public inbox for ipfire-scm@lists.ipfire.org
 help / color / mirror / Atom feed
From: Michael Tremer <git@ipfire.org>
To: ipfire-scm@lists.ipfire.org
Subject: [git.ipfire.org] IPFire 2.x development tree branch, next, updated. be93af142b427357666fae8232974864015402b2
Date: Thu, 19 Nov 2026 10:16:28 +0000 (UTC)	[thread overview]
Message-ID: <4jVZBN6xMDz2xVM@people01.haj.ipfire.org> (raw)

This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "IPFire 2.x development tree".

The branch, next has been updated
       via  be93af142b427357666fae8232974864015402b2 (commit)
       via  51765afa3316d9e5cce3f7913fc4c6cfe74fb6d3 (commit)
       via  146d6a8845ccb2da73d8b5701ccde94b45ced76c (commit)
      from  14463de2b723b2c07d19f9a1abe4dd8cc0d9f3bc (commit)

Those revisions listed above that are new to this repository have
not appeared on any other notification email; so we list those
revisions in full, below.

- Log -----------------------------------------------------------------
commit be93af142b427357666fae8232974864015402b2
Author: Matthias Fischer <matthias.fischer@ipfire.org>
Date:   Thu Nov 19 00:10:11 2026 +0100

    bind: Update to 9.20.31
    
    For details see:
    
    https://downloads.isc.org/isc/bind9/9.20.31/doc/arm/html/notes.html#notes-for-bind-9-20-31
    
    "Notes for BIND 9.20.31
    Removed Features
    
        Deprecate NXDOMAIN redirection.
    
        The nxdomain-redirect option and type redirect zones are now
        deprecated. A warning is logged when they are encountered in
        named.conf. [GL #6451]
    
        Deprecate NSEC3 salt.
    
        BCP 236 (RFC 9276) recommends using neither extra NSEC3 iterations nor
        salt. Extra iterations have been disallowed for some time; a warning is
        now also logged when nsec3param in a dnssec-policy sets a non-zero
        salt-length. [GL !13036]
    
    Feature Changes
    
        Minimize ANY responses over TCP when minimal-any is enabled.
    
        When minimal-any was enabled, named minimized ANY responses over UDP
        but returned all available RRsets over TCP. It now returns a single
        RRset, with its covering RRSIGs when requested, and omits unnecessary
        authority and additional data regardless of transport. [GL !13056]
    
        Keep accepting CN-only certificates in dig +https with OpenSSL 4.1.
    
        When verifying the hostname of a DNS-over-HTTPS server, dig +https
        accepts a certificate whose only name is in the Subject Common Name.
        OpenSSL 4.1 no longer consults the Common Name, so when BIND 9 was
        built against it, such certificates were rejected with a peer
        verification failure. dig +https now asks OpenSSL 4.1 and newer to
        check the Common Name explicitly, so these certificates are accepted as
        with earlier OpenSSL versions. With OpenSSL 4.1 and newer, the Common
        Name is now also consulted for certificates that contain a
        SubjectAltName. [GL !12944]
    
    Bug Fixes
    
        Prevent the server from stalling when reopening or rolling the dnstap
        output.
    
        Previously, reopening or rolling the dnstap-output file, either with
        rndc dnstap or automatically at its size limit, stopped all of named's
        processing until whatever was already running, such as an RPZ zone
        update, had finished and the old output had been replaced. The output
        is now switched over while the server keeps running; the few dnstap
        messages produced during the switch itself are dropped and counted in
        the dnstap statistics. Logging to a size-limited output also no longer
        checks the file on disk for every message. [GL #6461]
    
        Prevent automatic interface scanning from stopping after a burst of
        address changes.
    
