[PATCH] repack: add config to skip updating server info

[PATCH] repack: add config to skip updating server info

[Patrick Steinhardt]

[-- Attachment #1: Type: text/plain, Size: 4423 bytes --]

By default, git-repack(1) will update server info that is required by
the dumb HTTP transport. This can be skipped by passing the `-n` flag,
but what we're noticably missing is a config option to permanently
disable updating this information.

Add a new option "repack.updateServerInfo" which can be used to disable
the logic. Most hosting providers have turned off the dumb HTTP protocol
anyway, and on the client-side it woudln't typically be useful either.
Giving a persistent way to disable this feature thus makes quite some
sense to avoid wasting compute cycles and storage.

Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 Documentation/config/repack.txt |  3 +++
 builtin/repack.c                |  6 ++++-
 t/t7700-repack.sh               | 47 +++++++++++++++++++++++++++++++++
 3 files changed, 55 insertions(+), 1 deletion(-)

diff --git a/Documentation/config/repack.txt b/Documentation/config/repack.txt
index 9c413e177e..22bfc26afc 100644
--- a/Documentation/config/repack.txt
+++ b/Documentation/config/repack.txt
@@ -25,3 +25,6 @@ repack.writeBitmaps::
 	space and extra time spent on the initial repack.  This has
 	no effect if multiple packfiles are created.
 	Defaults to true on bare repos, false otherwise.
+
+repack.updateServerInfo::
+	If set to false, git-repack will not run git-update-server-info.
diff --git a/builtin/repack.c b/builtin/repack.c
index da1e364a75..3baa993da2 100644
--- a/builtin/repack.c
+++ b/builtin/repack.c
@@ -22,6 +22,7 @@ static int delta_base_offset = 1;
 static int pack_kept_objects = -1;
 static int write_bitmaps = -1;
 static int use_delta_islands;
+static int no_update_server_info = 0;
 static char *packdir, *packtmp_name, *packtmp;
 
 static const char *const git_repack_usage[] = {
@@ -54,6 +55,10 @@ static int repack_config(const char *var, const char *value, void *cb)
 		use_delta_islands = git_config_bool(var, value);
 		return 0;
 	}
+	if (strcmp(var, "repack.updateserverinfo") == 0) {
+		no_update_server_info = !git_config_bool(var, value);
+		return 0;
+	}
 	return git_default_config(var, value, cb);
 }
 
@@ -620,7 +625,6 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 	const char *unpack_unreachable = NULL;
 	int keep_unreachable = 0;
 	struct string_list keep_pack_list = STRING_LIST_INIT_NODUP;
-	int no_update_server_info = 0;
 	struct pack_objects_args po_args = {NULL};
 	int geometric_factor = 0;
 	int write_midx = 0;
diff --git a/t/t7700-repack.sh b/t/t7700-repack.sh
index 5922fb5bdd..f049120022 100755
--- a/t/t7700-repack.sh
+++ b/t/t7700-repack.sh
@@ -381,4 +381,51 @@ test_expect_success TTY '--quiet disables progress' '
 	test_must_be_empty stderr
 '
 
+test_expect_success 'updates server info by default' '
+	git init repo &&
+	test_when_finished "rm -rf repo" &&
+	test_commit -C repo message &&
+	test_path_is_missing repo/.git/objects/info/packs &&
+	test_path_is_missing repo/.git/info/refs &&
+	git -C repo repack &&
+	test_path_is_file repo/.git/objects/info/packs &&
+	test_path_is_file repo/.git/info/refs
+'
+
+test_expect_success '-n skips updating server info' '
+	git init repo &&
+	test_when_finished "rm -rf repo" &&
+	test_commit -C repo message &&
+	git -C repo repack -n &&
+	test_path_is_missing repo/.git/objects/info/packs &&
+	test_path_is_missing repo/.git/info/refs
+'
+
+test_expect_success 'repack.updateServerInfo=true updates server info' '
+	git init repo &&
+	test_when_finished "rm -rf repo" &&
+	test_commit -C repo message &&
+	git -C repo -c repack.updateServerInfo=true repack &&
+	test_path_is_file repo/.git/objects/info/packs &&
+	test_path_is_file repo/.git/info/refs
+'
+
+test_expect_success 'repack.updateServerInfo=false skips updating server info' '
+	git init repo &&
+	test_when_finished "rm -rf repo" &&
+	test_commit -C repo message &&
+	git -C repo -c repack.updateServerInfo=false repack &&
+	test_path_is_missing repo/.git/objects/info/packs &&
+	test_path_is_missing repo/.git/info/refs
+'
+
+test_expect_success '-n overrides repack.updateServerInfo=true' '
+	git init repo &&
+	test_when_finished "rm -rf repo" &&
+	test_commit -C repo message &&
+	git -C repo -c repack.updateServerInfo=true repack -n &&
+	test_path_is_missing repo/.git/objects/info/packs &&
+	test_path_is_missing repo/.git/info/refs
+'
+
 test_done
-- 
2.35.1


