From: Junio C Hamano <gitster@pobox.com>
To: Atharva Raykar <raykar.ath@gmail.com>
Cc: avarab@gmail.com, christian.couder@gmail.com,
emilyshaffer@google.com, git@vger.kernel.org, jrnieder@gmail.com,
kaartic.sivaraam@gmail.com, pc44800@gmail.com,
periperidip@gmail.com
Subject: Re: [PATCH v2 2/8] submodule--helper: get remote names from any repository
Date: Mon, 20 Sep 2021 09:52:54 -0700 [thread overview]
Message-ID: <xmqqr1djchp5.fsf@gitster.g> (raw)
In-Reply-To: <20210916103241.62376-3-raykar.ath@gmail.com> (Atharva Raykar's message of "Thu, 16 Sep 2021 16:02:35 +0530")
Atharva Raykar <raykar.ath@gmail.com> writes:
> `get_default_remote()` retrieves the name of a remote by resolving the
> refs from of the current repository's ref store.
>
> Thus in order to use it for retrieving the remote name of a submodule,
> we have to start a new subprocess which runs from the submodule
> directory.
>
> Let's instead introduce a function called `repo_get_default_remote()`
> which takes any repository object and retrieves the remote accordingly.
>
> `get_default_remote()` is then defined as a call to
> `repo_get_default_remote()` with 'the_repository' passed to it.
>
> Now that we have `repo_get_default_remote()`, we no longer have to start
> a subprocess that called `submodule--helper get-default-remote` from
> within the submodule directory.
>
> So let's make a function called `get_default_remote_submodule()` which
> takes a submodule path, and returns the default remote for that
> submodule, all within the same process.
>
> We can now use this function to save an unnecessary subprocess spawn in
> `sync_submodule()`, and also in the next patch, which will require this
> functionality.
>
> Mentored-by: Christian Couder <christian.couder@gmail.com>
> Mentored-by: Shourya Shukla <periperidip@gmail.com>
> Signed-off-by: Atharva Raykar <raykar.ath@gmail.com>
> ---
> builtin/submodule--helper.c | 34 +++++++++++++++++++++++-----------
> 1 file changed, 23 insertions(+), 11 deletions(-)
The above covers a lot of stuff, because this change does a lot of
things ;-)
The commit could be split into 3 logically distinct phases (two
independent, the third depends on the other two):
- extract the part that uses the run-command() API to run
"submodule--helper print-default-remote" from sync_submodule()
and create get_default_remote_submodule() function out of it.
- move bulk of get_default_remote() into repo_get_default_remote()
and reimplement the former as a thin wrapper of the latter.
- using the repo_get_default_remote() created in the second step,
update the get_default_remote_submodule() created in the first
step to work in-process without a subprocess.
but a bit larger granularity used here is probably OK.
> diff --git a/builtin/submodule--helper.c b/builtin/submodule--helper.c
> index 97512ccf0b..f6c4fc349b 100644
> --- a/builtin/submodule--helper.c
> +++ b/builtin/submodule--helper.c
> @@ -29,11 +29,10 @@
> typedef void (*each_submodule_fn)(const struct cache_entry *list_item,
> void *cb_data);
>
> -static char *get_default_remote(void)
> +static char *repo_get_default_remote(struct repository *repo, const char *refname)
> {
> char *dest = NULL, *ret;
> struct strbuf sb = STRBUF_INIT;
> - const char *refname = resolve_ref_unsafe("HEAD", 0, NULL, NULL);
>
> if (!refname)
> die(_("No such ref: %s"), "HEAD");
> @@ -46,7 +45,7 @@ static char *get_default_remote(void)
> die(_("Expecting a full ref name, got %s"), refname);
>
> strbuf_addf(&sb, "branch.%s.remote", refname);
> - if (git_config_get_string(sb.buf, &dest))
> + if (repo_config_get_string(repo, sb.buf, &dest))
> ret = xstrdup("origin");
> else
