Re: [PATCH v2 2/8] submodule--helper: get remote names from any repository

2 messages, 2 authors, 2021-10-03 · open the first message on its own page

Re: [PATCH v2 2/8] submodule--helper: get remote names from any repository

From: Junio C Hamano <hidden>
Date: 2021-09-20 16:54:50

Atharva Raykar [off-list ref] 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 [off-list ref]
Mentored-by: Shourya Shukla [off-list ref]
Signed-off-by: Atharva Raykar <redacted>
---
 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.
quoted hunk
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.

Re: [PATCH v2 2/8] submodule--helper: get remote names from any repository

From: Atharva Raykar <hidden>
Date: 2021-10-03 13:30:33

Junio C Hamano [off-list ref] writes:
Atharva Raykar [off-list ref] writes:
quoted
`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 [off-list ref]
Mentored-by: Shourya Shukla [off-list ref]
Signed-off-by: Atharva Raykar <redacted>
---
 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.
[...]
quoted
 	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().
Thanks, this was a good catch.
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.
My initial series had this patch split up similar to what you suggested,
but I ultimately chose to favour the larger granularity to make the
whole series a bit shorter. This patch definitely towed the line a bit.
Thanks.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help