2018-08-03 02:56:22 +05:30
|
|
|
|
Change Log
|
|
|
|
|
|
==========
|
|
|
|
|
|
|
|
|
|
|
|
All relevant changes to the project are documented in this file.
|
|
|
|
|
|
|
|
|
|
|
|
[v1.6][UNRELEASED]
|
|
|
|
|
|
------------------
|
|
|
|
|
|
|
|
|
|
|
|
### Changes
|
|
|
|
|
|
- IPv6 support forward ported from FreeBSD, by John Haxby <john.haxby@oracle.com>
|
2018-08-03 04:28:00 +05:30
|
|
|
|
- Built-in log rotation support from BusyBox syslogd, disabled by default
|
2018-09-24 23:35:19 +05:30
|
|
|
|
- Automatic compression (gzip) of rotated files from .1
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Only read /etc/services when needed, by Martin Schulze <joey@infodrom.org>
|
|
|
|
|
|
- Improved sleep/alarm/mark implementation,
|
|
|
|
|
|
by Alan Jenkins <alan-jenkins@tuffmail.co.uk>
|
|
|
|
|
|
- Move hostname setting code from `main()` into `init()`,
|
|
|
|
|
|
by Thomas Jarosch <thomas.jarosch@intra2net.com>
|
|
|
|
|
|
- Documentation update by Martin Schulze <joey@infodrom.org>
|
2018-09-24 23:35:19 +05:30
|
|
|
|
- Reindent code to Linux KNF
|
|
|
|
|
|
- Touch PID file on `SIGHUP`, for integration with Finit
|
|
|
|
|
|
- Add systemd unit files
|
|
|
|
|
|
- Add GNU configure & build system
|
|
|
|
|
|
- Add configure flags to enable features and control behavior
|
|
|
|
|
|
- Detect systemd PATHs
|
2018-08-03 02:56:22 +05:30
|
|
|
|
|
|
|
|
|
|
### Fixes
|
|
|
|
|
|
- Correct continuation line problems on 64bit architecture,
|
|
|
|
|
|
by David Couture <glowplugrelayw0rks@gmail.com>
|
|
|
|
|
|
- Bugfix against invalid PRI values (CVE-2014-3634), by mancha <mancha1@zoho.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
[v1.5.1][] - 2014-10-06
|
|
|
|
|
|
-----------------------
|
|
|
|
|
|
|
|
|
|
|
|
### Fixes
|
|
|
|
|
|
- Bugfix against invalid PRI values (CVE-2014-3634), by mancha <mancha1@zoho.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
[v1.5][] - 2007-07-27
|
|
|
|
|
|
---------------------
|
|
|
|
|
|
|
|
|
|
|
|
- Dmitry V. Levin <ldv@altlinux.org>
|
2005-01-16 00:44:21 +05:30
|
|
|
|
- Close file descriptor in FindSymbolFile() in ksym.c in order not to
|
|
|
|
|
|
leak file descriptors.
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Solar Designer <solar@openwall.com>
|
2005-01-16 00:44:21 +05:30
|
|
|
|
- improve crunch_list()
|
|
|
|
|
|
- Prevent potential buffer overflow in reading messages from the
|
|
|
|
|
|
kernel log ringbuffer.
|
|
|
|
|
|
- Ensure that "len" is not placed in a register, and that the
|
|
|
|
|
|
endtty() signal handler is not installed too early which could
|
|
|
|
|
|
cause a segmentation fault or worse.
