mirror of
https://github.com/netblue30/firejail.git
synced 2026-05-15 14:16:14 -06:00
[GH-ISSUE #6901] "Capability-limited" firejail packages to limit SUID privilege escalation risks #3411
Labels
No labels
LTS merge
LTS merge
bug
bug
converted-to-discussion
doc-todo
documentation
duplicate
enhancement
file-transfer
firecfg
firejail-in-firejail
firetools
graphics
help wanted
information_old
installation
invalid
modif
moved
needinfo
networking
notabug
notourbug
old-version
overlayfs
packaging
profile-request
pull-request
question
question_old
removal
runtime-permissions
sandbox-ipc
security
stale
wiki
wiki
wontfix
wordpress
workaround
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference: github-starred/firejail#3411
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @jerabaul29 on GitHub (Sep 19, 2025).
Original GitHub issue: https://github.com/netblue30/firejail/issues/6901
Some people / admins dont want to install firejail because it is a large SUID application and the associated perceived risk of privileges escalation, especially on many-user systems where not all users are fully trusted. Whether we agree with this viewpoint or not is another question, but this is the situation in practice.
My question: if I understand correctly, the risk of privileges escalation comes in theory typically from "misuse" of firejail and bug triggering, most likely this in theory could happen with weird / unusual / unexpected and specially crafted series of firejail calls with weird flag uses etc. To limit this, would it be doable / reasonable to provide for example (one large-encompassing or many small independent) debian packets that internally use firejail in a "constrained / safe(r) / standard / non malicious" way but without letting the user run arbitrary firejail commands themselves, that admins would be ok to install, and that may cover quite a few usecases?
For example, a common use is to run an app / a script with no network -
firejail --net=none CMD. Would it make sense to have firejail installed on the system but not usable per se by the user, instead the user would be only able to use e.g.firejail-netnonelikefirejail-netnone CMD, that defers to usingfirejailunder the hood to restrict network, butfirejailper se can not be called by the user? Would this be doable and reduce the risk of privilege escalation, since the user cannot do weird combinations offirejailcommands and flags, only use it in a few standard and well understood ways?Could providing this kind of packets that do not provide the possibility to run arbitrary
firejailcommands but that can run "under the hood"firejailwith default profiles for know apps, and / or a few typical cases (e.g. runfirejailwith either / or a combination of no net or only net allowed to a few IPs, no filesystem or only filesystem with an explicit whitelist of read only / read write files) be an option, either for the firejail project itself, or some downstream project?I dont know if there would be a way to provide this in a convenient way?
@rusty-snake commented on GitHub (Sep 19, 2025):
To all questions: Yes, but somebody would need to submit (large) patches.
Related: #5157, #510,