruft --expire-to=<dir>`
followed by `git prune` to achieve similar results within `git gc`.

By introducing `--expire-to=<dir>` directly into `git gc`,
we simplify the process for users who wish to manage their
repository's cleanup more efficiently. This change involves
passing the `--expire-to=<dir>` parameter through to `git repack`,
making it easier for users to set up a backup location for cruft
packs that will be pruned.

Signed-off-by: ZheNing Hu <adlternative@gmail.com>
---
 Documentation/git-gc.txt | 6 ++++++
 builtin/gc.c             | 6 +++++-
 t/t6500-gc.sh            | 6 ++++++
 3 files changed, 17 insertions(+), 1 deletion(-)

diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
index 370e22faaeb..b4c0cf02972 100644
--- a/Documentation/git-gc.txt
+++ b/Documentation/git-gc.txt
@@ -69,6 +69,12 @@ be performed as well.
 	the `--max-cruft-size` option of linkgit:git-repack[1] for
 	more.
 
+--expire-to=<dir>::
+	When packing unreachable objects into a cruft pack, write a cruft
+	pack containing pruned objects (if any) to the directory `<dir>`.
+	See the `--expire-to` option of linkgit:git-repack[1] for
+	more.
+
 --prune=<date>::
 	Prune loose objects older than date (default is 2 weeks ago,
 	overridable by the config variable `gc.pruneExpire`).
diff --git a/builtin/gc.c b/builtin/gc.c
index d52735354c9..77904694c9f 100644
--- a/builtin/gc.c
+++ b/builtin/gc.c
@@ -136,6 +136,7 @@ struct gc_config {
 	char *prune_worktrees_expire;
 	char *repack_filter;
 	char *repack_filter_to;
+	char *repack_expire_to;
 	unsigned long big_pack_threshold;
 	unsigned long max_delta_cache_size;
 };
@@ -441,6 +442,8 @@ static void add_repack_all_option(struct gc_config *cfg,
 		if (cfg->max_cruft_size)
 			strvec_pushf(&repack, "--max-cruft-size=%lu",
 				     cfg->max_cruft_size);
+		if (cfg->repack_expire_to)
+			strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);
 	} else {
 		strvec_push(&repack, "-A");
 		if (cfg->prune_expire)
@@ -675,7 +678,6 @@ struct repository *repo UNUSED)
 	const char *prune_expire_sentinel = "sentinel";
 	const char *prune_expire_arg = prune_expire_sentinel;
 	int ret;
-
 	struct option builtin_gc_options[] = {
 		OPT__QUIET(&quiet, N_("suppress progress reporting")),
 		{ OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),
@@ -694,6 +696,8 @@ struct repository *repo UNUSED)
 			   PARSE_OPT_NOCOMPLETE),
 		OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
 			 N_("repack all other packs except the largest pack")),
+		OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
+			   N_("pack prefix to store a pack containing pruned objects")),
 		OPT_END()
 	};
 
diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
index ee074b99b70..d4b0653a9b7 100755
--- a/t/t6500-gc.sh
+++ b/t/t6500-gc.sh
@@ -339,6 +339,12 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
 	test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
 '
 
+test_expect_success '--expire-to sets appropriate repack options' '
+	mkdir expired &&
+	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to=./expired/pack &&
+	test_subcommand $cruft_max_size_opts --expire-to=./expired/pack <trace2.txt
+'
+
 run_and_wait_for_gc () {
 	# We read stdout from gc for the side effect of waiting until the
 	# background gc process exits, closing its fd 9.  Furthermore, the
-- 
gitgitgadget

Re: [PATCH v2 1/2] gc: add `--expire-to` option

[ZheNing Hu]

ZheNing Hu via GitGitGadget <gitgitgadget@gmail.com> 于2024年12月31日周二 10:18写道：
>
> From: ZheNing Hu <adlternative@gmail.com>
>
> This commit extends the functionality of `git gc`
> by adding a new option, `--expire-to=<dir>`. Previously,
> this feature was implemented in `git repack` (see 91badeb),
> allowing users to specify a directory where unreachable and
> expired cruft packs are stored during garbage collection.
> However, users had to run `git repack --cruft --expire-to=<dir>`
> followed by `git prune` to achieve similar results within `git gc`.
>
> By introducing `--expire-to=<dir>` directly into `git gc`,
> we simplify the process for users who wish to manage their
> repository's cleanup more efficiently. This change involves
> passing the `--expire-to=<dir>` parameter through to `git repack`,
> making it easier for users to set up a backup location for cruft
> packs that will be pruned.
>
> Signed-off-by: ZheNing Hu <adlternative@gmail.com>
> ---
>  Documentation/git-gc.txt | 6 ++++++
>  builtin/gc.c             | 6 +++++-
>  t/t6500-gc.sh            | 6 ++++++
>  3 files changed, 17 insertions(+), 1 deletion(-)
>
> diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
> index 370e22faaeb..b4c0cf02972 100644
> --- a/Documentation/git-gc.txt
> +++ b/Documentation/git-gc.txt
> @@ -69,6 +69,12 @@ be performed as well.
>         the `--max-cruft-size` option of linkgit:git-repack[1] for
>         more.
>
> +--expire-to=<dir>::
> +       When packing unreachable objects into a cruft pack, write a cruft
> +       pack containing pruned objects (if any) to the directory `<dir>`.
> +       See the `--expire-to` option of linkgit:git-repack[1] for
> +       more.
> +
>  --prune=<date>::
>         Prune loose objects older than date (default is 2 weeks ago,
>         overridable by the config variable `gc.pruneExpire`).
> diff --git a/builtin/gc.c b/builtin/gc.c
> index d52735354c9..77904694c9f 100644
> --- a/builtin/gc.c
> +++ b/builtin/gc.c
> @@ -136,6 +136,7 @@ struct gc_config {
>         char *prune_worktrees_expire;
>         char *repack_filter;
>         char *repack_filter_to;
> +       char *repack_expire_to;
>         unsigned long big_pack_threshold;
>         unsigned long max_delta_cache_size;
>  };
> @@ -441,6 +442,8 @@ static void add_repack_all_option(struct gc_config *cfg,
>                 if (cfg->max_cruft_size)
>                         strvec_pushf(&repack, "--max-cruft-size=%lu",
>                                      cfg->max_cruft_size);
> +               if (cfg->repack_expire_to)
> +                       strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);
>         } else {
>                 strvec_push(&repack, "-A");
>                 if (cfg->prune_expire)
> @@ -675,7 +678,6 @@ struct repository *repo UNUSED)
>         const char *prune_expire_sentinel = "sentinel";
>         const char *prune_expire_arg = prune_expire_sentinel;
>         int ret;
> -
>         struct option builtin_gc_options[] = {
>                 OPT__QUIET(&quiet, N_("suppress progress reporting")),
>                 { OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),
> @@ -694,6 +696,8 @@ struct repository *repo UNUSED)
>                            PARSE_OPT_NOCOMPLETE),
>                 OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
>                          N_("repack all other packs except the largest pack")),
> +               OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
> +                          N_("pack prefix to store a pack containing pruned objects")),
>                 OPT_END()
>         };
>
> diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
> index ee074b99b70..d4b0653a9b7 100755
> --- a/t/t6500-gc.sh
> +++ b/t/t6500-gc.sh
> @@ -339,6 +339,12 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
>         test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
>  '
>
> +test_expect_success '--expire-to sets appropriate repack options' '
> +       mkdir expired &&
> +       GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to=./expired/pack &&
> +       test_subcommand $cruft_max_size_opts --expire-to=./expired/pack <trace2.txt
> +'
> +
>  run_and_wait_for_gc () {
>         # We read stdout from gc for the side effect of waiting until the
>         # background gc process exits, closing its fd 9.  Furthermore, the
> --
> gitgitgadget
>

Hi, Jeff King, could you come and help take a look at this patch?
I would be very grateful if you have time!