> ret = dest;
> @@ -55,6 +54,25 @@ static char *get_default_remote(void)
> return ret;
> }
>
> +static char *get_default_remote_submodule(const char *module_path)
> +{
> + const char *refname;
> + const struct submodule *sub;
> + struct repository subrepo;
> +
> + refname = refs_resolve_ref_unsafe(get_submodule_ref_store(module_path),
> + "HEAD", 0, NULL, NULL);
> + sub = submodule_from_path(the_repository, null_oid(), module_path);
> + repo_submodule_init(&subrepo, the_repository, sub);
> + return repo_get_default_remote(&subrepo, refname);
> +}
> +
> +static char *get_default_remote(void)
> +{
> + const char *refname = resolve_ref_unsafe("HEAD", 0, NULL, NULL);
> + return repo_get_default_remote(the_repository, refname);
> +}
> +
> static int print_default_remote(int argc, const char **argv, const char *prefix)
> {
> char *remote;
> @@ -1369,7 +1387,6 @@ static void sync_submodule(const char *path, const char *prefix,
> char *remote_key = NULL;
> char *sub_origin_url, *super_config_url, *displaypath;
> struct strbuf sb = STRBUF_INIT;
> - struct child_process cp = CHILD_PROCESS_INIT;
> char *sub_config_path = NULL;
>
> if (!is_submodule_active(the_repository, path))
> @@ -1418,14 +1435,9 @@ static void sync_submodule(const char *path, const char *prefix,
> if (!is_submodule_populated_gently(path, NULL))
> goto cleanup;
>
> - prepare_submodule_repo_env(&cp.env_array);
> - cp.git_cmd = 1;
> - cp.dir = path;
> - strvec_pushl(&cp.args, "submodule--helper",
> - "print-default-remote", NULL);
> -
> strbuf_reset(&sb);
> - if (capture_command(&cp, &sb, 0))
> + strbuf_addstr(&sb, get_default_remote_submodule(path));
> + if (!sb.buf)
> die(_("failed to get the default remote for submodule '%s'"),
> path);
There is this line after the post context presented here:
strbuf_strip_suffix(&sb, "\n");
remote_key = xstrfmt("remote.%s.url", sb.buf);
The LF was expected to come from the capture_command(), but it no
longer is needed after get_default_remote_submodule() switches to
use the in-process method. In fact, the use of sb for the purpose
of forming remote_key is not needed.
tmp = get_default_remote_submodule(path);
if (!tmp)
die(_("failed to get..."));
remote_key = xstrfmt("remote.%s.url", tmp);
free(tmp);
Yes, I think the new code leaks, and makes it impossible not to
leak, the returned value from get_default_remote_submodule() by
passing the call as an argument to strbuf_addstr().
Both of the two bugs pointed out here would have been easier to spot
if the commits were smaller, I would think, but as I said, a bit
larger granularity used here was still manageable.
Thanks.
next prev parent reply other threads:[~2021-09-20 16:55 UTC|newest]
Thread overview: 142+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-09-07 11:59 [PATCH 00/13] submodule: convert the rest of 'update' to C Atharva Raykar
2021-09-07 11:59 ` [PATCH 01/13] submodule--helper: split up ensure_core_worktree() Atharva Raykar
2021-09-07 11:59 ` [PATCH 02/13] submodule--helper: get remote names from any repository Atharva Raykar
2021-09-07 12:37 ` Ævar Arnfjörð Bjarmason
2021-09-07 13:33 ` Atharva Raykar
2021-09-07 11:59 ` [PATCH 03/13] submodule--helper: introduce get_default_remote_submodule() Atharva Raykar
2021-09-07 11:59 ` [PATCH 04/13] submodule--helper: rename helpers for update-clone Atharva Raykar
2021-09-07 11:59 ` [PATCH 05/13] submodule--helper: refactor get_submodule_displaypath() Atharva Raykar
2021-09-07 11:59 ` [PATCH 06/13] submodule: move core cmd_update() logic to C Atharva Raykar
2021-09-07 12:40 ` Ævar Arnfjörð Bjarmason
2021-09-07 11:59 ` [PATCH 07/13] submodule: remove fetch_in_submodule shell function Atharva Raykar
