| commit | dcc725bedfff11ab3265fa38fb36e15ea2261070 | |
| author | pmeenan <pmeenan@chromium.org> | |
| Wed, 22 Oct 2014 18:04:11 +0000 (22 11:04 -0700) | ||
| committer | Commit bot <commit-bot@chromium.org> | |
| Wed, 22 Oct 2014 18:04:35 +0000 (22 18:04 +0000) | ||
| tree | 111f80bfa86c352fc3878f451d06c59737bb58df | treesnapshot (tar.gz zip) |
| parent | 51530f07a1d3c34248944156556d7280c9b049e5 | commitdiff |
Fixed the reporting of paint times in window.chrome.loadTimes()
The paint events were updating the document_state of whatever the last
frame in the frame list was. In a case of "how did this ever work" it
must have been sheer luck that the last local frame in the list was the
main frame.
The update just checks to make sure that the local frame is actually
the main frame (no parent) when selecting the frame to update the
document_state of.
Could be that this broke in 38 when the proxy frames were added and
maybe they were only tested on pages with a single frame.
BUG=422913
Review URL: https://codereview.chromium.org/670513003
Cr-Commit-Position: refs/heads/master@{#300703}
The paint events were updating the document_state of whatever the last
frame in the frame list was. In a case of "how did this ever work" it
must have been sheer luck that the last local frame in the list was the
main frame.
The update just checks to make sure that the local frame is actually
the main frame (no parent) when selecting the frame to update the
document_state of.
Could be that this broke in 38 when the proxy frames were added and
maybe they were only tested on pages with a single frame.
BUG=422913
Review URL: https://codereview.chromium.org/670513003
Cr-Commit-Position: refs/heads/master@{#300703}
| content/renderer/render_view_impl.cc | diffblobblamehistory |