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Steve Grubb <linux_4ever@yahoo.com>
|
2005-01-16 00:44:21 +05:30
|
|
|
|
- fix memory calculation in crunch_list()
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Martin Schulze <joey@infodrom.org>
|
2005-01-16 00:44:21 +05:30
|
|
|
|
- klogd will reconnect to the logger (mostly syslogd) after it went
|
|
|
|
|
|
away
|
|
|
|
|
|
- On heavily loaded system syslog will not spit out error messages
|
|
|
|
|
|
anymore when recvfrom() results in EAGAIN
|
|
|
|
|
|
- Makefile improvements
|
|
|
|
|
|
- Local copy of module.h
|
|
|
|
|
|
- Improved sysklogd.8
|
|
|
|
|
|
- Always log with syslogd's timezone and locale
|
|
|
|
|
|
- Remove trailing newline when forwarding messages
|
2007-06-21 13:52:59 +05:30
|
|
|
|
- Continue working properly if /etc/service is missing and ignore
|
|
|
|
|
|
network activity
|
|
|
|
|
|
- Continue writing to log files as soon as space becomes available
|
|
|
|
|
|
again after a filled up disk
|
|
|
|
|
|
- Removed test to detect control characters > 0x20 as this prevented
|
|
|
|
|
|
characters encoded in UTF-8 to be properly passed through
|
|
|
|
|
|
- Only resolve the local domain when accepting messages from remote
|
|
|
|
|
|
- Properly accompany the MARK message with the facility
|
|
|
|
|
|
- Improved daemonise routine in klogd to stabilise startup
|
|
|
|
|
|
- klogd will not change the console log level anymore unless -c is given
|
|
|
|
|
|
- Added back /usr/src/linux/System.map as fall-back location
|
|
|
|
|
|
- Rewrote the module symbol parser to read from /proc/kallsyms
|
2007-07-29 23:07:07 +05:30
|
|
|
|
- Notify the waiting parent process if the client dies so it doesn't
|
2007-06-21 13:52:59 +05:30
|
|
|
|
wait the entire five minutes.
|
|
|
|
|
|
- Complete rewrite of the oops kernel module for Linux 2.6
|
|
|
|
|
|
- Only read kernel symbols from /proc/kallsyms if no System.map has been read
|
|
|
|
|
|
- Improved symbol lookup
|
2007-07-05 00:34:02 +05:30
|
|
|
|
- Prevent named pipes from becoming the controlling tty
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Jon Burgess <Jon_Burgess@eur.3com.com>
|
2003-09-29 23:17:51 +05:30
|
|
|
|
- Moved the installation of the signal handler up a little bit so it
|
|
|
|
|
|
guaranteed to be available when the child is forked, hence, fixing a
|
|
|
|
|
|
race condition. This used to create problems with UML and fast
|
|
|
|
|
|
machines.
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Greg Trounson <gregt@maths.otago.ac.nz>
|
2005-01-16 00:44:21 +05:30
|
|
|
|
- Improved README.linux
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Ulf H<>rnhammar <Ulf.Harnhammar.9485@student.uu.se>
|
2007-07-04 23:06:05 +05:30
|
|
|
|
- Boundary check for fscanf() in InitKsyms() and CheckMapVersion()
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Colin Phipps <cph@cph.demon.co.uk>
|
2007-07-29 23:07:07 +05:30
|
|
|
|
- Don't block on the network socket in case of packet loss
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Dirk Mueller <mueller@kde.org>
|
2005-01-16 00:44:21 +05:30
|
|
|
|
- Don't crash when filesize limit is reached (e.g. without LFS)
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Miquel van Smoorenburg <miquels@cistron.nl>
|
2007-06-21 13:52:59 +05:30
|
|
|
|
- Fix spurious hanging syslogd in connection with futex and NPTL
|
|
|
|
|
|
introduced in recent glibc versions and Linux 2.6
|
|
|
|
|
|
(Details: http://bugs.debian.org/301511)
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Eric Tucker <et@tallmaple.com>
|
2007-06-21 13:52:59 +05:30
|
|
|
|
- Improved syslog.conf(5) manpage
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Mike Frysinger <vapier@gentoo.org>
|
2007-06-21 13:52:59 +05:30
|
|
|
|
- use socklen_t where appropriate
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Kelledin <kelledin@skarpsey.dyndns.org>
|
2007-06-21 13:52:59 +05:30
|
|
|
|
- use newer query_module function rather than stepping through /dev/kmem.
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Matthew Fischer <futhark@vzavenue.net>
|
2007-06-21 13:52:59 +05:30
|
|
|
|
- Remove special treatment of the percent sign in klogd
|
2005-01-16 00:44:21 +05:30
|
|
|
|
|
2003-08-27 21:26:01 +05:30
|
|
|
|
|
2018-08-03 02:56:22 +05:30
|
|
|
|
[v1.4.1][] - 2001-03-11
|
|
|
|
|
|
-----------------------
|
|
|
|
|
|
|
. Removed kernel.patch since it is no longer required
. Changed SOCK_STREAM to SOCK_DGRAM in syslog.c
. klogd will only change the console log level if `-c' is supplied
. syslogd.c by Bill Nottingham <notting@redhat.com>
Um, if the directory is invalid, the bind() call in
create_unix_socket fails. Without the return -1, we return the
invalid fd that we just closed. When syslogd then starts
listening, select goes into a hard loop getting EBADF, IIRC.
