What's happening with the cpuidle code?

What's happening with the cpuidle code?

[Chuck Ebbert]

A look at the 'cpuidle' branch of git-acpi shows a commit
e40cede7d63a029e92712a3fe02faee60cc38fb4, "cpuidle: first
round of documentation updates" that doesn't show up in that
branch online. The entire Documentation/cpuidle directory
is missing from the tree when looking at the web pages, and
it's missing from git-acpi.patch in 2.6.23-rc4-mm1 (but the
patch shows up in the summary information in the patch
header.) Where did it go? And how can -mm be used to test
things if its patches don't even match their own headers?

Re: What's happening with the cpuidle code?

[Len Brown]

On Tuesday 04 September 2007 16:47, Chuck Ebbert wrote:
> A look at the 'cpuidle' branch of git-acpi shows a commit
> e40cede7d63a029e92712a3fe02faee60cc38fb4, "cpuidle: first
> round of documentation updates" that doesn't show up in that
> branch online. The entire Documentation/cpuidle directory
> is missing from the tree when looking at the web pages, and
> it's missing from git-acpi.patch in 2.6.23-rc4-mm1 (but the
> patch shows up in the summary information in the patch
> header.) Where did it go? And how can -mm be used to test
> things if its patches don't even match their own headers?

A later patch in that series, "cpuidle: re-write", reverted
the documentation from the intermediate patch that you refer to:

http://git.kernel.org/?p=linux/kernel/git/lenb/linux-acpi-2.6.git;a=commit;h=2305a5920fb8ee6ccec1c62ade05aa8351091d71

The cpuidle branch on git.kernel.org looks okay to me:

http://git.kernel.org/?p=linux/kernel/git/lenb/linux-acpi-2.6.git;a=shortlog;h=cpuidle

the top commit is this one:

commit 8975059a2c1e56cfe83d1bcf031bcf4cb39be743
Author: Adam Belay <abelay@novell.com>
Date:   Tue Aug 21 18:27:07 2007 -0400

    CPUIDLE: load ACPI properly when CPUIDLE is disabled

And this top patch is indeed included in the latest acpi test branch,
as well as the latest ACPI test patch here:

http://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/patches/test/2.6.23/acpi-test-20070126-2.6.23-rc5.diff.gz

Note that there were some merge conflicts when pulling cpuidle into 2.6.23,
so you are best off running either the acpi patch above on top of 2.6.23-rc5,
or the acpi test  branch, or the mm tree so you won't have to merge the
cpuidle branch onto your latest kernel again.

cheers,
-Len

Re: What's happening with the cpuidle code?

[Chuck Ebbert]

On 09/04/2007 05:43 PM, Len Brown wrote:
> On Tuesday 04 September 2007 16:47, Chuck Ebbert wrote:
>> A look at the 'cpuidle' branch of git-acpi shows a commit
>> e40cede7d63a029e92712a3fe02faee60cc38fb4, "cpuidle: first
>> round of documentation updates" that doesn't show up in that
>> branch online. The entire Documentation/cpuidle directory
>> is missing from the tree when looking at the web pages, and
>> it's missing from git-acpi.patch in 2.6.23-rc4-mm1 (but the
>> patch shows up in the summary information in the patch
>> header.) Where did it go? And how can -mm be used to test
>> things if its patches don't even match their own headers?
> 
> A later patch in that series, "cpuidle: re-write", reverted
> the documentation from the intermediate patch that you refer to:
> 
> http://git.kernel.org/?p=linux/kernel/git/lenb/linux-acpi-2.6.git;a=commit;h=2305a5920fb8ee6ccec1c62ade05aa8351091d71

It never occurred to me that a patch would just remove documentation.
Thanks for looking into that...