Description of problem: After starting an installation with any of the following commands a kernel panic takes place when the "northbridge chipset memory remapping" feature is enabled (2 memory blocks each of 2GB, total installed 4GB): -> linux -> linux mediacheck -> linux noprobe -> linux acpi=off -> linux pci=nommconf -> linux mem=3072MB (to) linux mem=2080MB In contrast, an installation with "linux mem=2048MB" and less memory succeeds! The installation succeeds too, if the "northbridge chipset memory remapping" feature is disabled (1 memory block of 3GB, total installed 4GB). Version-Release number of selected component (if applicable): Kernel 2.6.18-8.el5xen x86_64 How reproducible: Always Steps to Reproduce: 1.linux mediacheck or 1.linux Actual results: last regular message on the screen: io schedular cfq registered. an error message on the last screen: oops_end + 0x51/0x53 Expected results: Additional info: Hardware info: Asus P5B-E mobo (northbridge chipset: Intel P965) 4GB RAM (4x1GB) installed SATA HDD and SATA DVD-RAM NIC Intel e1000 Graphic card ATI x1950 or X1300 The system is currently running CentOS 4.4 x86_64 and Rocks 4.2.1 The installation of CentOS 5.0 x86_64 aborts with kernel panic.
change QA contact
This bug is very old now, and RHEL-5.0 is fairly old too. Can you give 5.1 a try and see if you still have the problem? If you do, please get a complete OOPS trace and attach it here (serial console is the easiest), and let us know the exact version of the kernel you tried. Thanks, Chris Lalancette
No response to need info for almost a year. Closing this out. Please reopen if you can reproduce the issue.