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH] repack: add config to skip updating server info

[Taylor Blau]

On Fri, Mar 11, 2022 at 12:09:30PM +0100, Patrick Steinhardt wrote:
> diff --git a/Documentation/config/repack.txt b/Documentation/config/repack.txt
> index 9c413e177e..22bfc26afc 100644
> --- a/Documentation/config/repack.txt
> +++ b/Documentation/config/repack.txt
> @@ -25,3 +25,6 @@ repack.writeBitmaps::
>  	space and extra time spent on the initial repack.  This has
>  	no effect if multiple packfiles are created.
>  	Defaults to true on bare repos, false otherwise.
> +
> +repack.updateServerInfo::
> +	If set to false, git-repack will not run git-update-server-info.

Can you clarify here what the default value of this config variable is,
and how it interacts with repack's `-n` flag? E.g., something along the
lines of:

    repack.updateServerInfo::
        If set to false, linkgit:git-repack[1] will not run
        linkgit:git-update-serve-info[1]. Defaults to true. Can be
        overridden when true by the `-n` option of
        linkgit:git-repack[1].

Perhaps a little verbose, but I think it leaves less ambiguity about
what this new configuration variable is for.

> diff --git a/builtin/repack.c b/builtin/repack.c
> index da1e364a75..3baa993da2 100644
> --- a/builtin/repack.c
> +++ b/builtin/repack.c
> @@ -22,6 +22,7 @@ static int delta_base_offset = 1;
>  static int pack_kept_objects = -1;
>  static int write_bitmaps = -1;
>  static int use_delta_islands;
> +static int no_update_server_info = 0;

Not the fault of this patch, but I wonder if this would be less
confusing if we stored `update_server_info` instead of
`no_update_server_info`. If you have time, I think it may be worth a
preparatory patch at the beginning to swap the two.

> +test_expect_success 'updates server info by default' '
> +	git init repo &&
> +	test_when_finished "rm -rf repo" &&
> +	test_commit -C repo message &&
> +	test_path_is_missing repo/.git/objects/info/packs &&
> +	test_path_is_missing repo/.git/info/refs &&
> +	git -C repo repack &&
> +	test_path_is_file repo/.git/objects/info/packs &&
> +	test_path_is_file repo/.git/info/refs
> +'

I wonder if this and the below tests might be cleaned up with a pair of
helper functions, perhaps:

    test_server_info_present () {
      test_path_is_file .git/objects/info/packs &&
      test_path_is_file .git/info/refs
    }

    test_server_info_missing () {
      test_path_is_missing .git/objects/info/packs &&
      test_path_is_missing .git/info/refs
    }

t7700 has a mix of styles, but it may shorten some of the lines to use a
subshell that is changed into the repo directory, e.g., the test above
would become:

    test_expect_success 'updates server info by default' '
      git init repo &&
      test_when_finished "rm -fr repo" &&
      (
        test_commit message &&
        test_server_info_missing &&
        git repack &&
        test_server_info_present
      )
    '

which reads a little more easily to me. It would be nice to avoid
creating the sub-repos at all, perhaps by removing these files
ourselves in between tests.

