1
0
Fork 0
mirror of git://slackware.nl/current.git synced 2025-01-22 07:27:59 +01:00
slackware-current/testing/source/PAM/a/pam/fedora-patches/pam-1.3.1-unix-checksalt_syslog.patch
Patrick J Volkerding 6a63f41b3b Wed Feb 12 05:05:50 UTC 2020
Hey folks! PAM has finally landed in /testing. Some here wanted it to go
right into the main tree immediately, and in a more normal development cycle
I'd have been inclined to agree (it is -current, after all). But it's
probably better for it to appear in /testing first, to make sure we didn't
miss any bugs and also to serve as a warning shot that we'll be shaking up
the tree pretty good over the next few weeks. I'd like to see this merged
into the main tree in a day or two, so any testing is greatly appreciated.
Switching to the PAM packages (or reverting from them) is as easy as
installing all of them with upgradepkg --install-new, and if reverting then
remove the three leftover _pam packages. After reverting, a bit of residue
will remain in /etc/pam.d/ and /etc/security/ which can either be manually
deleted or simply ignored. While there are many more features available in
PAM compared with plain shadow, out of the box about the only noticable
change is the use of cracklib and libpwquality to check the quality of a
user-supplied password. Hopefully having PAM and krb5 will get us on track
to having proper Active Directory integration as well as using code paths
that are likely better audited these days. The attack surface *might* be
bigger, but it's also a lot better scrutinized.
Thanks to Robby Workman and Vincent Batts who did most of the initial heavy
lifting on the core PAM packages as a side project for many years. Thanks
also to Phantom X whose PAM related SlackBuilds were a valuable reference.
And thanks as well to ivandi - I learned a lot from the SlackMATE build
scripts and was even occasionally thankful for the amusing ways you would
kick my ass on LQ. ;-) You're more than welcome to let us know where we've
messed up this time.
The binutils and glibc packages in /testing were removed and are off the
table for now. I'm not seeing much upside to heading down that rabbit hole
at the moment. Next we need to be looking at Xfce 4.14 and Plasma 5.18 LTS
and some other things that have been held back since KDE4 couldn't use them.
Cheers! :-)
a/kernel-generic-5.4.19-x86_64-1.txz:  Upgraded.
a/kernel-huge-5.4.19-x86_64-1.txz:  Upgraded.
a/kernel-modules-5.4.19-x86_64-1.txz:  Upgraded.
a/lvm2-2.03.08-x86_64-1.txz:  Upgraded.
a/shadow-4.8.1-x86_64-2.txz:  Rebuilt.
  Automatically backup /etc/login.defs and install the new version if
  incompatible PAM options are detected.
d/kernel-headers-5.4.19-x86-1.txz:  Upgraded.
k/kernel-source-5.4.19-noarch-1.txz:  Upgraded.
  VALIDATE_FS_PARSER y -> n
xap/mozilla-thunderbird-68.5.0-x86_64-1.txz:  Upgraded.
  This release contains security fixes and improvements.
  For more information, see:
    https://www.mozilla.org/en-US/thunderbird/68.5.0/releasenotes/
    https://www.mozilla.org/en-US/security/advisories/mfsa2020-07/
    https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-6793
    https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-6794
    https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-6795
    https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-6797
    https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-6798
    https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-6792
    https://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2020-6800
  (* Security fix *)
isolinux/initrd.img:  Rebuilt.
kernels/*:  Upgraded.
testing/packages/PAM/ConsoleKit2-1.2.1-x86_64-1_pam.txz:  Added.
testing/packages/PAM/at-3.2.1-x86_64-1_pam.txz:  Added.
testing/packages/PAM/cifs-utils-6.10-x86_64-2_pam.txz:  Added.
testing/packages/PAM/cracklib-2.9.7-x86_64-1_pam.txz:  Added.
testing/packages/PAM/cups-2.3.1-x86_64-1_pam.txz:  Added.
testing/packages/PAM/cyrus-sasl-2.1.27-x86_64-2_pam.txz:  Added.
testing/packages/PAM/dovecot-2.3.9.2-x86_64-1_pam.txz:  Added.
testing/packages/PAM/gnome-keyring-3.34.0-x86_64-1_pam.txz:  Added.
testing/packages/PAM/hplip-3.19.12-x86_64-2_pam.txz:  Added.
testing/packages/PAM/kde-workspace-4.11.22-x86_64-6_pam.txz:  Added.
testing/packages/PAM/libcap-2.31-x86_64-1_pam.txz:  Added.
testing/packages/PAM/libcgroup-0.41-x86_64-5_pam.txz:  Added.
testing/packages/PAM/libpwquality-1.4.2-x86_64-1_pam.txz:  Added.
testing/packages/PAM/mariadb-10.4.12-x86_64-1_pam.txz:  Added.
testing/packages/PAM/netatalk-3.1.12-x86_64-2_pam.txz:  Added.
testing/packages/PAM/netkit-rsh-0.17-x86_64-2_pam.txz:  Added.
testing/packages/PAM/openssh-8.1p1-x86_64-1_pam.txz:  Added.
testing/packages/PAM/openvpn-2.4.8-x86_64-1_pam.txz:  Added.
