| commit | 04504c246dd88763f14933199feb80cde09fd258 | |
| author | bartn@chromium.org <bartn@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | |
| Tue, 22 Jan 2013 21:08:55 +0000 (22 21:08 +0000) | ||
| committer | bartn@chromium.org <bartn@chromium.org@0039d316-1c4b-4281-b951-d872f2087c98> | |
| Tue, 22 Jan 2013 21:08:55 +0000 (22 21:08 +0000) | ||
| tree | 3ab20e7bd8493feeea31f0e34e3ea804a89ae12f | treesnapshot (tar.gz zip) |
| parent | 897cde9a5cdb6a534449ae15a2818efc94d712db | commitdiff |
By moving the call to Load() up in SearchProvider::Start(), we are giving a chance to initialize TemplateURLService as early as possible. Before the change, Load would be called only if there was a non-empty input. During the startup, there is at least 1 call with empty input.
There is a chance this CL may increase startup time, however I have not observed it while running performance UI tests. As discussed with pkasting@, I'm planning to submit this CL and watch performance graphs to make sure we don't regress.
BUG=167381
TEST=Manually ran performance_ui_tests.
Review URL: https://chromiumcodereview.appspot.com/11859042
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@178092 0039d316-1c4b-4281-b951-d872f2087c98
There is a chance this CL may increase startup time, however I have not observed it while running performance UI tests. As discussed with pkasting@, I'm planning to submit this CL and watch performance graphs to make sure we don't regress.
BUG=167381
TEST=Manually ran performance_ui_tests.
Review URL: https://chromiumcodereview.appspot.com/11859042
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@178092 0039d316-1c4b-4281-b951-d872f2087c98
| chrome/browser/autocomplete/search_provider.cc | diffblobblamehistory |