Just like for the Intel/ACPI cpufreq code, powernow''s dom_mask also
must not be confined to the number of CPUs in the system.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Index: 2008-09-01/xen/arch/x86/acpi/cpufreq/powernow.c
==================================================================---
2008-09-01.orig/xen/arch/x86/acpi/cpufreq/powernow.c 2008-09-16
15:55:06.000000000 +0200
+++ 2008-09-01/xen/arch/x86/acpi/cpufreq/powernow.c 2008-09-16
15:55:23.000000000 +0200
@@ -241,9 +241,8 @@ int powernow_cpufreq_init(void)
{
unsigned int i, ret = 0;
unsigned int dom, max_dom = 0;
- cpumask_t *pt, dom_mask;
-
- cpus_clear(dom_mask);
+ cpumask_t *pt;
+ unsigned long *dom_mask;
for_each_online_cpu(i) {
struct cpuinfo_x86 *c = &cpu_data[i];
@@ -258,20 +257,26 @@ int powernow_cpufreq_init(void)
}
if (ret)
return ret;
- cpu_set(processor_pminfo[i].perf.domain_info.domain, dom_mask);
if (max_dom < processor_pminfo[i].perf.domain_info.domain)
max_dom = processor_pminfo[i].perf.domain_info.domain;
}
max_dom++;
+ dom_mask = xmalloc_array(unsigned long, BITS_TO_LONGS(max_dom));
+ if (!dom_mask)
+ return -ENOMEM;
+ bitmap_zero(dom_mask, max_dom);
+
pt = xmalloc_array(cpumask_t, max_dom);
if (!pt)
return -ENOMEM;
memset(pt, 0, max_dom * sizeof(cpumask_t));
/* get cpumask of each psd domain */
- for_each_online_cpu(i)
+ for_each_online_cpu(i) {
+ __set_bit(processor_pminfo[i].perf.domain_info.domain, dom_mask);
cpu_set(i, pt[processor_pminfo[i].perf.domain_info.domain]);
+ }
for_each_online_cpu(i)
processor_pminfo[i].perf.shared_cpu_map =
@@ -289,8 +294,8 @@ int powernow_cpufreq_init(void)
}
/* setup ondemand cpufreq */
- for (dom=0; dom<max_dom; dom++) {
- if (!cpu_isset(dom, dom_mask))
+ for (dom = 0; dom < max_dom; dom++) {
+ if (!test_bit(dom, dom_mask))
continue;
i = first_cpu(pt[dom]);
ret = cpufreq_governor_dbs(&xen_px_policy[i], CPUFREQ_GOV_START);
@@ -300,6 +305,7 @@ int powernow_cpufreq_init(void)
cpufreq_init_out:
xfree(pt);
+ xfree(dom_mask);
return ret;
}
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel