We introduced the gdm-fingerprint.pam in 9d41fe6fcc.
We used the [upstream Arch config] as a template, which contains an extended control field that jumps over **one** immediately-following `auth` rule unless `pam_gdm.so` succeeds.
But we decided to not include `pam_gnome_keyring.so` so there was no rule to skip over, resulting in a broken control flow and the PAM module failing with “PAM bad jump in stack”, breaking the fingerprint authentication in GDM.
Let’s actually add `pam_gnome_keyring.so`, like the Arch config does. Because we are creating the PAM file using the `text` option, `security.pam.services.gdm-fingerprint.enableGnomeKeyring` does not do anything so we need to do it manually.
For the case where gnome-keyring is not enabled, we could add a no-op rule like `optional pam_permit.so` after `pam_gdm.so` so that the branching always has something to jump over but it will be simpler to just make the both conditional. There are no further `auth` rules that could benefit from `pam_gdm.so` doing something so it should be fine.
Unlike in Arch, we are not going to invoke `pam_gnome_keyring.so` in a `session` rule since that is already done by the included `login` module.
[upstream Arch config]: 81ee658c11/data/pam-arch/gdm-fingerprint.pam
The `optional pam_permit.so` comes from the [upstream Arch config] we used as a template in 9d41fe6fcc. But I do not think it does anything in this position – see also the discussion at https://bbs.archlinux.org/viewtopic.php?id=245892 – so let’s just remove it.
Let’s also add a comment about disabling `fprintAuth` and a blank line for clarity.
[upstream Arch config]: 81ee658c11/data/pam-arch/gdm-fingerprint.pam
My reasons following Mint are:
1. Geary signed https://stopthemingmy.app, per request we shouldn't pre-ship it under a themed desktop environment.
See also b7937b4509
2. Hexchat is still gtk2 and is not maintained anymore, Mint encourages switching to Matrix instead.
See also https://blog.linuxmint.com/?p=4675 ("Joining the Matrix")
- rename hardware.opengl to hardware.graphics
- remove hardware.opengl.driSupport, which does nothing
- remove hardware.opengl.setLdLibraryPath, which should never be done
- rename hardware.opengl.driSupport32Bit to hardware.graphics.enable32Bit
- lost of small docs / formatting cleanups
