Bug 1577910

Use the following template to for the 'fedpkg update' request to submit an
update for this issue as it contains the top-level parent bug(s) as well as
this tracking bug.  This will ensure that all associated bugs get updated
when new packages are pushed to stable.

=====

# bugfix, security, enhancement, newpackage (required)
type=security

# testing, stable
request=testing

# Bug numbers: 1234,9876
bugs=1577906,1577909,1577910

# Description of your update
notes=Security fix for [PUT CVEs HERE]

# Enable request automation based on the stable/unstable karma thresholds
autokarma=True
stable_karma=3
unstable_karma=-3

# Automatically close bugs when this marked as stable
close_bugs=True

# Suggest that users restart after update
suggest_reboot=False

======

Additionally, you may opt to use the bodhi web interface to submit updates:

https://bodhi.fedoraproject.org/updates/new

I'm not sure what evolution is supposed to do with it. The code itself resides in evolution-data-server/src/camel/, but that's only a nitpick. More importantly, there is called gpg in the background for PGP and there's used NSS for S/MIME. I'd  need an advice from someone whom knows these two enough to tell me whether there's any option/argument to avoid that CFB/CBC issue without breaking interoperability with other clients, but even that might not be enough, because it can be asked to decrypt old messages or messages sent by other unpatched clients. Thus I'm afraid this cannot be fixed in evolution, but in GPG and NSS first and then, if needed, a change in Evolution(-data-server) would be done.

Note that the https://efail.de/ itself claims that "The CBC/CFB Gadget Attack" is a problem in the specification of OpenPGP and S/MIME, not in the clients.

From the practical point of view, that "Direct Exfiltration" by prepending some unclosed HTML tag before encrypted text and then add another part at the end to close the unfinished tag would not work in evolution, because it first splits the parts into separates entities and works with them independently. But even if it did it, then the remote content is downloaded only if the user lets Evolution download it. It's a little different with DNS prefetching, because it's done in WebKitGTK+. Evolution does disable DNS prefetching for years, though had been an issue in WebKitGTK+ which could allow enabling the DNS prefetching in the HTML code, but that had been fixed more than two months ago in:
https://bugs.webkit.org/show_bug.cgi?id=182924

From the official statement on this of the GPG folks [1], there's no problem as long as the user uses new-enough gpg (all is fine in Fedora), and when the client properly rejects showing the modified decrypted content, which in case of evolution it does reject (which also the EFail page indicates with their list of "buggy email clients" at the end of https://efail.de in the section "Attacks on PGP clients:", there's Evolution as "green").

Thus there is really nothing to be done with GPG for Evolution.

S/MIME is still opened, I'm waiting for the NSS/Mozilla folks statement.

[1] https://lists.gnupg.org/pipermail/gnupg-users/2018-May/060334.html

I'm closing this, the S/MIME part is lower than in evolution-data-server, there's a protocol problem.