Re: linux-next: cpufreq-current tree build warning

[Stephen Rothwell]

Hi Dave,

On Wed, 16 Sep 2009 22:27:26 -0400 Dave Jones <davej@xxxxxxxxxx> wrote:
>
> On Thu, Sep 17, 2009 at 12:06:51PM +1000, Stephen Rothwell wrote:
>  > 
>  > Today's linux-next build (x86_64 allmodconfig) produced this warning:
>  > 
>  > arch/x86/kernel/cpu/cpufreq/powernow-k8.c: In function 'fill_powernow_table_fidvid':
>  > arch/x86/kernel/cpu/cpufreq/powernow-k8.c:1001: warning: passing argument 1 of 'invalidate_entry' from incompatible pointer type
>  > arch/x86/kernel/cpu/cpufreq/powernow-k8.c:608: note: expected 'struct cpufreq_frequency_table *' but argument is of type 'struct powernow_k8_data *'
>  > 
>  > Introduced by commit 7fc1b26613587ab4369615445b5485040c60fef1 ("[CPUFREQ]
>  > Fix NULL ptr regression in powernow-k8").
> 
> I'm very confused. I don't see this..
> 
> (22:22:25:davej:cpufreq)$ make arch/x86/kernel/cpu/cpufreq/powernow-k8.o
>   CHK     include/linux/version.h
>   CHK     include/linux/utsrelease.h
>   SYMLINK include/asm -> include/asm-x86
>   CALL    scripts/checksyscalls.sh
>   CC [M]  arch/x86/kernel/cpu/cpufreq/powernow-k8.o
> (22:22:30:davej:cpufreq)$ 
> 
> Also, line 1001 is ..
> 
> 	if (data->acpi_data.state_count)

That is line 1022 in my tree.

>From line 986 (in fucntion fill_powernow_table_fidvid):

		/* verify only 1 entry from the lo frequency table */
		if (fid < HI_FID_TABLE_BOTTOM) {
			if (cntlofreq) {
				/* if both entries are the same,
				 * ignore this one ... */
				if ((freq != powernow_table[cntlofreq].frequency) ||
				    (index != powernow_table[cntlofreq].index)) {
					printk(KERN_ERR PFX
						"Too many lo freq table "
						"entries\n");
					return 1;
				}

				dprintk("double low frequency table entry, "
						"ignoring it.\n");
				invalidate_entry(data, i);
				continue;
			} else
				cntlofreq = i;
		}

> So I'm not sure what your gcc is talking about.

This is gcc 4.4.0 if that is relevant.

> Does next contain additional patches to this file from outside the cpufreq.git tree?

$ git log origin/master.. arch/x86/kernel/cpu/cpufreq/powernow-k8.c
commit 7fc1b26613587ab4369615445b5485040c60fef1
Author: Kurt Roeckx <kurt@xxxxxxxxx>
Date:   Wed Sep 16 11:09:32 2009 -0400

    [CPUFREQ] Fix NULL ptr regression in powernow-k8
    
    Fixes bugzilla #13780
    
    From: Kurt Roeckx <kurt@xxxxxxxxx>
    Signed-off-by: Dave Jones <davej@xxxxxxxxxx>

so, no.
(origin/master is Linus' tree of about 2 hours ago, commit
ab86e5765d41a5eb4239a1c04d613db87bea5ed8 "Merge
git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core-2.6")

Looking further, this is the cpureq-current tree I am complaining about,
in the cpufreq tree, you also have:

commit db39d5529d347de5e2eec1a72d67fcfacae6c5a2
Author: Mark Langsdorf <mark.langsdorf@xxxxxxx>
Date:   Fri Aug 21 19:15:28 2009 -0500

    [CPUFREQ] Powernow-k8: Enable more than 2 low P-states
    
    Remove an obsolete check that used to prevent there being more
    than 2 low P-states.  Now that low-to-low P-states changes are
    enabled, it prevents otherwise workable configurations with
    multiple low P-states.
    
    Signed-off-by: Mark Langsdorf <mark.langsdorf@xxxxxxx>
    Tested-by: Krists Krilovs <pow@xxxxxxxxxx>
    Signed-off-by: Dave Jones <davej@xxxxxxxxxx>

Which removes the above code ...

-- 
Cheers,
Stephen Rothwell                    sfr@xxxxxxxxxxxxxxxx
http://www.canb.auug.org.au/~sfr/

Attachment: pgp00000.pgp
Description: PGP signature