From: Bruce Ashfield <bruce.ashfield@windriver.com>
To: Darren Hart <dvhart@linux.intel.com>
Cc: yocto@yoctoproject.org
Subject: Re: [PATCH 0/7][KERNEL] add usb and mei features
Date: Sun, 6 May 2012 00:53:19 -0400 [thread overview]
Message-ID: <4FA603BF.5060506@windriver.com> (raw)
In-Reply-To: <4FA47B9A.1090109@linux.intel.com>
On 12-05-04 9:00 PM, Darren Hart wrote:
>
>
> On 05/04/2012 05:06 PM, Tom Zanussi wrote:
>> On Fri, 2012-05-04 at 16:04 -0700, Darren Hart wrote:
>>>
>>> On 05/04/2012 03:07 PM, tom.zanussi@intel.com wrote:
>>>> From: Tom Zanussi<tom.zanussi@intel.com>
>>>>
>>>> This adds a few new features, one each for usb/[xeou]hci-hcd and
>>>> another for amt/mei, and refactors existing config options into a new
>>>> usb/base, which are then used in crownbay.
>>>>
>>>> v2: after suggestions from Bruce and Darren of a preference for a more
>>>> flattened feature tree structure, moved all the features into a single
>>>> usb/ feature dir. I did however retain the common 'base' feature
>>>> because inlining that in each feature would mean that the same common
>>>> set would be repeated in multiple config fragments if more than one usb
>>>> feature was included in a BSP.
>>>
>>> As things stand now, a BSP would have to include both the base and the
>>> *hcd right?
>>>
>>> include usb/base.scc
>>> include ehci-hcd.scc
>>>
>>> Seems to me we should manage the dependencies wherever we can. When I've
>>> created fragments I've been including their requirements as well to make
>>> it easier on the integrator. Since you've put the dependencies in their
>>> own fragment, why not include the fragment in the *hcd.scc files?
>>>
>>> +++ b/meta/cfg/kernel-cache/features/usb/ehci-hcd.scc
>>> @@ -0,0 +1,4 @@
>>> +define KFEATURE_DESCRIPTION "Enable options for ehci (USB 2.0)"
>>> +define KFEATURE_COMPATIBILITY board
>>> +
>>> + include base.scc
>>> +kconf hardware ehci-hcd.cfg
>>>
>>> Now if someone need only include the one line for the *hcd. Since it's
>>> only a total of 3 lines of CONFIG settings we're packaging up here, it
>>> seems to me we should be able to enable it with only 1 line of meta data.
>>>
>>
>> Yes, this is exactly what I did in the previous patchset.
>>
>> The problem is that if we do this for each of the *hcd fragments, we end
>> up with the options in base.cfg repeated for each, which may or may not
>> produce a config-check warning.
>
> "May or may not"? If our tools complain about setting a config to the
> same value, I would prefer to address that in the tools.
>
> Bruce, any comment on the intended behavior of the tools?
That warning is supposed to only trigger if you literally are in the
same fragment. The switch to a new fragment should allow the duplication.
There is a secondary warning that yells if an option is redefined by
ANY fragment. That mode is to support a configuration scheme that says
all fragments should be exclusive and additive. I made that into a
flag some time ago.
That being said, the switch to merge_config (the upstream script), caused
a few things to be more difficult. This is one of them. I have some
changes to this already queued, and I'm not seeing anything like this
in my tests. I've got something that allows the multiple includes to
work fine (and stops short of a new keyword (although that it tempting
to reduce complexity)).
I'll run some configuration tests here when I merge the series, and see
if a regression snuck in.
Cheers,
Bruce
>
> --
> Darren
>
>>
>> To avoid that, we need to be able to add the base.cfg items once
>> followed by the individual *hcd items as I did with the crownbay
>> example:
>>
>> +include features/usb/base.scc
>> +include features/usb/ehci-hcd.scc
>> +include features/usb/ohci-hcd.scc
>> +include features/usb/uhci-hcd.scc
>>
>>
>> Tom
>>
>>> --
>>> Darren
>>>
>>>>
>>>> Please pull into linux-yocto-3.2.