ZheNing Hu

Re: [PATCH v2 1/2] gc: add `--expire-to` option

[ZheNing Hu]

This patch has been sitting for weeks with no review. Does anyone want
to help take a look?

ZheNing Hu via GitGitGadget <gitgitgadget@gmail.com> 于2024年12月31日周二 10:18写道：
>
> From: ZheNing Hu <adlternative@gmail.com>
>
> This commit extends the functionality of `git gc`
> by adding a new option, `--expire-to=<dir>`. Previously,
> this feature was implemented in `git repack` (see 91badeb),
> allowing users to specify a directory where unreachable and
> expired cruft packs are stored during garbage collection.
> However, users had to run `git repack --cruft --expire-to=<dir>`
> followed by `git prune` to achieve similar results within `git gc`.
>
> By introducing `--expire-to=<dir>` directly into `git gc`,
> we simplify the process for users who wish to manage their
> repository's cleanup more efficiently. This change involves
> passing the `--expire-to=<dir>` parameter through to `git repack`,
> making it easier for users to set up a backup location for cruft
> packs that will be pruned.
>
> Signed-off-by: ZheNing Hu <adlternative@gmail.com>
> ---
>  Documentation/git-gc.txt | 6 ++++++
>  builtin/gc.c             | 6 +++++-
>  t/t6500-gc.sh            | 6 ++++++
>  3 files changed, 17 insertions(+), 1 deletion(-)
>
> diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
> index 370e22faaeb..b4c0cf02972 100644
> --- a/Documentation/git-gc.txt
> +++ b/Documentation/git-gc.txt
> @@ -69,6 +69,12 @@ be performed as well.
>         the `--max-cruft-size` option of linkgit:git-repack[1] for
>         more.
>
> +--expire-to=<dir>::
> +       When packing unreachable objects into a cruft pack, write a cruft
> +       pack containing pruned objects (if any) to the directory `<dir>`.
> +       See the `--expire-to` option of linkgit:git-repack[1] for
> +       more.
> +
>  --prune=<date>::
>         Prune loose objects older than date (default is 2 weeks ago,
>         overridable by the config variable `gc.pruneExpire`).
> diff --git a/builtin/gc.c b/builtin/gc.c
> index d52735354c9..77904694c9f 100644
> --- a/builtin/gc.c
> +++ b/builtin/gc.c
> @@ -136,6 +136,7 @@ struct gc_config {
>         char *prune_worktrees_expire;
>         char *repack_filter;
>         char *repack_filter_to;
> +       char *repack_expire_to;
>         unsigned long big_pack_threshold;
>         unsigned long max_delta_cache_size;
>  };
> @@ -441,6 +442,8 @@ static void add_repack_all_option(struct gc_config *cfg,
>                 if (cfg->max_cruft_size)
>                         strvec_pushf(&repack, "--max-cruft-size=%lu",
>                                      cfg->max_cruft_size);
> +               if (cfg->repack_expire_to)
> +                       strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);
>         } else {
>                 strvec_push(&repack, "-A");
>                 if (cfg->prune_expire)
> @@ -675,7 +678,6 @@ struct repository *repo UNUSED)
>         const char *prune_expire_sentinel = "sentinel";
>         const char *prune_expire_arg = prune_expire_sentinel;
>         int ret;
> -
>         struct option builtin_gc_options[] = {
>                 OPT__QUIET(&quiet, N_("suppress progress reporting")),
>                 { OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),
> @@ -694,6 +696,8 @@ struct repository *repo UNUSED)
>                            PARSE_OPT_NOCOMPLETE),
>                 OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
>                          N_("repack all other packs except the largest pack")),
> +               OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
> +                          N_("pack prefix to store a pack containing pruned objects")),
>                 OPT_END()
>         };
>
> diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
> index ee074b99b70..d4b0653a9b7 100755
> --- a/t/t6500-gc.sh
> +++ b/t/t6500-gc.sh
> @@ -339,6 +339,12 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
>         test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
>  '
>
> +test_expect_success '--expire-to sets appropriate repack options' '
> +       mkdir expired &&
> +       GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to=./expired/pack &&
> +       test_subcommand $cruft_max_size_opts --expire-to=./expired/pack <trace2.txt
> +'
> +
>  run_and_wait_for_gc () {
>         # We read stdout from gc for the side effect of waiting until the
>         # background gc process exits, closing its fd 9.  Furthermore, the
> --
> gitgitgadget
>

[PATCH v2 2/2] fix(gc): make --prune=now compatible with --expire-to

[ZheNing Hu via GitGitGadget]

From: ZheNing Hu <adlternative@gmail.com>

The original `git gc --prune=now` attempted to delete all
unreachable objects. However, after the introduction of
`--cruft` and `--expire-to=<dir>` in git gc, `--prune=now`
can now compress unreachable objects into a cruft pack and
store them in the specified <dir> instead of deleting them
directly. This is beneficial for recovery in case of data
corruption during repository GC. Therefore, update the
handling logic of `--prune=now` in gc so that `-a` parameter
is only passed to the repack command when neither `--cruft`
nor `--expire-to` are used.

Signed-off-by: ZheNing Hu <adlternative@gmail.com>
---
 builtin/gc.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/builtin/gc.c b/builtin/gc.c
index 77904694c9f..8656e1caff0 100644
--- a/builtin/gc.c
+++ b/builtin/gc.c
@@ -433,7 +433,8 @@ static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
 static void add_repack_all_option(struct gc_config *cfg,
 				  struct string_list *keep_pack)
 {
-	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
+	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
+		&& !(cfg->cruft_packs && cfg->repack_expire_to))
 		strvec_push(&repack, "-a");
 	else if (cfg->cruft_packs) {
 		strvec_push(&repack, "--cruft");
-- 
gitgitgadget

Re: [PATCH v2 2/2] fix(gc): make --prune=now compatible with --expire-to

[Jeff King]

On Tue, Dec 31, 2024 at 02:18:33AM +0000, ZheNing Hu via GitGitGadget wrote:

> diff --git a/builtin/gc.c b/builtin/gc.c
> index 77904694c9f..8656e1caff0 100644
> --- a/builtin/gc.c
> +++ b/builtin/gc.c
> @@ -433,7 +433,8 @@ static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
>  static void add_repack_all_option(struct gc_config *cfg,
>  				  struct string_list *keep_pack)
>  {
> -	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
> +	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
> +		&& !(cfg->cruft_packs && cfg->repack_expire_to))
>  		strvec_push(&repack, "-a");

I expected to see a mention of repack_expire_to here, but not
cfg->cruft_packs. These two are AND-ed together so we are only disabling
"repack -a" when both options ("--expire-to" and "--cruft") are passed.
Can we --expire-to without cruft? I.e., what should happen with:

  git gc --expire-to=some-path --prune=now --no-cruft

Looking at the underlying git-repack, it seems that we only respect
--expire-to at all when used with "--cruft", and don't otherwise
consider it. Which is what the manpage says ("Only useful with --cruft
-d").