2021-09-07 12:44 ` Ævar Arnfjörð Bjarmason
2021-09-07 11:59 ` [PATCH 08/13] submodule--helper: remove update-clone subcommand Atharva Raykar
2021-09-07 12:46 ` Ævar Arnfjörð Bjarmason
2021-09-07 11:59 ` [PATCH 09/13] submodule--helper: remove update-module-mode subcommand Atharva Raykar
2021-09-07 12:49 ` Ævar Arnfjörð Bjarmason
2021-09-07 13:50 ` Atharva Raykar
2021-09-07 11:59 ` [PATCH 10/13] submodule--helper: remove shell interface to ensure_core_worktree() Atharva Raykar
2021-09-07 11:59 ` [PATCH 11/13] submodule--helper: remove print-default-remote subcommand Atharva Raykar
2021-09-07 11:59 ` [PATCH 12/13] submodule--helper: remove relative-path subcommand Atharva Raykar
2021-09-07 11:59 ` [PATCH 13/13] submodule--helper: remove run-update-procedure subcommand Atharva Raykar
2021-09-07 12:34 ` [PATCH 00/13] submodule: convert the rest of 'update' to C Ævar Arnfjörð Bjarmason
2021-09-07 12:53 ` Atharva Raykar
2021-09-16 10:32 ` [PATCH v2 0/8] " Atharva Raykar
2021-09-16 10:32 ` [PATCH v2 1/8] submodule--helper: split up ensure_core_worktree() Atharva Raykar
2021-09-16 10:32 ` [PATCH v2 2/8] submodule--helper: get remote names from any repository Atharva Raykar
2021-09-20 16:52 ` Junio C Hamano [this message]
2021-10-03 13:22 ` Atharva Raykar
2021-09-20 21:28 ` Junio C Hamano
2021-09-21 16:33 ` Jonathan Tan
2021-09-16 10:32 ` [PATCH v2 3/8] submodule--helper: rename helpers for update-clone Atharva Raykar
2021-09-16 10:32 ` [PATCH v2 4/8] submodule--helper: refactor get_submodule_displaypath() Atharva Raykar
2021-09-16 10:32 ` [PATCH v2 5/8] submodule: move core cmd_update() logic to C Atharva Raykar
2021-09-20 17:13 ` Junio C Hamano
2021-10-03 10:38 ` Atharva Raykar
2021-09-20 19:58 ` Junio C Hamano
2021-09-20 21:28 ` Junio C Hamano
2021-09-16 10:32 ` [PATCH v2 6/8] submodule--helper: remove update-clone subcommand Atharva Raykar
2021-09-16 10:32 ` [PATCH v2 7/8] submodule--helper: remove unused helpers Atharva Raykar
2021-09-20 17:19 ` Junio C Hamano
2021-09-16 10:32 ` [PATCH v2 8/8] submodule--helper: rename helper functions Atharva Raykar
2021-09-20 17:19 ` Junio C Hamano
2021-10-13 5:17 ` [PATCH v3 0/9] submodule: convert the rest of 'update' to C Atharva Raykar
2021-10-13 5:17 ` [PATCH v3 1/9] submodule--helper: split up ensure_core_worktree() Atharva Raykar
2021-10-13 5:17 ` [PATCH v3 2/9] submodule--helper: get remote names from any repository Atharva Raykar
2021-10-13 5:17 ` [PATCH v3 3/9] submodule--helper: rename helpers for update-clone Atharva Raykar
2021-10-13 5:18 ` [PATCH v3 4/9] submodule--helper: refactor get_submodule_displaypath() Atharva Raykar
2021-10-13 5:18 ` [PATCH v3 5/9] submodule--helper: allow setting superprefix for init_submodule() Atharva Raykar
2021-10-13 5:18 ` [PATCH v3 6/9] submodule--helper: run update using child process struct Atharva Raykar
2021-10-13 5:18 ` [PATCH v3 7/9] submodule: move core cmd_update() logic to C Atharva Raykar
2021-10-13 5:18 ` [PATCH v3 8/9] submodule--helper: remove unused helpers Atharva Raykar
2021-10-13 5:18 ` [PATCH v3 9/9] submodule--helper: rename helper functions Atharva Raykar
2021-10-14 0:05 ` [PATCH v3 0/9] submodule: convert the rest of 'update' to C Junio C Hamano
2021-10-14 20:46 ` Emily Shaffer
2021-12-03 19:00 ` Junio C Hamano
2021-12-03 20:15 ` Ævar Arnfjörð Bjarmason
2021-12-04 10:38 ` Atharva Raykar
2022-01-27 16:22 ` [PATCH v4 0/7] " Ævar Arnfjörð Bjarmason
2022-01-27 16:22 ` [PATCH v4 1/7] submodule--helper: get remote names from any repository Ævar Arnfjörð Bjarmason
2022-01-27 18:45 ` Glen Choo
2022-01-27 16:22 ` [PATCH v4 2/7] submodule--helper: refactor get_submodule_displaypath() Ævar Arnfjörð Bjarmason
