Refactor site_servo, site_vblock and site_gpio module into client/cros package(directory).

Refactor site_servo, site_vblock and site_gpio module into client/cros package(directory).

Change-Id: Iafa730a226541a20d628bb6619b2717c91ca8fbf

BUG=chromium-os:12633
TEST=grep the entire autotest repo and find out all source code which has site_servo, site_vblock, site_gpio and changed it.

I had performed some extra test steps to ensure factory image was properly
built.

1. cros_workon start autotest autotest-deps autotest-tests autotest-factory
2. build_packages && build_image
3. ./image_to_usb.sh --from=... --factory --force_copy --to /tmp/factory.bin
4. image_to_usb.sh factory.bin to an usb.
5. boot from usb and watch the factory test screen, this is my first time to see
this screen, impressive!
6. Back to chroot and ./mount_gpt_image.sh --from
../build/images/x86-generic/latest --image=chromiumos_factory_image.bin
7. then inspect /tmp/s/dev_image/usr/lo
cal/autotest/cros directory and
/tmp/s/dev_image/usr/local/autotest/site_tests/hardware_Components/hardware_Components.py.

Committed: http://chrome-svn/viewvc/chromeos?view=rev&revision=e60f0e6

[Sameer Nanda, 2011-02-25 22:32:29]

While the change itself looks good, can you explain the background for this
move?  There are several other chromeos specific libraries in there
(client/common_lib/site*) -- are you intending to move them to the cros
directory as well?

[ericli, 2011-02-25 22:51:27]

Yes, slowly and eventually.

The plan is to migrate all chromeos site_* related modules into cros
directory from a pure maintainability perspective.

I have communicated my plan to the team several times,  but it was couple of
month ago and this activity is always a very low priority task for me, and
some time I purposefully do it very slowly in order to avoid any disruption
to the team. But the side effect was consistently explaining to people what
I intend to do. :)


On Fri, Feb 25, 2011 at 2:32 PM, <snanda@chromium.org> wrote:

> While the change itself looks good, can you explain the background for this
> move?  There are several other chromeos specific libraries in there
> (client/common_lib/site*) -- are you intending to move them to the cros
> directory as well?
>
>
> http://codereview.chromium.org/6596018/
>



-- 
Eric Li
李咏竹
Google Kirkland

[Sameer Nanda, 2011-02-25 22:55:31]

LGTM. And thanks for the explanation!

On 2011/02/25 22:51:27, ericli wrote:
> Yes, slowly and eventually.
> 
> The plan is to migrate all chromeos site_* related modules into cros
> directory from a pure maintainability perspective.
> 
> I have communicated my plan to the team several times,  but it was couple of
> month ago and this activity is always a very low priority task for me, and
> some time I purposefully do it very slowly in order to avoid any disruption
> to the team. But the side effect was consistently explaining to people what
> I intend to do. :)
> 
> 
> On Fri, Feb 25, 2011 at 2:32 PM, <mailto:snanda@chromium.org> wrote:
> 
> > While the change itself looks good, can you explain the background for this
> > move?  There are several other chromeos specific libraries in there
> > (client/common_lib/site*) -- are you intending to move them to the cros
> > directory as well?
> >
> >
> > http://codereview.chromium.org/6596018/
> >
> 
> 
> 
> -- 
> Eric Li
> 李咏竹
> Google Kirkland

[ericli, 2011-02-25 22:58:26]

Thanks, also seek an LTGM from hungte.