But if we look at this proposed patch for example:

  https://lore.kernel.org/git/48438876fb42a889110e100a6c42ca84e93aac49.1733011259.git.me@ttaylorr.com/

then it is expanding how --expire-to is used during the pruning step.
OTOH, I think the way your patch 1 is structured means that we'd always
pass --expire-to to git-repack anyway, and I _think_ even with the patch
linked above that "repack -a -d --expire-to=whatever" would do the right
thing.

In which case the problem really is the combination of cruft packs and
expire-to. Just cruft packs by themselves do not need to override using
"-a" for "--prune=now" because we know that any such cruft pack would be
empty.

So I think this logic is correct. Taylor might have more thoughts,
though (and ideas on whether he intends to revisit that earlier patch).

I do think this change should probably be done as part of patch 1,
rather than introducing a buggy state and then fixing it in patch 2.

-Peff

Re: [PATCH v2 2/2] fix(gc): make --prune=now compatible with --expire-to

[ZheNing Hu]

Jeff King <peff@peff.net> 于2025年1月13日周一 17:17写道：
>
> On Tue, Dec 31, 2024 at 02:18:33AM +0000, ZheNing Hu via GitGitGadget wrote:
>
> > diff --git a/builtin/gc.c b/builtin/gc.c
> > index 77904694c9f..8656e1caff0 100644
> > --- a/builtin/gc.c
> > +++ b/builtin/gc.c
> > @@ -433,7 +433,8 @@ static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
> >  static void add_repack_all_option(struct gc_config *cfg,
> >                                 struct string_list *keep_pack)
> >  {
> > -     if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
> > +     if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
> > +             && !(cfg->cruft_packs && cfg->repack_expire_to))
> >               strvec_push(&repack, "-a");
>
> I expected to see a mention of repack_expire_to here, but not
> cfg->cruft_packs. These two are AND-ed together so we are only disabling
> "repack -a" when both options ("--expire-to" and "--cruft") are passed.
> Can we --expire-to without cruft? I.e., what should happen with:
>
>   git gc --expire-to=some-path --prune=now --no-cruft
>
> Looking at the underlying git-repack, it seems that we only respect
> --expire-to at all when used with "--cruft", and don't otherwise
> consider it. Which is what the manpage says ("Only useful with --cruft
> -d").
>

Yes, this is the current state of git-repack. The --expire-to option can
only be used with --cruft, which is why I use cruft_packs && repack_expire_to
as a double safeguard.

When using --no-cruft, the option --expire-to becomes irrelevant.
So leaving `git gc --prune=now` as is at this point: passing -a as a
parameter to repack seems reasonable.

> But if we look at this proposed patch for example:
>
>   https://lore.kernel.org/git/48438876fb42a889110e100a6c42ca84e93aac49.1733011259.git.me@ttaylorr.com/
>
> then it is expanding how --expire-to is used during the pruning step.
> OTOH, I think the way your patch 1 is structured means that we'd always
> pass --expire-to to git-repack anyway, and I _think_ even with the patch
> linked above that "repack -a -d --expire-to=whatever" would do the right
> thing.
>

I've taken a look at the patch, and I believe Taylor's changes are primarily
aimed at extending the --expire-to functionality within the --cruft feature,
rather than expecting --expire-to to be used on its own.

> In which case the problem really is the combination of cruft packs and
> expire-to. Just cruft packs by themselves do not need to override using
> "-a" for "--prune=now" because we know that any such cruft pack would be
> empty.
>
> So I think this logic is correct. Taylor might have more thoughts,
> though (and ideas on whether he intends to revisit that earlier patch).
>
> I do think this change should probably be done as part of patch 1,
> rather than introducing a buggy state and then fixing it in patch 2.
>

Yes, I agree with that, and perhaps a single patch will suffice.

> -Peff

- ZheNing Hu

[PATCH v3] gc: add `--expire-to` option

[ZheNing Hu via GitGitGadget]

From: ZheNing Hu <adlternative@gmail.com>

This commit extends the functionality of `git gc`
by adding a new option, `--expire-to=<dir>`. Previously,
this feature was implemented in `git repack` (see 91badeb),
allowing users to specify a directory where unreachable and
expired cruft packs are stored during garbage collection.
However, users had to run `git repack --cruft --expire-to=<dir>`
followed by `git prune` to achieve similar results within `git gc`.

By introducing `--expire-to=<dir>` directly into `git gc`,
we simplify the process for users who wish to manage their
repository's cleanup more efficiently. This change involves
passing the `--expire-to=<dir>` parameter through to `git repack`,
making it easier for users to set up a backup location for cruft
packs that will be pruned.

Note: When git-gc is used with both `--cruft` and `--expire-to`,
it does not pass `-a` to git-repack to delete all unreachable
objects as `git gc --prune=now` originally did. Instead, it
generates a cruft pack in the directory specified by expire-to.

Signed-off-by: ZheNing Hu <adlternative@gmail.com>
---
    gc: add --expire-to option
    
    I want to perform a "safe" garbage collection for the Git repository on
    the server, which avoids data corruption issues caused by concurrent
    pushes during git GC. To achieve this, I currently need to use git
    repack --cruft --expire-to=<dir> and git prune in combination. However,
    it would be simpler if we could directly use --expire-to=<dir> with the
    git-gc command.
    
    v1: add --expire-to option to gc v1 -> v2: fix git gc --prune=now with
    --expire-to v2 -> v3: squash two patch into one patch

Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-1843%2Fadlternative%2Fzh%2Fgc-expire-to-v3
Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-1843/adlternative/zh/gc-expire-to-v3
Pull-Request: https://github.com/gitgitgadget/git/pull/1843

Range-diff vs v2:

 1:  14e94bf04e5 ! 1:  0842ec34948 gc: add `--expire-to` option
     @@ Commit message
          making it easier for users to set up a backup location for cruft
          packs that will be pruned.
      
