This adds a recovery mode to the diagnostics

This adds a recovery mode to the diagnostics so that we can
attempt to recover from a corrupted profile.

Currently, the only recovery that is attempted is recovering from a
corrupted NSS database on ChromeOS.

(TBR'ing trivial OWNERS review for added switch constants in chromeos dir)

TBR=stevenjb@chromium.org
BUG=chromium:236093
TEST=ran against corrupted nss db files.  Ran unit test.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=216089

[cpu_(ooo_6.6-7.5), 2013-07-25 19:40:55]

https://codereview.chromium.org/16948012/diff/35001/chrome/app/chrome_main_de...
File chrome/app/chrome_main_delegate.cc (right):

https://codereview.chromium.org/16948012/diff/35001/chrome/app/chrome_main_de...
chrome/app/chrome_main_delegate.cc:472: }
shouldn't we return here? in what case it is ok to continue?

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
File chrome/browser/diagnostics/diagnostics_controller_unittest.cc (right):

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
chrome/browser/diagnostics/diagnostics_controller_unittest.cc:79:
file_util::WriteFile(path, bogus_data.c_str(), bogus_data.size());
bogus_data[] = "wwZ2uNYNuyUVzFbDm3DL";

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
File chrome/browser/diagnostics/recon_diagnostics.cc (right):

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
chrome/browser/diagnostics/recon_diagnostics.cc:426: true /* critical */);
please remove /* critical */ and everywhere or use an enum if you want to make
this more readable, your choice.

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
File chrome/browser/diagnostics/sqlite_diagnostics.cc (right):

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
chrome/browser/diagnostics/sqlite_diagnostics.cc:219: return new
SqliteIntegrityTest(true,  // critical
same with these, either gone or switch to enums.

[Greg Spencer (Chromium), 2013-07-26 18:51:30]

https://codereview.chromium.org/16948012/diff/35001/chrome/app/chrome_main_de...
File chrome/app/chrome_main_delegate.cc (right):

https://codereview.chromium.org/16948012/diff/35001/chrome/app/chrome_main_de...
chrome/app/chrome_main_delegate.cc:472: }
On 2013/07/25 19:40:56, cpu wrote:
> shouldn't we return here? in what case it is ok to continue?

No, in fact, we want to continue here because the recovery is complete, and we
can now continue to start up and recreate things that have been deleted by the
recovery code.  This is soon enough in the startup sequence that nothing from
the profile should have been initialized yet.

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
File chrome/browser/diagnostics/diagnostics_controller_unittest.cc (right):

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
chrome/browser/diagnostics/diagnostics_controller_unittest.cc:79:
file_util::WriteFile(path, bogus_data.c_str(), bogus_data.size());
On 2013/07/25 19:40:56, cpu wrote:
> bogus_data[] = "wwZ2uNYNuyUVzFbDm3DL";

Done.

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
File chrome/browser/diagnostics/recon_diagnostics.cc (right):

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
chrome/browser/diagnostics/recon_diagnostics.cc:426: true /* critical */);
On 2013/07/25 19:40:56, cpu wrote:
> please remove /* critical */ and everywhere or use an enum if you want to make
> this more readable, your choice.

Done.

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
File chrome/browser/diagnostics/sqlite_diagnostics.cc (right):

https://codereview.chromium.org/16948012/diff/35001/chrome/browser/diagnostic...
chrome/browser/diagnostics/sqlite_diagnostics.cc:219: return new
SqliteIntegrityTest(true,  // critical
On 2013/07/25 19:40:56, cpu wrote:
> same with these, either gone or switch to enums.

Done.

[cpu_(ooo_6.6-7.5), 2013-07-26 21:30:51]

yeah it reads nicer.

LGTM

[commit-bot: I haz the power, 2013-08-06 00:00:11]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gspencer@chromium.org/16948012/60001

[commit-bot: I haz the power, 2013-08-06 02:03:04]

Retried try job too often on mac_rel for step(s) unit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_rel&nu...

[commit-bot: I haz the power, 2013-08-06 15:36:51]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gspencer@chromium.org/16948012/60001

[commit-bot: I haz the power, 2013-08-06 16:38:56]

Retried try job too often on mac_rel for step(s) unit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_rel&nu...

[commit-bot: I haz the power, 2013-08-06 22:48:21]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gspencer@chromium.org/16948012/99001

[commit-bot: I haz the power, 2013-08-06 22:57:42]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gspencer@chromium.org/16948012/97002

[commit-bot: I haz the power, 2013-08-06 23:08:09]

Step "update" is always a major failure.
Look at the try server FAQ for more details.
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android_db...

[commit-bot: I haz the power, 2013-08-06 23:15:26]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/gspencer@chromium.org/16948012/95003

[commit-bot: I haz the power, 2013-08-07 05:20:31]

Change committed as 216089