TPEPlatform/2013-11-28: Difference between revisions
Jump to navigation
Jump to search
(→Media) |
(→Media) |
||
(10 intermediate revisions by 6 users not shown) | |||
Line 10: | Line 10: | ||
* {{Bug|898949}} - [B2G getUserMedia] Display front/back camera list on permission prompt \ SC | * {{Bug|898949}} - [B2G getUserMedia] Display front/back camera list on permission prompt \ SC | ||
** rebase to latest m-c and do the integration test with other dependent gecko/gaia patches. | ** rebase to latest m-c and do the integration test with other dependent gecko/gaia patches. | ||
'''Media Recording''' | |||
*{{Bug|879669}} Support Video Encoder module in MediaEncoder framewrok /rlin | *{{Bug|879669}} Support Video Encoder module in MediaEncoder framewrok /rlin | ||
** MediaEncoder 1r+, 2r-, need some work | ** MediaEncoder 1r+, 2r-, need some work | ||
Line 20: | Line 21: | ||
*{{Bug|926746}} - HARDWARE_UNAVAILABLE is already thrown while GetUserMedia for audio \ slee | *{{Bug|926746}} - HARDWARE_UNAVAILABLE is already thrown while GetUserMedia for audio \ slee | ||
** Waiting for mwu for landing | ** Waiting for mwu for landing | ||
* Test cached memory copy speed by using vendor's camera.msm7627a.so /slee | |||
' | ** The speed is worse than the original version. Waiting for vendor's reply. | ||
*{{Bug|920934}} - [MediaEncoder] Implement AudioOmxEncoder on B2G / John Lin | *{{Bug|920934}} - [MediaEncoder] Implement AudioOmxEncoder on B2G / John Lin | ||
** upload patches for review. | ** upload patches for review. | ||
** revise OMX codec wrapper to address review comments. | ** revise OMX codec wrapper to address review comments. | ||
** merge audio track encoder with video track encoder. | ** merge audio track encoder with video track encoder. | ||
*{{Bug|879669}} - Support Video Encoder module in MediaEncoder framewrok / Shelly | |||
** Part1: Refactor of TrackEncoder, review+ | |||
** Part2: Add VideoTrackEncoder, review+ | |||
*{{Bug|879668}} - [MediaEncoder] Implement VideoOmxEncoder on B2G / Shelly | |||
** Part 2: Add OmxTrackEncoder, review- | |||
*{{Bug|941283}} - OpusTrackEncoder writing pre-skip at the original samplerate / Shelly | |||
** review+, but need nit fix. | |||
'''Web Audio''' | |||
* review code about how audio nodes integrated into MediaStreamGraph / JW | |||
** {{Bug|804387}} - Make the mediastreams backend awesome enough for us to hook the initial web audio APIs into | |||
** {{Bug|830707}} - Make MediaStream AudioSegments not have a fixed number of channels | |||
* {{Bug|916399}} - Intermittent test_timeupdate_small_files.html | Test timed out / JW | |||
** debug: it looks like there is a long latency between when gecko fires the event and when gaia receive the event | |||
==Streaming Media== | ==Streaming Media== | ||
Line 33: | Line 48: | ||
**{{Bug|898866}} RTSP: the seek function usually fails | **{{Bug|898866}} RTSP: the seek function usually fails | ||
*** Debugging: thread issue. | *** Debugging: thread issue. | ||
* WebRTC performance measurement \slee | |||
** profile.sh can retrieve only main thread. | |||
** jld's branch has compiling problem. Trying to figure out why. | |||
* {{Bug|911046}} - WebRTC OmxCodec coding path support / JW | |||
** review patch, trace code to see how to integrate h264 encoder/decoder | |||
==Graphic == | ==Graphic == | ||
* {{Bug|933711}} - [B2G][Helix][stability][fengximing]When the Firefox phone goes to sleep, wait for long time, if you wake up the phone, it took too much time to enter the lock screen. /Morris | * {{Bug|933711}} - [B2G][Helix][stability][fengximing]When the Firefox phone goes to sleep, wait for long time, if you wake up the phone, it took too much time to enter the lock screen. /Morris | ||