> +test_expect_success '-n skips updating server info' '
> +test_expect_success 'repack.updateServerInfo=true updates server info' '
> +test_expect_success 'repack.updateServerInfo=false skips updating server info' '
> +test_expect_success '-n overrides repack.updateServerInfo=true' '

Great, these four and the above together cover all of the cases I think
we'd be interested in.

Thanks,
Taylor

Re: [PATCH] repack: add config to skip updating server info

[Patrick Steinhardt]

[-- Attachment #1: Type: text/plain, Size: 2078 bytes --]

On Fri, Mar 11, 2022 at 11:20:41AM -0500, Taylor Blau wrote:
> On Fri, Mar 11, 2022 at 12:09:30PM +0100, Patrick Steinhardt wrote:
> > diff --git a/Documentation/config/repack.txt b/Documentation/config/repack.txt
> > index 9c413e177e..22bfc26afc 100644
> > --- a/Documentation/config/repack.txt
> > +++ b/Documentation/config/repack.txt
> > @@ -25,3 +25,6 @@ repack.writeBitmaps::
> >  	space and extra time spent on the initial repack.  This has
> >  	no effect if multiple packfiles are created.
> >  	Defaults to true on bare repos, false otherwise.
> > +
> > +repack.updateServerInfo::
> > +	If set to false, git-repack will not run git-update-server-info.
> 
> Can you clarify here what the default value of this config variable is,
> and how it interacts with repack's `-n` flag? E.g., something along the
> lines of:
> 
>     repack.updateServerInfo::
>         If set to false, linkgit:git-repack[1] will not run
>         linkgit:git-update-serve-info[1]. Defaults to true. Can be
>         overridden when true by the `-n` option of
>         linkgit:git-repack[1].
> 
> Perhaps a little verbose, but I think it leaves less ambiguity about
> what this new configuration variable is for.

Makes sense.

> > diff --git a/builtin/repack.c b/builtin/repack.c
> > index da1e364a75..3baa993da2 100644
> > --- a/builtin/repack.c
> > +++ b/builtin/repack.c
> > @@ -22,6 +22,7 @@ static int delta_base_offset = 1;
> >  static int pack_kept_objects = -1;
> >  static int write_bitmaps = -1;
> >  static int use_delta_islands;
> > +static int no_update_server_info = 0;
> 
> Not the fault of this patch, but I wonder if this would be less
> confusing if we stored `update_server_info` instead of
> `no_update_server_info`. If you have time, I think it may be worth a
> preparatory patch at the beginning to swap the two.
[snip]

I indeed first had a look at how to do this, but didn't find a negated
`OPT_BOOL()`. I had another look now though, and it seems like this is
typically solved via `OPT_NEGBIT()`.

Thanks!

Patrick

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[PATCH v2 0/2] repack: add config to skip updating server info

[Patrick Steinhardt]

[-- Attachment #1: Type: text/plain, Size: 1038 bytes --]

Hi,

this is the second version of my patch series which adds a new config to
allow users and admins to permanently skip updating server info required
for the dumb HTTP protocol in git-repack(1).

The following things changed in comparison to v1:

    - I've split out a preparatory patch which refactors the logic to
      avoid double-negation of the `no_update_server_info` field.

    - I've updated documentation to be more explicit with regards to the
      default of the new config and to mention that the `-n` flag
      overrides it.

    - I've refactored tests to be less verbose.

Thanks for your feedback, Taylor!

Patrick

Patrick Steinhardt (2):
  repack: refactor to avoid double-negation of update-server-info
  repack: add config to skip updating server info

 Documentation/config/repack.txt |  5 ++++
 builtin/repack.c                | 12 +++++---
 t/t7700-repack.sh               | 50 +++++++++++++++++++++++++++++++++
 3 files changed, 63 insertions(+), 4 deletions(-)

-- 
2.35.1


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[PATCH v2 1/2] repack: refactor to avoid double-negation of update-server-info

[Patrick Steinhardt]

[-- Attachment #1: Type: text/plain, Size: 3321 bytes --]

By default, git-repack(1) runs `update_server_info()` to generate info
required for the dumb HTTP protocol. This can be disabled via the `-n`
flag, which then sets the `no_update_server_info` flag. Further down the
code this leads to some double-negation logic, which is about to become
more confusing as we're about to add a new config which allows the user
to permanently disable generation of the info.

