Fix problem whereby the "-full" magic value is removed from the "ap" value wh...

Fix problem whereby the "-full" magic value is removed from the "ap" value when a differential update for CF fails (it should remain unless the update succeeds).

Also, fix problem with installer return codes being squashed. This was a regression introduced in http://src.chromium.org/viewvc/chrome?view=rev&revision=41322.

BUG=46051, 40607
TEST=Cause a differential update to fail, observe that the "ap" value contains a "-full".



Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=49346

[Paweł Hajdan Jr., 2010-06-09 06:59:15]

Drive-by with a minor test comment.

http://codereview.chromium.org/2773002/diff/10001/11008
File chrome/installer/util/google_update_settings_unittest.cc (right):

http://codereview.chromium.org/2773002/diff/10001/11008#newcode277
chrome/installer/util/google_update_settings_unittest.cc:277: if
(!CreateApKey(work_item_list.get(), L""))
nit: Why not just ASSERT_TRUE? You can still add the comment to it via <<. Also
applies to other places in this file.

[robertshield, 2010-06-09 23:25:41]

Thanks Pawel!

http://codereview.chromium.org/2773002/diff/10001/11008
File chrome/installer/util/google_update_settings_unittest.cc (right):

http://codereview.chromium.org/2773002/diff/10001/11008#newcode277
chrome/installer/util/google_update_settings_unittest.cc:277: if
(!CreateApKey(work_item_list.get(), L""))
On 2010/06/09 06:59:15, Paweł Hajdan Jr. wrote:
> nit: Why not just ASSERT_TRUE? You can still add the comment to it via <<.
Also
> applies to other places in this file.

Done.

[huanr, 2010-06-09 23:46:11]

LGTM

http://codereview.chromium.org/2773002/diff/12004/23001
File chrome/installer/setup/setup_main.cc (right):

http://codereview.chromium.org/2773002/diff/12004/23001#newcode745
chrome/installer/setup/setup_main.cc:745: return_code =
dist->GetInstallReturnCode(install_status);
So this is a regression by r41322 3 months ago 
http://src.chromium.org/viewvc/chrome?view=rev&revision=41322

Did we have any related user bug report?

[robertshield, 2010-06-10 00:30:30]

http://codereview.chromium.org/2773002/diff/12004/23001
File chrome/installer/setup/setup_main.cc (right):

http://codereview.chromium.org/2773002/diff/12004/23001#newcode745
chrome/installer/setup/setup_main.cc:745: return_code =
dist->GetInstallReturnCode(install_status);
On 2010/06/09 23:46:11, huanr wrote:
> So this is a regression by r41322 3 months ago 
> http://src.chromium.org/viewvc/chrome?view=rev&revision=41322
> 
> Did we have any related user bug report?

Yes, it is a regression. I believe this could impact http://crbug.com/40607

[huanr, 2010-06-10 00:35:24]

The current behavior is that we alway report success. How is that same as
40607?

Huan

On Wed, Jun 9, 2010 at 5:30 PM, <robertshield@chromium.org> wrote:

>
> http://codereview.chromium.org/2773002/diff/12004/23001
> File chrome/installer/setup/setup_main.cc (right):
>
> http://codereview.chromium.org/2773002/diff/12004/23001#newcode745
> chrome/installer/setup/setup_main.cc:745: return_code =
> dist->GetInstallReturnCode(install_status);
> On 2010/06/09 23:46:11, huanr wrote:
>
>> So this is a regression by r41322 3 months ago
>> http://src.chromium.org/viewvc/chrome?view=rev&revision=41322
>>
>
>  Did we have any related user bug report?
>>
>
> Yes, it is a regression. I believe this could impact
> http://crbug.com/40607
>
>
> http://codereview.chromium.org/2773002/show
>

[huanr, 2010-06-10 00:36:45]

On Wed, Jun 9, 2010 at 5:35 PM, Huan Ren <huanr@google.com> wrote:

> The current behavior is that we alway report success. How is that same as
> 40607?


Never mind. I misread the bug.


>
> Huan
>
>
> On Wed, Jun 9, 2010 at 5:30 PM, <robertshield@chromium.org> wrote:
>
>>
>> http://codereview.chromium.org/2773002/diff/12004/23001
>> File chrome/installer/setup/setup_main.cc (right):
>>
>> http://codereview.chromium.org/2773002/diff/12004/23001#newcode745
>> chrome/installer/setup/setup_main.cc:745: return_code =
>> dist->GetInstallReturnCode(install_status);
>> On 2010/06/09 23:46:11, huanr wrote:
>>
>>> So this is a regression by r41322 3 months ago
>>> http://src.chromium.org/viewvc/chrome?view=rev&revision=41322
>>>
>>
>>  Did we have any related user bug report?
>>>
>>
>> Yes, it is a regression. I believe this could impact
>> http://crbug.com/40607
>>
>>
>> http://codereview.chromium.org/2773002/show
>>
>
>