[PATCH 0/4] NR_CPUS: non-x86 arch specific reduction of NR_CPUS usage

[PATCH 0/4] NR_CPUS: non-x86 arch specific reduction of NR_CPUS usage

[Mike Travis]

Here's another round of removing static allocations of arrays
using NR_CPUS to size the length.  The change is to use PER_CPU
variables in place of the static tables.

Based on linux-2.6.git + x86.git

Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: cpufreq@lists.linux.org.uk
Cc: Len Brown <len.brown@intel.com>
Cc: linux-acpi@vger.kernel.org
Cc: Philippe Elie <phil.el@wanadoo.fr>
Cc: oprofile-list@lists.sf.net

Signed-off-by: Mike Travis <travis@sgi.com>
---
(1 - if modules enabled, does not complete boot even
     without this patch)

x86_64 configs built and booted:

    ingo-stress-test(1)
    defconfi
    nonuma
    nosmp
    bigsmp (NR_CPUS=1024, 1024 possible, 8 real)

Other configs built:

    arm-default
    i386-default
    i386-single
    i386-smp
    ppc-pmac32
    ppc-smp
    sparc64-default
    sparc64-smp
    x86_64-8psmp
    x86_64-debug
    x86_64-default
    x86_64-numa
    x86_64-single
    x86_64-allmodconfig
    x86_64-allyesconfig
    x86_64-maxsmp (NR_CPUS=4096 MAXNODES=512)

Configs not built due to prior errors:

    ia64-sn2
    ia64-default
    ia64-nosmp
    ia64-zx1
    s390-default
    sparc-default

Memory effects using x86_64-maxsmp

    Removes 1MB from permanant data adding 440 bytes to percpu area.

4k-cpus-before                      4k-cpus-after
   3463036 .bss                         -98304 -2%
   6067456 .data.cacheline_alig       -1048576 -17%
     48712 .data.percpu                   +440 +0%
  14275505 .text                          +336 +0%

  14275505 Text                           +336 +0%
   8521495 Data                           -336 +0%
   3463036 Bss                          -98304 -2%
  10974520 OtherData                  -1048576 -9%
     48712 PerCpu                         +440 +0%
  39275796 Total                      -1146104 -2%

-- 

[PATCH 1/4] cpufreq: change cpu freq tables to per_cpu variables

[Mike Travis]

[-- Attachment #1: nr_cpus-in-cpufreq --]
[-- Type: text/plain, Size: 5159 bytes --]

Change cpu frequency tables from arrays to per_cpu variables.

Based on linux-2.6.git + x86.git

Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: cpufreq@lists.linux.org.uk
Signed-off-by: Mike Travis <travis@sgi.com>
---
 drivers/cpufreq/cpufreq_userspace.c |   71 +++++++++++++++++++-----------------
 1 file changed, 39 insertions(+), 32 deletions(-)

--- a/drivers/cpufreq/cpufreq_userspace.c
+++ b/drivers/cpufreq/cpufreq_userspace.c
@@ -30,11 +30,11 @@
 /**
  * A few values needed by the userspace governor
  */
-static unsigned int	cpu_max_freq[NR_CPUS];
-static unsigned int	cpu_min_freq[NR_CPUS];
-static unsigned int	cpu_cur_freq[NR_CPUS]; /* current CPU freq */
-static unsigned int	cpu_set_freq[NR_CPUS]; /* CPU freq desired by userspace */
-static unsigned int	cpu_is_managed[NR_CPUS];
+DEFINE_PER_CPU(int, cpu_max_freq);
+DEFINE_PER_CPU(int, cpu_min_freq);
+DEFINE_PER_CPU(int, cpu_cur_freq); /* current CPU freq */
+DEFINE_PER_CPU(int, cpu_set_freq); /* CPU freq desired by userspace */
+DEFINE_PER_CPU(int, cpu_is_managed);
 
 static DEFINE_MUTEX	(userspace_mutex);
 static int cpus_using_userspace_governor;
@@ -48,12 +48,12 @@ userspace_cpufreq_notifier(struct notifi
 {
         struct cpufreq_freqs *freq = data;
 