Refactor the code to avoid the double-negation and add some tests which
verify that the flag continues to work as expected.

Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 builtin/repack.c  |  8 ++++----
 t/t7700-repack.sh | 32 ++++++++++++++++++++++++++++++++
 2 files changed, 36 insertions(+), 4 deletions(-)

diff --git a/builtin/repack.c b/builtin/repack.c
index da1e364a75..f2ac8ad14b 100644
--- a/builtin/repack.c
+++ b/builtin/repack.c
@@ -620,7 +620,7 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 	const char *unpack_unreachable = NULL;
 	int keep_unreachable = 0;
 	struct string_list keep_pack_list = STRING_LIST_INIT_NODUP;
-	int no_update_server_info = 0;
+	int run_update_server_info = 1;
 	struct pack_objects_args po_args = {NULL};
 	int geometric_factor = 0;
 	int write_midx = 0;
@@ -637,8 +637,8 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 				N_("pass --no-reuse-delta to git-pack-objects")),
 		OPT_BOOL('F', NULL, &po_args.no_reuse_object,
 				N_("pass --no-reuse-object to git-pack-objects")),
-		OPT_BOOL('n', NULL, &no_update_server_info,
-				N_("do not run git-update-server-info")),
+		OPT_NEGBIT('n', NULL, &run_update_server_info,
+				N_("do not run git-update-server-info"), 1),
 		OPT__QUIET(&po_args.quiet, N_("be quiet")),
 		OPT_BOOL('l', "local", &po_args.local,
 				N_("pass --local to git-pack-objects")),
@@ -939,7 +939,7 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 			prune_shallow(PRUNE_QUICK);
 	}
 
-	if (!no_update_server_info)
+	if (run_update_server_info)
 		update_server_info(0);
 	remove_temporary_files();
 
diff --git a/t/t7700-repack.sh b/t/t7700-repack.sh
index 5922fb5bdd..6b387bbdbe 100755
--- a/t/t7700-repack.sh
+++ b/t/t7700-repack.sh
@@ -381,4 +381,36 @@ test_expect_success TTY '--quiet disables progress' '
 	test_must_be_empty stderr
 '
 
+test_expect_success 'setup for update-server-info' '
+	git init update-server-info &&
+	test_commit -C update-server-info message
+'
+
+test_server_info_present () {
+	test_path_is_file update-server-info/.git/objects/info/packs &&
+	test_path_is_file update-server-info/.git/info/refs
+}
+
+test_server_info_missing () {
+	test_path_is_missing update-server-info/.git/objects/info/packs &&
+	test_path_is_missing update-server-info/.git/info/refs
+}
+
+test_server_info_cleanup () {
+	rm -f update-server-info/.git/objects/info/packs update-server-info/.git/info/refs &&
+	test_server_info_missing
+}
+
+test_expect_success 'updates server info by default' '
+	test_server_info_cleanup &&
+	git -C update-server-info repack &&
+	test_server_info_present
+'
+
+test_expect_success '-n skips updating server info' '
+	test_server_info_cleanup &&
+	git -C update-server-info repack -n &&
+	test_server_info_missing
+'
+
 test_done
-- 
2.35.1


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

[PATCH v2 2/2] repack: add config to skip updating server info

[Patrick Steinhardt]

[-- Attachment #1: Type: text/plain, Size: 3514 bytes --]

By default, git-repack(1) will update server info that is required by
the dumb HTTP transport. This can be skipped by passing the `-n` flag,
but what we're noticably missing is a config option to permanently
disable updating this information.

Add a new option "repack.updateServerInfo" which can be used to disable
the logic. Most hosting providers have turned off the dumb HTTP protocol
anyway, and on the client-side it woudln't typically be useful either.
Giving a persistent way to disable this feature thus makes quite some
sense to avoid wasting compute cycles and storage.

Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 Documentation/config/repack.txt |  5 +++++
 builtin/repack.c                |  6 +++++-
 t/t7700-repack.sh               | 18 ++++++++++++++++++
 3 files changed, 28 insertions(+), 1 deletion(-)

diff --git a/Documentation/config/repack.txt b/Documentation/config/repack.txt
index 9c413e177e..41ac6953c8 100644
--- a/Documentation/config/repack.txt
+++ b/Documentation/config/repack.txt
@@ -25,3 +25,8 @@ repack.writeBitmaps::
 	space and extra time spent on the initial repack.  This has
 	no effect if multiple packfiles are created.
 	Defaults to true on bare repos, false otherwise.
+
+repack.updateServerInfo::
+	If set to false, linkgit:git-repack[1] will not run
+	linkgit:git-update-server-info[1]. Defaults to true. Can be overridden
+	when true by the `-n` option of linkgit:git-repack[1].
diff --git a/builtin/repack.c b/builtin/repack.c
index f2ac8ad14b..d1a563d5b6 100644
--- a/builtin/repack.c
+++ b/builtin/repack.c
@@ -22,6 +22,7 @@ static int delta_base_offset = 1;
 static int pack_kept_objects = -1;
 static int write_bitmaps = -1;
 static int use_delta_islands;
+static int run_update_server_info = 1;
 static char *packdir, *packtmp_name, *packtmp;
 
 static const char *const git_repack_usage[] = {
@@ -54,6 +55,10 @@ static int repack_config(const char *var, const char *value, void *cb)
 		use_delta_islands = git_config_bool(var, value);
 		return 0;
 	}
+	if (strcmp(var, "repack.updateserverinfo") == 0) {
+		run_update_server_info = git_config_bool(var, value);
+		return 0;
+	}
 	return git_default_config(var, value, cb);
 }
 
@@ -620,7 +625,6 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 	const char *unpack_unreachable = NULL;
 	int keep_unreachable = 0;
 	struct string_list keep_pack_list = STRING_LIST_INIT_NODUP;
-	int run_update_server_info = 1;
 	struct pack_objects_args po_args = {NULL};
 	int geometric_factor = 0;
 	int write_midx = 0;
diff --git a/t/t7700-repack.sh b/t/t7700-repack.sh
index 6b387bbdbe..770d143204 100755
--- a/t/t7700-repack.sh
+++ b/t/t7700-repack.sh
@@ -413,4 +413,22 @@ test_expect_success '-n skips updating server info' '
 	test_server_info_missing
 '
 
+test_expect_success 'repack.updateServerInfo=true updates server info' '
+	test_server_info_cleanup &&
+	git -C update-server-info -c repack.updateServerInfo=true repack &&
+	test_server_info_present
+'
+
+test_expect_success 'repack.updateServerInfo=false skips updating server info' '
+	test_server_info_cleanup &&
+	git -C update-server-info -c repack.updateServerInfo=false repack &&
+	test_server_info_missing
+'
+
+test_expect_success '-n overrides repack.updateServerInfo=true' '
+	test_server_info_cleanup &&
+	git -C update-server-info -c repack.updateServerInfo=true repack -n &&
+	test_server_info_missing
+'
+
 test_done
-- 
2.35.1


[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH v2 2/2] repack: add config to skip updating server info

[Junio C Hamano]

Patrick Steinhardt <ps@pks.im> writes:

> By default, git-repack(1) will update server info that is required by
> the dumb HTTP transport. This can be skipped by passing the `-n` flag,
> but what we're noticably missing is a config option to permanently
> disable updating this information.
>
> Add a new option "repack.updateServerInfo" which can be used to disable
> the logic. Most hosting providers have turned off the dumb HTTP protocol
> anyway, and on the client-side it woudln't typically be useful either.
> Giving a persistent way to disable this feature thus makes quite some
> sense to avoid wasting compute cycles and storage.

Makes sense.  Perhaps we can leave a NEEDSWORK comment somewhere to
flip the default to false in two years or so.

Will queue.

Thanks.