File: day_503__security_hole_part_5.mdwn

package info (click to toggle)
git-annex 8.20210223-2
  • links: PTS, VCS
  • area: main
  • in suites: bullseye
  • size: 68,764 kB
  • sloc: haskell: 70,359; javascript: 9,103; sh: 1,304; makefile: 212; perl: 136; ansic: 44
file content (19 lines) | stat: -rw-r--r-- 995 bytes parent folder | download | duplicates (6)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
Started testing that the security fix will build everywhere on
release day. This is being particularly painful for the android build,
which has very old libraries and needed http-client updated, with many
follow-on changes, and is not successfully building yet after 5 hours. 
I really need to finish deprecating the android build.

Pretty exhausted from all this, and thinking what to do about
external special remotes, I elaborated on an idea that Daniel Dent had
raised in discussions about vulnerability, and realized that git-annex
has a second, worse vulnerability. This new one could be used to trick a
git-annex user into decrypting gpg encrypted data that they had 
never stored in git-annex. The attacker needs to have control of both an
encrypted special remote and a git remote, so it's not an easy exploit to
pull off, but it's still super bad.

This week is going to be a lot longer than I thought, and it's already
feeling kind of endless..

[[!meta date="June 19 2018 8:00 pm"]]