Skip to contentSkip to conversation

repack: call prune_packed_objects() and update_server_info() directly

PATCH In review René Scharfe 4 participants 9 messages Jun 15, 2016

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 repack: call prune_packed_objects() and update_server_info() directly

Call the functions behind git prune-packed and git update-server-info directly instead of using run_command(). This is shorter, easier and quicker.

---
 builtin/repack.c | 23 ++++++-----------------
 1 file changed, 6 insertions(+), 17 deletions(-)
builtin/repack.c +6 -17
index fc088db..2aae05d 100644
@@ -377,6 +377,7 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 	/* End of pack replacement. */
 
 	if (delete_redundant) {
+		int opts = 0;
 		sort_string_list(&names);
 		for_each_string_list_item(item, &existing_packs) {
 			char *sha1;
@@ -387,25 +388,13 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 			if (!string_list_has_string(&names, sha1))
 				remove_redundant_pack(packdir, item->string);
 		}
-		argv_array_push(&cmd_args, "prune-packed");
-		if (quiet)
-			argv_array_push(&cmd_args, "--quiet");
-
-		memset(&cmd, 0, sizeof(cmd));
-		cmd.argv = cmd_args.argv;
-		cmd.git_cmd = 1;
-		run_command(&cmd);
-		argv_array_clear(&cmd_args);
+		if (!quiet && isatty(2))
+			opts |= PRUNE_PACKED_VERBOSE;
+		prune_packed_objects(opts);
 	}
 
-	if (!no_update_server_info) {
-		argv_array_push(&cmd_args, "update-server-info");
-		memset(&cmd, 0, sizeof(cmd));
-		cmd.argv = cmd_args.argv;
-		cmd.git_cmd = 1;
-		run_command(&cmd);
-		argv_array_clear(&cmd_args);
-	}
+	if (!no_update_server_info)
+		update_server_info(0);
 	remove_temporary_files();
 	string_list_clear(&names, 0);
 	string_list_clear(&rollback, 0);
Signature · 1 line
2.1.0
#1 lore raw
Quoted 3 lines · René Scharfe

Call the functions behind git prune-packed and git update-server-info directly instead of using run_command(). This is shorter, easier and quicker.

It can also introduce bugs, since a lot of git code assumes it is running in a single process and can die() or mark up global variables at will. :)

I gave a quick read-through of the code and I think these calls are OK. The two things I noticed were:

1. We might die on a malloc failure that would otherwise go unnoticed in a sub-process. That's probably OK.

2. The info/packs file is generated from our internal packed_git list. This list can get crufty if you have a long-running process that accesses objects and other processes are repacking. I think that's OK here; the parent repack process is not very long-lived.

I did, however, notice that the code we are calling has some problems of its own. :) Here are some fixes:

[1/3]: prune-packed: fix minor memory leak [2/3]: make update-server-info more robust [3/3]: server-info: clean up after writing info/packs

-Peff

#2 lore raw

PATCH 1/3 prune-packed: fix minor memory leak

We form all of our directories in a strbuf, but never release it.

---
 builtin/prune-packed.c | 1 +
 1 file changed, 1 insertion(+)
builtin/prune-packed.c +1 -0
index 6879468..d430731 100644
@@ -68,6 +68,7 @@ void prune_packed_objects(int opts)
 		rmdir(pathname.buf);
 	}
 	stop_progress(&progress);
+	strbuf_release(&pathname);
 }
 
 int cmd_prune_packed(int argc, const char **argv, const char *prefix)
Signature · 1 line
2.1.0.373.g91ca799
#3 lore raw

PATCH 2/3 make update-server-info more robust

In reply to Jeff King

Since "git update-server-info" may be called automatically as part of a push or a "gc --auto", we should be robust against two processes trying to update it simultaneously. However, we currently use a fixed tempfile, which means that two simultaneous writers may step on each other's toes and end up renaming junk into place.

Let's instead switch to using a unique tempfile via mkstemp. We do not want to use a lockfile here, because it's OK for two writers to simultaneously update (one will "win" the rename race, but that's OK; they should be writing the same information).

