Skip to content

GitLab

Switch branch/tag
  1. 12 Jun, 2015 1 commit
  3. 27 May, 2015 5 commits
  5. 30 Apr, 2015 2 commits
  7. 20 Apr, 2015 1 commit
  9. 17 Apr, 2015 1 commit
    • Johan Redestig's avatar
      Fix build with prebuilt webview · 1ec638c7
      Johan Redestig authored 
      generate_trig_table and system_icu was removed for the regular
build in earlier commits.

Change-Id: I69535f135c30b75b7329f756bcec981b70ae966c
      1ec638c7
  11. 02 Apr, 2015 1 commit
  13. 24 Mar, 2015 1 commit
    • Mikhail Naganov's avatar
      Cherry-pick "[Android WebView] Provide user-initiated provisional load detection" · 520a78ad
      Mikhail Naganov authored 
      Original CL:
    https://crrev.com/7a3234c9a19a38726a4a984ff12e2b406423c1d3

Original description:
    [Android WebView] Provide user-initiated provisional load detection

    Do not synthesize page loading events on DOM modification, if the provisional
    load has been started from the API side.

    It appears that a lot of apps tend to use the following scenario:

    webView.loadUrl(...);
    webView.loadUrl('javascript:...');

    Which was triggering page loading events to be emitted. This scenario is
    dubious, as no one guarantees that loading will actually finish prior to
    executing javascript. But for compatibility reasons we must take it into
    account and not emit page loading events for "about:blank", as it seems
    that some apps do unexpected things when they receive it.

    BUG=458569,469099

Note that this patch also contains bits from this commit required for changes
in AwWebContentsObserver to work:
    https://crrev.com/5668d12d29ab659ad01dc0e0ffdf24f2eb5f187a

Bug: 19729876
Change-Id: I12c9610a13a382d725cc9a14fcb9879c92f1f461
      520a78ad
  15. 12 Mar, 2015 2 commits
  17. 10 Mar, 2015 2 commits
    • Torne (Richard Coles)'s avatar
      Disable orientation locking support in WebView. · 1cf285d3
      Torne (Richard Coles) authored 
      The orientation lock code was accidentally inherited from Chrome and
doesn't actually work in WebView. Make it actually fail when JS calls
it, instead of claiming to work but doing nothing, by disabling the
screen orientation provider. This puts WebView in the same state as
desktop chrome which also doesn't set the provider. Tested that this
results in correctly reported errors when JS tries to use the API.

Bug: 19672531
Change-Id: I66a142fc50fa12db418b9465c6f29914633bd9b2
      1cf285d3
    • Torne (Richard Coles)'s avatar
      Fix the battery status API. · 3dbfef37
      Torne (Richard Coles) authored 
      Battery status was crashing when called because the Java code in the
component was not being compiled into the webview in the android build
system. Add the missing java source directory. Battery status now works.
http://crbug.com/465366

Change-Id: I6e0f7d8d4bd0f8a284ac3e6a2a7c9e7ec1585f51
      3dbfef37
  19. 04 Mar, 2015 1 commit
    • Mikhail Naganov's avatar
      Back-port: "[Android WebView] Synthesize a fake page loading event on page... · 7699f2d5
      Mikhail Naganov authored 
      Back-port: "[Android WebView] Synthesize a fake page loading event on page source modification (Re-land)"

Back-port of https://codereview.chromium.org/970883002/ for m40.

Original description:

    [Android WebView] Synthesize a fake page loading event on page source modification (Re-land)

    When a script modifies page source of a non-committed page, we need to
    notify clients, so they can update the URL bar to avoid confusion.

    New logic since the last attempt (https://codereview.chromium.org/924833003/):
    distinguish between a "vanilla" WebView state (basically, a newly created
    WebView, where no loading attempts have been made) and an "attempted to
    navigate" state. In the "vanilla" state, don't fire synthesized page loading
    events to avoid confusing clients. This is safe, as WebView is guaranteed to be
    on a blank page.

