headers
Gabriel A. Devenyi | 3 Nov 2005 02:55
Picon

[PATCH] arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c

Thanks to LinuxICC (http://linuxicc.sf.net), a comparison of a u32 less than 0 was found, this patch
changes the variable to a signed int so that comparison is meaningful.

This patch applies to linus' git tree as of 02.11.2005

Signed-off-by: Gabriel A. Devenyi <ace <at> staticwave.ca>

--
diff --git a/arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c b/arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c
index 04a4053..039297b 100644
--- a/arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c
+++ b/arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c
 <at>  <at>  -177,9 +177,9  <at>  <at>  static unsigned int nforce2_fsb_read(int
  */
 static int nforce2_set_fsb(unsigned int fsb)
 {
-	u32 pll, temp = 0;
+	u32 temp = 0;
 	unsigned int tfsb;
-	int diff;
+	int diff, pll = 0;

 	if ((fsb > max_fsb) || (fsb < NFORCE2_MIN_FSB)) {
 		printk(KERN_ERR "cpufreq: FSB %d is out of range!\n", fsb);

--

-- 
Gabriel A. Devenyi
ace <at> staticwave.ca
Permalink | Reply | 0 comments |
headers
bugme-daemon | 5 Nov 2005 01:19

[Bug 5553] New: speedstep-smi fails to load if CONFIG_CPU_FREQ_DEBUG is not set

http://bugzilla.kernel.org/show_bug.cgi?id=5553

           Summary: speedstep-smi fails to load if CONFIG_CPU_FREQ_DEBUG is
                    not set
    Kernel Version: 2.6.14
            Status: NEW
          Severity: normal
             Owner: cpufreq <at> www.linux.org.uk
         Submitter: tm <at> del.bg

Most recent kernel where this bug did not occur:
 2.6.14
Distribution:
 Debian GNU/Linux Etch
Hardware Environment:
 IBM ThinkPad T22, Pentium III (Coppermine), i440BX
Software Environment:
Problem Description:

cpufreq used to work with 2.6.8 and 2.6.11 kernels, but doesn't work with 2.6.12
and 2.6.14 for me. I've found that it actually works with 2.6.14 but only if
kernel is compiled with CONFIG_CPU_FREQ_DEBUG option set.

Steps to reproduce:

Compile kernel with all allmodconfig and clear the CONFIG_CPU_FREQ_DEBUG.

Then the following happens (otherwise it works):

# modprobe speedstep-lib relaxed_check=1
(Continue reading)

Permalink | Reply | 0 comments |
headers
Pierre Ossman | 7 Nov 2005 09:36

sleeping function called from cpufreq

As of lately I've been getting tonnes of these:

[  610.185635] Debug: sleeping function called from invalid context at 
include/linux/rwsem.h:43
[  610.185647] in_atomic():1, irqs_disabled():0
[  610.185653]  [<c01041be>] dump_stack+0x1e/0x20
[  610.185667]  [<c0119b62>] __might_sleep+0xa2/0xc0
[  610.185678]  [<c029de86>] cpufreq_notify_transition+0x46/0x220
[  610.185690]  [<e09d08fc>] centrino_target+0xfc/0x130 [speedstep_centrino]
[  610.185708]  [<c029f17f>] __cpufreq_driver_target+0x5f/0x70
[  610.185718]  [<c02a029d>] cpufreq_set+0x7d/0xa0
[  610.185728]  [<c02a0339>] store_speed+0x49/0x50
[  610.185737]  [<c029e6c6>] store+0x46/0x60
[  610.185745]  [<c01a5f27>] flush_write_buffer+0x37/0x40
[  610.185754]  [<c01a5f98>] sysfs_write_file+0x68/0x90
[  610.185763]  [<c01639b8>] vfs_write+0xa8/0x190
[  610.185773]  [<c0163b57>] sys_write+0x47/0x70
[  610.185781]  [<c01032bb>] sysenter_past_esp+0x54/0x75

Ideas on solving it?

