CVE-2019-9948: [2.7] bpo-35907: Avoid file reading as disallowing the unnecessary URL scheme in urllib (GH-11842) by push0ebp · Pull Request #11842 · python/cpython

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Conversation 29 Commits 6 Checks 0 Files changed

Conversation

Hello, and thanks for your contribution!

I’m a bot set up to make sure that the project can legally accept your contribution by verifying you have signed the PSF contributor agreement (CLA).

Our records indicate we have not received your CLA. For legal reasons we need you to sign this before we can look at your contribution. Please follow the steps outlined in the CPython devguide to rectify this issue.

If you have recently signed the CLA, please wait at least one business day
before our records are updated.

You can check yourself to see if the CLA has been received.

Thanks again for your contribution, we look forward to reviewing it!

push0ebp changed the title bpo-35907: Avoid file reading as disallowing the unnecessary URL scheme in urllib [2.7] bpo-35907: Avoid file reading as disallowing the unnecessary URL scheme in urllib (GH-11842)

Feb 13, 2019

Is this patch the accepted resolution for CVE-2019-9948? If so, when is it expected to be merged?

# bpo-35907: # disallow the file reading with the type not allowed

if not hasattr(self, name) or \

(self == _urlopener and name == ‘open_local_file’):

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please simplify the fix. There is no need to check for _urlopener here. Just block all access to local_file:// schema. We don’t need special cases for subclasses or special instances.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If it does not check the instance type of self, all of overridden open_local_file method will be blocked.
will be okay?
like this

 self.assertRaises(IOError, DummyURLopener().open
            'local_file://example')
 self.assertRaises(IOError, DummyURLopener().open
            'local-file://example')

I think that it needs to check open_local_file is overridden.
What do you think?

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think it has to expect this test result.

def test_local_file_open(self):
        class DummyURLopener(urllib.URLopener):
            def open_local_file(self, url):
                return url

        class DummyURLopener2(urllib.URLopener):
            def open_test(self, url):
                return url

        opener = DummyURLopener()
        opener2 = DummyURLopener2()

        for url in ('local-file://example', 'local_file://example'):    
            self.assertEqual(opener.open(url), '//example')
            self.assertRaises(IOError, urllib.urlopen, url)
            self.assertRaises(IOError, opener2.open, url)
        self.assertEqual(opener2.open('test://example'), '//example')

but Considering the overriding method and instance type check was complex.
If you have a better idea. let me know.

A Python core developer has requested some changes be made to your pull request before we can consider merging it. If you could please address their requests along with any other requests in other reviews from core developers that would be appreciated.

Once you have made the requested changes, please leave a comment on this pull request containing the phrase I have made the requested changes; please review again. I will then notify any core developers who have left a review that you’re ready for them to take another look at this pull request.

…sies in DummyURLopener test, and simplify mitigation

Copy link

Contributor Author

** push0ebp left a comment**

I have made the requested changes; please review again.

# bpo-35907: # disallow the file reading with the type not allowed

if not hasattr(self, name) or \

(self == _urlopener and name == ‘open_local_file’):

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If it does not check the instance type of self, all of overridden open_local_file method will be blocked.
will be okay?
like this

 self.assertRaises(IOError, DummyURLopener().open
            'local_file://example')
 self.assertRaises(IOError, DummyURLopener().open
            'local-file://example')

I think that it needs to check open_local_file is overridden.
What do you think?

# bpo-35907: # disallow the file reading with the type not allowed

if not hasattr(self, name) or \

(self == _urlopener and name == ‘open_local_file’):

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think it has to expect this test result.

def test_local_file_open(self):
        class DummyURLopener(urllib.URLopener):
            def open_local_file(self, url):
                return url

        class DummyURLopener2(urllib.URLopener):
            def open_test(self, url):
                return url

        opener = DummyURLopener()
        opener2 = DummyURLopener2()

        for url in ('local-file://example', 'local_file://example'):    
            self.assertEqual(opener.open(url), '//example')
            self.assertRaises(IOError, urllib.urlopen, url)
            self.assertRaises(IOError, opener2.open, url)
        self.assertEqual(opener2.open('test://example'), '//example')

but Considering the overriding method and instance type check was complex.
If you have a better idea. let me know.

# bpo-35907: # disallow the file reading with the type not allowed

if not hasattr(self, name) or \

getattr(self, name) == self.open_local_file:

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Another fix is to rename the open_local_file() as _open_local_file(). But… In Python 2.7, there is a risk that the method is called directly in 3rd party code for whatever reason :-(

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why do you compare function objects instead of simply name == "open_local_file"?

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@tiran
If someone rename open_local_file() method and also forgets to modify name == "open_local_file", it may be bypassed.
So I made the mitigation compare the object to raise syntax error.
Should I do change back?

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I concur with Christian, please test directly the name, remove getattr().

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@vstinner
I agree with you. I also considered renaming the open_local_file(). I think that this fix can guarantees availability to the developers who want to override this method. But I chose another fix to minimize side effects for third party codes.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Please remove getattr().

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Yes. I will do it. I want suggest you other fix.

        if not hasattr(self, name):
            if proxy:
                return self.open_unknown_proxy(proxy, fullurl, data)
            else:
                return self.open_unknown(fullurl, data)
        if name == 'open_local_file': 
                raise IOError, ('url error', 'invalid url type', urltype)
        try:

I think that it is better to raise exception with explicit error message for developers to overriding open_local_file().
they can easily recognize local_file is invalid type.
What do you think about it?
or should I use open_unknown()?

if not hasattr(self, name) or \
    name == 'open_local_file':

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

# bpo-35907: disallow the file reading with the type not allowed 
if not hasattr(self, name) or name == 'open_local_file': ...

LGTM.

        if name == 'open_local_file': 
                raise IOError, ('url error', 'invalid url type', urltype)

That’s wrong exception type, don’t do that.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have changed it. Please review.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have made the requested changes; please review again.

Copy link

Member

** vstinner left a comment**

Modify the object to string in check method name.

Copy link

Member

** vstinner left a comment**

LGTM. Just curious about the # which looks like a typo.

@@ -203,7 +203,9 @@ def open(self, fullurl, data=None):

name = ‘open_’ + urltype

self.type = urltype

name = name.replace('-', ‘_’)

if not hasattr(self, name):

# bpo-35907: # disallow the file reading with the type not allowed

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don’t understand the # in "# bpo-35907: # disallow …". Is it a typo?

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Oh Sorry, my mistake. I will correct it.

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I have fixed the typo.

Copy link

Contributor Author

** push0ebp commented May 22, 2019**

Thank you for your accept.
related to #13474