    Implementation note: we detect navigation attempts using
    didStartProvisionalLoadForFrame WebContentsObserver event on the Java side. As
    for popups AwWebContentsObserver gets re-attached from the original popup
    WebView to the one provided by the client, notifications issued inbetween can be
    missed on the Java side. To work around this, we assume that WebViews opened as
    popups can never be in "vanilla" state (as they are anyway opened as a result of
    navigation).

    BUG=458569, 462213
    TBR=davidben@chromium.org,tedchoc@chromium.org

Bug: 19325392
Change-Id: I3df0484f00d2b816cb0293bec432b2fa6ac5e05b
      7699f2d5
  21. 03 Mar, 2015 2 commits
  23. 26 Feb, 2015 1 commit
  25. 25 Feb, 2015 1 commit
    • Gustav Sennton's avatar
      Revert of [Clank IME] Make keyCode detection sensitive to autocomplete=on|off. · 4922e3cf
      Gustav Sennton authored 
      Reason for revert:
Probably the cause of a bug where characters are not properly deleted when
using a Samsung keyboard.

Original issue's description:
> [Clank IME] Make keyCode detection sensitive to autocomplete=on|off.
>
> Fixing Stable blocker for Clank IME.
>
> The old keyCode detection heuristics would emit KEYCODE_DEL when the last
> character of composition is deleted by backspace. This causes the bug where
> the composition gets reset (Japanese keyboard). We fix the the problem by
> emitting keyCode 229 when composition is updated, if autocomplete=on. If
> autocomplete=off then we can can emit the keyCode, provided that we have
> single-character commits. Due to unfortunate collision with
> https://codereview.chromium.org/834133004 , we have to to define constant
> textInputFlagAutocompleteOff for compatibility with trunk and Beta.
>
> The companion WebKit CL is https://codereview.chromium.org/797243003/
>
> BUG=422685
> Committed: https://crrev.com/1a8a145d5ea8289f1e527c3de95f233e4a659cde
> Cr-Commit-Position: refs/heads/master@{#311245}

Bug: 19363073
Change-Id: I25a945f797ffdb5959fd277ad8f86f149d79f8c1
      4922e3cf
  27. 24 Feb, 2015 3 commits
  29. 19 Feb, 2015 1 commit
    • Bo Liu's avatar
      Fork: Skip sharing GL_LUMINANCE textures · 262f5c6c
      Bo Liu authored 
      Due bugs in some drivers, sharing GL_LUMINANCE fails and causes causes a
NPE crash. WebRTC is the only feature using GL_LUMINANCE textures which
do not require to be shared, so skip them altogether.

Forking because situation in trunk has changed significantly and this is
no longer appropriate.

BUG: 19366032
Change-Id: I09c076068e1f38daa344009ca7b12cf30ead6df1
      262f5c6c
  31. 18 Feb, 2015 3 commits
  33. 12 Feb, 2015 1 commit
  35. 11 Feb, 2015 1 commit
  37. 09 Feb, 2015 6 commits
  39. 06 Feb, 2015 1 commit
  41. 03 Feb, 2015 2 commits
  43. 30 Jan, 2015 1 commit
    • Hui Shu's avatar
      Cherry pick: Cache viewport rect for tile priority in UpdateTiles · a8a07571
      Hui Shu authored 
      This is a cherry pick from chromium:
https://crrev.com/9178c77a0167ee7c73364b21a6f3a633029b7a44

Original CL description:
PictureLayerImpls in a tree should continue using the same viewport rect
for tile priority until UpdateTiles() is called with the new value of
viewport rect for tile priority. Otherwise, a PictureLayerImpl could be
in an inconsistent state that could lead to a wrong decision about
AllTilesRequiredForActivationAreReadyToDraw.

This CL caches the viewport rect for tile priority in PictureLayerImpl.

Conflicts:
  cc/layers/picture_layer_impl.h

BUG: 18915361
Change-Id: I7738cf793484fa46670388e0cf0e1d2cefbbddd4
      a8a07571