While we're there, let's clean up a few other things:

1. Detect write errors. Report them and abort the update if any are found.

2. Free path memory rather than leaking it (and clean up the tempfile when necessary).

3. Use the pathdup functions consistently rather than static buffers or manually calculated lengths.

This last one fixes a potential overflow of "infofile" in update_info_packs (e.g., by putting large junk into $GIT_OBJECT_DIRECTORY). However, this overflow was probably not an interesting attack vector for two reasons:

a. The attacker would need to control the environment to
do this, in which case it was already game-over.

b. During its setup phase, git checks that the directory
actually exists, which means it is probably shorter than PATH_MAX anyway.

Because both update_info_refs and update_info_packs share these same failings (and largely duplicate each other), this patch factors out the improved error-checking version into a helper function.

---
I guess point (b) may not apply on systems that have a really small PATH_MAX that does not reflect what you can actually create in the filesystem (Windows?). But I think point (a) still applies, so this is not really a big deal security-wise (though it is certainly a bugfix for such systems).

 server-info.c | 116 +++++++++++++++++++++++++++++++++++-----------------------
 1 file changed, 71 insertions(+), 45 deletions(-)
server-info.c +71 -45
index 9ec744e..d54a3d6 100644
@@ -4,45 +4,80 @@
 #include "commit.h"
 #include "tag.h"
 
-/* refs */
-static FILE *info_ref_fp;
+/*
+ * Create the file "path" by writing to a temporary file and renaming
+ * it into place. The contents of the file come from "generate", which
+ * should return non-zero if it encounters an error.
+ */
+static int update_info_file(char *path, int (*generate)(FILE *))
+{
+	char *tmp = mkpathdup("%s_XXXXXX", path);
+	int ret = -1;
+	int fd = -1;
+	FILE *fp = NULL;
+
+	safe_create_leading_directories(path);
+	fd = mkstemp(tmp);
+	if (fd < 0)
+		goto out;
+	fp = fdopen(fd, "w");
+	if (!fp)
+		goto out;
+	ret = generate(fp);
+	if (ret)
+		goto out;
+	if (fclose(fp))
+		goto out;
+	if (adjust_shared_perm(tmp) < 0)
+		goto out;
+	if (rename(tmp, path) < 0)
+		goto out;
+	ret = 0;
+
+out:
+	if (ret) {
+		error("unable to update %s: %s", path, strerror(errno));
+		if (fp)
+			fclose(fp);
+		else if (fd >= 0)
+			close(fd);
+		unlink(tmp);
+	}
+	free(tmp);
+	return ret;
+}
 
 static int add_info_ref(const char *path, const unsigned char *sha1, int flag, void *cb_data)
 {
+	FILE *fp = cb_data;
 	struct object *o = parse_object(sha1);
 	if (!o)
 		return -1;
 
-	fprintf(info_ref_fp, "%s	%s\n", sha1_to_hex(sha1), path);
+	if (fprintf(fp, "%s	%s\n", sha1_to_hex(sha1), path) < 0)
+		return -1;
+
 	if (o->type == OBJ_TAG) {
 		o = deref_tag(o, path, 0);
 		if (o)
-			fprintf(info_ref_fp, "%s	%s^{}\n",
-				sha1_to_hex(o->sha1), path);
+			if (fprintf(fp, "%s	%s^{}\n",
+				sha1_to_hex(o->sha1), path) < 0)
+				return -1;
 	}
 	return 0;
 }
 
+static int generate_info_refs(FILE *fp)
+{
+	return for_each_ref(add_info_ref, fp);
+}
+
 static int update_info_refs(int force)
 {
-	char *path0 = git_pathdup("info/refs");
-	int len = strlen(path0);
-	char *path1 = xmalloc(len + 2);
-
-	strcpy(path1, path0);
-	strcpy(path1 + len, "+");
-
-	safe_create_leading_directories(path0);
-	info_ref_fp = fopen(path1, "w");
-	if (!info_ref_fp)
-		return error("unable to update %s", path1);
-	for_each_ref(add_info_ref, NULL);
-	fclose(info_ref_fp);
-	adjust_shared_perm(path1);
-	rename(path1, path0);
-	free(path0);
-	free(path1);
-	return 0;
+	char *path = git_pathdup("info/refs");
+	int ret = update_info_file(path, generate_info_refs);
+	free(path);
+	return ret;
 }
 
