Feed on
Posts
Comments

Snappy Jan12

January 12th, 2012 by

The most user facing fix has been discovery and removal of some sneaky cache IO on the main thread.

Saptashi did some analysis on the impact of running sqlite in async mode on mobile. Turns out it’s only a win for DELETEs. Expect a blog post from him soon.

Dave discovered that we sometimes wait on locks on the main thread.

Jeff and Bas are looking into diagnosing when d2d causes a slowdown.

There was discussion of 4x reduction in cycle collection times landing soon, focusing on having cycle collector run less, etc. Lots of work(chromehang, profiler, …) is continuing from last week.

4 Comments »

4 Responses to “Snappy Jan12”

  1. on 12 Jan 2012 at 9:13 pm   Kyle Huey

    Er, of course we sometimes wait on locks on the main thread. That’s how locks are used for synchronization, after all. The bug here is that some code (the cache, I guess?) is doing I/O while holding a lock, so the time the lock is held is unbounded. From a performance perspective, in the worst case that’s equivalent to doing the I/O on the main thread in the first place.

  2. on 12 Jan 2012 at 11:03 pm   tglek

    Kyle, right.

  3. on 13 Jan 2012 at 8:36 am   Ed

    Tara you got my sessionstore.js yet?

  4. on 13 Jan 2012 at 10:17 am   Hera

    “Jeff and Bas are looking into diagnosing when d2d causes a slowdown.”

    One thing I noticed is that, for a very long time, there hasn’t really been any work on optimizing D2D (aside from crash fixing and such). FX D2D implementation is really CPU heavy and any optimization is welcome!