** At nsFrameMessageManager::ReceiveMessage, we have about 600 event listeners at most. And bug 900221 would cause the time of 600 listeners iteration costs about 500ms which will slow down wake up procedure. I revert bug 900221 and looks like problem resolved. | ** At nsFrameMessageManager::ReceiveMessage, we have about 600 event listeners at most. And bug 900221 would cause the time of 600 listeners iteration costs about 500ms which will slow down wake up procedure. I revert bug 900221 and looks like problem resolved. | ||
* {{Bug|709490}} - Run WebGL on Web Worker with Off-Main-Thread-Compositing / Morris | * {{Bug|709490}} - Run WebGL on Web Worker with Off-Main-Thread-Compositing / Morris | ||
** Update to address comment as well | ** Update to address comment as well | ||
* {{Bug|920921}} - [Flatfish]: Flatfish has bad performance on Homescreen wiping /Jerry | * {{Bug|920921}} - [Flatfish]: Flatfish has bad performance on Homescreen wiping /Jerry | ||
**Trace Layer inactive/active mechanism | **Trace Layer inactive/active mechanism | ||
**Test {{Bug|940842}} for pre-render. Compare BenWa's implement with mine. | **Test {{Bug|940842}} for pre-render. Compare BenWa's implement with mine. | ||
* {{Bug|939962}} - Gallery app will get OOM process killed upon saving an edited photo after flashing or resetting the device | * {{Bug|939962}} - Gallery app will get OOM process killed upon saving an edited photo after flashing or resetting the device | ||
**Dump memory footprint when using gallery. | **Dump memory footprint when using gallery. | ||
***We always get pmem OOM on buri device and fall back to ashmem. | ***We always get pmem OOM on buri device and fall back to ashmem. | ||
* Flatfish performance followup /peter | * Flatfish performance followup /peter | ||
** https://etherpad.mozilla.org/hdperf | ** https://etherpad.mozilla.org/hdperf | ||
* KitKat graphic porting plan /peter | * KitKat graphic porting plan /peter | ||
* {{Bug|929432}} - [B2G][Helix][Browser][ChenHoulai] Open youtube in the browser, there is an extra horizontal line on the top when in full-screen mode /Peter | * {{Bug|929432}} - [B2G][Helix][Browser][ChenHoulai] Open youtube in the browser, there is an extra horizontal line on the top when in full-screen mode /Peter | ||
** Found the problem and work on patch now | ** Found the problem and work on patch now | ||
* {{Bug|944564}} - [B2G] Page transition in Homescreen can't reach 60fps stably. /vlin | |||
** Breaking down 3 problems cause this. | |||
==Input== | ==Input== | ||
* {{Bug|924692}} - Add support for selecting using touch / Phoebe | |||
**append a fixed position AnonymousContent in CanvasFrame as CaretHandleFrame. | |||
**study CSS pseudo stylecontext setup and modification. | |||
**debug contenteditable focus on caretHandle assertion. | |||
==Stability== | ==Stability== | ||
Line 67: | Line 84: | ||
==Study== | ==Study== | ||
* nsClipboard study / slee | |||
*{{Bug|854753}} - [B2G][Audio] Implement SpeakerManager API /rlin | |||
**landed |
Latest revision as of 02:32, 2 December 2013
Media
GetUserMedia
- bug 940045 - The origin displayed in the persistent notification matches the location bar, not content \ SC
- We should send page URL directly from content process.
- patch was uploaded and get 3/4 r+'ed.
- bug 940075 - crash in mozilla::dom::PBrowserChild::SendRecordingDeviceEvents(nsString const&, bool const&, bool const&) \ SC
- TabChild is garbage collected after loading other page in the same process. We cannot use it to notify gUM is shutdown for that window.
- We should switch back to PContent for sending recording status, which is always existed during the entire content process life cycle.
- patch was merged to bug 940045.
- bug 898949 - [B2G getUserMedia] Display front/back camera list on permission prompt \ SC
- rebase to latest m-c and do the integration test with other dependent gecko/gaia patches.
