Media Gallery: Register isolated file systems on a per RenderProcessHost basis. Otherwise when one…

Media Gallery: Register isolated file system on a per RenderProcessHost basis. Otherwise when one RenderProcessHost closed, it revokes the file system for all RenderProcessHosts.

BUG=none
TEST=none

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=141629

[kinuko, 2012-06-11 12:21:06]

http://codereview.chromium.org/10537092/diff/7001/chrome/browser/media_galler...
File chrome/browser/media_gallery/media_file_system_registry.cc (right):

http://codereview.chromium.org/10537092/diff/7001/chrome/browser/media_galler...
chrome/browser/media_gallery/media_file_system_registry.cc:43:
ChildIdToMediaFSMap::iterator child_it = media_fs_map_.find(rph);
Can we simply call insert() here and check the second of the returned value to
see if it was succeeded (therefore we need to RegisterForRPH... etc)?

http://codereview.chromium.org/10537092/diff/7001/chrome/browser/media_galler...
chrome/browser/media_gallery/media_file_system_registry.cc:53:
child_it->second.insert(std::make_pair(pictures_path, fsid));
nit:
child_it->second[pictures_path] = fsid;
might be slightly easier to read?

[Lei Zhang, 2012-06-11 19:59:07]

https://chromiumcodereview.appspot.com/10537092/diff/7001/chrome/browser/medi...
File chrome/browser/media_gallery/media_file_system_registry.cc (right):

https://chromiumcodereview.appspot.com/10537092/diff/7001/chrome/browser/medi...
chrome/browser/media_gallery/media_file_system_registry.cc:43:
ChildIdToMediaFSMap::iterator child_it = media_fs_map_.find(rph);
On 2012/06/11 12:21:06, kinuko wrote:
> Can we simply call insert() here and check the second of the returned value to
> see if it was succeeded (therefore we need to RegisterForRPH... etc)?

Done.

https://chromiumcodereview.appspot.com/10537092/diff/7001/chrome/browser/medi...
chrome/browser/media_gallery/media_file_system_registry.cc:53:
child_it->second.insert(std::make_pair(pictures_path, fsid));
On 2012/06/11 12:21:06, kinuko wrote:
> nit:
> child_it->second[pictures_path] = fsid;
> might be slightly easier to read?

Yes, but "foo[bar] = qux" is less efficient. My C++ readability reviewer said
google3 prefers insert() over operator[].

[kinuko, 2012-06-12 04:40:11]

lgtm, though I'm not very familiar with the notification code.  If you're
comfortable with it please go ahead.

https://chromiumcodereview.appspot.com/10537092/diff/7001/chrome/browser/medi...
File chrome/browser/media_gallery/media_file_system_registry.cc (right):

https://chromiumcodereview.appspot.com/10537092/diff/7001/chrome/browser/medi...
chrome/browser/media_gallery/media_file_system_registry.cc:53:
child_it->second.insert(std::make_pair(pictures_path, fsid));
On 2012/06/11 19:59:07, Lei Zhang wrote:
> On 2012/06/11 12:21:06, kinuko wrote:
> > nit:
> > child_it->second[pictures_path] = fsid;
> > might be slightly easier to read?
> 
> Yes, but "foo[bar] = qux" is less efficient. My C++ readability reviewer said
> google3 prefers insert() over operator[].

Ok, in chromium review I was once told I should use foo[bar] over .insert for
readability, but I'm ok with either of them :)

[commit-bot: I haz the power, 2012-06-12 05:29:20]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/thestig@chromium.org/10537092/2004

[commit-bot: I haz the power, 2012-06-12 08:34:50]

Change committed as 141629