 /* packs */
@@ -198,36 +233,27 @@ static void init_pack_info(const char *infofile, int force)
 		info[i]->new_num = i;
 }
 
-static void write_pack_info_file(FILE *fp)
+static int write_pack_info_file(FILE *fp)
 {
 	int i;
-	for (i = 0; i < num_pack; i++)
-		fprintf(fp, "P %s\n", info[i]->p->pack_name + objdirlen + 6);
-	fputc('\n', fp);
+	for (i = 0; i < num_pack; i++) {
+		if (fprintf(fp, "P %s\n", info[i]->p->pack_name + objdirlen + 6) < 0)
+			return -1;
+	}
+	if (fputc('\n', fp) == EOF)
+		return -1;
+	return 0;
 }
 
 static int update_info_packs(int force)
 {
-	char infofile[PATH_MAX];
-	char name[PATH_MAX];
-	int namelen;
-	FILE *fp;
-
-	namelen = sprintf(infofile, "%s/info/packs", get_object_directory());
-	strcpy(name, infofile);
-	strcpy(name + namelen, "+");
+	char *infofile = mkpathdup("%s/info/packs", get_object_directory());
+	int ret;
 
 	init_pack_info(infofile, force);
-
-	safe_create_leading_directories(name);
-	fp = fopen(name, "w");
-	if (!fp)
-		return error("cannot open %s", name);
-	write_pack_info_file(fp);
-	fclose(fp);
-	adjust_shared_perm(name);
-	rename(name, infofile);
-	return 0;
+	ret = update_info_file(infofile, write_pack_info_file);
+	free(infofile);
+	return ret;
 }
 
 /* public */
Signature · 1 line
2.1.0.373.g91ca799
#4 lore raw

Re: [PATCH 2/3] make update-server-info more robust

Am 13.09.2014 um 22:19 schrieb Jeff King:

Quoted 41 lines

Since "git update-server-info" may be called automatically as part of a push or a "gc --auto", we should be robust against two processes trying to update it simultaneously. However, we currently use a fixed tempfile, which means that two simultaneous writers may step on each other's toes and end up renaming junk into place.

Let's instead switch to using a unique tempfile via mkstemp. We do not want to use a lockfile here, because it's OK for two writers to simultaneously update (one will "win" the rename race, but that's OK; they should be writing the same information).

While we're there, let's clean up a few other things:

1. Detect write errors. Report them and abort the update if any are found.

2. Free path memory rather than leaking it (and clean up the tempfile when necessary).

3. Use the pathdup functions consistently rather than static buffers or manually calculated lengths.

This last one fixes a potential overflow of "infofile" in update_info_packs (e.g., by putting large junk into $GIT_OBJECT_DIRECTORY). However, this overflow was probably not an interesting attack vector for two reasons:

a. The attacker would need to control the environment to
do this, in which case it was already game-over.

b. During its setup phase, git checks that the directory
actually exists, which means it is probably shorter than PATH_MAX anyway.

Because both update_info_refs and update_info_packs share these same failings (and largely duplicate each other), this patch factors out the improved error-checking version into a helper function.

Quoted 5 lines

---
I guess point (b) may not apply on systems that have a really small PATH_MAX that does not reflect what you can actually create in the filesystem (Windows?).

It's the other way around: PATH_MAX is an actual limit basically only on Windows [1] unless you avoid using the Windows API [2].

Regardless of the security implications, getting rid of more PATH_MAX buffers is a good move.

And I looked only briefly at your patch, but I like the three bullet points above. :)

René

[1] http://insanecoding.blogspot.de/2007/11/pathmax-simply-isnt.html
[2] http://msdn.microsoft.com/en-us/library/windows/desktop/aa365247(v=vs.85).aspx

#5 lore raw

Re: [PATCH 2/3] make update-server-info more robust

In reply to Jeff King
Quoted 6 lines · Jeff King

Since "git update-server-info" may be called automatically as part of a push or a "gc --auto", we should be robust against two processes trying to update it simultaneously. However, we currently use a fixed tempfile, which means that two simultaneous writers may step on each other's toes and end up renaming junk into place.