     +    Note: When git-gc is used with both `--cruft` and `--expire-to`,
     +    it does not pass `-a` to git-repack to delete all unreachable
     +    objects as `git gc --prune=now` originally did. Instead, it
     +    generates a cruft pack in the directory specified by expire-to.
     +
          Signed-off-by: ZheNing Hu <adlternative@gmail.com>
      
       ## Documentation/git-gc.txt ##
     @@ builtin/gc.c: struct gc_config {
       	unsigned long big_pack_threshold;
       	unsigned long max_delta_cache_size;
       };
     +@@ builtin/gc.c: static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
     + static void add_repack_all_option(struct gc_config *cfg,
     + 				  struct string_list *keep_pack)
     + {
     +-	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
     ++	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
     ++		&& !(cfg->cruft_packs && cfg->repack_expire_to))
     + 		strvec_push(&repack, "-a");
     + 	else if (cfg->cruft_packs) {
     + 		strvec_push(&repack, "--cruft");
      @@ builtin/gc.c: static void add_repack_all_option(struct gc_config *cfg,
       		if (cfg->max_cruft_size)
       			strvec_pushf(&repack, "--max-cruft-size=%lu",
 2:  579757957d2 < -:  ----------- fix(gc): make --prune=now compatible with --expire-to


 Documentation/git-gc.txt | 6 ++++++
 builtin/gc.c             | 9 +++++++--
 t/t6500-gc.sh            | 6 ++++++
 3 files changed, 19 insertions(+), 2 deletions(-)

diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
index 370e22faaeb..b4c0cf02972 100644
--- a/Documentation/git-gc.txt
+++ b/Documentation/git-gc.txt
@@ -69,6 +69,12 @@ be performed as well.
 	the `--max-cruft-size` option of linkgit:git-repack[1] for
 	more.
 
+--expire-to=<dir>::
+	When packing unreachable objects into a cruft pack, write a cruft
+	pack containing pruned objects (if any) to the directory `<dir>`.
+	See the `--expire-to` option of linkgit:git-repack[1] for
+	more.
+
 --prune=<date>::
 	Prune loose objects older than date (default is 2 weeks ago,
 	overridable by the config variable `gc.pruneExpire`).
diff --git a/builtin/gc.c b/builtin/gc.c
index d52735354c9..8656e1caff0 100644
--- a/builtin/gc.c
+++ b/builtin/gc.c
@@ -136,6 +136,7 @@ struct gc_config {
 	char *prune_worktrees_expire;
 	char *repack_filter;
 	char *repack_filter_to;
+	char *repack_expire_to;
 	unsigned long big_pack_threshold;
 	unsigned long max_delta_cache_size;
 };
@@ -432,7 +433,8 @@ static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
 static void add_repack_all_option(struct gc_config *cfg,
 				  struct string_list *keep_pack)
 {
-	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
+	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
+		&& !(cfg->cruft_packs && cfg->repack_expire_to))
 		strvec_push(&repack, "-a");
 	else if (cfg->cruft_packs) {
 		strvec_push(&repack, "--cruft");
@@ -441,6 +443,8 @@ static void add_repack_all_option(struct gc_config *cfg,
 		if (cfg->max_cruft_size)
 			strvec_pushf(&repack, "--max-cruft-size=%lu",
 				     cfg->max_cruft_size);
+		if (cfg->repack_expire_to)
+			strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);
 	} else {
 		strvec_push(&repack, "-A");
 		if (cfg->prune_expire)
@@ -675,7 +679,6 @@ struct repository *repo UNUSED)
 	const char *prune_expire_sentinel = "sentinel";
 	const char *prune_expire_arg = prune_expire_sentinel;
 	int ret;
-
 	struct option builtin_gc_options[] = {
 		OPT__QUIET(&quiet, N_("suppress progress reporting")),
 		{ OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),
@@ -694,6 +697,8 @@ struct repository *repo UNUSED)
 			   PARSE_OPT_NOCOMPLETE),
 		OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
 			 N_("repack all other packs except the largest pack")),
+		OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
+			   N_("pack prefix to store a pack containing pruned objects")),
 		OPT_END()
 	};
 
diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
index ee074b99b70..d4b0653a9b7 100755
--- a/t/t6500-gc.sh
+++ b/t/t6500-gc.sh
@@ -339,6 +339,12 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
 	test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
 '
 
+test_expect_success '--expire-to sets appropriate repack options' '
+	mkdir expired &&
+	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to=./expired/pack &&
+	test_subcommand $cruft_max_size_opts --expire-to=./expired/pack <trace2.txt
+'
+
 run_and_wait_for_gc () {
 	# We read stdout from gc for the side effect of waiting until the
 	# background gc process exits, closing its fd 9.  Furthermore, the

base-commit: 92999a42db1c5f43f330e4f2bca4026b5b81576f
-- 
gitgitgadget

Re: [PATCH v3] gc: add `--expire-to` option

[Junio C Hamano]

"ZheNing Hu via GitGitGadget" <gitgitgadget@gmail.com> writes:

> From: ZheNing Hu <adlternative@gmail.com>
>
> This commit extends the functionality of `git gc`
> by adding a new option, `--expire-to=<dir>`. Previously,
> this feature was implemented in `git repack` (see 91badeb),
> allowing users to specify a directory where unreachable and
> expired cruft packs are stored during garbage collection.
> However, users had to run `git repack --cruft --expire-to=<dir>`
> followed by `git prune` to achieve similar results within `git gc`.
>
> By introducing `--expire-to=<dir>` directly into `git gc`,
> we simplify the process for users who wish to manage their
> repository's cleanup more efficiently. This change involves
> passing the `--expire-to=<dir>` parameter through to `git repack`,
> making it easier for users to set up a backup location for cruft
> packs that will be pruned.

Today I do not have enough time to do my usual commit log message
critique.  Please use "git show -s --format=reference" when
referring to an earlier commit.

> Note: When git-gc is used with both `--cruft` and `--expire-to`,
> it does not pass `-a` to git-repack to delete all unreachable
> objects as `git gc --prune=now` originally did. Instead, it
> generates a cruft pack in the directory specified by expire-to.

Is this less important than "we added --expire-to to gc that is
passed down to underlying repack" in the previous paragraph?

Not removing the unreachables too early with "repack -a" is an
essential part of the design of this new feature to allow us not to
lose the cruft objects, so I was a bit surprised that this was
described as a "Note:".

> diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
> index 370e22faaeb..b4c0cf02972 100644
> --- a/Documentation/git-gc.txt
> +++ b/Documentation/git-gc.txt
> @@ -69,6 +69,12 @@ be performed as well.
>  	the `--max-cruft-size` option of linkgit:git-repack[1] for
>  	more.
>  
> +--expire-to=<dir>::
> +	When packing unreachable objects into a cruft pack, write a cruft
> +	pack containing pruned objects (if any) to the directory `<dir>`.
> +	See the `--expire-to` option of linkgit:git-repack[1] for
> +	more.

Does "When packing unreachable objects into a cruft pack" mean that
this option is only meaningful with "--cruft"?  As "--cruft" is on
by default, is it an error to pass "--no-cruft" when you use this
option?

"for more" -> "for more information" or something?

