Bug 1868877

Summary: Enabling the smack feature on the host may cause the container to fail to start
Product: Red Hat Enterprise Linux 8 Reporter: Simon <simon.wy>
Component: systemdAssignee: systemd-maint
Status: CLOSED ERRATA QA Contact: Frantisek Sumsal <fsumsal>
Severity: medium Docs Contact:
Priority: unspecified    
Version: 8.0CC: jamacku, systemd-maint-list
Target Milestone: rcKeywords: FutureFeature
Target Release: 8.0   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: systemd-239-42.el8 Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2021-05-18 14:53:56 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:

Description Simon 2020-08-14 06:17:42 UTC
For security reasons, we enabled the smack feature on the host machine (for example, adding: security=smack in grub), but the customer's container did not set the "container=docker" environment variable, which caused the container to fail to start.

We observe the following code:

static const MountPoint mount_table[] = {
        { "sysfs",       "/sys",                      "sysfs",      NULL,                      MS_NOSUID|MS_NOEXEC|MS_NODEV,
          NULL,          MNT_FATAL|MNT_IN_CONTAINER },
…
#if ENABLE_SMACK
        { "smackfs",     "/sys/fs/smackfs",           "smackfs",    "smackfsdef=*",            MS_NOSUID|MS_NOEXEC|MS_NODEV,
          mac_smack_use, MNT_FATAL                  },
        { "tmpfs",       "/dev/shm",                  "tmpfs",      "mode=1777,smackfsroot=*", MS_NOSUID|MS_NODEV|MS_STRICTATIME,
          mac_smack_use, MNT_FATAL                  },
#endif
...
}

static int mount_one(const MountPoint *p, bool relabel) {
…
        /* Skip securityfs in a container */
        if (!(p->mode & MNT_IN_CONTAINER) && detect_container(NULL) > 0)
                return 0;
…
}

MNT_IN_CONTAINER is not set in the smackfs item, but it is still mounted in the container. The reason is that the "container=docker" environment variable is not set, which causes the detect_container to fail to detect the container environment.

We need to make some enhancements to the detect_container() function. Even if the "container=docker" environment variable is not set, we hope it can still detect the container environment and enhance the robustness of systemd.


Finally, we found that the following patch can solve this problem:
https://github.com/systemd/systemd/commit/342bed02084c4396dd2f1054bd559bfb2699cfcb

Comment 3 Lukáš Nykrýn 2020-11-02 13:44:34 UTC
fix merged to github master branch -> https://github.com/systemd-rhel/rhel-8/pull/115

Comment 4 Lukáš Nykrýn 2020-11-02 14:55:25 UTC
fix merged to github master branch -> https://github.com/systemd-rhel/rhel-8/pull/128

Comment 10 errata-xmlrpc 2021-05-18 14:53:56 UTC
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (Moderate: systemd security, bug fix, and enhancement update), and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2021:1611