        When many addresses were removed and added again in quick succession,
        as happens with netplan apply, the kernel could drop some of the
        resulting interface change notifications. named then gave up on
        automatic interface scanning, logging automatic interface scanning
        terminated: not enough free resources, and stopped listening on some
        addresses until it was restarted. Lost notifications now trigger a
        fresh scan of the interfaces once the burst is over. named also asks
        the kernel to report only address changes and to queue more of them, so
        that notifications are lost less often. [GL #6389]
    
        Try the next primary server instead of aborting the zone transfer.
    
        If a misconfigured primary server returned, for example, a zone with
        too many records or a zone that failed verification, the secondary
        server could abort the zone transfer altogether instead of trying the
        next configured primary. This has been fixed. [GL #6352]
    
        Fix Offline KSK signing when the KSK and a ZSK share a key tag.
    
        With offline-ksk in dnssec-policy, the KSK and the ZSKs are generated
        separately, so a ZSK can have the same key tag as the KSK. named then
        mistook the ZSK for the KSK when signing the zone; this meant the
        DNSKEY RRset was left without the KSK signature from the imported SKR,
        or, depending on the signing algorithm, the import of the SKR failed
        with a not exact error. This has been fixed. [GL #5540]
    
        Sign NOTIFY messages retried over TCP with the configured TSIG key.
    
        When a NOTIFY message to an also-notify target with an explicit TSIG
        key failed over UDP and was retried over TCP, the key was lost. The
        retry then used the key from a matching server statement or, if there
        was none, was sent unsigned. This has been fixed. [GL #6240]
    
        Answer DS queries for CNAME names from the cache.
    
        Previously, a resolver answering a DS query for a name that was a CNAME
        (for example, www.example.com pointing to example.com) ignored the
        CNAME it had already cached and sent the query upstream every time, so
        such answers were never served from the cache. The cached CNAME is now
        used for DS queries as it is for any other type, while a DS record or a
        negative DS answer cached at the same name still takes precedence. [GL
        #6455]
    
        Omit unvalidated data from type-ANY cache responses.
    
        When answering a type-ANY query from the cache, named could include
        data that had not yet been validated. Such data is now only included in
        the response if the query has the CD bit set. [GL #5984]
    
        Fix slow rndc responses under a high request rate.
    
        The rndc command could run slowly if a large number of requests was
        made over a short period of time. This has been fixed. [GL #3698]
    
        Prevent rndc from accepting stale or mismatched signed responses.
    
        The rndc command could accept reflected requests as answers, as well as
        replayed responses. This has been fixed. [GL #6101]
    
        Make rndc -r work when rndc and named are from different builds.
    
        Previously, rndc -r was only guaranteed to display the correct result
        code when rndc and named were from the same build. This has been fixed.
        [GL #6102]"
    
    Signed-off-by: Matthias Fischer <matthias.fischer@ipfire.org>
    Signed-off-by: Michael Tremer <michael.tremer@ipfire.org>

commit 51765afa3316d9e5cce3f7913fc4c6cfe74fb6d3
Merge: 14463de2b 146d6a884
Author: Michael Tremer <michael.tremer@ipfire.org>
Date:   Thu Nov 19 10:15:44 2026 +0000

    Merge branch 'master' into next

-----------------------------------------------------------------------

Summary of changes:
 config/rootfiles/common/bind                       | 10 ++--
 .../207 => oldcore/206}/filelists/aarch64/linux    |  0
 .../206}/filelists/loongarch64/linux               |  0
 .../207 => oldcore/206}/filelists/riscv64/linux    |  0
 .../207 => oldcore/206}/filelists/x86_64/linux     |  0
 config/rootfiles/oldcore/206/update.sh             | 63 +++++++++++++++++++++-
 lfs/bind                                           |  4 +-
 7 files changed, 69 insertions(+), 8 deletions(-)
 copy config/rootfiles/{core/207 => oldcore/206}/filelists/aarch64/linux (100%)
 copy config/rootfiles/{core/207 => oldcore/206}/filelists/loongarch64/linux (100%)
 copy config/rootfiles/{core/207 => oldcore/206}/filelists/riscv64/linux (100%)
 copy config/rootfiles/{core/207 => oldcore/206}/filelists/x86_64/linux (100%)