-	if (!cpu_is_managed[freq->cpu])
+	if (!per_cpu(cpu_is_managed, freq->cpu))
 		return 0;
 
 	dprintk("saving cpu_cur_freq of cpu %u to be %u kHz\n",
 			freq->cpu, freq->new);
-	cpu_cur_freq[freq->cpu] = freq->new;
+	per_cpu(cpu_cur_freq, freq->cpu) = freq->new;
 
         return 0;
 }
@@ -77,15 +77,15 @@ static int cpufreq_set(unsigned int freq
 	dprintk("cpufreq_set for cpu %u, freq %u kHz\n", policy->cpu, freq);
 
 	mutex_lock(&userspace_mutex);
-	if (!cpu_is_managed[policy->cpu])
+	if (!per_cpu(cpu_is_managed, policy->cpu))
 		goto err;
 
-	cpu_set_freq[policy->cpu] = freq;
+	per_cpu(cpu_set_freq, policy->cpu) = freq;
 
-	if (freq < cpu_min_freq[policy->cpu])
-		freq = cpu_min_freq[policy->cpu];
-	if (freq > cpu_max_freq[policy->cpu])
-		freq = cpu_max_freq[policy->cpu];
+	if (freq < per_cpu(cpu_min_freq, policy->cpu))
+		freq = per_cpu(cpu_min_freq, policy->cpu);
+	if (freq > per_cpu(cpu_max_freq, policy->cpu))
+		freq = per_cpu(cpu_max_freq, policy->cpu);
 
 	/*
 	 * We're safe from concurrent calls to ->target() here
@@ -105,7 +105,7 @@ static int cpufreq_set(unsigned int freq
 /************************** sysfs interface ************************/
 static ssize_t show_speed (struct cpufreq_policy *policy, char *buf)
 {
-	return sprintf (buf, "%u\n", cpu_cur_freq[policy->cpu]);
+	return sprintf (buf, "%u\n", per_cpu(cpu_cur_freq, policy->cpu));
 }
 
 static ssize_t
@@ -154,12 +154,16 @@ static int cpufreq_governor_userspace(st
 		}
 		cpus_using_userspace_governor++;
 
-		cpu_is_managed[cpu] = 1;
-		cpu_min_freq[cpu] = policy->min;
-		cpu_max_freq[cpu] = policy->max;
-		cpu_cur_freq[cpu] = policy->cur;
-		cpu_set_freq[cpu] = policy->cur;
-		dprintk("managing cpu %u started (%u - %u kHz, currently %u kHz)\n", cpu, cpu_min_freq[cpu], cpu_max_freq[cpu], cpu_cur_freq[cpu]);
+		per_cpu(cpu_is_managed, cpu) = 1;
+		per_cpu(cpu_min_freq, cpu) = policy->min;
+		per_cpu(cpu_max_freq, cpu) = policy->max;
+		per_cpu(cpu_cur_freq, cpu) = policy->cur;
+		per_cpu(cpu_set_freq, cpu) = policy->cur;
+		dprintk("managing cpu %u started "
+			"(%u - %u kHz, currently %u kHz)\n", cpu,
+				per_cpu(cpu_min_freq, cpu),
+				per_cpu(cpu_max_freq, cpu),
+				per_cpu(cpu_cur_freq, cpu));
 start_out:
 		mutex_unlock(&userspace_mutex);
 		break;
@@ -172,11 +176,12 @@ start_out:
 					CPUFREQ_TRANSITION_NOTIFIER);
 		}
 