. klogd.c by Troels Walsted Hansen <troels@thule.no>
I found a bug in the sysklogd package version 1.4. When it
encounters a zero byte in the kernel logging output, the text
parser enters a busy loop. I came upon it when the 3c59x driver
from kernel 2.4.0 started outputting two zero bytes for the product
code of my laptop's 3Com card. It could be argued that the kernel
should never output zero bytes in the logging info, but obviously
that will happen from time to time.
I fear this bug might be considered a security issue as well, if
the kernel can be coerced to output a zero byte somehow, all kernel
logging will stop.
Wolfgang Oertl <Wolfgang.Oertl@uibk.ac.at> had a similar bugfix
idea
. klogd.c by Thomas Roessler <roessler@does-not-exist.org>
Additionally, the patch prevents LogLine from being invoked with a
negative counter as an argument.
2001-03-12 01:05:52 +05:30
|
|
|
|
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- klogd will set the console log level only if `-c' is given on the
|
. Removed kernel.patch since it is no longer required
. Changed SOCK_STREAM to SOCK_DGRAM in syslog.c
. klogd will only change the console log level if `-c' is supplied
. syslogd.c by Bill Nottingham <notting@redhat.com>
Um, if the directory is invalid, the bind() call in
create_unix_socket fails. Without the return -1, we return the
invalid fd that we just closed. When syslogd then starts
listening, select goes into a hard loop getting EBADF, IIRC.
. klogd.c by Troels Walsted Hansen <troels@thule.no>
I found a bug in the sysklogd package version 1.4. When it
encounters a zero byte in the kernel logging output, the text
parser enters a busy loop. I came upon it when the 3c59x driver
from kernel 2.4.0 started outputting two zero bytes for the product
code of my laptop's 3Com card. It could be argued that the kernel
should never output zero bytes in the logging info, but obviously
that will happen from time to time.
I fear this bug might be considered a security issue as well, if
the kernel can be coerced to output a zero byte somehow, all kernel
logging will stop.
Wolfgang Oertl <Wolfgang.Oertl@uibk.ac.at> had a similar bugfix
idea
. klogd.c by Thomas Roessler <roessler@does-not-exist.org>
Additionally, the patch prevents LogLine from being invoked with a
negative counter as an argument.
2001-03-12 01:05:52 +05:30
|
|
|
|
commandline, not overwriting local settings in `/etc/sysctl.conf'.
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Bugfix: klogd will use SOCK_DGRM as well, re-enables kernel logging
|
|
|
|
|
|
- Bugfix: Don't make syslogd fail with broken `-a'
|
|
|
|
|
|
- Bugfix: klogd will skip zero bytes and not enter a busy loop anymore
|
|
|
|
|
|
- Thomas Roessler <roessler@does-not-exist.org>
|
. Removed kernel.patch since it is no longer required
. Changed SOCK_STREAM to SOCK_DGRAM in syslog.c
. klogd will only change the console log level if `-c' is supplied
. syslogd.c by Bill Nottingham <notting@redhat.com>
Um, if the directory is invalid, the bind() call in
create_unix_socket fails. Without the return -1, we return the
invalid fd that we just closed. When syslogd then starts
listening, select goes into a hard loop getting EBADF, IIRC.
. klogd.c by Troels Walsted Hansen <troels@thule.no>
I found a bug in the sysklogd package version 1.4. When it
encounters a zero byte in the kernel logging output, the text
parser enters a busy loop. I came upon it when the 3c59x driver
from kernel 2.4.0 started outputting two zero bytes for the product
code of my laptop's 3Com card. It could be argued that the kernel
should never output zero bytes in the logging info, but obviously
that will happen from time to time.
I fear this bug might be considered a security issue as well, if
the kernel can be coerced to output a zero byte somehow, all kernel
logging will stop.
Wolfgang Oertl <Wolfgang.Oertl@uibk.ac.at> had a similar bugfix
idea
. klogd.c by Thomas Roessler <roessler@does-not-exist.org>
Additionally, the patch prevents LogLine from being invoked with a
negative counter as an argument.