Thanks. I'll queue these clean-ups but we may want to start thinking about deprecating and removing the dumb http support.

#6 lore raw

Re: [PATCH 2/3] make update-server-info more robust

Quoted 11 lines · Junio C Hamano
Quoted 6 lines · Jeff King

Since "git update-server-info" may be called automatically as part of a push or a "gc --auto", we should be robust against two processes trying to update it simultaneously. However, we currently use a fixed tempfile, which means that two simultaneous writers may step on each other's toes and end up renaming junk into place.

Thanks. I'll queue these clean-ups but we may want to start thinking about deprecating and removing the dumb http support.

Yeah, I have often thought about that (especially the push support, which has always been flaky and underused). However, some possible schemes for resumable clone could be easily implemented by shunting the cloner to a dumb-http conversation. So it may be worth keeping at least the fetch side around for the time being. Food for thought.

-Peff

#7 lore raw

PATCH 3/3 server-info: clean up after writing info/packs

In reply to Jeff King

We allocate pack information in a static global list but never clean it up. This leaks memory, and means that calling update_server_info twice will generate a buggy file (it will have duplicate entries).

---
 server-info.c | 9 +++++++++
 1 file changed, 9 insertions(+)
server-info.c +9 -0
index d54a3d6..31f4a74 100644
@@ -233,6 +233,14 @@ static void init_pack_info(const char *infofile, int force)
 		info[i]->new_num = i;
 }
 
+static void free_pack_info(void)
+{
+	int i;
+	for (i = 0; i < num_pack; i++)
+		free(info[i]);
+	free(info);
+}
+
 static int write_pack_info_file(FILE *fp)
 {
 	int i;
@@ -252,6 +260,7 @@ static int update_info_packs(int force)
 
 	init_pack_info(infofile, force);
 	ret = update_info_file(infofile, write_pack_info_file);
+	free_pack_info();
 	free(infofile);
 	return ret;
 }
Signature · 1 line
2.1.0.373.g91ca799
Reviewed-by on v1 1/1
#8 lore raw
In reply to René Scharfe
Quoted 5 lines · René Scharfe

Call the functions behind git prune-packed and git update-server-info directly instead of using run_command(). This is shorter, easier and quicker.

Thanks for cleaning up the literal rewrite of the shell script and making it look more like a C program.

Quoted 48 lines
---
 builtin/repack.c | 23 ++++++-----------------
 1 file changed, 6 insertions(+), 17 deletions(-)
builtin/repack.c +6 -17
index fc088db..2aae05d 100644
@@ -377,6 +377,7 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 	/* End of pack replacement. */
 
 	if (delete_redundant) {
+		int opts = 0;
 		sort_string_list(&names);
 		for_each_string_list_item(item, &existing_packs) {
 			char *sha1;
@@ -387,25 +388,13 @@ int cmd_repack(int argc, const char **argv, const char *prefix)
 			if (!string_list_has_string(&names, sha1))
 				remove_redundant_pack(packdir, item->string);
 		}
-		argv_array_push(&cmd_args, "prune-packed");
-		if (quiet)
-			argv_array_push(&cmd_args, "--quiet");
-
-		memset(&cmd, 0, sizeof(cmd));
-		cmd.argv = cmd_args.argv;
-		cmd.git_cmd = 1;
-		run_command(&cmd);
-		argv_array_clear(&cmd_args);
+		if (!quiet && isatty(2))
+			opts |= PRUNE_PACKED_VERBOSE;
+		prune_packed_objects(opts);
 	}
 
-	if (!no_update_server_info) {
-		argv_array_push(&cmd_args, "update-server-info");
-		memset(&cmd, 0, sizeof(cmd));
-		cmd.argv = cmd_args.argv;
-		cmd.git_cmd = 1;
-		run_command(&cmd);
-		argv_array_clear(&cmd_args);
-	}
+	if (!no_update_server_info)
+		update_server_info(0);
 	remove_temporary_files();
 	string_list_clear(&names, 0);
 	string_list_clear(&rollback, 0);