Difference in files:
diff --git a/config/rootfiles/common/bind b/config/rootfiles/common/bind
index 6959e4233..2cb6d34b7 100644
--- a/config/rootfiles/common/bind
+++ b/config/rootfiles/common/bind
@@ -238,18 +238,18 @@ usr/bin/nsupdate
 #usr/include/ns/types.h
 #usr/include/ns/update.h
 #usr/include/ns/xfrout.h
-usr/lib/libdns-9.20.30.so
+usr/lib/libdns-9.20.31.so
 #usr/lib/libdns.la
 #usr/lib/libdns.so
-usr/lib/libisc-9.20.30.so
+usr/lib/libisc-9.20.31.so
 #usr/lib/libisc.la
 #usr/lib/libisc.so
-usr/lib/libisccc-9.20.30.so
+usr/lib/libisccc-9.20.31.so
 #usr/lib/libisccc.la
 #usr/lib/libisccc.so
-usr/lib/libisccfg-9.20.30.so
+usr/lib/libisccfg-9.20.31.so
 #usr/lib/libisccfg.la
 #usr/lib/libisccfg.so
-usr/lib/libns-9.20.30.so
+usr/lib/libns-9.20.31.so
 #usr/lib/libns.la
 #usr/lib/libns.so
diff --git a/config/rootfiles/oldcore/206/filelists/aarch64/linux b/config/rootfiles/oldcore/206/filelists/aarch64/linux
new file mode 120000
index 000000000..3a2532bc7
--- /dev/null
+++ b/config/rootfiles/oldcore/206/filelists/aarch64/linux
@@ -0,0 +1 @@
+../../../../common/aarch64/linux
\ No newline at end of file
diff --git a/config/rootfiles/oldcore/206/filelists/loongarch64/linux b/config/rootfiles/oldcore/206/filelists/loongarch64/linux
new file mode 120000
index 000000000..31b2bcffe
--- /dev/null
+++ b/config/rootfiles/oldcore/206/filelists/loongarch64/linux
@@ -0,0 +1 @@
+../../../../common/loongarch64/linux
\ No newline at end of file
diff --git a/config/rootfiles/oldcore/206/filelists/riscv64/linux b/config/rootfiles/oldcore/206/filelists/riscv64/linux
new file mode 120000
index 000000000..c8e8350ca
--- /dev/null
+++ b/config/rootfiles/oldcore/206/filelists/riscv64/linux
@@ -0,0 +1 @@
+../../../../common/riscv64/linux
\ No newline at end of file
diff --git a/config/rootfiles/oldcore/206/filelists/x86_64/linux b/config/rootfiles/oldcore/206/filelists/x86_64/linux
new file mode 120000
index 000000000..0615b5b9a
--- /dev/null
+++ b/config/rootfiles/oldcore/206/filelists/x86_64/linux
@@ -0,0 +1 @@
+../../../../common/x86_64/linux
\ No newline at end of file
diff --git a/config/rootfiles/oldcore/206/update.sh b/config/rootfiles/oldcore/206/update.sh
index 454bcf620..8a5515d44 100644
--- a/config/rootfiles/oldcore/206/update.sh
+++ b/config/rootfiles/oldcore/206/update.sh
@@ -26,13 +26,65 @@
 
 core=206
 
+exit_with_error() {
+	# Set last succesfull installed core.
+	echo $(($core-1)) > /opt/pakfire/db/core/mine
+	# force fsck at next boot, this may fix free space on xfs
+	touch /forcefsck
+	# don't start pakfire again at error
+	killall -KILL pak_update
+	/usr/bin/logger -p syslog.emerg -t ipfire \
+		"core-update-${core}: $1"
+	exit $2
+}
+
 # Remove old core updates from pakfire cache to save space...
 for (( i=1; i<=$core; i++ )); do
 	rm -f /var/cache/pakfire/core-upgrade-*-$i.ipfire
 done
 
