Merge 8401001 r107836 into branch 912: Fix history importing by delaying DownloadManager creation.

Merge 8401001 r107836 into branch 912: Fix history importing by delaying DownloadManager creation. Replace GetNextIdThunkType with DownloadIdFactory (RefCountedThreadSafe, created by DownloadService).

DownloadService uses the same DownloadIdFactory for an OTR profile as its original profile.

DownloadService passes the DownloadIdFactory into the DownloadManager so that the DownloadManager can allocate new valid ids for items loaded from the history or downloads started on the ui thread.

Since the DownloadService precedes and outlives its DownloadManager, DownloadManager does not have a scoped_refptr<DownloadIdFactory>. Objects that do have a scoped_refptr<DownloadIdFactory>: DownloadService, ProfileIOData, ShellBrowserContext, ShellResourceContext.

The DownloadIdFactory must be RefCountedThreadSafe because ProfileIOData outlives Profile and because it's used in both the OTR and original profiles.

Longer term, the import process should strictly precede profile initialization, and the next_download_id counter should be loaded from the History db strictly before DownloadService is created and creates a DownloadIdFactory.

BUG=98966
Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=108346

[benjhayden, 2011-11-02 19:20:35]

PTAL and add appropriate reviewers.

What is the correct value for svn:eol-style, and how do I set it?

How do I tell gcl try to use 912? `gcl try --upstream_branch=912 k2s8` said
invalid changelist name '--upstream_branch=912', and `gcl try k2s8
--upstream_branch=912` asked for my password a dozen times without telling me
which password before I rage-killed it.

[Randy Smith (Not in Mondays), 2011-11-02 19:34:03]

On 2011/11/02 19:20:35, b s h wrote:
> PTAL and add appropriate reviewers.

Usually merges are self-reviewed.  This one's complex enough that I think having
reviewers make sense, but I don't think we need anyone besides Miranda and I.

> What is the correct value for svn:eol-style, and how do I set it?

Leaving this one for Miranda.

> How do I tell gcl try to use 912? `gcl try --upstream_branch=912 k2s8` said
> invalid changelist name '--upstream_branch=912', and `gcl try k2s8
> --upstream_branch=912` asked for my password a dozen times without telling me
> which password before I rage-killed it.

Personal opinion: If it builds in your sandbox, that's good enough for a merge;
I'm not sure you *can* run try bots on branch SBs.  It's worth watching the
(beta) tree after you commit to make sure that nothing bad happened, but I
wouldn't worry about running try bots.  I do think it's pretty important to try
building it locally, though.  Miranda, what are your thoughts?

[Miranda Callahan, 2011-11-02 19:37:34]

On 2011/11/02 19:20:35, b s h wrote:
> PTAL and add appropriate reviewers.
> 
> What is the correct value for svn:eol-style, and how do I set it?

Correct value for .cc, .h, and .gypi files is

svn:eol-style=LF

(see http://src.chromium.org/viewvc/chrome/trunk/tools/build/slave/config)

you can set individual files thusly:

$ svn ps svn:eol-style LF content/browser/resource_context.h


> 
> How do I tell gcl try to use 912? `gcl try --upstream_branch=912 k2s8` said
> invalid changelist name '--upstream_branch=912', and `gcl try k2s8
> --upstream_branch=912` asked for my password a dozen times without telling me
> which password before I rage-killed it.

good question -- but I was able to do 

gcl try -R http://codereview.chromium.org/8372073 --upstream_branch=912 --email
benjhayden@chromium.org

with no problems. You probably need your svn password, actually --
https://chromium-access.appspot.com/ should give you the correct password.

Thanks, Ben!!

[Randy Smith (Not in Mondays), 2011-11-02 19:44:28]

LGTM presuming it compiles.

[benjhayden, 2011-11-02 19:49:25]

Yep, it compiles and runs on my linux machine. Running browser tests locally
now.

On 2011/11/02 19:37:34, Miranda Callahan wrote:
> On 2011/11/02 19:20:35, b s h wrote:
> > PTAL and add appropriate reviewers.
> > 
> > What is the correct value for svn:eol-style, and how do I set it?
> 
> Correct value for .cc, .h, and .gypi files is
> 
> svn:eol-style=LF

Good, it was already right.