-		cpu_is_managed[cpu] = 0;
-		cpu_min_freq[cpu] = 0;
-		cpu_max_freq[cpu] = 0;
-		cpu_set_freq[cpu] = 0;
-		sysfs_remove_file (&policy->kobj, &freq_attr_scaling_setspeed.attr);
+		per_cpu(cpu_is_managed, cpu) = 0;
+		per_cpu(cpu_min_freq, cpu) = 0;
+		per_cpu(cpu_max_freq, cpu) = 0;
+		per_cpu(cpu_set_freq, cpu) = 0;
+		sysfs_remove_file (&policy->kobj,
+					&freq_attr_scaling_setspeed.attr);
 		dprintk("managing cpu %u stopped\n", cpu);
 		mutex_unlock(&userspace_mutex);
 		break;
@@ -185,22 +190,24 @@ start_out:
 		dprintk("limit event for cpu %u: %u - %u kHz,"
 			"currently %u kHz, last set to %u kHz\n",
 			cpu, policy->min, policy->max,
-			cpu_cur_freq[cpu], cpu_set_freq[cpu]);
-		if (policy->max < cpu_set_freq[cpu]) {
+			per_cpu(cpu_cur_freq, cpu),
+			per_cpu(cpu_set_freq, cpu));
+		if (policy->max < per_cpu(cpu_set_freq, cpu)) {
 			__cpufreq_driver_target(policy, policy->max,
 						CPUFREQ_RELATION_H);
 		}
-		else if (policy->min > cpu_set_freq[cpu]) {
+		else if (policy->min > per_cpu(cpu_set_freq, cpu)) {
 			__cpufreq_driver_target(policy, policy->min,
 						CPUFREQ_RELATION_L);
 		}
 		else {
-			__cpufreq_driver_target(policy, cpu_set_freq[cpu],
+			__cpufreq_driver_target(policy,
+						per_cpu(cpu_set_freq, cpu),
 						CPUFREQ_RELATION_L);
 		}
-		cpu_min_freq[cpu] = policy->min;
-		cpu_max_freq[cpu] = policy->max;
-		cpu_cur_freq[cpu] = policy->cur;
+		per_cpu(cpu_min_freq, cpu) = policy->min;
+		per_cpu(cpu_max_freq, cpu) = policy->max;
+		per_cpu(cpu_cur_freq, cpu) = policy->cur;
 		mutex_unlock(&userspace_mutex);
 		break;
 	}

-- 

[PATCH 2/4] acpi: change cpufreq tables to per_cpu variables

[Mike Travis]

[-- Attachment #1: nr_cpus-in-acpi-driver --]
[-- Type: text/plain, Size: 2537 bytes --]

Change cpufreq tables from arrays to per_cpu variables in
drivers/acpi/processor_thermal.c

Based on linux-2.6.git + x86.git

Cc: Len Brown <len.brown@intel.com>
Cc: linux-acpi@vger.kernel.org
Signed-off-by: Mike Travis <travis@sgi.com>
---
 drivers/acpi/processor_thermal.c |   21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)

--- a/drivers/acpi/processor_thermal.c
+++ b/drivers/acpi/processor_thermal.c
@@ -93,7 +93,7 @@ static int acpi_processor_apply_limit(st
  * _any_ cpufreq driver and not only the acpi-cpufreq driver.
  */
 
-static unsigned int cpufreq_thermal_reduction_pctg[NR_CPUS];
+static DEFINE_PER_CPU(unsigned int, cpufreq_thermal_reduction_pctg);
 static unsigned int acpi_thermal_cpufreq_is_init = 0;
 
 static int cpu_has_cpufreq(unsigned int cpu)
@@ -109,8 +109,8 @@ static int acpi_thermal_cpufreq_increase
 	if (!cpu_has_cpufreq(cpu))
 		return -ENODEV;
 
-	if (cpufreq_thermal_reduction_pctg[cpu] < 60) {
-		cpufreq_thermal_reduction_pctg[cpu] += 20;
+	if (per_cpu(cpufreq_thermal_reduction_pctg, cpu) < 60) {
+		per_cpu(cpufreq_thermal_reduction_pctg, cpu) += 20;
 		cpufreq_update_policy(cpu);
 		return 0;
 	}
@@ -123,13 +123,13 @@ static int acpi_thermal_cpufreq_decrease
 	if (!cpu_has_cpufreq(cpu))
 		return -ENODEV;
 
-	if (cpufreq_thermal_reduction_pctg[cpu] > 20)
-		cpufreq_thermal_reduction_pctg[cpu] -= 20;
+	if (per_cpu(cpufreq_thermal_reduction_pctg, cpu) > 20)
+		per_cpu(cpufreq_thermal_reduction_pctg, cpu) -= 20;
 	else
-		cpufreq_thermal_reduction_pctg[cpu] = 0;
+		per_cpu(cpufreq_thermal_reduction_pctg, cpu) = 0;
 	cpufreq_update_policy(cpu);
 	/* We reached max freq again and can leave passive mode */
-	return !cpufreq_thermal_reduction_pctg[cpu];
+	return !per_cpu(cpufreq_thermal_reduction_pctg, cpu);
 }
 
 static int acpi_thermal_cpufreq_notifier(struct notifier_block *nb,
@@ -143,7 +143,7 @@ static int acpi_thermal_cpufreq_notifier
 
 	max_freq =
 	    (policy->cpuinfo.max_freq *
-	     (100 - cpufreq_thermal_reduction_pctg[policy->cpu])) / 100;
+	     (100 - per_cpu(cpufreq_thermal_reduction_pctg, policy->cpu))) / 100;
 
 	cpufreq_verify_within_limits(policy, 0, max_freq);
 
