[PATCH] fix PMU initialization on pseries lpar

STALE7310d

2 messages, 2 authors, 2006-08-02 · open the first message on its own page

[PATCH] fix PMU initialization on pseries lpar

From: Sonny Rao <hidden>
Date: 2006-08-02 04:20:14

We should not be calling power4_enable_pmcs() in
pseries_lpar_enable_pmcs(); just doing the hypercall is sufficient. 
Prior to 2.6.15 we did not call power4_enable_pmcs() for an lpar.

power4_enable_pmcs() tries to read the hid0 register which is no
longer legal for an lpar in newer Power processors.
--- a/arch/powerpc/platforms/pseries/setup.c	2006-04-23 01:45:09.000000000 -0500
+++ b/arch/powerpc/platforms/pseries/setup.c~pmcfix	2006-08-01 23:12:39.000000000 -0500
@@ -182,8 +182,6 @@ static void pseries_lpar_enable_pmcs(voi
 {
 	unsigned long set, reset;
 
-	power4_enable_pmcs();
-
 	set = 1UL << 63;
 	reset = 0;
 	plpar_hcall_norets(H_PERFMON, set, reset);

Re: [PATCH] fix PMU initialization on pseries lpar

From: Michael Ellerman <hidden>
Date: 2006-08-02 04:29:33

On Wed, 2006-08-02 at 00:20 -0400, Sonny Rao wrote:
We should not be calling power4_enable_pmcs() in
pseries_lpar_enable_pmcs(); just doing the hypercall is sufficient. 
Prior to 2.6.15 we did not call power4_enable_pmcs() for an lpar.

power4_enable_pmcs() tries to read the hid0 register which is no
longer legal for an lpar in newer Power processors.
Argh, how did I miss that. Mea culpa. 

cheers

-- 
Michael Ellerman
IBM OzLabs

wwweb: http://michael.ellerman.id.au
phone: +61 2 6212 1183 (tie line 70 21183)

We do not inherit the earth from our ancestors,
we borrow it from our children. - S.M.A.R.T Person
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help