IndexedDB: Close open databases when user requests browsing data deletion

IndexedDB: Close open databases when user requests browsing data deletion

BUG=64050
TEST=


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=155075

[jsbell, 2012-07-11 20:55:23]

Chromium side of a fix for http://crbug.com/64050 - can't delete browsing data
in Chromium port when a connection is open. WebKit side is:
http://webkit.org/b/91010

This patch plumbs through and utilizes a new API that the back end can call on
each connection to force it to close. 

Still TODO: Tests (not sure where these should live). And the naming sucks -
suggestions welcome.

[michaeln, 2012-07-11 23:33:37]

lgtm, names seem ok to me?

[dgrogan, 2012-07-17 20:20:20]

lgtm

Michael, how did you write tests for websql's force close mechanism?

https://chromiumcodereview.appspot.com/10695158/diff/1/content/browser/in_pro...
File content/browser/in_process_webkit/indexed_db_context_impl.h (right):

https://chromiumcodereview.appspot.com/10695158/diff/1/content/browser/in_pro...
content/browser/in_process_webkit/indexed_db_context_impl.h:128: std::map<GURL,
ConnectionSet > connections_;
extra space

[michaeln, 2012-07-17 20:53:30]

On 2012/07/17 20:20:20, dgrogan wrote:
> lgtm
> 
> Michael, how did you write tests for websql's force close mechanism?

There's a unittest for the DatabaseTracker's ability to delete an open database
which triggers the force close mechanism. I believe that test verifius the
tracker spits out the expected notifications (which result in sending ipcs to
renderers), and then verifies that when the renderer responds with a 'closed'
message, files are deleted and such.

I don't think there's an automated integration test running in any test suite?
But i had put together a 'thrasher' test page that would repeatedly
open/write/read a database from a doc+iframe+worker, and ran that against a
local build that was modified to induce forcible closures. That page definitely
uncovered cases to take care.

[michaeln, 2012-07-17 20:54:40]

> I don't think there's an automated integration test running in any test suite?
> But i had put together a 'thrasher' test page that would repeatedly
> open/write/read a database from a doc+iframe+worker, and ran that against a
> local build that was modified to induce forcible closures. That page
definitely
> uncovered cases to take care.

The trasher page is in here...
http://www.corp.google.com/~michaeln/chrome/bugs/dom_11/

[jsbell, 2012-08-28 21:39:07]

I added a basic unit test to verify forceClose is called on open connections
when an origin is deleted.

michaeln@ - can you take a quick look?

I did some local thrashing the last time I had this patch "hot", and will do
more before landing.

[michaeln, 2012-08-29 21:03:38]

still lgtm!

[commit-bot: I haz the power, 2012-09-05 18:55:25]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jsbell@chromium.org/10695158/14001

[commit-bot: I haz the power, 2012-09-05 20:57:20]

Try job failure for 10695158-14001 (retry) on android for steps "compile, build"
(clobber build).
It's a second try, previously, steps "compile, build" failed.
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=android&nu...

[commit-bot: I haz the power, 2012-09-05 21:56:38]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/jsbell@chromium.org/10695158/25001

[commit-bot: I haz the power, 2012-09-06 00:42:42]

Change committed as 155075