Created attachment 1929393 [details] screenshot Description of problem: It is not possible to login to RHEL when p11-uri is specified in sssd.conf and smart card is required for auth (i.e. authselect profile is 'sssd with-smartcard with-smartcard required'). This applies for GDM as well as console login. Version-Release number of selected component (if applicable): RHEL-8.7.0-20221012.5 sssd-2.7.3-4.el8.x86_64 How reproducible: always Steps to Reproduce: 1. get fresh system; smart card and reader 2. `useradd testuser` 3. configure system for smart-card auth: 3.1: create sssd.conf file with matchrule for user 3.2: get p11_uri (p11tool --list-tokens) of card and add it to sssd.conf 3.3: add rootCA.cert to /etc/sssd/pki/sssd_auth_ca_db.pem 3.4: `chmod 0700 /etc/sssd` `chmod 0600 /etc/sssd/sssd.conf` `chmod 0644 /etc/sssd/pki/sssd_auth_ca_db.pem` `rm -rf /var/lib/sss/{mc,db}/*` 3.5: `systemctl restart sssd` 3.6: `authselect select sssd with-smartcard with-smartcard-required --force` 4. Open GDM login screen 5. Insert card into the reader Actual results: 1) system reognises smart card 2) user can pick certificate object 3) user can insert pin 4) "Sorry, smart card authentication didn't work. Please try again." appears in GDM (see attached screenshot) Expected results: 1) system reognises smart card 2) user can pick certificate object 3) user can insert pin 4) login is successful Additional info: This bug is present on both RHEL-8 and RHEL-9. It is manifested in GDM and also in console login. It is possible that this scenario was always broken as it was most likely tested only with "with-smartcard" profile of sssd.
Created attachment 1929406 [details] p11_child.log
Created attachment 1929409 [details] sssd_pam.log
Created attachment 1929410 [details] sssd_shadowutils.log
Hi, there are various odd items in the logs. Can you set 'debug_level = 9' in the [domain/...] section as well? There is '[pam] [child_sig_handler] (0x0020): [CID#14] child [9412] was terminated by signal [11].', can you check, e.g. with coredumpctl if there is a core dump and sent the backtrace? Can you send your ssssd.conf as well? Does the test work if 'p11_uri' is not set in sssd.conf? bye, Sumit
Pushed PR: https://github.com/SSSD/sssd/pull/6480 * `master` * aac303e84b71325d3c45fa7a22f83f7f54d4b7a2 - p11: fix size of argument array * `sssd-2-8` * 20037ae5354a874f04802844c930c6b52704c5c7 - p11: fix size of argument array
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 (sssd 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/RHBA-2023:2986