Rgds
Pierre
Permalink | Reply | 0 comments |
headers
Zwane Mwaikambo | 7 Nov 2005 09:57
Picon

Re: sleeping function called from cpufreq

On Mon, 7 Nov 2005, Pierre Ossman wrote:

> As of lately I've been getting tonnes of these:
> 
> [  610.185635] Debug: sleeping function called from invalid context at
> include/linux/rwsem.h:43
> [  610.185647] in_atomic():1, irqs_disabled():0
> [  610.185653]  [<c01041be>] dump_stack+0x1e/0x20
> [  610.185667]  [<c0119b62>] __might_sleep+0xa2/0xc0
> [  610.185678]  [<c029de86>] cpufreq_notify_transition+0x46/0x220
> [  610.185690]  [<e09d08fc>] centrino_target+0xfc/0x130 [speedstep_centrino]
> [  610.185708]  [<c029f17f>] __cpufreq_driver_target+0x5f/0x70
> [  610.185718]  [<c02a029d>] cpufreq_set+0x7d/0xa0
> [  610.185728]  [<c02a0339>] store_speed+0x49/0x50
> [  610.185737]  [<c029e6c6>] store+0x46/0x60
> [  610.185745]  [<c01a5f27>] flush_write_buffer+0x37/0x40
> [  610.185754]  [<c01a5f98>] sysfs_write_file+0x68/0x90
> [  610.185763]  [<c01639b8>] vfs_write+0xa8/0x190
> [  610.185773]  [<c0163b57>] sys_write+0x47/0x70
> [  610.185781]  [<c01032bb>] sysenter_past_esp+0x54/0x75
> 
> Ideas on solving it?

There is a fix for it in 2.6.14-mm1 but i think it also requires the 
following patch in order to link. I'm not actually advocating the hotplug 
in progress flag and would prefer using a previously posted 
preempt_disable/enable method.

--- linux-mm/include/linux/cpu.h.old	2005-11-06 22:08:39.000000000 -0500
+++ linux-mm/include/linux/cpu.h	2005-11-06 22:41:17.000000000 -0500
(Continue reading)

Permalink | Reply | 0 comments |
headers
Adrian Bunk | 7 Nov 2005 21:03
Picon

[2.6 patch] drivers/cpufreq/cpufreq.c: fix a compile warning with SMP=n

This patch fixes the following compile warning with CONFIG_SMP=n:

<--  snip  -->

...
  CC      drivers/cpufreq/cpufreq.o
drivers/cpufreq/cpufreq.c: In function 'cpufreq_remove_dev':
drivers/cpufreq/cpufreq.c:696: warning: unused variable 'cpu_sys_dev'
...

<--  snip  -->

Signed-off-by: Adrian Bunk <bunk <at> stusta.de>

--- linux-2.6.14-mm1-full/drivers/cpufreq/cpufreq.c.old	2005-11-07 19:27:33.000000000 +0100
+++ linux-2.6.14-mm1-full/drivers/cpufreq/cpufreq.c	2005-11-07 19:27:50.000000000 +0100
 <at>  <at>  -693,8 +693,8  <at>  <at> 
 	unsigned int cpu = sys_dev->id;
 	unsigned long flags;
 	struct cpufreq_policy *data;
-	struct sys_device *cpu_sys_dev;
 #ifdef CONFIG_SMP
+	struct sys_device *cpu_sys_dev;
 	unsigned int j;
 #endif
Permalink | Reply | 0 comments |
headers
Gabriel A. Devenyi | 8 Nov 2005 13:38
Picon

[RESEND] [PATCH] arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c

Thanks to LinuxICC (http://linuxicc.sf.net), a comparison of a u32 less than 0 was found, this patch
changes the variable to a signed int so that comparison is meaningful.