>>>>
>>>> Thanks,
>>>>
>>>> Tom
>>>>
>>>> The following changes since commit b14a08f5c7b469a5077c10942f4e1aec171faa9d:
>>>> Yang Shi (1):
>>>> meta: Clean up BSPs kernel config
>>>>
>>>> are available in the git repository at:
>>>>
>>>> git://git.yoctoproject.org/linux-yocto-2.6.37-contrib.git tzanussi/xhcd-mei-features
>>>> http://git.yoctoproject.org/cgit.cgi//log/?h=tzanussi/xhcd-mei-features
>>>>
>>>> Tom Zanussi (7):
>>>> meta: add usb/base feature
>>>> meta: add usb/xhci-hcd feature
>>>> meta: add usb/ehci-hcd feature
>>>> meta: add usb/ohci-hcd feature
>>>> meta: add usb/uhci-hcd feature
>>>> meta/crownbay: use usb features
>>>> meta: add mei feature
>>>>
>>>> meta/cfg/kernel-cache/bsp/crownbay/crownbay.cfg | 6 ------
>>>> meta/cfg/kernel-cache/bsp/crownbay/crownbay.scc | 5 +++++
>>>> meta/cfg/kernel-cache/features/amt/mei/mei.cfg | 3 +++
>>>> meta/cfg/kernel-cache/features/amt/mei/mei.scc | 4 ++++
>>>> meta/cfg/kernel-cache/features/usb/base.cfg | 3 +++
>>>> meta/cfg/kernel-cache/features/usb/base.scc | 4 ++++
>>>> meta/cfg/kernel-cache/features/usb/ehci-hcd.cfg | 1 +
>>>> meta/cfg/kernel-cache/features/usb/ehci-hcd.scc | 4 ++++
>>>> meta/cfg/kernel-cache/features/usb/ohci-hcd.cfg | 1 +
>>>> meta/cfg/kernel-cache/features/usb/ohci-hcd.scc | 4 ++++
>>>> meta/cfg/kernel-cache/features/usb/uhci-hcd.cfg | 1 +
>>>> meta/cfg/kernel-cache/features/usb/uhci-hcd.scc | 4 ++++
>>>> meta/cfg/kernel-cache/features/usb/xhci-hcd.cfg | 1 +
>>>> meta/cfg/kernel-cache/features/usb/xhci-hcd.scc | 4 ++++
>>>> 14 files changed, 39 insertions(+), 6 deletions(-)
>>>> create mode 100644 meta/cfg/kernel-cache/features/amt/mei/mei.cfg
>>>> create mode 100644 meta/cfg/kernel-cache/features/amt/mei/mei.scc
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/base.cfg
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/base.scc
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/ehci-hcd.cfg
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/ehci-hcd.scc
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/ohci-hcd.cfg
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/ohci-hcd.scc
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/uhci-hcd.cfg
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/uhci-hcd.scc
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/xhci-hcd.cfg
>>>> create mode 100644 meta/cfg/kernel-cache/features/usb/xhci-hcd.scc
>>>>
>>>
>>
>>
>
next prev parent reply other threads:[~2012-05-06 4:53 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-05-04 22:07 [PATCH 0/7][KERNEL] add usb and mei features tom.zanussi
2012-05-04 22:07 ` [PATCH 1/7][KERNEL] meta: add usb/base feature tom.zanussi
2012-05-04 22:07 ` [PATCH 2/7][KERNEL] meta: add usb/xhci-hcd feature tom.zanussi
2012-05-04 22:07 ` [PATCH 3/7][KERNEL] meta: add usb/ehci-hcd feature tom.zanussi
2012-05-04 22:07 ` [PATCH 4/7][KERNEL] meta: add usb/ohci-hcd feature tom.zanussi
2012-05-04 22:07 ` [PATCH 5/7][KERNEL] meta: add usb/uhci-hcd feature tom.zanussi
2012-05-04 22:07 ` [PATCH 6/7][KERNEL] meta/crownbay: use usb features tom.zanussi
2012-05-04 22:07 ` [PATCH 7/7][KERNEL] meta: add mei feature tom.zanussi
2012-05-04 23:04 ` [PATCH 0/7][KERNEL] add usb and mei features Darren Hart
2012-05-05 0:06 ` Tom Zanussi
2012-05-05 1:00 ` Darren Hart
2012-05-05 1:14 ` Tom Zanussi
2012-05-06 4:53 ` Bruce Ashfield [this message]
-- strict thread matches above, loose matches on Subject: below --
2012-05-06 16:53 tom.zanussi
2012-05-07 3:48 ` Bruce Ashfield
2012-05-07 13:00 ` Bruce Ashfield
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4FA603BF.5060506@windriver.com \
--to=bruce.ashfield@windriver.com \
--cc=dvhart@linux.intel.com \
--cc=yocto@yoctoproject.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.