[Android] Fix BufferUnderflowException in webrtc buffer

[Android] Fix BufferUnderflowException in webrtc buffer

The image buffer may not contain all bytes in the last row, causing an
exception. Fix by only reading up to what's remaining() in the buffer.

BUG=458701
Committed: https://crrev.com/e64698bbd6c20bfdad48db99d65a04c25fb02834
Cr-Commit-Position: refs/heads/master@{#319401}

[boliu, 2015-03-04 01:56:06]

boliu@chromium.org changed reviewers:
+ mcasas@chromium.org

[boliu, 2015-03-04 01:56:06]

ptal, thanks

[boliu, 2015-03-05 01:04:45]

friendly ping?

[boliu, 2015-03-06 00:37:25]

boliu@chromium.org changed reviewers:
+ perkj@chromium.org

[boliu, 2015-03-06 00:37:26]

+perkj for review?

[mcasas, 2015-03-06 02:11:48]

Sorry for the delay.

https://codereview.chromium.org/975913003/diff/1/media/base/android/java/src/...
File media/base/android/java/src/org/chromium/media/VideoCaptureCamera2.java
(right):

https://codereview.chromium.org/975913003/diff/1/media/base/android/java/src/...
media/base/android/java/src/org/chromium/media/VideoCaptureCamera2.java:268:
buffer.get(rowData, 0, Math.min(rowStride, buffer.remaining()));
I'm not super in love with adding an extra math call in 
every iteration, plus not explicitly stating why we do
this.

What about:

for (int row = 0; row < planeHeight - 1; ++row) {
   ...
}
// Last row is special in some devices and comes incomplete.
// See http://crbug.com/458701.
buffer.get(rowData, 0, Math.min(rowStride, buffer.remaining()));
for (int col = 0; col < planeWidth; ++col) {
    data[offset++] = rowData[col * pixelStride];
}

[boliu, 2015-03-06 02:22:29]

https://codereview.chromium.org/975913003/diff/1/media/base/android/java/src/...
File media/base/android/java/src/org/chromium/media/VideoCaptureCamera2.java
(right):

https://codereview.chromium.org/975913003/diff/1/media/base/android/java/src/...
media/base/android/java/src/org/chromium/media/VideoCaptureCamera2.java:268:
buffer.get(rowData, 0, Math.min(rowStride, buffer.remaining()));
On 2015/03/06 02:11:48, mcasas wrote:
> I'm not super in love with adding an extra math call in 
> every iteration, plus not explicitly stating why we do
> this.
> 
> What about:
> 
> for (int row = 0; row < planeHeight - 1; ++row) {
>    ...
> }
> // Last row is special in some devices and comes incomplete.
> // See http://crbug.com/458701.
> buffer.get(rowData, 0, Math.min(rowStride, buffer.remaining()));
> for (int col = 0; col < planeWidth; ++col) {
>     data[offset++] = rowData[col * pixelStride];
> } 

Done

Will need to remember to make that bug public at some point..

[mcasas, 2015-03-06 02:26:12]

lgtm

[boliu, 2015-03-06 02:28:29]

Thanks :)

[boliu, 2015-03-06 02:28:37]

The CQ bit was checked by boliu@chromium.org

[commit-bot: I haz the power, 2015-03-06 02:29:36]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/975913003/20001

[commit-bot: I haz the power, 2015-03-06 03:19:16]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-03-06 03:29:21]

Patchset 2 (id:??) landed as
https://crrev.com/e64698bbd6c20bfdad48db99d65a04c25fb02834
Cr-Commit-Position: refs/heads/master@{#319401}