> diff --git a/builtin/gc.c b/builtin/gc.c
> index d52735354c9..8656e1caff0 100644
> --- a/builtin/gc.c
> +++ b/builtin/gc.c
> @@ -136,6 +136,7 @@ struct gc_config {
>  	char *prune_worktrees_expire;
>  	char *repack_filter;
>  	char *repack_filter_to;
> +	char *repack_expire_to;
>  	unsigned long big_pack_threshold;
>  	unsigned long max_delta_cache_size;
>  };
> @@ -432,7 +433,8 @@ static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
>  static void add_repack_all_option(struct gc_config *cfg,
>  				  struct string_list *keep_pack)
>  {
> -	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
> +	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
> +		&& !(cfg->cruft_packs && cfg->repack_expire_to))
>  		strvec_push(&repack, "-a");

Hmph.  When "--expire-to=<there>" is given, we are dropping these
unreachable objects right away, but we said "--no-cruft", then we
say "repack -a".  If we have both "--cruft" and "--expire-to=<there>",
then ...

>  	else if (cfg->cruft_packs) {
>  		strvec_push(&repack, "--cruft");
> @@ -441,6 +443,8 @@ static void add_repack_all_option(struct gc_config *cfg,
>  		if (cfg->max_cruft_size)
>  			strvec_pushf(&repack, "--max-cruft-size=%lu",
>  				     cfg->max_cruft_size);
> +		if (cfg->repack_expire_to)
> +			strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);

... we do the usual "repack --cruft --expire-to=<there>" in the next
block.

> @@ -675,7 +679,6 @@ struct repository *repo UNUSED)
>  	const char *prune_expire_sentinel = "sentinel";
>  	const char *prune_expire_arg = prune_expire_sentinel;
>  	int ret;
> -
>  	struct option builtin_gc_options[] = {
>  		OPT__QUIET(&quiet, N_("suppress progress reporting")),
>  		{ OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),

OK.

> @@ -694,6 +697,8 @@ struct repository *repo UNUSED)
>  			   PARSE_OPT_NOCOMPLETE),
>  		OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
>  			 N_("repack all other packs except the largest pack")),
> +		OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
> +			   N_("pack prefix to store a pack containing pruned objects")),
>  		OPT_END()
>  	};

OK.

> diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
> index ee074b99b70..d4b0653a9b7 100755
> --- a/t/t6500-gc.sh
> +++ b/t/t6500-gc.sh
> @@ -339,6 +339,12 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
>  	test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
>  '
>  
> +test_expect_success '--expire-to sets appropriate repack options' '
> +	mkdir expired &&
> +	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to=./expired/pack &&
> +	test_subcommand $cruft_max_size_opts --expire-to=./expired/pack <trace2.txt
> +'

As "--cruft" is on by default, the command line does not have to
have it, but being explicit is good.

Should we also see what happens when "--no-cruft" is given?

Thanks.

Re: [PATCH v3] gc: add `--expire-to` option

[ZheNing Hu]

Junio C Hamano <gitster@pobox.com> 于2025年1月17日周五 02:23写道：
>
> "ZheNing Hu via GitGitGadget" <gitgitgadget@gmail.com> writes:
>
> > From: ZheNing Hu <adlternative@gmail.com>
> >
> > This commit extends the functionality of `git gc`
> > by adding a new option, `--expire-to=<dir>`. Previously,
> > this feature was implemented in `git repack` (see 91badeb),
> > allowing users to specify a directory where unreachable and
> > expired cruft packs are stored during garbage collection.
> > However, users had to run `git repack --cruft --expire-to=<dir>`
> > followed by `git prune` to achieve similar results within `git gc`.
> >
> > By introducing `--expire-to=<dir>` directly into `git gc`,
> > we simplify the process for users who wish to manage their
> > repository's cleanup more efficiently. This change involves
> > passing the `--expire-to=<dir>` parameter through to `git repack`,
> > making it easier for users to set up a backup location for cruft
> > packs that will be pruned.
>
> Today I do not have enough time to do my usual commit log message
> critique.  Please use "git show -s --format=reference" when
> referring to an earlier commit.
>

Okay, I will change to using this format.

> > Note: When git-gc is used with both `--cruft` and `--expire-to`,
> > it does not pass `-a` to git-repack to delete all unreachable
> > objects as `git gc --prune=now` originally did. Instead, it
> > generates a cruft pack in the directory specified by expire-to.
>
> Is this less important than "we added --expire-to to gc that is
> passed down to underlying repack" in the previous paragraph?
>

I had thought that adding --expire-to to gc was key in this patch,
but the change to the implementation of --prune=now should
indeed be mentioned more.

> Not removing the unreachables too early with "repack -a" is an
> essential part of the design of this new feature to allow us not to
> lose the cruft objects, so I was a bit surprised that this was
> described as a "Note:".
>

You're right. This section shouldn't use a note; it should provide
a more detailed explanation instead.

> > diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
> > index 370e22faaeb..b4c0cf02972 100644
> > --- a/Documentation/git-gc.txt
> > +++ b/Documentation/git-gc.txt
> > @@ -69,6 +69,12 @@ be performed as well.
> >       the `--max-cruft-size` option of linkgit:git-repack[1] for
> >       more.
> >
> > +--expire-to=<dir>::
> > +     When packing unreachable objects into a cruft pack, write a cruft
> > +     pack containing pruned objects (if any) to the directory `<dir>`.
> > +     See the `--expire-to` option of linkgit:git-repack[1] for
> > +     more.
>
> Does "When packing unreachable objects into a cruft pack" mean that
> this option is only meaningful with "--cruft"?  As "--cruft" is on
> by default, is it an error to pass "--no-cruft" when you use this
> option?
>

It (--expired-to) can currently only be used together with --cruft.
Using --no-cruft together with --expire-to will not result in an error,
but --expired-to will not take effect either.

I should mention in the document that --expire-to and --cruft
need to be used together, otherwise --expire-to will not
have any effect.

> "for more" -> "for more information" or something?
>

OK,  "for more information".

