Patchwork errata#89 patch for Family 0Fh Prozessors

login
register
about
Submitter Josef Kellermann
Date 2011-02-02 17:39:07
Message ID <4D4996BB.7080105@arcor.de>
Download mbox | patch
Permalink /patch/2601/
State Superseded
Headers show

Comments

Josef Kellermann - 2011-02-02 17:39:07
This patch fixes a potential system hang.
Don't know why this patch isn't included if 'CONFIG_K8_REV_F_SUPPORT == 1'.

Signed-off-by: Josef Kellermann  <mailto://seppk@arcor.de><seppk@arcor.de>  <mailto://seppk@arcor.de>
Alexandru Gagniuc - 2011-02-02 19:11:08
Erratum 89 is already handled on line 390:

if (!is_cpu_pre_b3()) {

    /* Erratum 89 ... */
    msr = rdmsr(NB_CFG_MSR);
    msr.lo |= 1 << 3;

If this is also needed when CONFIG_K8_REV_F_SUPPORT == 1, then we should
move the existing one out of the
#if CONFIG_K8_REV_F_SUPPORT == 0
so it gets executed in both cases, as opposed to duplicating it.

Alex

On 02/02/2011 07:39 PM, Josef Kellermann wrote:
> This patch fixes a potential system hang.
> Don't know why this patch isn't included if 'CONFIG_K8_REV_F_SUPPORT == 1'.
> 
> Signed-off-by: Josef Kellermann  <mailto://seppk@arcor.de><seppk@arcor.de> <mailto://seppk@arcor.de>
>

Patch

Index: src/cpu/amd/model_fxx/model_fxx_init.c
===================================================================
--- src/cpu/amd/model_fxx/model_fxx_init.c	(Revision 6326)
+++ src/cpu/amd/model_fxx/model_fxx_init.c	(Arbeitskopie)
@@ -452,6 +452,10 @@ 
 	wrmsr(HWCR_MSR, msr);
 
 #if CONFIG_K8_REV_F_SUPPORT == 1
+	/* Erratum 89 */
+	msr = rdmsr(NB_CFG_MSR);
+	msr.lo |= 1 << 3; // Erratum 89
+	wrmsr(NB_CFG_MSR, msr);	
 	/* Erratum 131... */
 	msr = rdmsr(NB_CFG_MSR);
 	msr.lo |= 1 << 20;