Hello
please find below the summary of the status report for wk03.2012. The full report with details is in
https://wiki.linaro.org/WorkingGroups/Middleware/Multimedia/WeeklyReport
Latest weekly meeting notes: https://wiki.linaro.org/WorkingGroups/Middleware/Multimedia/Notes/2012-01-17
= Highlights = - 12.01 release. Some statistics: + started with 16 Blueprints. Due to splitting the total number of blueprints became 20 out of which: + Completed: 3 blueprints + In good progress - may finalise in 12.01: 5 blueprints + Rescheduled for 12.02: 6 blueprints + Moved back to backlog: 5 blueprints + Finally 1 Blueprint remains to be determined if it will go to backlog or 12.02 Also worked on 1 extra technical debt blueprint (CMA Debug) The development was plagued by lots of issues (see below), which resulted in a lot of work in progress (team moving about trying to find how to best proceed). Despite the statistics, the cycle was a success since many of the issues were unearthed and pushed for fixing, hopefully improving the chances for good future releases.
- Lots of troubleshooting for audio issues - Completed CMA debug blueprint - technical debt from 11.12. One of the issues there was that the CMA lava test is too long to run (> 3 hours), which means short time to fix any issue (only 2-3 attempts are possible each day) - libjpeg-turbo: pulled in qualcomm 565 and 8888 updates, improved skia_bench time by 100%, pushed to git, integrated in linaro Android builds, knocked off open ubuntu libjpeg8 bugs, update pushed and in archive, put together draft lava testcase for tjunittest - For DRM : dri2video cleanup and patches - Continued tinyalsa/tinyhardware work for UCM - Continued e2eaudiotest - wrote tinyalsa based testfreq - Libav AAC decoder optimisation in progress, NEON code reviews for Libav and x264
= Issues = - Bugs + #893402 is still an issue for end-to-end audio testing, it mostly works with the right configuration but there should configuration fixed on the rootfs and also there are cases where bus exceptions can appear (see bug commentary) + #911860 Android NEON detection is busticated - Triaged + #912035 - Android: skia-bench segvs - Fix Committed + #898395 - jpegint.h missing in -dev package - Fix Committed + #889844 - Audio on HDMI is not working on trackin-panda - Fix Released from Landing Team + #908957 - ICS: All arches: No aplay / arecord or equivalent + #880173 - Audio does not work on Upstream Panda - Other issues + skia_bench timers inconsistent, it is difficult to evaluate performance improvement + Remaining Realvideo work on hold pending completion of upstream changes to decoder + LAVA: what is the best way to verify testcases?
Best regards,