Bug 123422
Summary: | on AMD64, the events shown are suitable for Athlon not Athlon64 | ||
---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | David Monniaux <david.monniaux> |
Component: | oprofile | Assignee: | William Cohen <wcohen> |
Status: | CLOSED RAWHIDE | QA Contact: | |
Severity: | medium | Docs Contact: | |
Priority: | medium | ||
Version: | 1 | ||
Target Milestone: | --- | ||
Target Release: | --- | ||
Hardware: | x86_64 | ||
OS: | Linux | ||
Whiteboard: | |||
Fixed In Version: | Doc Type: | Bug Fix | |
Doc Text: | Story Points: | --- | |
Clone Of: | Environment: | ||
Last Closed: | 2004-05-20 12:57:42 UTC | Type: | --- |
Regression: | --- | Mount Type: | --- |
Documentation: | --- | CRM: | |
Verified Versions: | Category: | --- | |
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
Cloudforms Team: | --- | Target Upstream Version: | |
Embargoed: |
Description
David Monniaux
2004-05-18 13:55:02 UTC
The events for the AMD64 are a pure superset of the Athlon events. One can still get measurements even with the processor being misidentified as a regular Athlon. However oprofile will not have access to the AMD64 specific events such as the ones related to the hypertransport channels. This problem is due to the kernel misidenting the processor. I have verified this problem has been fixed in the kernel in Rawhide (FC2-test3). I'm interested in FPU-related events. Where is the misidentification problem? If this helps, the kernel correctly prints in /proc/cpuinfo: model name : AMD Athlon(tm) 64 Processor 3200+ The oprofile kernel driver has its own cpu identification code; it is separate from the /proc/cpuinfo. In the case of the x86_64 it is going to be in arch/i386/oprofile/nmi_int.c of the kernel. The oprofile driver provide the cpu identification information in /dev/oprofile/cpu_type. |