@@ -159,8 +159,9 @@ void acpi_thermal_cpufreq_init(void)
 {
 	int i;
 
-	for (i = 0; i < NR_CPUS; i++)
-		cpufreq_thermal_reduction_pctg[i] = 0;
+	for (i = 0; i < nr_cpu_ids; i++)
+		if (cpu_present(i))
+			per_cpu(cpufreq_thermal_reduction_pctg, i) = 0;
 
 	i = cpufreq_register_notifier(&acpi_thermal_cpufreq_notifier_block,
 				      CPUFREQ_POLICY_NOTIFIER);

-- 

[PATCH 3/4] oprofile: change cpu_buffer from array to per_cpu variable

[Mike Travis]

[-- Attachment #1: nr_cpus-in-cpu_buffer --]
[-- Type: text/plain, Size: 4277 bytes --]

Change cpu_buffer from array to per_cpu variable in
oprofile functions.

Based on linux-2.6.git + x86.git

Cc: Philippe Elie <phil.el@wanadoo.fr>
Cc: oprofile-list@lists.sf.net
Signed-off-by: Mike Travis <travis@sgi.com>
---
 drivers/oprofile/buffer_sync.c    |    2 +-
 drivers/oprofile/cpu_buffer.c     |   16 ++++++++--------
 drivers/oprofile/cpu_buffer.h     |    3 ++-
 drivers/oprofile/oprofile_stats.c |    4 ++--
 4 files changed, 13 insertions(+), 12 deletions(-)

--- a/drivers/oprofile/buffer_sync.c
+++ b/drivers/oprofile/buffer_sync.c
@@ -494,7 +494,7 @@ typedef enum {
  */
 void sync_buffer(int cpu)
 {
-	struct oprofile_cpu_buffer * cpu_buf = &cpu_buffer[cpu];
+	struct oprofile_cpu_buffer * cpu_buf = &per_cpu(cpu_buffer, cpu);
 	struct mm_struct *mm = NULL;
 	struct task_struct * new;
 	unsigned long cookie = 0;
--- a/drivers/oprofile/cpu_buffer.c
+++ b/drivers/oprofile/cpu_buffer.c
@@ -27,7 +27,7 @@
 #include "buffer_sync.h"
 #include "oprof.h"
 
-struct oprofile_cpu_buffer cpu_buffer[NR_CPUS] __cacheline_aligned;
+DEFINE_PER_CPU_SHARED_ALIGNED(struct oprofile_cpu_buffer, cpu_buffer);
 
 static void wq_sync_buffer(struct work_struct *work);
 
@@ -39,7 +39,7 @@ void free_cpu_buffers(void)
 	int i;
  
 	for_each_online_cpu(i)
-		vfree(cpu_buffer[i].buffer);
+		vfree(per_cpu(cpu_buffer, i).buffer);
 }
 
 int alloc_cpu_buffers(void)
@@ -49,7 +49,7 @@ int alloc_cpu_buffers(void)
 	unsigned long buffer_size = fs_cpu_buffer_size;
  
 	for_each_online_cpu(i) {
-		struct oprofile_cpu_buffer * b = &cpu_buffer[i];
+		struct oprofile_cpu_buffer * b = &per_cpu(cpu_buffer, i);
  
 		b->buffer = vmalloc_node(sizeof(struct op_sample) * buffer_size,
 			cpu_to_node(i));
@@ -83,7 +83,7 @@ void start_cpu_work(void)
 	work_enabled = 1;
 
 	for_each_online_cpu(i) {
-		struct oprofile_cpu_buffer * b = &cpu_buffer[i];
+		struct oprofile_cpu_buffer * b = &per_cpu(cpu_buffer, i);
 