On Fri, Feb 25, 2011 at 2:55 PM, <snanda@chromium.org> wrote:

> LGTM. And thanks for the explanation!
>
>
> On 2011/02/25 22:51:27, ericli wrote:
>
>> Yes, slowly and eventually.
>>
>
>  The plan is to migrate all chromeos site_* related modules into cros
>> directory from a pure maintainability perspective.
>>
>
>  I have communicated my plan to the team several times,  but it was couple
>> of
>> month ago and this activity is always a very low priority task for me, and
>> some time I purposefully do it very slowly in order to avoid any
>> disruption
>> to the team. But the side effect was consistently explaining to people
>> what
>> I intend to do. :)
>>
>
>
>  On Fri, Feb 25, 2011 at 2:32 PM, <mailto:snanda@chromium.org> wrote:
>>
>
>  > While the change itself looks good, can you explain the background for
>> this
>> > move?  There are several other chromeos specific libraries in there
>> > (client/common_lib/site*) -- are you intending to move them to the cros
>> > directory as well?
>> >
>> >
>> > http://codereview.chromium.org/6596018/
>> >
>>
>
>
>
>  --
>> Eric Li
>> 李咏竹
>> Google Kirkland
>>
>
>
>
> http://codereview.chromium.org/6596018/
>



-- 
Eric Li
李咏竹
Google Kirkland

[Hung-Te, 2011-03-01 04:50:17]

Hi Eric, thanks for your code refine.

The site_vblock is used by
src/third_party/autotest/files/client/site_tests/hardware_Components/hardware_Components.py,
so please also fix it.

Another issue is, would you mind to add the reference to the CL that you're
going to add "cros" into built images (or there is no need?)?

Please try to build a factory package by:
 build_packages && build_images &&
 image_to_usb.sh --factory --force_copy --to some_file.bin
And then mount/install from some_file.bin to check if it has the proper files
(/usr/local/autotest/cros/*) files, thanks!

[Hung-Te, 2011-03-01 04:54:42]

And please file an issue page for that - so we can track how many files  should
be also moved -- like site_fmap, flashrom_util, gbb_util.

Those are also dedicated to cros-only -- as far as I know.

[ericli, 2011-03-01 22:05:55]

Hi HungTe,

Thanks for your advice, PLAL.

I had performed some extra test steps to ensure factory image was properly
built.

1. cros_workon start autotest autotest-deps autotest-tests autotest-factory
2. build_packages && build_image
3. ./image_to_usb.sh --from=... --factory --force_copy --to /tmp/factory.bin
4. image_to_usb.sh factory.bin to an usb.
5. boot from usb and watch the factory test screen, this is my first time to see
this screen, impressive!
6. Back to chroot and ./mount_gpt_image.sh --from
../build/images/x86-generic/latest --image=chromiumos_factory_image.bin
7. then inspect /tmp/s/dev_image/usr/lo
cal/autotest/cros directory and
/tmp/s/dev_image/usr/local/autotest/site_tests/hardware_Components/hardware_Components.py.

I will follow up with other requests from you.

[ericli, 2011-03-01 22:08:34]

On 2011/03/01 04:50:17, Hung-Te wrote:
> Hi Eric, thanks for your code refine.
> 
> The site_vblock is used by
>
src/third_party/autotest/files/client/site_tests/hardware_Components/hardware_Components.py,
> so please also fix it.
> 
> Another issue is, would you mind to add the reference to the CL that you're
> going to add "cros" into built images (or there is no need?)?
cros dir is already on the factory image, verified by:

./mount_gpt_image.sh --from ../build/images/x86-generic/latest
--image=chromiumos_factory_image.bin

ls /tmp/s/dev_image/usr/local/autotest/cros/

> 
> Please try to build a factory package by:
>  build_packages && build_images &&
>  image_to_usb.sh --factory --force_copy --to some_file.bin
> And then mount/install from some_file.bin to check if it has the proper files
> (/usr/local/autotest/cros/*) files, thanks!

[Hung-Te, 2011-03-02 04:05:18]

> cros dir is already on the factory image, verified by:
> ls /tmp/s/dev_image/usr/local/autotest/cros/
   That also works - dev_image/usr/local will be mounted at
(/mnt/stateful_partition)/usr/local for a live factory test system.

LGTM with some nits in CL description:

  Please update your BUG=chromiumos:12633 to BUG=chromium-os:12633 so that we
can have the link activated in review system.
  Also please remove "BTW, I could not find any usage of vblock.",
  and update with your test results, like "cros directory exists in a factory
test image" :)

[ericli, 2011-03-02 18:02:39]

Done. Pushing, thanks.