Add `--` in some commands that receive user input
and if interpreted as options could lead to remote
code execution (RCE).

There may be more commands that could benefit from `--`
so the input is never interpreted as an option,
but most of those aren't dangerous.

Fixed commands:

- push
- pull
- fetch
- clone/clone_from and friends
- archive (not sure if this one can be exploited, but it doesn't hurt
  adding `--` :))

For anyone using GitPython and exposing any of the GitPython methods to users,
make sure to always validate the input (like if starts with `--`).
And for anyone allowing users to pass arbitrary options, be aware
that some options may lead fo RCE, like `--exc`, `--upload-pack`,
`--receive-pack`, `--config` (gitpython-developers#1516).

Ref gitpython-developers#1517

All versions of package gitpython are vulnerable to Remote Code Execution
(RCE) due to improper user input validation, which makes it possible to
inject a maliciously crafted remote URL into the clone command. Exploiting
this vulnerability is possible because the library makes external calls to
git without sufficient sanitization of input arguments.

CVE: CVE-2022-24439

Upstream-Status: Backport

Reference:
gitpython-developers/GitPython#1529
gitpython-developers/GitPython#1518
gitpython-developers/GitPython#1521

(From OE-Core rev: 55f93e3786290dfa5ac72b5969bb2793f6a98bde)

Signed-off-by: Narpat Mali <narpat.mali@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>

Source: poky
MR: 124663
Type: Integration
Disposition: Merged from poky
ChangeID: 0721360
Description:

All versions of package gitpython are vulnerable to Remote Code Execution
(RCE) due to improper user input validation, which makes it possible to
inject a maliciously crafted remote URL into the clone command. Exploiting
this vulnerability is possible because the library makes external calls to
git without sufficient sanitization of input arguments.

CVE: CVE-2022-24439

Upstream-Status: Backport

Reference:
gitpython-developers/GitPython#1529
gitpython-developers/GitPython#1518
gitpython-developers/GitPython#1521

(From OE-Core rev: 55f93e3786290dfa5ac72b5969bb2793f6a98bde)

Signed-off-by: Narpat Mali <narpat.mali@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Jeremy A. Puhlman <jpuhlman@mvista.com>

This other GitCommandError on Windows is not related to
IndexFile.from_tree whose 8 related failing tests were marked
xfail in the preceding commit.

Also, test_clone_command_injection should not be confused with
test_clone_from_command_injection, which passes on all platforms.

The problem here appears to be that, on Windows, the path of the
directory GitPython is intended to clone to (when the possible
security vulnerability this test checks for is *absent*) is not
valid. Although this suggest the bug may only be in the test and
that the code under test may be working on Windows, but the test
does not establish that, for which it would need to test with a
payload clearly capable of creating a file unexpected_path points
to when run on its own. I am unsure if that is the case, given
that the "touch" command is used.

This doesn't appear to be reported as a bug, but some general
context about the implementation can be examined in gitpython-developers#1518 where it
was introduced, and gitpython-developers#1531 where it was modified.

This other GitCommandError on Windows is not related to
IndexFile.from_tree whose 8 related failing tests were marked xfail
in the preceding commit.

Also, test_clone_command_injection should not be confused with
test_clone_from_command_injection, which passes on all platforms.

The problem here appears to be that, on Windows, the path of the
directory GitPython is intended to clone to -- when the possible
security vulnerability this test checks for is absent -- is not
valid. This suggests the bug may only be in the test and that the
code under test may be working on Windows. But the test does not
establish that, for which it would need to test with a payload
clearly capable of creating the file unexpected_path refers to when
run on its own. (The "\" characters in the path seem to be treated
as escape characters rather than literally. Also, "touch" is not a
native Windows command, and the "touch" command in Git for Windows
maps disallowed occurrences of ":" in filenames to a separate code
point in the Private Use Area of the Basic Multilingual Plane.)

This doesn't currently seem to be reported as a bug, but some
general context about the implementation can be examined in gitpython-developers#1518
where it was introduced, and gitpython-developers#1531 where it was modified.

This other GitCommandError on Windows is not related to
IndexFile.from_tree whose 8 related failing tests were marked xfail
in the preceding commit.

Also, test_clone_command_injection should not be confused with
test_clone_from_command_injection, which passes on all platforms.

The problem here appears to be that, on Windows, the path of the
directory GitPython is intended to clone to -- when the possible
security vulnerability this test checks for is absent -- is not
valid. This suggests the bug may only be in the test and that the
code under test may be working on Windows. But the test does not
establish that, for which it would need to test with a payload
clearly capable of creating the file unexpected_path refers to when
run on its own.

This doesn't currently seem to be reported as a bug, but some
general context about the implementation can be examined in gitpython-developers#1518
where it was introduced, and gitpython-developers#1531 where it was modified.