 		/*
 		 * Spread the work by 1 jiffy per cpu so they dont all
@@ -100,7 +100,7 @@ void end_cpu_work(void)
 	work_enabled = 0;
 
 	for_each_online_cpu(i) {
-		struct oprofile_cpu_buffer * b = &cpu_buffer[i];
+		struct oprofile_cpu_buffer * b = &per_cpu(cpu_buffer, i);
 
 		cancel_delayed_work(&b->work);
 	}
@@ -227,7 +227,7 @@ static void oprofile_end_trace(struct op
 void oprofile_add_ext_sample(unsigned long pc, struct pt_regs * const regs,
 				unsigned long event, int is_kernel)
 {
-	struct oprofile_cpu_buffer * cpu_buf = &cpu_buffer[smp_processor_id()];
+	struct oprofile_cpu_buffer * cpu_buf = &__get_cpu_var(cpu_buffer);
 
 	if (!backtrace_depth) {
 		log_sample(cpu_buf, pc, is_kernel, event);
@@ -254,13 +254,13 @@ void oprofile_add_sample(struct pt_regs 
 
 void oprofile_add_pc(unsigned long pc, int is_kernel, unsigned long event)
 {
-	struct oprofile_cpu_buffer * cpu_buf = &cpu_buffer[smp_processor_id()];
+	struct oprofile_cpu_buffer * cpu_buf = &__get_cpu_var(cpu_buffer);
 	log_sample(cpu_buf, pc, is_kernel, event);
 }
 
 void oprofile_add_trace(unsigned long pc)
 {
-	struct oprofile_cpu_buffer * cpu_buf = &cpu_buffer[smp_processor_id()];
+	struct oprofile_cpu_buffer * cpu_buf = &__get_cpu_var(cpu_buffer);
 
 	if (!cpu_buf->tracing)
 		return;
--- a/drivers/oprofile/cpu_buffer.h
+++ b/drivers/oprofile/cpu_buffer.h
@@ -14,6 +14,7 @@
 #include <linux/spinlock.h>
 #include <linux/workqueue.h>
 #include <linux/cache.h>
+#include <linux/sched.h>
  
 struct task_struct;
  
@@ -47,7 +48,7 @@ struct oprofile_cpu_buffer {
 	struct delayed_work work;
 } ____cacheline_aligned;
 
-extern struct oprofile_cpu_buffer cpu_buffer[];
+DECLARE_PER_CPU(struct oprofile_cpu_buffer, cpu_buffer);
 
 void cpu_buffer_reset(struct oprofile_cpu_buffer * cpu_buf);
 
--- a/drivers/oprofile/oprofile_stats.c
+++ b/drivers/oprofile/oprofile_stats.c
@@ -23,7 +23,7 @@ void oprofile_reset_stats(void)
 	int i;
  
 	for_each_possible_cpu(i) {
-		cpu_buf = &cpu_buffer[i]; 
+		cpu_buf = &per_cpu(cpu_buffer, i);
 		cpu_buf->sample_received = 0;
 		cpu_buf->sample_lost_overflow = 0;
 		cpu_buf->backtrace_aborted = 0;
@@ -49,7 +49,7 @@ void oprofile_create_stats_files(struct 
 		return;
 
 	for_each_possible_cpu(i) {
-		cpu_buf = &cpu_buffer[i]; 
+		cpu_buf = &per_cpu(cpu_buffer, i);
 		snprintf(buf, 10, "cpu%d", i);
 		cpudir = oprofilefs_mkdir(sb, dir, buf);
  

-- 

[PATCH 4/4] x86: minor cleanup of comments in processor.h

[Mike Travis]

[-- Attachment #1: cleanup-processor.h --]
[-- Type: text/plain, Size: 521 bytes --]

Removal of trivial comments in processor.h

Based on linux-2.6.git + x86.git

Signed-off-by: Mike Travis <travis@sgi.com>
---
 include/asm-x86/processor.h |    4 ----
 1 file changed, 4 deletions(-)

--- a/include/asm-x86/processor.h
+++ b/include/asm-x86/processor.h
@@ -302,10 +302,6 @@ union i387_union {
 };
 
 #ifdef CONFIG_X86_32
-/*
- * the following now lives in the per cpu area:
- * extern	int cpu_llc_id[NR_CPUS];
- */
 DECLARE_PER_CPU(u8, cpu_llc_id);
 #else
 DECLARE_PER_CPU(struct orig_ist, orig_ist);

-- 

Re: [PATCH 1/4] cpufreq: change cpu freq tables to per_cpu variables

[Dave Jones]

On Fri, Feb 08, 2008 at 03:37:39PM -0800, Mike Travis wrote:
 > Change cpu frequency tables from arrays to per_cpu variables.
 > 
 > Based on linux-2.6.git + x86.git

Looks ok to me.   Would you like me to push this though cpufreq.git,
or do you want the series to go through all in one?