> > diff --git a/builtin/gc.c b/builtin/gc.c
> > index d52735354c9..8656e1caff0 100644
> > --- a/builtin/gc.c
> > +++ b/builtin/gc.c
> > @@ -136,6 +136,7 @@ struct gc_config {
> >       char *prune_worktrees_expire;
> >       char *repack_filter;
> >       char *repack_filter_to;
> > +     char *repack_expire_to;
> >       unsigned long big_pack_threshold;
> >       unsigned long max_delta_cache_size;
> >  };
> > @@ -432,7 +433,8 @@ static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
> >  static void add_repack_all_option(struct gc_config *cfg,
> >                                 struct string_list *keep_pack)
> >  {
> > -     if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
> > +     if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
> > +             && !(cfg->cruft_packs && cfg->repack_expire_to))
> >               strvec_push(&repack, "-a");
>
> Hmph.  When "--expire-to=<there>" is given, we are dropping these
> unreachable objects right away, but we said "--no-cruft", then we
> say "repack -a".  If we have both "--cruft" and "--expire-to=<there>",
> then ...
>
> >       else if (cfg->cruft_packs) {
> >               strvec_push(&repack, "--cruft");
> > @@ -441,6 +443,8 @@ static void add_repack_all_option(struct gc_config *cfg,
> >               if (cfg->max_cruft_size)
> >                       strvec_pushf(&repack, "--max-cruft-size=%lu",
> >                                    cfg->max_cruft_size);
> > +             if (cfg->repack_expire_to)
> > +                     strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);
>
> ... we do the usual "repack --cruft --expire-to=<there>" in the next
> block.
>
> > @@ -675,7 +679,6 @@ struct repository *repo UNUSED)
> >       const char *prune_expire_sentinel = "sentinel";
> >       const char *prune_expire_arg = prune_expire_sentinel;
> >       int ret;
> > -
> >       struct option builtin_gc_options[] = {
> >               OPT__QUIET(&quiet, N_("suppress progress reporting")),
> >               { OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),
>
> OK.
>
> > @@ -694,6 +697,8 @@ struct repository *repo UNUSED)
> >                          PARSE_OPT_NOCOMPLETE),
> >               OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
> >                        N_("repack all other packs except the largest pack")),
> > +             OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
> > +                        N_("pack prefix to store a pack containing pruned objects")),
> >               OPT_END()
> >       };
>
> OK.
>
> > diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
> > index ee074b99b70..d4b0653a9b7 100755
> > --- a/t/t6500-gc.sh
> > +++ b/t/t6500-gc.sh
> > @@ -339,6 +339,12 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
> >       test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
> >  '
> >
> > +test_expect_success '--expire-to sets appropriate repack options' '
> > +     mkdir expired &&
> > +     GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to=./expired/pack &&
> > +     test_subcommand $cruft_max_size_opts --expire-to=./expired/pack <trace2.txt
> > +'
>
> As "--cruft" is on by default, the command line does not have to
> have it, but being explicit is good.
>
> Should we also see what happens when "--no-cruft" is given?
>

 --expire-to with --no-cruft  will still run repack -a, I will add
corresponding tests.

> Thanks.

Thanks.

[PATCH v4] gc: add `--expire-to` option

[ZheNing Hu via GitGitGadget]

From: ZheNing Hu <adlternative@gmail.com>

This commit extends the functionality of `git gc`
by adding a new option, `--expire-to=<dir>`. Previously,
this feature was implemented in 91badeba32 (builtin/repack.c:
implement `--expire-to` for storing pruned objects, 2022-10-24),
which allowing users to specify a directory where unreachable
and expired cruft packs are stored during garbage collection.
However, users had to run `git repack --cruft --expire-to=<dir>`
followed by `git prune` to achieve similar results within `git gc`.

By introducing `--expire-to=<dir>` directly into `git gc`,
we simplify the process for users who wish to manage their
repository's cleanup more efficiently. This change involves
passing the `--expire-to=<dir>` parameter through to `git repack`,
making it easier for users to set up a backup location for cruft
packs that will be pruned.

Due to the original `git gc --prune=now` deleting all unreachable
objects by passing the `-a` parameter to git repack. With the
addition of the `--cruft` and `--expire-to` options, it is necessary
to modify this default behavior: instead of deleting these
unreachable objects, they should be merged into a cruft pack and
collected in a specified directory. Therefore, we do not pass `-a`
to the repack command but instead pass `--cruft`, `--expire-to`,
and `--cruft-expiration=now` to repack.

Signed-off-by: ZheNing Hu <adlternative@gmail.com>
---
    gc: add --expire-to option
    
    I want to perform a "safe" garbage collection for the Git repository on
    the server, which avoids data corruption issues caused by concurrent
    pushes during git GC. To achieve this, I currently need to use git
    repack --cruft --expire-to=<dir> and git prune in combination. However,
    it would be simpler if we could directly use --expire-to=<dir> with the
    git-gc command.
    
    v1: add --expire-to option to gc v1 -> v2: fix git gc --prune=now with
    --expire-to v2 -> v3: squash two patch into one patch v3 -> v4: modify
    docs, commit message, and give more tests

Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-1843%2Fadlternative%2Fzh%2Fgc-expire-to-v4
Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-1843/adlternative/zh/gc-expire-to-v4
Pull-Request: https://github.com/gitgitgadget/git/pull/1843

Range-diff vs v3:

 1:  0842ec34948 ! 1:  6946ccde275 gc: add `--expire-to` option
     @@ Commit message
      
          This commit extends the functionality of `git gc`
          by adding a new option, `--expire-to=<dir>`. Previously,
     -    this feature was implemented in `git repack` (see 91badeb),
     -    allowing users to specify a directory where unreachable and
     -    expired cruft packs are stored during garbage collection.
     +    this feature was implemented in 91badeba32 (builtin/repack.c:
     +    implement `--expire-to` for storing pruned objects, 2022-10-24),
     +    which allowing users to specify a directory where unreachable
     +    and expired cruft packs are stored during garbage collection.
          However, users had to run `git repack --cruft --expire-to=<dir>`
          followed by `git prune` to achieve similar results within `git gc`.
      
     @@ Commit message
          making it easier for users to set up a backup location for cruft
          packs that will be pruned.
      
     -    Note: When git-gc is used with both `--cruft` and `--expire-to`,
     -    it does not pass `-a` to git-repack to delete all unreachable
     -    objects as `git gc --prune=now` originally did. Instead, it
     -    generates a cruft pack in the directory specified by expire-to.
     +    Due to the original `git gc --prune=now` deleting all unreachable
     +    objects by passing the `-a` parameter to git repack. With the
     +    addition of the `--cruft` and `--expire-to` options, it is necessary
     +    to modify this default behavior: instead of deleting these
     +    unreachable objects, they should be merged into a cruft pack and
     +    collected in a specified directory. Therefore, we do not pass `-a`
     +    to the repack command but instead pass `--cruft`, `--expire-to`,
     +    and `--cruft-expiration=now` to repack.
      
