Here is a list of what people in the ART optimizing compiler team are working on:
*Zheng Xu:*
- Use LDP/STP on arm64 for slow paths (landed). https://android-review.googlesource.com/148028
- Speedup div/rem by constants on arm32 and arm64 (landed). https://android-review.googlesource.com/150710
- Use LDP/STP and 32bit moves in the parallel move resolver on arm64 (waiting for internal review). https://dev-private-review.linaro.org/122
*Serban Constantinescu:* - Working on the team's transition to Linaro. ART team wiki page: https://wiki.linaro.org/Internal/ARTIntroduction
- Move xSELF from x18 to x19 (waiting for Google's review) https://android-review.googlesource.com/#/c/148873/ https://dev-private-review.linaro.org/124
*Alexandre Rames:*
- Rebasing the architecture-specific optimisation framework (will be discussed face to face with Google).
- Extending VIXL literal pool API to enable patching the literal pool after the code has been finalized.
- Adding disassembly support to the .cfg output. (waiting for Google's review) https://android-review.googlesource.com/148699 https://dev-private-review.linaro.org/118