	Dave

-- 
http://www.codemonkey.org.uk

Re: [PATCH 1/4] cpufreq: change cpu freq tables to per_cpu variables

[Mike Travis]

Dave Jones wrote:
> On Fri, Feb 08, 2008 at 03:37:39PM -0800, Mike Travis wrote:
>  > Change cpu frequency tables from arrays to per_cpu variables.
>  > 
>  > Based on linux-2.6.git + x86.git
> 
> Looks ok to me.   Would you like me to push this though cpufreq.git,
> or do you want the series to go through all in one?
> 
> 	Dave
> 

Thanks Dave.  The patches are pretty much independent but it is
easier to keep track of them if they go in together.  Btw, I have
another set coming shortly that I'm testing now.  It should remove
most of the remaining references to NR_CPUS.

Thanks again,
Mike

Re: [PATCH 1/4] cpufreq: change cpu freq tables to per_cpu variables

[Dave Jones]

On Mon, Feb 11, 2008 at 09:32:19AM -0800, Mike Travis wrote:
 > Dave Jones wrote:
 > > On Fri, Feb 08, 2008 at 03:37:39PM -0800, Mike Travis wrote:
 > >  > Change cpu frequency tables from arrays to per_cpu variables.
 > >  > 
 > >  > Based on linux-2.6.git + x86.git
 > > 
 > > Looks ok to me.   Would you like me to push this though cpufreq.git,
 > > or do you want the series to go through all in one?
 > 
 > Thanks Dave.  The patches are pretty much independent but it is
 > easier to keep track of them if they go in together.

No problem.  Feel free to add my
Signed-off-by: Dave Jones <davej@redhat.com>

 >  Btw, I have
 > another set coming shortly that I'm testing now.  It should remove
 > most of the remaining references to NR_CPUS.

Cool! As a distro kernel maintainer, this is appreciated.
Keeping everyone happy isn't easy, and work like this definitly
goes a long way towards that goal.

	Dave

-- 
http://www.codemonkey.org.uk

Re: [PATCH 2/4] acpi: change cpufreq tables to per_cpu variables

[Andrew Morton]

On Fri, 08 Feb 2008 15:37:40 -0800
Mike Travis <travis@sgi.com> wrote:

> Change cpufreq tables from arrays to per_cpu variables in
> drivers/acpi/processor_thermal.c
> 
> Based on linux-2.6.git + x86.git

I fixed a bunch of rejects in "[PATCH 1/4] cpufreq: change cpu freq tables
to per_cpu variables" and it compiles OK.  But this one was beyond my
should-i-repair-it threshold, sorry.

Re: [PATCH 2/4] acpi: change cpufreq tables to per_cpu variables

[Mike Travis]

Andrew Morton wrote:
> On Fri, 08 Feb 2008 15:37:40 -0800
> Mike Travis <travis@sgi.com> wrote:
> 
>> Change cpufreq tables from arrays to per_cpu variables in
>> drivers/acpi/processor_thermal.c
>>
>> Based on linux-2.6.git + x86.git
> 
> I fixed a bunch of rejects in "[PATCH 1/4] cpufreq: change cpu freq tables
> to per_cpu variables" and it compiles OK.  But this one was beyond my
> should-i-repair-it threshold, sorry.

Should I rebase all the pending patches on 2.6.25-rc1 or 2.6.24-mm1
(or some other combination)?

Thanks,
Mike

Re: [PATCH 2/4] acpi: change cpufreq tables to per_cpu variables

[Andrew Morton]

On Wed, 13 Feb 2008 10:10:00 -0800
Mike Travis <travis@sgi.com> wrote:

> Andrew Morton wrote:
> > On Fri, 08 Feb 2008 15:37:40 -0800
> > Mike Travis <travis@sgi.com> wrote:
> > 
> >> Change cpufreq tables from arrays to per_cpu variables in
> >> drivers/acpi/processor_thermal.c
> >>
> >> Based on linux-2.6.git + x86.git
> > 
> > I fixed a bunch of rejects in "[PATCH 1/4] cpufreq: change cpu freq tables
> > to per_cpu variables" and it compiles OK.  But this one was beyond my
> > should-i-repair-it threshold, sorry.
> 
> Should I rebase all the pending patches on 2.6.25-rc1 or 2.6.24-mm1
> (or some other combination)?
> 

That depends on whether you have other things queued in one of the git
trees.  If not, against current mainline (which is later than 2.6.25-rc1!)
would suit.