2001-03-12 01:05:52 +05:30
|
|
|
|
- Patch to prevent LogLine() from being invoked with a negative
|
|
|
|
|
|
counter as an argument.
|
|
|
|
|
|
|
2009-12-25 00:50:49 +05:30
|
|
|
|
|
2018-08-03 02:56:22 +05:30
|
|
|
|
[v1.4][] - 2000-09-19
|
|
|
|
|
|
---------------------
|
1999-08-21 15:58:44 +05:30
|
|
|
|
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Skip newline when reading in klog messages
|
|
|
|
|
|
- Use lseek64() instead of llseek() which is deprecated these days
|
|
|
|
|
|
- Close symbol file before returning with 0 when an error occurred
|
2000-09-13 03:23:31 +05:30
|
|
|
|
while reading it. This will enable systems to umount that
|
|
|
|
|
|
partition with no open file descriptor left over.
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Solar Designer <solar@false.com>
|
2000-09-18 00:19:39 +05:30
|
|
|
|
- printline() fixes
|
2000-09-18 13:04:11 +05:30
|
|
|
|
- priority decoding fix
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Daniel Jacobowitz <dan@debian.org>
|
2000-09-18 12:43:08 +05:30
|
|
|
|
- printchopped() fix
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Keith Owens <kaos@ocs.com.au>
|
2000-09-13 02:45:28 +05:30
|
|
|
|
- Fixed bug that caused klogd to die if there is no sym_array available.
|
|
|
|
|
|
- When symbols are expanded, print the line twice. Once with
|
|
|
|
|
|
addresses converted to symbols, once with the raw text. Allows
|
|
|
|
|
|
external programs such as ksymoops do their own processing on the
|
|
|
|
|
|
original data.
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Olaf Kirch <okir@caldera.de>
|
2000-09-18 19:11:33 +05:30
|
|
|
|
- Remove Unix Domain Sockets and switch to Datagram Unix Sockets
|
2018-08-03 02:56:22 +05:30
|
|
|
|
- Several bugfixes and improvements, please refer to the .c files
|
* New upstream source (closes: Bug#87819, Bug#70790, Bug#32580,
Bug#62358, Bug#71631)
* Upstream: Doesn't re-set log-level if not requested (closes:
Bug#76170, Bug#76170, Bug#85289)
* Upstream: Ignore zero bytes (closes: Bug#85478, Bug#85478, Bug#41068)
* Upstream: Corrected documentation for `-s' (closes: Bug#87020)
* Upstream: test for existence of syslogd-listfiles before calling
them. This got lost due to 1.4.0 brokennes which was packaged and
removed some hours later (closes: Bug#84872, Bug#66712)
* Applied patch by Tommi Virtanen <tv@debian.org> splitting the package
into `sysklogd' and `klogd' (closes:Bug#35586, Bug#72043, Bug#74864,
Bug#72122)
* Provide / depend on virtual packages system-log-daemon
and linux-kernel-log-daemon (closes: Bug#67604)
* Applied patch from Tim Janik <timj@gtk.org> to support `-s pattern' in
syslogd-listfiles
* Transition to FHS, i.e. /usr/share/doc instead of /usr/doc and
/usr/share/man instead of /usr/man (closes: Bug#79250, Bug#80771)
* Use --exec for stopping services (closes: Bug#76757)
* Corrected broken character in klogd.8 (cloes: Bug#75932)
* Only rotate logfiles with size greater than zero. This got lost due
to 1.4.0 brokennes which was packaged and removed some hours later
(closes: Bug#74993, Bug#49824)
* Added another note about modificability of cronjobs (closes:
Bug#88741)
* Since klogd replaces parts of sysklogd a proper Replaces line is there
* Added final newline to CHANGES
2001-03-12 04:30:51 +05:30
|
|
|
|
|
2003-08-27 21:26:01 +05:30
|
|
|
|
|
2018-08-03 02:56:22 +05:30
|
|
|
|
[UNRELEASED]: https://github.com/troglobit/sysklogd/compare/v1.5...HEAD
|
2018-09-24 23:35:19 +05:30
|
|
|
|
[v1.6]: https://github.com/troglobit/sysklogd/compare/v1.5...v1.6
|
2018-08-03 02:56:22 +05:30
|
|
|
|
[v1.5.1]: https://github.com/troglobit/sysklogd/compare/v1.5...v1.5.1
|
|
|
|
|
|
[v1.5]: https://github.com/troglobit/sysklogd/compare/v1.4...v1.5
|
|
|
|
|
|
[v1.4]: https://github.com/troglobit/sysklogd/compare/v1.3...v1.4
|