          Signed-off-by: ZheNing Hu <adlternative@gmail.com>
      
     @@ Documentation/git-gc.txt: be performed as well.
      +--expire-to=<dir>::
      +	When packing unreachable objects into a cruft pack, write a cruft
      +	pack containing pruned objects (if any) to the directory `<dir>`.
     ++	This option only has an effect when used together with `--cruft`.
      +	See the `--expire-to` option of linkgit:git-repack[1] for
     -+	more.
     ++	more information.
      +
       --prune=<date>::
       	Prune loose objects older than date (default is 2 weeks ago,
     @@ t/t6500-gc.sh: test_expect_success 'gc.maxCruftSize sets appropriate repack opti
       	test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
       '
       
     -+test_expect_success '--expire-to sets appropriate repack options' '
     ++test_expect_success '--expire-to sets repack --expire-to' '
     ++	rm -rf expired &&
      +	mkdir expired &&
     -+	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to=./expired/pack &&
     -+	test_subcommand $cruft_max_size_opts --expire-to=./expired/pack <trace2.txt
     ++	expire_to="$(pwd)/expired/pack" &&
     ++	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to="$expire_to" &&
     ++	test_subcommand $cruft_max_size_opts --expire-to="$expire_to" <trace2.txt
     ++'
     ++
     ++test_expect_success '--expire-to with --prune=now sets repack --expire-to' '
     ++	rm -rf expired &&
     ++	mkdir expired &&
     ++	expire_to="$(pwd)/expired/pack" &&
     ++	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --prune=now --expire-to="$expire_to" &&
     ++	test_subcommand git repack -d -l --cruft --cruft-expiration=now --expire-to="$expire_to" <trace2.txt
     ++'
     ++
     ++
     ++test_expect_success '--expire-to with --no-cruft sets repack -A' '
     ++	rm -rf expired &&
     ++	mkdir expired &&
     ++	expire_to="$(pwd)/expired/pack" &&
     ++	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --no-cruft --expire-to="$expire_to" &&
     ++	test_subcommand git repack -d -l -A --unpack-unreachable=2.weeks.ago <trace2.txt
     ++'
     ++
     ++test_expect_success '--expire-to with --no-cruft sets repack -a' '
     ++	rm -rf expired &&
     ++	mkdir expired &&
     ++	expire_to="$(pwd)/expired/pack" &&
     ++	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --no-cruft --prune=now --expire-to="$expire_to" &&
     ++	test_subcommand git repack -d -l -a <trace2.txt
      +'
      +
       run_and_wait_for_gc () {


 Documentation/git-gc.txt |  7 +++++++
 builtin/gc.c             |  9 +++++++--
 t/t6500-gc.sh            | 33 +++++++++++++++++++++++++++++++++
 3 files changed, 47 insertions(+), 2 deletions(-)

diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
index 370e22faaeb..0eac8e85f08 100644
--- a/Documentation/git-gc.txt
+++ b/Documentation/git-gc.txt
@@ -69,6 +69,13 @@ be performed as well.
 	the `--max-cruft-size` option of linkgit:git-repack[1] for
 	more.
 
+--expire-to=<dir>::
+	When packing unreachable objects into a cruft pack, write a cruft
+	pack containing pruned objects (if any) to the directory `<dir>`.
+	This option only has an effect when used together with `--cruft`.
+	See the `--expire-to` option of linkgit:git-repack[1] for
+	more information.
+
 --prune=<date>::
 	Prune loose objects older than date (default is 2 weeks ago,
 	overridable by the config variable `gc.pruneExpire`).
diff --git a/builtin/gc.c b/builtin/gc.c
index d52735354c9..8656e1caff0 100644
--- a/builtin/gc.c
+++ b/builtin/gc.c
@@ -136,6 +136,7 @@ struct gc_config {
 	char *prune_worktrees_expire;
 	char *repack_filter;
 	char *repack_filter_to;
+	char *repack_expire_to;
 	unsigned long big_pack_threshold;
 	unsigned long max_delta_cache_size;
 };
@@ -432,7 +433,8 @@ static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
 static void add_repack_all_option(struct gc_config *cfg,
 				  struct string_list *keep_pack)
 {
-	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
+	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
+		&& !(cfg->cruft_packs && cfg->repack_expire_to))
 		strvec_push(&repack, "-a");
 	else if (cfg->cruft_packs) {
 		strvec_push(&repack, "--cruft");
@@ -441,6 +443,8 @@ static void add_repack_all_option(struct gc_config *cfg,
 		if (cfg->max_cruft_size)
 			strvec_pushf(&repack, "--max-cruft-size=%lu",
 				     cfg->max_cruft_size);
+		if (cfg->repack_expire_to)
+			strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);
 	} else {
 		strvec_push(&repack, "-A");
 		if (cfg->prune_expire)
@@ -675,7 +679,6 @@ struct repository *repo UNUSED)
 	const char *prune_expire_sentinel = "sentinel";
 	const char *prune_expire_arg = prune_expire_sentinel;
 	int ret;
-
 	struct option builtin_gc_options[] = {
 		OPT__QUIET(&quiet, N_("suppress progress reporting")),
 		{ OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),
@@ -694,6 +697,8 @@ struct repository *repo UNUSED)
 			   PARSE_OPT_NOCOMPLETE),
 		OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
 			 N_("repack all other packs except the largest pack")),
+		OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
+			   N_("pack prefix to store a pack containing pruned objects")),
 		OPT_END()
 	};
 
diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
index ee074b99b70..74f7bd09046 100755
--- a/t/t6500-gc.sh
+++ b/t/t6500-gc.sh
@@ -339,6 +339,39 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
 	test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
 '
 
+test_expect_success '--expire-to sets repack --expire-to' '
+	rm -rf expired &&
+	mkdir expired &&
+	expire_to="$(pwd)/expired/pack" &&
+	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to="$expire_to" &&
+	test_subcommand $cruft_max_size_opts --expire-to="$expire_to" <trace2.txt
+'
+
+test_expect_success '--expire-to with --prune=now sets repack --expire-to' '
+	rm -rf expired &&
+	mkdir expired &&
+	expire_to="$(pwd)/expired/pack" &&
+	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --prune=now --expire-to="$expire_to" &&
+	test_subcommand git repack -d -l --cruft --cruft-expiration=now --expire-to="$expire_to" <trace2.txt
+'
+
+
+test_expect_success '--expire-to with --no-cruft sets repack -A' '
+	rm -rf expired &&
+	mkdir expired &&
+	expire_to="$(pwd)/expired/pack" &&
+	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --no-cruft --expire-to="$expire_to" &&
+	test_subcommand git repack -d -l -A --unpack-unreachable=2.weeks.ago <trace2.txt
+'
+
+test_expect_success '--expire-to with --no-cruft sets repack -a' '
+	rm -rf expired &&
+	mkdir expired &&
+	expire_to="$(pwd)/expired/pack" &&
+	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --no-cruft --prune=now --expire-to="$expire_to" &&
+	test_subcommand git repack -d -l -a <trace2.txt
+'
+
 run_and_wait_for_gc () {
 	# We read stdout from gc for the side effect of waiting until the
 	# background gc process exits, closing its fd 9.  Furthermore, the

base-commit: 92999a42db1c5f43f330e4f2bca4026b5b81576f
-- 
gitgitgadget

Re: [PATCH v4] gc: add `--expire-to` option

[Junio C Hamano]

"ZheNing Hu via GitGitGadget" <gitgitgadget@gmail.com> writes:

> From: ZheNing Hu <adlternative@gmail.com>
>
> This commit extends the functionality of `git gc`
> by adding a new option, `--expire-to=<dir>`. Previously,
> this feature was implemented in 91badeba32 (builtin/repack.c:
> implement `--expire-to` for storing pruned objects, 2022-10-24),
> which allowing users to specify a directory where unreachable
> and expired cruft packs are stored during garbage collection.
> However, users had to run `git repack --cruft --expire-to=<dir>`
> followed by `git prune` to achieve similar results within `git gc`.
>
> By introducing `--expire-to=<dir>` directly into `git gc`,
> we simplify the process for users who wish to manage their
> repository's cleanup more efficiently. This change involves
> passing the `--expire-to=<dir>` parameter through to `git repack`,
> making it easier for users to set up a backup location for cruft
> packs that will be pruned.
>
> Due to the original `git gc --prune=now` deleting all unreachable
> objects by passing the `-a` parameter to git repack. With the
> addition of the `--cruft` and `--expire-to` options, it is necessary
> to modify this default behavior: instead of deleting these
> unreachable objects, they should be merged into a cruft pack and
> collected in a specified directory. Therefore, we do not pass `-a`
> to the repack command but instead pass `--cruft`, `--expire-to`,
> and `--cruft-expiration=now` to repack.
>
> Signed-off-by: ZheNing Hu <adlternative@gmail.com>
> ---

This hasn't seen any reaction for a while.

