Important: kernel security and bug fix update

Updated kernel packages that resolve several security issues and fix
various bugs are now available for Red Hat Enterprise Linux 4.

This update has been rated as having important security impact by the Red
Hat Security Response Team.

The kernel packages contain the Linux kernel, the core of any Linux
operating system.

This update addresses the following security issues:

* a buffer overflow was found in the Linux kernel Partial Reliable Stream
Control Transmission Protocol (PR-SCTP) implementation. This could,
potentially, lead to a denial of service if a Forward-TSN chunk is received
with a large stream ID. (CVE-2009-0065, Important)

* a memory leak was found in keyctl handling. A local, unprivileged user
could use this flaw to deplete kernel memory, eventually leading to a
denial of service. (CVE-2009-0031, Important)

* a deficiency was found in the Remote BIOS Update (RBU) driver for Dell
systems. This could allow a local, unprivileged user to cause a denial of
service by reading zero bytes from the image_type or packet_size file in
"/sys/devices/platform/dell_rbu/". (CVE-2009-0322, Important)

* a deficiency was found in the libATA implementation. This could,
potentially, lead to a denial of service. Note: by default, "/dev/sg*"
devices are accessible only to the root user. (CVE-2008-5700, Low)

This update also fixes the following bugs:

* when the hypervisor changed a page table entry (pte) mapping from
read-only to writable via a make_writable hypercall, accessing the changed
page immediately following the change caused a spurious page fault. When
trying to install a para-virtualized Red Hat Enterprise Linux 4 guest on a
Red Hat Enterprise Linux 5.3 dom0 host, this fault crashed the installer
with a kernel backtrace. With this update, the "spurious" page fault is
handled properly. (BZ#483748)

* net_rx_action could detect its cpu poll_list as non-empty, but have that
same list reduced to empty by the poll_napi path. This resulted in garbage
data being returned when net_rx_action calls list_entry, which subsequently
resulted in several possible crash conditions. The race condition in the
network code which caused this has been fixed. (BZ#475970, BZ#479681 &
BZ#480741)

* a misplaced memory barrier at unlock_buffer() could lead to a concurrent
h_refcounter update which produced a reference counter leak and, later, a
double free in ext3_xattr_release_block(). Consequent to the double free,
ext3 reported an error

ext3_free_blocks_sb: bit already cleared for block [block number]

and mounted itself as read-only. With this update, the memory barrier is
now placed before the buffer head lock bit, forcing the write order and
preventing the double free. (BZ#476533)

* when the iptables module was unloaded, it was assumed the correct entry
for removal had been found if "wrapper->ops->pf" matched the value passed
in by "reg->pf". If several ops ranges were registered against the same
protocol family, however, (which was likely if you had both ip_conntrack
and ip_contrack_* loaded) this assumption could lead to NULL list pointers
and cause a kernel panic. With this update, "wrapper->ops" is matched to
pointer values "reg", which ensures the correct entry is removed and
results in no NULL list pointers. (BZ#477147)

* when the pidmap page (used for tracking process ids, pids) incremented to
an even page (ie the second, fourth, sixth, etc. pidmap page), the
alloc_pidmap() routine skipped the page. This resulted in "holes" in the
allocated pids. For example, after pid 32767, you would expect 32768 to be
allocated. If the page skipping behavior presented, however, the pid
allocated after 32767 was 65536. With this update, alloc_pidmap() no longer
skips alternate pidmap pages and allocated pid holes no longer occur. This
fix also corrects an error which allowed pid_max to be set higher than the
pid_max limit has been corrected. (BZ#479182)

All Red Hat Enterprise Linux 4 users should upgrade to these updated
packages, which contain backported patches to resolve these issues. The
system must be rebooted for this update to take effect.

Before applying this update, make sure that all previously-released
errata relevant to your system have been applied.

This update is available via Red Hat Network. Details on how to use
the Red Hat Network to apply this update are available at
http://kbase.redhat.com/faq/docs/DOC-11259

474495 - CVE-2008-5700 kernel: enforce a minimum SG_IO timeout
475970 - oops in e1000_clean (list corruption due to race with e1000_down)
476533 - Read-only filesystem after 'ext3_free_blocks_sb: bit already cleared for block' errors
477147 - Kernel panic when unloading ip conntrack modules
478800 - CVE-2009-0065 kernel: sctp: memory overflow when FWD-TSN chunk is received with bad stream ID
479182 - RHEL4 64 bit skips all pids with bit 15 set (32768-65535, 98304-131071 etc)
479681 - oops in net_rx_action on double free of dev->poll_list
480592 - CVE-2009-0031 kernel: local denial of service in keyctl_join_session_keyring
480741 - RHEL4.8 kernel crashed in net_rx_action() on IA64 machine in RHTS connectathon test
482866 - CVE-2009-0322 kernel: dell_rbu local oops
483748 - rhel4 PV guest installations busted on rhel 5.3 i386 intel dom0

http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2008-5700
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2009-0031
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2009-0065
http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2009-0322
http://www.redhat.com/security/updates/classification/#important