From Bugzilla Helper: User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7.12) Gecko/20050915 Description of problem: Currently, RHEL 3.0 only supports a maximum of eight logical processors for the 64-bit x86_64 (AMD64/EM64T) architecture. (On the 32-bit x86 architecture, 32 logical processors are supported.) These limits are outlined on the following Red Hat page: http://www.redhat.com/en_us/USA/rhel/details/limits/ With the release of new multi-core processors from Intel and AMD, and new hardware from vendors supporting large numbers of processor cores, this limit is quickly becoming problematic. For example, a multi-processor system containing 4 Intel dual core processors (e.g., Paxville) with hyperthreading enabled will show up as 16 logical processors under Linux. On x86, this works fine, but on x86_64, only 8 logical processors show up, and hyperthreading is effectively disabled. Worse yet, on a multi-processor system containing 8 AMD dual core processors you effectively lose access to half the cores in the system under x86_64. Intel and systems vendors assure me that Red Hat is aware of this problem and that fixes are being planned, but I've been unable to locate a bugzilla bug which is tracking this need. Thus, I'm opening this bug. If there is some internal/hidden bug to track this development, please feel free to dupe the bug. (Note that this bug is similar/related to but NOT a duplicate of Bug 147823.) Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. Install RHEL 3.0 x86_64 on a system with more than eight logical processors. 2. cat /proc/cpuinfo Actual Results: Only eight logical processors will be listed. Expected Results: All available logical processors should be listed. Additional info:
I believe the real limit is 32. Tim, is this a documentation bug?
Two notes: 1) I was slightly incorrect in my description of the problem. In RHEL *4.0* only the first eight logical processors are listed (and I've separately opened Bug 174760 regarding that. In RHEL 3.0 the problem is more severe when there are more than eight logical processors: The kernel won't boot. I've opened Bug 174991 for that bug. If this feature is implemented, it will resolve that bug; if not, the failure to boot will need to be addressed separately. 2) I'm fairly certain this is NOT a documentation bug. (I've tested it and Sun, Dell, and Intel have all indicated to me that it's a problem.)
Reassigning to Jim.
User jparadis's account has been closed