> 
> (see http://src.chromium.org/viewvc/chrome/trunk/tools/build/slave/config)
> 
> you can set individual files thusly:
> 
> $ svn ps svn:eol-style LF content/browser/resource_context.h
> 
> 
> > 
> > How do I tell gcl try to use 912? `gcl try --upstream_branch=912 k2s8` said
> > invalid changelist name '--upstream_branch=912', and `gcl try k2s8
> > --upstream_branch=912` asked for my password a dozen times without telling
me
> > which password before I rage-killed it.
> 
> good question -- but I was able to do 
> 
> gcl try -R http://codereview.chromium.org/8372073 --upstream_branch=912
--email
> mailto:benjhayden@chromium.org
> 
> with no problems. You probably need your svn password, actually --
> https://chromium-access.appspot.com/ should give you the correct password.

Looks like that didn't sync to 912:
http://build.chromium.org/p/tryserver.chromium/builders/linux/builds/842/step...

I entered my chromium-access svn password a few times for `gcl try  k2s8
--upstream_branch=912` and it kept asking. Oh well, I won't bother with the try
bots. I'll commit as soon as the browser tests are done.


> 
> Thanks, Ben!!

[Miranda Callahan, 2011-11-02 19:55:21]

On 2011/11/02 19:49:25, b s h wrote:
> Yep, it compiles and runs on my linux machine. Running browser tests locally
> now.
> 
> On 2011/11/02 19:37:34, Miranda Callahan wrote:
> > On 2011/11/02 19:20:35, b s h wrote:
> > > PTAL and add appropriate reviewers.
> > > 
> > > What is the correct value for svn:eol-style, and how do I set it?
> > 
> > Correct value for .cc, .h, and .gypi files is
> > 
> > svn:eol-style=LF
> 
> Good, it was already right.
> 
> > 
> > (see http://src.chromium.org/viewvc/chrome/trunk/tools/build/slave/config)
> > 
> > you can set individual files thusly:
> > 
> > $ svn ps svn:eol-style LF content/browser/resource_context.h
> > 
> > 
> > > 
> > > How do I tell gcl try to use 912? `gcl try --upstream_branch=912 k2s8`
said
> > > invalid changelist name '--upstream_branch=912', and `gcl try k2s8
> > > --upstream_branch=912` asked for my password a dozen times without telling
> me
> > > which password before I rage-killed it.
> > 
> > good question -- but I was able to do 
> > 
> > gcl try -R http://codereview.chromium.org/8372073 --upstream_branch=912
> --email
> > mailto:benjhayden@chromium.org
> > 
> > with no problems. You probably need your svn password, actually --
> > https://chromium-access.appspot.com/ should give you the correct password.
> 
> Looks like that didn't sync to 912:
>
http://build.chromium.org/p/tryserver.chromium/builders/linux/builds/842/step...
> 
> I entered my chromium-access svn password a few times for `gcl try  k2s8
> --upstream_branch=912` and it kept asking. Oh well, I won't bother with the
try
> bots. I'll commit as soon as the browser tests are done.
> 
> 
> > 
> > Thanks, Ben!!

Agreed that running local tests should be sufficient. LGTM!

[benjhayden, 2011-11-02 20:41:38]

Where is the branch waterfall? Three tests are failing:
BrowserTest.TestNewTabExitsFullscreen,
BrowserTest.TestTabExitsItselfFromFullscreen,
BrowserTest.TestFullscreenBubbleMouseLockState. They all time out in
base::MessagePumpGtk::RunOnce().

[benjhayden, 2011-11-02 21:17:31]

Presubmit says I need an LGTM from an owner for  content/browser/*.
I ran content_unittests, browser_tests, and unit_tests locally and they all
passed. (The fullscreen browser tests failed on xmonad but passed on gnome.)
Also, it looks like the branch is locked for windows interactive tests.
And I need to go home now.
I'll see if the branch is unlocked sometime tonight.

[Avi (use Gerrit), 2011-11-02 21:21:27]

owner rubberstamp lgtm

[benjhayden, 2011-11-02 21:25:14]

Committed to 912 with --no_presubmit. I'll check the waterfall when I get home
in 20 mins.
Apologies for the noise.
Maybe presubmit could ignore the main tree if it's going to commit to a branch.
:-)