+KVER="@KERNEL_RELEASE@"
+
+# Backup uEnv.txt if exist
+if [ -e /boot/uEnv.txt ]; then
+	cp -vf /boot/uEnv.txt /boot/uEnv.txt.org
+fi
+
+# Do some sanity checks prior to the kernel update
+case $(uname -r) in
+	*-ipfire*)
+		# Ok.
+		;;
+	*)
+		exit_with_error "ERROR cannot update. No IPFire Kernel." 1
+		;;
+esac
+
+# Check diskspace on root and size of boot
+ROOTSPACE=$( df / -Pk | sed "s| * | |g" | cut -d" " -f4 | tail -n 1 )
+if [ $ROOTSPACE -lt 200000 ]; then
+	exit_with_error "ERROR cannot update because not enough free space on root." 2
+fi
+BOOTSIZE=$( df /boot -Pk | sed "s| * | |g" | cut -d" " -f2 | tail -n 1 )
+if [ $BOOTSIZE -lt 100000 ]; then
+	exit_with_error "ERROR cannot update. BOOT partition is to small." 3
+fi
+
 # Stop services
 
+# Remove the old kernel
+rm -rvf \
+	/boot/System.map-* \
+	/boot/config-* \
+	/boot/ipfirerd-* \
+	/boot/initramfs-* \
+	/boot/vmlinuz-* \
+	/boot/uImage-* \
+	/boot/zImage-* \
+	/boot/uInit-* \
+	/boot/dtb-* \
+	/lib/modules
+
 # Remove files
 rm -rfv \
 	/etc/strongswan.d/charon/gcrypt.conf \
@@ -117,7 +169,6 @@ fi
 
 # Build initial ramdisks
 dracut --regenerate-all --force
-KVER="@KERNEL_RELEASE@"
 case "$(uname -m)" in
 	aarch64)
 		mkimage -A arm64 -T ramdisk -C lzma -d /boot/initramfs-${KVER}.img /boot/uInit-${KVER}
@@ -128,6 +179,16 @@ esac
 # This update needs a reboot...
 touch /var/run/need_reboot
 
+# Update kernel version in uEnv.txt
+if [ -e /boot/uEnv.txt ]; then
+	sed -i -e "s/KVER=.*/KVER=${KVER}/g" /boot/uEnv.txt
+fi
+
+# Call user update script (needed for some ARM boards)
+if [ -e /boot/pakfire-kernel-update ]; then
+	/boot/pakfire-kernel-update ${KVER}
+fi
+
 # Finish
 /etc/init.d/fireinfo start
 sendprofile
diff --git a/lfs/bind b/lfs/bind
index 08fd4e3cd..84e262ff8 100644
--- a/lfs/bind
+++ b/lfs/bind
@@ -25,7 +25,7 @@
 
 include Config
 
-VER        = 9.20.30
+VER        = 9.20.31
 
 THISAPP    = bind-$(VER)
 DL_FILE    = $(THISAPP).tar.xz
@@ -43,7 +43,7 @@ objects = $(DL_FILE)
 
 $(DL_FILE) = $(DL_FROM)/$(DL_FILE)
 
-$(DL_FILE)_BLAKE2 = 5ab0bba53c631817984d03a630f2f71a52a06a5a17bae34229af723e9acddcd13d12c4f40a90ecdb1967ce2fa8406977a84590985a143b38dc93c5894839424a
+$(DL_FILE)_BLAKE2 = 36bfa28837eb2c0d7aeaf93ed0c10589671e1d8447ef59e8c2b29c329752e0cbdd8604ed514f2cf2f3b5caa8362f7033c5a872de9ecd734b7a830b42ca472639
 
 install : $(TARGET)
 


hooks/post-receive
--
IPFire 2.x development tree


                 reply	other threads:[~2026-11-19 10:16 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=4jVZBN6xMDz2xVM@people01.haj.ipfire.org \
    --to=git@ipfire.org \
    --cc=ipfire-scm@lists.ipfire.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox