(UPDATE): Re: [WOLK-devel] [ANNOUNCE] WOLK v4.4s FINAL - Server Edition

SourceForge Headquarters 1320 Columbia Street Suite 310 San Diego, CA 92101 +1 (858) 422-6466

=2D----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Wednesday 30 July 2003 08:47, Marc-Christian Petersen wrote:

Hi all,

due to some problems in the previous 4.4 update patch, I've rediffed it aga=
in,=20
uploaded new patches and therefore also a new announce with 2 missing=20
changelog entries too and right md5sums. Sorry all!

=2D -------

Changelog from v4.3s -> v4.4s
=2D -----------------------------
o   added:    High Performance Packet Classification (nf-hipac) v0.8 rev2
o   added:    boost the copy-user asm.
o   added:    merged IO-Stall fixes from 2.4.22-pre3/4/5/6/7
o   added:    q->full, defaults to off and keeps the elvtune changes.
                So to turn on the q->full low latency fixes, you need to:
                "elvtune -b 1 /dev/xxxx" . Note that for lvm and md, you ne=
ed
                to elvtune each underlying device. Running it on an lvm/md
                device doesn't do anything.
o   added:    low latency / normal / max throughput - I/O elevator selectio=
n.
o   fixed:    CAN-2003-0461: /proc/tty/driver/serial reveals the exact
                character counts for serial links. This could be used by a
                local attacker to infer password lengths and inter-keystroke
                timings during password entry.
o   fixed:    CAN-2003-0462: Paul Starzetz discovered a file read race
                condition existing in the execve() system call, which could
                cause a local crash.
o   fixed:    CAN-2003-0476: The execve system call in Linux 2.4.x records
                the file descriptor of the executable process in the file
                table of the calling process, allowing local users to gain
                read access to restricted file descriptors.
o   fixed:    CAN-2003-0501: The /proc filesystem in Linux allows local use=
rs
                to obtain sensitive information by opening various entries =
in
                /proc/self before executing a setuid program. This causes t=
he
                program to fail to change the ownership and permissions of
                already opened entries.
o   fixed:    CAN-2003-0550: The STP protocol is known to have no security,
                which could allow attackers to alter the bridge topology.
                STP is now turned off by default.
o   fixed:    CAN-2003-0551: STP input processing was lax in its length
                checking, which could lead to a denial of service.
o   fixed:    CAN-2003-0552: Jerry Kreuscher discovered that the Forwarding
                table could be spoofed by sending forged packets with bogus
                source addresses the same as the local host.
o   fixed:    reduced the number of requests during seeks (the latency times
                increased slightly during seeks with pre5/pre6).
o   fixed:    reserved some spare request for reads. This is been measured
                to avoid some waiting for reads and it's beneficial in the
                common case
o   fixed:    copy-namespace
o   fixed:    RMAP: refill free list in batches, in fixup_freespace
o   fixed:    RMAP: only wake up kswapd when low on free+clean pages
o   fixed:    RMAP: in __alloc_pages_limit, first do rmqueue (to use per-cpu
                      freelist) and fall back to direct_reclaim when needed
o   fixed:    RMAP: make sure the first stage of page allocation doesn't
                      upset the zone balancing
o   fixed:    RANDOM: another SMP deadlock. I really wonder how many (read:
                      tons * 2^99) deadlocks we have in the Linux kernel!
                 I am _shouting_ out for a lock free kernel!! ;)
o   updated:  Oracle Cluster FileSystem (OCFS) v1.0.9-pre Jul 17 2003
o   updated:  AIO: allow aio on blkdevices too
o   updated:  smp-timers: merged an anti deadlock fix from lcm, 2.5 probably
                needs it too. In short the theory that mod_timer is the only
                thing that can run in parallel was wrong, add_timer and
                del_timer/del_timer_sync can too. Having already fixed
                mod_timer in a backwards compatible way before merging the
                smp-timers in -aa, made it easy to fix those further windows
                too.
o   updated:  ksoftirqd: merged a fix from Philip Craig to be sure to make
                the anti-DoS logic effective. He wrote and verified the cod=
e.
                It makes perfect sense so it's applied. Normal usages
                shouldn't notice the difference, especially with the
                max-loop logic.
o   updated:  Intel E100 driver v2.3.18-k1 + bugfixes from .22-BK
o   updated:  Intel E1000 driver v5.1.11-k1 + bugfixes from .22-BK
o   updated:  Broadcom BCM5700 driver v6.2.11
o   updated:  Broadcom Tigon3 v1.6
o   updated:  SysKonnect SK-98xx driver v6.12
o   updated:  HP CISS Driver v2.4.47
o   updated:  Compaq SMART2 Driver v2.4.25
o   updated:  raw vary-io 21, including more SCSI driver support
o   updated:  Super FreeS/WAN v1.99.8 Final
o   updated:  Qlogic QLA 2x00 v6 FC SCSI support v6.05.60
o   updated:  3com 3c59x v0.99Za (vanilla ones is over 2 years old)
o   removed:  LUFS: braindamaged, non-working. Get rid of that shit.
o   changed:  DRM v4.3 is now seperate from DRM 4.0/4.1/4.2

md5sums:
=2D --------
6fbff6c3c6c0e298eb203d6c2af36772 *linux-2.4.20-wolk4.3s-to-4.4s.patch.bz2
48ee6b8c23aa9d137ae5cc02c0b9882d *linux-2.4.20-wolk4.3s-to-4.4s.patch.gz

=2D --
Kind regards
        Marc-Christian Petersen

http://sourceforge.net/projects/wolk

PGP/GnuPG Key: 1024D/569DE2E3DB441A16
=46ingerprint: 3469 0CF8 CA7E 0042 7824 080A 569D E2E3 DB44 1A16
Key available at www.keyserver.net. Encrypted e-mail preferred.
=2D----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)
Comment: !! No Risk - No Fun !! - Try to crack this ;-)

iD8DBQE/J8I2Vp3i49tEGhYRAgInAJ4ynv/moRIQpFaQe4qP+a6LqnwdTQCg3NmZ
6Mh0MHHzcH6VPhNJ4GeuMRk=3D
=3DCq3M
=2D----END PGP SIGNATURE-----