Bug 447097

Summary: warning in mtrr_trim_uncached_memory
Product: [Fedora] Fedora Reporter: Jon Stanley <jonstanley>
Component: kernelAssignee: Kernel Maintainer List <kernel-maint>
Status: CLOSED CANTFIX QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: low Docs Contact:
Priority: low    
Version: rawhideCC: hayden.james, nhorman
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2008-05-17 21:38:55 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 Jon Stanley 2008-05-17 21:26:30 UTC 
I've been having this for the entire cycle of F9, and it's present in rawhide as
well.  On a VMWare VM, there's always an early warning, which I assume is a
buggy emulated BIOS, but I'm not sure:

WARNING: strange, CPU MTRRs all blank?
------------[ cut here ]------------
WARNING: at arch/x86/kernel/cpu/mtrr/main.c:696
mtrr_trim_uncached_memory+0x190/0x195() (Not tainted)
Modules linked in:
Pid: 0, comm: swapper Not tainted 2.6.25.2-5.fc10.i686 #1
 [<c04279c5>] warn_on_slowpath+0x47/0x73
 [<c04f4d58>] ? vsnprintf+0x3e7/0x420
 [<c062bc01>] ? _spin_unlock_irqrestore+0x10/0x14
 [<c0427e89>] ? release_console_sem+0x195/0x19d
 [<c0428370>] ? vprintk+0x301/0x330
 [<c0428370>] ? vprintk+0x301/0x330
 [<c074f94e>] ? do_early_param+0x30/0x6f
 [<c0757e26>] mtrr_trim_uncached_memory+0x190/0x195
 [<c0755d5b>] setup_arch+0x27b/0x449
 [<c074f63f>] start_kernel+0x66/0x345
 =======================
---[ end trace ca143223eefdc828 ]---
Comment 1 Dave Jones 2008-05-17 21:38:55 UTC 
This spectacular display of fail is VMWares bug..

It should either a) not advertise MTRR support or b) set at least one MTRR to
cover the main memory as write-back.
Comment 2 Chuck Ebbert 2008-05-19 04:51:14 UTC 
Eh, our own VM solution does the same exact thing (no MTRRs) and we work around
it by not emitting the error message. The real solution is to just punt the
error message completely.