https://github.com/git/git
Revision 0383bbb9015898cbc79abd7b64316484d7713b44 authored by Jeff King on 30 April 2018, 07:25:25 UTC, committed by Jeff King on 22 May 2018, 03:50:11 UTC
Submodule "names" come from the untrusted .gitmodules file, but we blindly append them to $GIT_DIR/modules to create our on-disk repo paths. This means you can do bad things by putting "../" into the name (among other things). Let's sanity-check these names to avoid building a path that can be exploited. There are two main decisions: 1. What should the allowed syntax be? It's tempting to reuse verify_path(), since submodule names typically come from in-repo paths. But there are two reasons not to: a. It's technically more strict than what we need, as we really care only about breaking out of the $GIT_DIR/modules/ hierarchy. E.g., having a submodule named "foo/.git" isn't actually dangerous, and it's possible that somebody has manually given such a funny name. b. Since we'll eventually use this checking logic in fsck to prevent downstream repositories, it should be consistent across platforms. Because verify_path() relies on is_dir_sep(), it wouldn't block "foo\..\bar" on a non-Windows machine. 2. Where should we enforce it? These days most of the .gitmodules reads go through submodule-config.c, so I've put it there in the reading step. That should cover all of the C code. We also construct the name for "git submodule add" inside the git-submodule.sh script. This is probably not a big deal for security since the name is coming from the user anyway, but it would be polite to remind them if the name they pick is invalid (and we need to expose the name-checker to the shell anyway for our test scripts). This patch issues a warning when reading .gitmodules and just ignores the related config entry completely. This will generally end up producing a sensible error, as it works the same as a .gitmodules file which is missing a submodule entry (so "submodule update" will barf, but "git clone --recurse-submodules" will print an error but not abort the clone. There is one minor oddity, which is that we print the warning once per malformed config key (since that's how the config subsystem gives us the entries). So in the new test, for example, the user would see three warnings. That's OK, since the intent is that this case should never come up outside of malicious repositories (and then it might even benefit the user to see the message multiple times). Credit for finding this vulnerability and the proof of concept from which the test script was adapted goes to Etienne Stalmans. Signed-off-by: Jeff King <peff@peff.net>
1 parent 42e6fde
Tip revision: 0383bbb9015898cbc79abd7b64316484d7713b44 authored by Jeff King on 30 April 2018, 07:25:25 UTC
submodule-config: verify submodule names as paths
submodule-config: verify submodule names as paths
Tip revision: 0383bbb
connected.c
#include "cache.h"
#include "run-command.h"
#include "sigchain.h"
#include "connected.h"
#include "transport.h"
/*
* If we feed all the commits we want to verify to this command
*
* $ git rev-list --objects --stdin --not --all
*
* and if it does not error out, that means everything reachable from
* these commits locally exists and is connected to our existing refs.
* Note that this does _not_ validate the individual objects.
*
* Returns 0 if everything is connected, non-zero otherwise.
*/
int check_connected(sha1_iterate_fn fn, void *cb_data,
struct check_connected_options *opt)
{
struct child_process rev_list = CHILD_PROCESS_INIT;
struct check_connected_options defaults = CHECK_CONNECTED_INIT;
char commit[41];
unsigned char sha1[20];
int err = 0;
struct packed_git *new_pack = NULL;
struct transport *transport;
size_t base_len;
if (!opt)
opt = &defaults;
transport = opt->transport;
if (fn(cb_data, sha1)) {
if (opt->err_fd)
close(opt->err_fd);
return err;
}
if (transport && transport->smart_options &&
transport->smart_options->self_contained_and_connected &&
transport->pack_lockfile &&
strip_suffix(transport->pack_lockfile, ".keep", &base_len)) {
struct strbuf idx_file = STRBUF_INIT;
strbuf_add(&idx_file, transport->pack_lockfile, base_len);
strbuf_addstr(&idx_file, ".idx");
new_pack = add_packed_git(idx_file.buf, idx_file.len, 1);
strbuf_release(&idx_file);
}
if (opt->shallow_file) {
argv_array_push(&rev_list.args, "--shallow-file");
argv_array_push(&rev_list.args, opt->shallow_file);
}
argv_array_push(&rev_list.args,"rev-list");
argv_array_push(&rev_list.args, "--objects");
argv_array_push(&rev_list.args, "--stdin");
argv_array_push(&rev_list.args, "--not");
argv_array_push(&rev_list.args, "--all");
argv_array_push(&rev_list.args, "--quiet");
if (opt->progress)
argv_array_pushf(&rev_list.args, "--progress=%s",
_("Checking connectivity"));
rev_list.git_cmd = 1;
rev_list.env = opt->env;
rev_list.in = -1;
rev_list.no_stdout = 1;
if (opt->err_fd)
rev_list.err = opt->err_fd;
else
rev_list.no_stderr = opt->quiet;
if (start_command(&rev_list))
return error(_("Could not run 'git rev-list'"));
sigchain_push(SIGPIPE, SIG_IGN);
commit[40] = '\n';
do {
/*
* If index-pack already checked that:
* - there are no dangling pointers in the new pack
* - the pack is self contained
* Then if the updated ref is in the new pack, then we
* are sure the ref is good and not sending it to
* rev-list for verification.
*/
if (new_pack && find_pack_entry_one(sha1, new_pack))
continue;
memcpy(commit, sha1_to_hex(sha1), 40);
if (write_in_full(rev_list.in, commit, 41) < 0) {
if (errno != EPIPE && errno != EINVAL)
error_errno(_("failed write to rev-list"));
err = -1;
break;
}
} while (!fn(cb_data, sha1));
if (close(rev_list.in))
err = error_errno(_("failed to close rev-list's stdin"));
sigchain_pop(SIGPIPE);
return finish_command(&rev_list) || err;
}
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...