testing/packages/PAM/pam-1.3.1-x86_64-1_pam.txz:  Added.
testing/packages/PAM/polkit-0.116-x86_64-1_pam.txz:  Added.
testing/packages/PAM/popa3d-1.0.3-x86_64-3_pam.txz:  Added.
testing/packages/PAM/ppp-2.4.7-x86_64-3_pam.txz:  Added.
testing/packages/PAM/proftpd-1.3.6b-x86_64-1_pam.txz:  Added.
testing/packages/PAM/samba-4.11.6-x86_64-1_pam.txz:  Added.
testing/packages/PAM/screen-4.8.0-x86_64-1_pam.txz:  Added.
testing/packages/PAM/shadow-4.8.1-x86_64-2_pam.txz:  Added.
testing/packages/PAM/sudo-1.8.31-x86_64-1_pam.txz:  Added.
testing/packages/PAM/system-config-printer-1.5.12-x86_64-2_pam.txz:  Added.
testing/packages/PAM/util-linux-2.35.1-x86_64-1_pam.txz:  Added.
testing/packages/PAM/vsftpd-3.0.3-x86_64-5_pam.txz:  Added.
testing/packages/PAM/xdm-1.1.11-x86_64-9_pam.txz:  Added.
testing/packages/PAM/xlockmore-5.62-x86_64-1_pam.txz:  Added.
testing/packages/PAM/xscreensaver-5.43-x86_64-1_pam.txz:  Added.
testing/packages/binutils-2.34-x86_64-1.txz:  Removed.
testing/packages/glibc-2.31-x86_64-1.txz:  Removed.
testing/packages/glibc-i18n-2.31-x86_64-1.txz:  Removed.
testing/packages/glibc-profile-2.31-x86_64-1.txz:  Removed.
testing/packages/glibc-solibs-2.31-x86_64-1.txz:  Removed.
usb-and-pxe-installers/usbboot.img:  Rebuilt.
2020-02-12 17:59:48 +01:00

73 lines
2.8 KiB
Diff

From 86eed7ca01864b9fd17099e57f10f2b9b6b568a1 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Bj=C3=B6rn=20Esser?= <besser82@fedoraproject.org>
Date: Mon, 26 Nov 2018 22:33:17 +0100
Subject: [PATCH] pam_unix: Report unusable hashes found by checksalt to
syslog.
libxcrypt can be build-time configured to support (or not support)
various hashing methods. Future versions will also have support for
runtime configuration by the system's vendor and/or administrator.
For that reason adminstrator should be notified by pam if users cannot
log into their account anymore because of such a change in the system's
configuration of libxcrypt.
Also check for malformed hashes, like descrypt hashes starting with
"$2...", which might have been generated by unsafe base64 encoding
functions as used in glibc <= 2.16.
Such hashes are likely to be rejected by many recent implementations
of libcrypt.
* modules/pam_unix/passverify.c (verify_pwd_hash): Report unusable
hashes found by checksalt to syslog.
---
modules/pam_unix/passverify.c | 36 +++++++++++++++++++++++++++++++++++
1 file changed, 36 insertions(+)
diff --git a/modules/pam_unix/passverify.c b/modules/pam_unix/passverify.c
index eb2444bb..2c808eb5 100644
--- a/modules/pam_unix/passverify.c
+++ b/modules/pam_unix/passverify.c
@@ -103,6 +103,42 @@ verify_pwd_hash(const char *p, char *hash, unsigned int nullok)
* Ok, we don't know the crypt algorithm, but maybe
* libcrypt knows about it? We should try it.
*/
+#if defined(CRYPT_CHECKSALT_AVAILABLE) && CRYPT_CHECKSALT_AVAILABLE
+ /* Get the status of the hash from checksalt */
+ int retval_checksalt = crypt_checksalt(hash);
+
+ /*
+ * Check for hashing methods that are disabled by
+ * libcrypt configuration and/or system preset.
+ */
+ if (retval_checksalt == CRYPT_SALT_METHOD_DISABLED) {
+ /*
+ * pam_syslog() needs a pam handle,
+ * but that's not available here.
+ */
+ helper_log_err(LOG_ERR,
+ "pam_unix(verify_pwd_hash): The method "
+ "for computing the hash \"%.6s\" has been "
+ "disabled in libcrypt by the preset from "
+ "the system's vendor and/or administrator.",
+ hash);
+ }
+ /*
+ * Check for malformed hashes, like descrypt hashes
+ * starting with "$2...", which might have been
+ * generated by unsafe base64 encoding functions
+ * as used in glibc <= 2.16.
+ * Such hashes are likely to be rejected by many
+ * recent implementations of libcrypt.
+ */
+ if (retval_checksalt == CRYPT_SALT_INVALID) {
+ helper_log_err(LOG_ERR,
+ "pam_unix(verify_pwd_hash): The hash \"%.6s\""
+ "does not use a method known by the version "
+ "of libcrypt this system is supplied with.",
+ hash);
+ }
+#endif
#ifdef HAVE_CRYPT_R
struct crypt_data *cdata;
cdata = malloc(sizeof(*cdata));