Media Recording
- bug 879669 Support Video Encoder module in MediaEncoder framewrok /rlin
- MediaEncoder 1r+, 2r-, need some work
- bug 891705 [MediaEncoder] Implement WebM 1.0 container writer /rlin
- provide patch for reviewing, will leverage libvpx library directly
- bug 935774 "Assertion failure: meta" in mozilla::MediaEncoder::GetEncodedData
- landed and verify
- bug 936981 - Build up MediaEncoder GTest framework /rlin
- framework provided, start to write a test.
- bug 926746 - HARDWARE_UNAVAILABLE is already thrown while GetUserMedia for audio \ slee
- Waiting for mwu for landing
- Test cached memory copy speed by using vendor's camera.msm7627a.so /slee
- The speed is worse than the original version. Waiting for vendor's reply.
- bug 920934 - [MediaEncoder] Implement AudioOmxEncoder on B2G / John Lin
- upload patches for review.
- revise OMX codec wrapper to address review comments.
- merge audio track encoder with video track encoder.
- bug 879669 - Support Video Encoder module in MediaEncoder framewrok / Shelly
- Part1: Refactor of TrackEncoder, review+
- Part2: Add VideoTrackEncoder, review+
- bug 879668 - [MediaEncoder] Implement VideoOmxEncoder on B2G / Shelly
- Part 2: Add OmxTrackEncoder, review-
- bug 941283 - OpusTrackEncoder writing pre-skip at the original samplerate / Shelly
- review+, but need nit fix.
Web Audio
- review code about how audio nodes integrated into MediaStreamGraph / JW
- bug 804387 - Make the mediastreams backend awesome enough for us to hook the initial web audio APIs into
- bug 830707 - Make MediaStream AudioSegments not have a fixed number of channels
- bug 916399 - Intermittent test_timeupdate_small_files.html | Test timed out / JW
- debug: it looks like there is a long latency between when gecko fires the event and when gaia receive the event
Streaming Media
- RTSP \bechen
- bug 938512 [RTSP] HTMLMediaElement does not fire loadedmetadata event for RTSP streaming.
- Patch reviewing.
- bug 898866 RTSP: the seek function usually fails
- Debugging: thread issue.
- bug 938512 [RTSP] HTMLMediaElement does not fire loadedmetadata event for RTSP streaming.
- WebRTC performance measurement \slee
- profile.sh can retrieve only main thread.
- jld's branch has compiling problem. Trying to figure out why.
- bug 911046 - WebRTC OmxCodec coding path support / JW
- review patch, trace code to see how to integrate h264 encoder/decoder
Graphic
- bug 933711 - [B2G][Helix][stability][fengximing]When the Firefox phone goes to sleep, wait for long time, if you wake up the phone, it took too much time to enter the lock screen. /Morris
- At nsFrameMessageManager::ReceiveMessage, we have about 600 event listeners at most. And bug 900221 would cause the time of 600 listeners iteration costs about 500ms which will slow down wake up procedure. I revert bug 900221 and looks like problem resolved.
- bug 709490 - Run WebGL on Web Worker with Off-Main-Thread-Compositing / Morris
- Update to address comment as well
- bug 920921 - [Flatfish]: Flatfish has bad performance on Homescreen wiping /Jerry
- Trace Layer inactive/active mechanism
- Test bug 940842 for pre-render. Compare BenWa's implement with mine.
- bug 939962 - Gallery app will get OOM process killed upon saving an edited photo after flashing or resetting the device
- Dump memory footprint when using gallery.
- We always get pmem OOM on buri device and fall back to ashmem.
- Dump memory footprint when using gallery.
- Flatfish performance followup /peter
- KitKat graphic porting plan /peter
- bug 929432 - [B2G][Helix][Browser][ChenHoulai] Open youtube in the browser, there is an extra horizontal line on the top when in full-screen mode /Peter
- Found the problem and work on patch now
- bug 944564 - [B2G] Page transition in Homescreen can't reach 60fps stably. /vlin
- Breaking down 3 problems cause this.
Input
- bug 924692 - Add support for selecting using touch / Phoebe
- append a fixed position AnonymousContent in CanvasFrame as CaretHandleFrame.
- study CSS pseudo stylecontext setup and modification.
- debug contenteditable focus on caretHandle assertion.
Stability
Free Agent
Study
- nsClipboard study / slee
- bug 854753 - [B2G][Audio] Implement SpeakerManager API /rlin
- landed