kernel-ml

Kernel-ml for Enterprise Linux 6.

The kernel configuration is based upon the default RHEL-6 configuration with added functionality enabled as appropriate. The packages are intentionally named kernel-ml so as not to conflict with the RHEL-6 kernels and, as such, they may be installed and updated alongside the regular kernel.

You can also Download manually from http://elrepo.org/linux/kernel/el6/ (or from a mirror site near you).

Notes

These packages are provided As-Is with no implied warranty or support. Using the kernel-ml may expose your system to security, performance and/or data corruption issues. Since timely updates may not be available from the ELRepo Project, the end user has the ultimate responsibility for deciding whether to continue using the kernel-ml packages in regular service.

If a bug is found when using these kernels, the end user is encouraged to report it upstream to the Linux Kernel Bug Tracker and, for our reference, to the ELRepo bug tracker. By taking such action, the reporter will be assisting the kernel developers, Red Hat and the Open Source Community as a whole.

Known Issues

Currently none are known.

---

Kernel-ml for Enterprise Linux 5.

The kernel configuration is based upon the default RHEL-5 configuration with added functionality enabled as appropriate. The packages are intentionally named kernel-ml so as not to conflict with the RHEL-5 kernels and, as such, they may be installed and updated alongside the regular kernel.

To install kernel-ml you will need elrepo-release-5-1.el5.elrepo (or newer). Run:

yum --enablerepo=elrepo-kernel install kernel-ml

You can also Download manually from http://elrepo.org/linux/kernel/el5/ (or from a mirror site near you).

Notes

These packages are provided As-Is with no implied warranty or support. Using the kernel-ml may expose your system to security, performance and/or data corruption issues. Since timely updates may not be available from the ELRepo Project, the end user has the ultimate responsibility for deciding whether to continue using the kernel-ml packages in regular service.

If a bug is found when using these kernels, the end user is encouraged to report it upstream to the Linux Kernel Bug Tracker and, for our reference, to the ELRepo bug tracker. By taking such action, the reporter will be assisting the kernel developers, Red Hat and the Open Source Community as a whole.

Known Issues

A set of known issues with mainline kernels are listed below. If you have a solution or workaround for one of these issues, please let us know.

HAL does not provide battery information (affects eg. Gnome)

This is a regression in the Linux kernel (since 2.6.36) and is tracked in our bug-tracker at: http://elrepo.org/bugs/view.php?id=91

Hardware clock (/dev/rtc) not working

This is a regression in the Linux kernel and is tracked in our bug-tracker at: http://elrepo.org/bugs/view.php?id=92

Depending upon the hardware being used, the RTC may not be accessible. This can be checked by executing /sbin/hwclock --debug. If the hwclock binary reports that the RTC is not accessible by any means, the patch below fixes the problem after a reboot. A simple temporary work-around is to: ln -sf /dev/rtc0 /dev/rtc

--- rc.sysinit.distro   2012-02-11 11:03:13.000000000 +0000
+++ rc.sysinit  2012-02-11 11:11:05.000000000 +0000
@@ -269,6 +269,19 @@ if [ "$PROMPT" != "no" ]; then
  echo
 fi
 
+### ELREPO: See if the RTC driver is built-in, otherwise load it
+if [ ! -f /proc/driver/rtc ]; then
+    action $"Loading rtc_cmos driver: " /sbin/modprobe rtc_cmos
+fi
+
+### ELREPO: Make the newer device nodes to accomodate hwclock
+RTC_MAJOR_NO=`/bin/awk '/rtc/ { print $1 }' /proc/devices`
+if [ -n "$RTC_MAJOR_NO" ]; then
+    action $"Creating /dev/rtc0: " /bin/mknod /dev/rtc0 c $RTC_MAJOR_NO 0
+    action $"Creating /dev/rtc: " /bin/ln -sf /dev/rtc0 /dev/rtc
+    [ -n "$SELINUX_STATE" ] && restorecon /dev/rtc0 /dev/rtc >/dev/null 2>&1
+fi
+
 # Set the system clock.
 update_boot_stage RCclock
 ARC=0

mcelog reports incorrect record size

Only 64-bit systems are affected. The userland tool needs to be revised to distinguish between the mcelog record size of a distro kernel and a more modern kernel. As a work-around to avoid hourly e-mail messsages from the cron daemon, you can append 2>&1 to the last line of the /etc/cron.hourly/mcelog.cron file.

See also: RHbz #507026 mcelog: warning: record length longer than expected.

NMI watchdog failed to create perf event

When booting kernel-2.6.36-0.el5.elrepo on a 32-bit uniprocessor system and using a boot line of "ro root=/dev/VolGroup00/LogVol00 nosmp quiet nompath nodmraid", a harmless message "NMI watchdog failed to create perf event on CPU01: ffffffa1" is displayed on the system console.

Missing dm-region-hash module

A warning message will be displayed on the system console at boot time about a missing dm-region-hash module.

One workaround to silence the message is to add the nodmraid kernel parameter as part of your kernel options in /etc/grub.conf. Another option is to remove the dmraid modules from the initrd if you do not use dmraid, however the mkinitrd options --without-dmraid and --omit-raid-modules do not work. You will have to comment out three findmodule lines in /sbin/mkinitrd that reference dm-mirror, dm-zero and dm-snapshot and then recreate your initrd. (Search for "132001" in the script)

See also: RHbz #132001 add dm-mirror and dm-snapshot if dm-mod is brought in

Some iptables modules fail causing iptables firewall to become effectively disabled.

Some iptables modules do not function correctly. This has a knock-on effect that iptables-restore fails to load any rules if any rule should fail, thus rendering the firewall effectively disabled. Please check your iptables rules carefully to ensure there are no errors and the rules are correctly loaded.

See Bug: http://elrepo.org/bugs/view.php?id=97

---