RHBA-2010:0461 - Bug Fix Advisory

Topic

Updated kernel packages that resolve several issues are now available for Red
Hat Enterprise Linux 5.4 Extended Update Support.

Description

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

These updated kernel packages provide fixes for the following bugs:

• on systems using the Intel igb driver, the boot process halted and dom0, the

first domain started by the Xen hypervisor on boot, stopped responding. This was
resolved by enabling and disabling interrupt bits for the igb physical function
driver. With this update, a dom0 host using the Intel igb driver completes the
boot process successfully and without becoming unresponsive. (BZ#568750)

• in rare circumstances, when using kdump and booting a kernel with

"crashkernel=128M@16M", the kdump kernel did not boot after a crash. (BZ#580837)

• iSCSI failover times were slower than in Red Hat Enterprise Linux 5.3.

(BZ#583899)

• the RHSA-2009:0225 update fixed a bug where, in some cases, on systems with

the kdump service enabled, pressing Alt+SysRq+C to trigger a crash resulted in a
system hang; therefore, the system did not restart and boot the dump-capture
kernel as expected; no vmcore file was logged; and the following message was
displayed on the console:

BUG: warning at arch/[arch]/kernel/crash.c:[xxx]/nmi_shootdown_cpus() (Not
tainted)

The RHSA-2009:0225 update resolved this issue by not calling printk() during a
crash. It was later discovered that this fix did not resolve the issue in all
cases, since there was one condition where printk() was still being called: at a
warning condition inside the mdelay() call.

This update replaces mdelay() calls with udelay(), where such a warning
condition does not exist, which fully resolves this issue, allowing Alt+SysRq+C
to work as expected. (BZ#588215)

• when an SFQ (Stochastic Fair Queuing) qdisc that limited the queue size

to two packets was added to a network interface (for example, via "tc
qdisc add"), sending traffic through that interface resulted in a kernel
crash. With this update, such a qdisc no longer results in a kernel crash
when sending traffic. (BZ#594053)

• when a system was configured using channel bonding in "mode=0" (round-robin

balancing) with multicast, IGMP traffic was transmitted via a single interface.
If that interface failed (due to a port, NIC or cable failure, for example),
IGMP was not transmitted via another port in the group, thus resulting in
packets for the previously-registered multicast group not being routed
correctly. (BZ#594056)

• on NFS, the read(2) system call could have returned an unexpected EIO

(input/ouput error) value. (BZ#594060)

Users should upgrade to these updated packages, which contain backported patches
to correct these issues. The system must be rebooted for this update to take
effect.

Solution

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

This update is available via the 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

Affected Products

• Red Hat Enterprise Linux for x86_64 - Extended Update Support 5.4 x86_64
• Red Hat Enterprise Linux for x86_64 - Extended Update Support 5.4 ia64
• Red Hat Enterprise Linux for x86_64 - Extended Update Support 5.4 i386
• Red Hat Enterprise Linux for IBM z Systems - Extended Update Support 5.4 s390x
• Red Hat Enterprise Linux for Power, big endian - Extended Update Support 5.4 ppc

Fixes

• BZ - 580837 - [5.4]System panic occurred during boot sequence with the server which carries 256GMB physical memory. [rhel-5.4.z]
• BZ - 583899 - REGRESSION: Fix iscsi failover time [rhel-5.4.z]
• BZ - 588215 - [RHEL5.3 GA] The kernel stalls before starting second kernel when pressing Alt+SysRq+c in graphical console [rhel-5.4.z]
• BZ - 594053 - [5.5] SFQ qdisc crashes with limit of 2 packets [rhel-5.4.z]
• BZ - 594056 - [RHEL5] bonding mode 0 doesn't resend IGMP after a failure [rhel-5.4.z]
• BZ - 594060 - nfs: sys_read sometimes returns -EIO [rhel-5.4.z]

CVEs

(none)

References

(none)