Skip to contentSkip to conversation

cpufreq: unbreak build for !CONFIG_SMP

PATCH New Daniel Mack 2 participants 3 messages Jul 8, 2009 – Jul 9, 2009 CPU frequency scaling framework

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 cpufreq: unbreak build for !CONFIG_SMP

'managed_policy' is used for the non-smp case in line 941.

Remove #ifdef CONFIG_SMP for its declaration in order to make cpufreq.c build on systems with CONFIG_SMP=n.

Cc Mathieu Desnoyers <[email protected]>
Cc Venkatesh Pallipadi <[email protected]>
Cc Shaohua Li <[email protected]>
Cc Pekka Enberg <[email protected]>
Cc Dave Young <[email protected]>
Cc "Rafael J. Wysocki" <[email protected]>
Cc Rusty Russell <[email protected]>
Cc Thomas Renninger <[email protected]>
Cc Dave Jones <[email protected]>
---
 drivers/cpufreq/cpufreq.c |    2 --
 1 files changed, 0 insertions(+), 2 deletions(-)
drivers/cpufreq/cpufreq.c +0 -2
index c668ac8..d160867 100644
@@ -776,9 +776,7 @@ static int cpufreq_add_dev(struct sys_device *sys_dev)
 	struct sys_device *cpu_sys_dev;
 	unsigned long flags;
 	unsigned int j;
-#ifdef CONFIG_SMP
 	struct cpufreq_policy *managed_policy;
-#endif
 
 	if (cpu_is_offline(cpu))
 		return 0;
Signature · 1 line
1.6.3.1
#1 lore raw

PATCH 1/1 cpufreq: Unbreak build for !SMP: symlinked cpus cannot happen for !SMP

Or like that:
Either all #ifdef CONFIG_SMP related conditionals to managed cpus should vanish or all should be covered. This one takes the latter approach.

---
'managed_policy' is used for the non-smp case in line 941.

Make the symlink code conditional to #ifdef CONFIG_SMP as already done at other places.

drivers/cpufreq/cpufreq.c +2 -0
Index: cpufreq/drivers/cpufreq/cpufreq.c
===================================================================
@@ -930,6 +930,7 @@ static int cpufreq_add_dev(struct sys_de
 	}
 	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
 
+#ifdef CONFIG_SMP
 	/* symlink affected CPUs */
 	for_each_cpu(j, policy->cpus) {
 		if (j == cpu)
@@ -947,6 +948,7 @@ static int cpufreq_add_dev(struct sys_de
 			goto err_out_unregister;
 		}
 	}
+#endif
 
 	policy->governor = NULL; /* to assure that the starting sequence is
 				  * run in cpufreq_set_policy */
#2 lore raw

Re: [PATCH] cpufreq: Unbreak build for !SMP: symlinked cpus cannot happen for !SMP

Quoted 4 lines · Thomas Renninger

Or like that:
Either all #ifdef CONFIG_SMP related conditionals to managed cpus should vanish or all should be covered. This one takes the latter approach.

Dave Jones fixed it already, see

http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=5e1596f75395e7a402e1059c518e633d2732dcf8

Daniel

Quoted 29 lines

---
'managed_policy' is used for the non-smp case in line 941.

Make the symlink code conditional to #ifdef CONFIG_SMP as already done at other places.

Index: cpufreq/drivers/cpufreq/cpufreq.c
===================================================================

drivers/cpufreq/cpufreq.c +2 -0
@@ -930,6 +930,7 @@ static int cpufreq_add_dev(struct sys_de
 	}
 	spin_unlock_irqrestore(&cpufreq_driver_lock, flags);
 
+#ifdef CONFIG_SMP
 	/* symlink affected CPUs */
 	for_each_cpu(j, policy->cpus) {
 		if (j == cpu)
@@ -947,6 +948,7 @@ static int cpufreq_add_dev(struct sys_de
 			goto err_out_unregister;
 		}
 	}
+#endif
 
 	policy->governor = NULL; /* to assure that the starting sequence is
 				  * run in cpufreq_set_policy */