<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=UTF-8" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
<pre>Oracle is pleased to announce the general availability of Enterprise Linux Release 5 Update 2 for x86 (32 bit) and x86_64 (64 Bit) architectures.
This update includes the following kernel/driver changes:
</pre>
<tt>Bug fixes added by Oracle</tt> <br>
<br>
<tt>* Add entropy module option to e1000 and bnx2 [ORA
6045759]<br>
* Fix bad unlock_page() in error case [ORA 6263574]<br>
* Fix error-path crashes in dio [ORA 6242289]<br>
* Fix netpoll race condition<br>
</tt><br>
<div style="margin-left: 80px;">
</div>
<pre>Fixes/Additions from upstream distribution provider
* Fixed issue with executing binaries with more than 2GB of debug information.
* All allocated hugepages are getting released on database shutdown.
* Fixed soft lockup warnings caused when allocating memory on a system with
large memory and invalidate_mapping_pages() calls
* 64-bit inode numbers are properly processed by 32-bit NFS clients.
* Enumeration order of PCI devices has changed to have NICs appear in the
order they are labeled on the chassis and how the BIOS numbers them.
Impacted platforms are:
o Dell PowerEdge R900
o HP ProLiant DL385 G2
o HP ProLiant DL585 G2
o HP Proliant DL580 G5
Other platforms may see similar behavior if there is a combination of PCI
and PCI Express devices (e.g. Intel's e1000 & e1000e drivers). Use kernel parameter
pci=nobfsort to use old enumeration order.
* Added following two new kernel parameters
o /proc/sys/vm/flush_mmap_pages (kupdate to flush or not to flush
memory-mapped file pages should be flushed to disk while the memory
map is active. Possible value 0 ,1 . 0 means not to flush)
o /proc/sys/kernel/sched_interactivity (allows interactive processes
to acquire more CPU time without causing CPU starvation in other
processes. 2 fully active, 1 weaker affinity, 0 disabled).
* Core dump masking is supported (core dump process to skip the shared
memory segments of a process when creating a core dump file). Use
/proc/[pid]/coredump_filter to set bitmask of memory types. Supported
memory types are:
o 0x0 anonymous private memory
o 0x1 anonymous shared memory
o 0x2 file-backed private memory
o 0x3 file-backed shared memory
* For systems with large number of CPUs, maximum soft lockup timeout is now
increased from 60 seconds to 300 seconds.
* MSI-X is disabled by default. To enable it use the kernel module
parameter ql2xenablemsix.
* Support added for lock contention tracing and lockdep.
* CONFIG_IPMI_PANIC_EVENT & CONFIG_IPMI_PANIC_STRING kernel configuration
flags are enabled to enhance kernel panic handling for IPMI (Intelligent
Platform Management Interface)
* Support added for Intel's dynamic acceleration technology.
* Boot command size is increased to 2048.
* Driver Updates:
o bnx2x (Included support for bcm5710 hardware)
o e1000 driver support Intel 82540 – 82547 network cards, while e1000e driver support the following network cards:
* Intel 82571 – 82573
* Intel 82562
* Intel 82566
* Intel 80003eslan
o tg3 (Included BroadCom 5761 & 5784 devices support)
o Added PCI IDs to support Intel E7221 graphic controller
o Added sub-device IDs to support PCI-Express icom WAN adapter
o AMD SBX00 SMBus supported by enabling i2c-piix4 kernel module
o tlclk ( support Intel MPCBL0050 systems)
o aacraid (version 1.1.5-2453)
o lpfc ( version 8.2.0.22, Included support for NPIV virtual ports )
o qla2xxx (Version 8.02.00-k5, Included support for 8 Gb/s adapters, NPIV)
o mpt fusion (Version 3.04.05)
o megaraid_sas (version 3.15)
<u><b>Software Accessibility</b></u>
All packages are available via ULN (<a set="yes" linkindex="6"
href="http://linux.oracle.com">http://linux.oracle.com</a>).
Installable binary and source ISO images will be available at <a
linkindex="7" href="http://edelivery.oracle.com">http://edelivery.oracle.com</a> shortly.
If ISO images are needed before they are available on eDelivery, please request these via a Metalink service request.
For more information please visit <a set="yes" linkindex="8"
href="http://linux.oracle.com">http://linux.oracle.com</a>
Thank you
Sincerely,
Enterprise Linux Team</pre>
</body>
</html>