This patch applies to linus' git tree as of 02.11.2005

Signed-off-by: Gabriel A. Devenyi <ace <at> staticwave.ca>

--
diff --git a/arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c b/arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c
index 04a4053..039297b 100644
--- a/arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c
+++ b/arch/i386/kernel/cpu/cpufreq/cpufreq-nforce2.c
 <at>  <at>  -177,9 +177,9  <at>  <at>  static unsigned int nforce2_fsb_read(int
   */
  static int nforce2_set_fsb(unsigned int fsb)
  {
-	u32 pll, temp = 0;
+	u32 temp = 0;
  	unsigned int tfsb;
-	int diff;
+	int diff, pll = 0;

  	if ((fsb > max_fsb) || (fsb < NFORCE2_MIN_FSB)) {
  		printk(KERN_ERR "cpufreq: FSB %d is out of range!\n", fsb);

--

-- 
Gabriel A. Devenyi
ace <at> staticwave.ca
Permalink | Reply | 1 comment |
headers
bugme-daemon | 9 Nov 2005 23:53

[Bug 4059] cpufreq longhaul for Via Nehemiah C (C5N) freezes whole system

http://bugzilla.kernel.org/show_bug.cgi?id=4059

------- Additional Comments From ulrich <at> holeschak.de  2005-11-09 14:48 -------
I had the same problem. I only occurs if DMA is activated for IDE disk. I expect
that the systems dies when the DMA occurs while the systems is halted ...

In newer kernel versions (2.6.14) i new bug has been introduced:
The modification of the pci bus master mode prevents the 8139too network driver
from working. Removing the following lines in longhaul.c solves this problem
(but not the DMA crash!)

       /*
         * get current pci bus master state for all devices
         * and clear bus master bit
         */
        dev = NULL;
        i = 0;
        do {
                dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev);
                if (dev != NULL) {
                        pci_read_config_word(dev, PCI_COMMAND, &pci_cmd);
                        cmd_state[i++] = pci_cmd;
                        pci_cmd &= ~PCI_COMMAND_MASTER;
                        pci_write_config_word(dev, PCI_COMMAND, pci_cmd);
                }
        } while (dev != NULL);

...

        /* restore pci bus master state for all devices */
(Continue reading)

Permalink | Reply | 0 comments |
headers
Langsdorf, Mark | 10 Nov 2005 23:52
Picon
Favicon

[PATCH] 1/2 Improve Error reporting

This patch cleans up some error messages in the 
powernow-k8 driver and makes them more understandable.

Signed-off-by: Jacob Shin <jacob.shin <at> amd.com>
Signed-off-by: Mark Langsdorf <mark.langsdorf <at> amd.com>
Attachment (pn-cleanup.patch): application/octet-stream, 2900 bytes
_______________________________________________
Cpufreq mailing list
Cpufreq <at> lists.linux.org.uk
http://lists.linux.org.uk/mailman/listinfo/cpufreq
Permalink | Reply | 0 comments |
headers
Langsdorf, Mark | 10 Nov 2005 23:53
Picon
Favicon

[PATCH] 2/2 fix driver removal code to clean up all governors

This patch fixes a bug that causes a segfault when the
powernow-k8 driver is removed while the ondemand 
governor is running.

Signed-off-by: Jacob Shin <jacob.shin <at> amd.com>
Signed-off-by: Mark Langsdorf <mark.langsdorf <at> amd.com>
Attachment (cpufreq-gov_stop-fix.patch): application/octet-stream, 353 bytes
_______________________________________________
Cpufreq mailing list
Cpufreq <at> lists.linux.org.uk
http://lists.linux.org.uk/mailman/listinfo/cpufreq
Permalink | Reply | 1 comment |
headers
bugme-daemon | 11 Nov 2005 10:39

[Bug 5553] speedstep-smi fails to load if CONFIG_CPU_FREQ_DEBUG is not set

http://bugzilla.kernel.org/show_bug.cgi?id=5553

akpm <at> osdl.org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |akpm <at> osdl.org

------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.
Permalink | Reply | 0 comments |

Gmane