Don't freeze cgroup on update for systemd cgroup v2#3067
Don't freeze cgroup on update for systemd cgroup v2#3067cyphar merged 2 commits intoopencontainers:masterfrom
Conversation
3f36eeb to
125f33b
Compare
|
Indeed, I reinstated the freeze because of #3014, which is v1 only bug. |
kolyshkin
left a comment
There was a problem hiding this comment.
Second commit looks good; I don't like code duplication in the first one.
Run device update tests on cgroup v2, and add a test verifying that we don't allow access to devices when we don't intend to. Signed-off-by: Odin Ugedal <odin@uged.al>
Since device updates in cgroup v2 are atomic for systemd, there is no need to freeze the processes before running the updates. Signed-off-by: Odin Ugedal <odin@uged.al>
125f33b to
f33be7c
Compare
|
Thanks @odinuge for the update. I treat this as an improvement (rather than a bug fix), and so this is not a candidate for 1.0 backport. Let me know if you disagree. |
|
@cyphar @AkihiroSuda @mrunalp PTAL |
|
OK, I found out this is actually a bug fix -- it fixes the inability to freeze the container/cgroup via cgroup manager's While I find using Set for freezing cgroup questionable (as the freezer is kind of special -- and so we have This PR fixes the ability to freeze systemd/v2 cgroup via So, this is a bug, but I still don't think it calls for 1.0 backport. |
1.0 backport: #3092