Re: [PATCH 7/8] cpufreq: Preserve policy structure across suspend/resume

["Srivatsa S. Bhat" <srivatsa.bhat@linux.vnet.ibm.com>]

On 07/15/2013 03:51 PM, Viresh Kumar wrote:
> On 15 July 2013 15:35, Srivatsa S. Bhat
> <srivatsa.bhat@linux.vnet.ibm.com> wrote:
>> Actually even I was wondering about this while writing the patch and
>> I even tested shutdown after multiple suspend/resume cycles, to verify that
>> the refcount is messed up. But surprisingly, things worked just fine.
> 
> What kind of system have you tested it on?
> 

The system has 2 sockets with 8 cores each, and has Intel Sandybridge
CPUs. I had used a local patch to simulate CPU hotplug in the suspend-to-ram
path using the freeze state of pm_test (because I had other problems in
using the 'processors' state of pm_test). The patch is shown below:


diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c
index ece0422..fe07b77 100644
--- a/kernel/power/suspend.c
+++ b/kernel/power/suspend.c
@@ -342,8 +342,13 @@ static int enter_state(suspend_state_t state)
 	if (error)
 		goto Unlock;
 
-	if (suspend_test(TEST_FREEZER))
+	if (suspend_test(TEST_FREEZER)) {
+		pr_debug("Disabling nonboot CPUs\n");
+		disable_nonboot_cpus();
+		pr_debug("Enabling nonboot CPUs\n");
+		enable_nonboot_cpus();
 		goto Finish;
+	}
 
 	pr_debug("PM: Entering %s sleep\n", pm_states[state]);
 	pm_restrict_gfp_mask();


 
Regards,
Srivatsa S. Bhat