2022-01-27 16:22 ` [PATCH v4 3/7] submodule--helper: allow setting superprefix for init_submodule() Ævar Arnfjörð Bjarmason
2022-01-27 16:22 ` [PATCH v4 4/7] submodule--helper: run update using child process struct Ævar Arnfjörð Bjarmason
2022-01-27 16:22 ` [PATCH v4 5/7] builtin/submodule--helper.c: reformat designated initializers Ævar Arnfjörð Bjarmason
2022-01-27 16:22 ` [PATCH v4 6/7] builtin/submodule--helper.c: rename "suc" variable to "opt" Ævar Arnfjörð Bjarmason
2022-01-27 16:22 ` [PATCH v4 7/7] submodule: move core cmd_update() logic to C Ævar Arnfjörð Bjarmason
2022-01-27 21:55 ` Glen Choo
2022-01-28 12:56 ` [PATCH v5 0/9] submodule: convert the rest of 'update' " Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 1/9] submodule--helper: get remote names from any repository Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 2/9] submodule--helper: refactor get_submodule_displaypath() Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 3/9] submodule--helper: allow setting superprefix for init_submodule() Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 4/9] submodule--helper: run update using child process struct Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 5/9] builtin/submodule--helper.c: reformat designated initializers Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 6/9] builtin/submodule--helper.c: rename option variables to "opt" Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 7/9] submodule--helper: don't use bitfield indirection for parse_options() Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 8/9] submodule tests: test for init and update failure output Ævar Arnfjörð Bjarmason
2022-01-28 12:56 ` [PATCH v5 9/9] submodule: move core cmd_update() logic to C Ævar Arnfjörð Bjarmason
2022-02-03 0:18 ` Glen Choo
2022-02-03 2:26 ` Ævar Arnfjörð Bjarmason
2022-02-03 8:15 ` Ævar Arnfjörð Bjarmason
2022-02-03 17:35 ` Glen Choo
2022-02-08 8:39 ` [PATCH v6 00/16] submodule: convert the rest of 'update' " Glen Choo
2022-02-08 8:39 ` [PATCH v6 01/16] submodule--helper: get remote names from any repository Glen Choo
2022-02-08 8:39 ` [PATCH v6 02/16] submodule--helper: refactor get_submodule_displaypath() Glen Choo
2022-02-08 8:39 ` [PATCH v6 03/16] submodule--helper: allow setting superprefix for init_submodule() Glen Choo
2022-02-08 8:39 ` [PATCH v6 04/16] submodule--helper: run update using child process struct Glen Choo
2022-02-08 8:39 ` [PATCH v6 05/16] builtin/submodule--helper.c: reformat designated initializers Glen Choo
2022-02-08 8:39 ` [PATCH v6 06/16] builtin/submodule--helper.c: rename option variables to "opt" Glen Choo
2022-02-08 8:39 ` [PATCH v6 07/16] submodule--helper: don't use bitfield indirection for parse_options() Glen Choo
2022-02-08 8:39 ` [PATCH v6 08/16] submodule tests: test for init and update failure output Glen Choo
2022-02-08 8:39 ` [PATCH v6 09/16] submodule--helper: remove update-module-mode Glen Choo
2022-02-08 8:39 ` [PATCH v6 10/16] submodule--helper: reorganize code for sh to C conversion Glen Choo
2022-02-08 8:39 ` [PATCH v6 11/16] submodule--helper run-update-procedure: remove --suboid Glen Choo
2022-02-08 8:39 ` [PATCH v6 12/16] submodule--helper run-update-procedure: learn --remote Glen Choo
2022-02-08 8:39 ` [PATCH v6 13/16] submodule--helper: remove ensure-core-worktree Glen Choo
2022-02-08 8:39 ` [PATCH v6 14/16] submodule--helper update-clone: learn --init Glen Choo
2022-02-08 8:39 ` [PATCH v6 15/16] submodule--helper: move functions around Glen Choo
2022-02-08 8:39 ` [PATCH v6 16/16] submodule: move core cmd_update() logic to C Glen Choo
2022-02-10 9:28 ` [PATCH v7 00/20] submodule: convert the rest of 'update' " Glen Choo
2022-02-10 9:28 ` [PATCH v7 01/20] submodule--helper: get remote names from any repository Glen Choo
2022-02-10 9:28 ` [PATCH v7 02/20] submodule--helper: refactor get_submodule_displaypath() Glen Choo
2022-02-12 14:24 ` Ævar Arnfjörð Bjarmason
2022-02-10 9:28 ` [PATCH v7 03/20] submodule--helper: allow setting superprefix for init_submodule() Glen Choo
2022-02-12 14:30 ` Ævar Arnfjörð Bjarmason
2022-02-10 9:28 ` [PATCH v7 04/20] submodule--helper: run update using child process struct Glen Choo
2022-02-12 14:33 ` Ævar Arnfjörð Bjarmason
2022-02-10 9:28 ` [PATCH v7 05/20] builtin/submodule--helper.c: reformat designated initializers Glen Choo
2022-02-10 9:28 ` [PATCH v7 06/20] builtin/submodule--helper.c: rename option variables to "opt" Glen Choo
2022-02-10 9:28 ` [PATCH v7 07/20] submodule--helper: don't use bitfield indirection for parse_options() Glen Choo
2022-02-10 9:28 ` [PATCH v7 08/20] submodule tests: test for init and update failure output Glen Choo
2022-02-10 9:28 ` [PATCH v7 09/20] submodule--helper: remove update-module-mode Glen Choo
2022-02-12 14:35 ` Ævar Arnfjörð Bjarmason
2022-02-10 9:28 ` [PATCH v7 10/20] submodule--helper: reorganize code for sh to C conversion Glen Choo
2022-02-10 9:28 ` [PATCH v7 11/20] submodule--helper run-update-procedure: remove --suboid Glen Choo
2022-02-10 9:28 ` [PATCH v7 12/20] submodule--helper run-update-procedure: learn --remote Glen Choo
2022-02-12 14:38 ` Ævar Arnfjörð Bjarmason
2022-02-10 9:28 ` [PATCH v7 13/20] submodule--helper: remove ensure-core-worktree Glen Choo
2022-02-10 9:28 ` [PATCH v7 14/20] submodule--helper update-clone: learn --init Glen Choo
2022-02-10 9:28 ` [PATCH v7 15/20] submodule--helper: move functions around Glen Choo
2022-02-12 14:41 ` Ævar Arnfjörð Bjarmason
2022-02-10 9:28 ` [PATCH v7 16/20] submodule--helper: reduce logic in run_update_procedure() Glen Choo
2022-02-10 9:28 ` [PATCH v7 17/20] submodule: move core cmd_update() logic to C Glen Choo
2022-02-10 9:34 ` Glen Choo
2022-02-10 9:28 ` [PATCH v7 18/20] fixup! submodule--helper run-update-procedure: remove --suboid Glen Choo
2022-02-12 14:41 ` Ævar Arnfjörð Bjarmason
2022-02-10 9:28 ` [PATCH v7 19/20] fixup! submodule--helper run-update-procedure: learn --remote Glen Choo
2022-02-12 14:43 ` Ævar Arnfjörð Bjarmason
2022-02-10 9:28 ` [PATCH v7 20/20] fixup! submodule: move core cmd_update() logic to C Glen Choo
2022-02-12 14:45 ` [PATCH v7 00/20] submodule: convert the rest of 'update' " Ævar Arnfjörð Bjarmason
2022-02-17 5:44 ` Glen Choo
2022-02-17 9:17 ` Ævar Arnfjörð Bjarmason
2022-02-17 16:14 ` Glen Choo
2022-02-13 5:54 ` Junio C Hamano
2022-02-13 6:14 ` Junio C Hamano
2022-02-14 16:37 ` Glen Choo
2022-02-14 17:19 ` Ævar Arnfjörð Bjarmason
2022-02-15 9:34 ` Glen Choo
2022-02-14 17:34 ` Junio C Hamano
2022-02-15 9:31 ` Glen Choo
2022-02-15 9:47 ` Glen Choo
2021-10-14 21:50 ` [PATCH 00/13] " Glen Choo
2021-10-15 9:13 ` Atharva Raykar
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
List information: http://vger.kernel.org/majordomo-info.html
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=xmqqr1djchp5.fsf@gitster.g \
--to=gitster@pobox.com \
--cc=avarab@gmail.com \
--cc=christian.couder@gmail.com \
--cc=emilyshaffer@google.com \
--cc=git@vger.kernel.org \
--cc=jrnieder@gmail.com \
--cc=kaartic.sivaraam@gmail.com \
--cc=pc44800@gmail.com \
--cc=periperidip@gmail.com \
--cc=raykar.ath@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
Code repositories for project(s) associated with this public inbox
https://80x24.org/mirrors/git.git
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).