Does anybody have further comments?  Otherwise let's mark it for
'next'.

Thanks.

>  Documentation/git-gc.txt |  7 +++++++
>  builtin/gc.c             |  9 +++++++--
>  t/t6500-gc.sh            | 33 +++++++++++++++++++++++++++++++++
>  3 files changed, 47 insertions(+), 2 deletions(-)
>
> diff --git a/Documentation/git-gc.txt b/Documentation/git-gc.txt
> index 370e22faaeb..0eac8e85f08 100644
> --- a/Documentation/git-gc.txt
> +++ b/Documentation/git-gc.txt
> @@ -69,6 +69,13 @@ be performed as well.
>  	the `--max-cruft-size` option of linkgit:git-repack[1] for
>  	more.
>  
> +--expire-to=<dir>::
> +	When packing unreachable objects into a cruft pack, write a cruft
> +	pack containing pruned objects (if any) to the directory `<dir>`.
> +	This option only has an effect when used together with `--cruft`.
> +	See the `--expire-to` option of linkgit:git-repack[1] for
> +	more information.
> +
>  --prune=<date>::
>  	Prune loose objects older than date (default is 2 weeks ago,
>  	overridable by the config variable `gc.pruneExpire`).
> diff --git a/builtin/gc.c b/builtin/gc.c
> index d52735354c9..8656e1caff0 100644
> --- a/builtin/gc.c
> +++ b/builtin/gc.c
> @@ -136,6 +136,7 @@ struct gc_config {
>  	char *prune_worktrees_expire;
>  	char *repack_filter;
>  	char *repack_filter_to;
> +	char *repack_expire_to;
>  	unsigned long big_pack_threshold;
>  	unsigned long max_delta_cache_size;
>  };
> @@ -432,7 +433,8 @@ static int keep_one_pack(struct string_list_item *item, void *data UNUSED)
>  static void add_repack_all_option(struct gc_config *cfg,
>  				  struct string_list *keep_pack)
>  {
> -	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now"))
> +	if (cfg->prune_expire && !strcmp(cfg->prune_expire, "now")
> +		&& !(cfg->cruft_packs && cfg->repack_expire_to))
>  		strvec_push(&repack, "-a");
>  	else if (cfg->cruft_packs) {
>  		strvec_push(&repack, "--cruft");
> @@ -441,6 +443,8 @@ static void add_repack_all_option(struct gc_config *cfg,
>  		if (cfg->max_cruft_size)
>  			strvec_pushf(&repack, "--max-cruft-size=%lu",
>  				     cfg->max_cruft_size);
> +		if (cfg->repack_expire_to)
> +			strvec_pushf(&repack, "--expire-to=%s", cfg->repack_expire_to);
>  	} else {
>  		strvec_push(&repack, "-A");
>  		if (cfg->prune_expire)
> @@ -675,7 +679,6 @@ struct repository *repo UNUSED)
>  	const char *prune_expire_sentinel = "sentinel";
>  	const char *prune_expire_arg = prune_expire_sentinel;
>  	int ret;
> -
>  	struct option builtin_gc_options[] = {
>  		OPT__QUIET(&quiet, N_("suppress progress reporting")),
>  		{ OPTION_STRING, 0, "prune", &prune_expire_arg, N_("date"),
> @@ -694,6 +697,8 @@ struct repository *repo UNUSED)
>  			   PARSE_OPT_NOCOMPLETE),
>  		OPT_BOOL(0, "keep-largest-pack", &keep_largest_pack,
>  			 N_("repack all other packs except the largest pack")),
> +		OPT_STRING(0, "expire-to", &cfg.repack_expire_to, N_("dir"),
> +			   N_("pack prefix to store a pack containing pruned objects")),
>  		OPT_END()
>  	};
>  
> diff --git a/t/t6500-gc.sh b/t/t6500-gc.sh
> index ee074b99b70..74f7bd09046 100755
> --- a/t/t6500-gc.sh
> +++ b/t/t6500-gc.sh
> @@ -339,6 +339,39 @@ test_expect_success 'gc.maxCruftSize sets appropriate repack options' '
>  	test_subcommand $cruft_max_size_opts --max-cruft-size=3145728 <trace2.txt
>  '
>  
> +test_expect_success '--expire-to sets repack --expire-to' '
> +	rm -rf expired &&
> +	mkdir expired &&
> +	expire_to="$(pwd)/expired/pack" &&
> +	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --expire-to="$expire_to" &&
> +	test_subcommand $cruft_max_size_opts --expire-to="$expire_to" <trace2.txt
> +'
> +
> +test_expect_success '--expire-to with --prune=now sets repack --expire-to' '
> +	rm -rf expired &&
> +	mkdir expired &&
> +	expire_to="$(pwd)/expired/pack" &&
> +	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --cruft --prune=now --expire-to="$expire_to" &&
> +	test_subcommand git repack -d -l --cruft --cruft-expiration=now --expire-to="$expire_to" <trace2.txt
> +'
> +
> +
> +test_expect_success '--expire-to with --no-cruft sets repack -A' '
> +	rm -rf expired &&
> +	mkdir expired &&
> +	expire_to="$(pwd)/expired/pack" &&
> +	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --no-cruft --expire-to="$expire_to" &&
> +	test_subcommand git repack -d -l -A --unpack-unreachable=2.weeks.ago <trace2.txt
> +'
> +
> +test_expect_success '--expire-to with --no-cruft sets repack -a' '
> +	rm -rf expired &&
> +	mkdir expired &&
> +	expire_to="$(pwd)/expired/pack" &&
> +	GIT_TRACE2_EVENT=$(pwd)/trace2.txt git -C cruft--max-size gc --no-cruft --prune=now --expire-to="$expire_to" &&
> +	test_subcommand git repack -d -l -a <trace2.txt
> +'
> +
>  run_and_wait_for_gc () {
>  	# We read stdout from gc for the side effect of waiting until the
>  	# background gc process exits, closing its fd 9.  Furthermore, the
>
> base-commit: 92999a42db1c5f43f330e4f2bca4026b5b81576f