On modern versions of Android running in AOT mode
FloatMath is slower than Math. Calls to Math.sqrt(),
etc. are replaced by intrinsics which can be as small
as a single CPU opcode.
When running in interpreted mode the new
implementation is unfortunately slower, but I'm
judging this acceptable and likely to be improved
over time. This change saves a small amount of native
code.
Example timings:
Mako AOSP AOT:
Method: Original / New / Direct call to Math
ceil: 596ns / 146.ns / 111ns
sqrt: 694ns / 56ns / 25ns
Mako AOSP interpreted:
Method: Original / New / Direct call to Math
ceil: 1900ns / 2307ns / 1485ns
sqrt: 1998ns / 2603ns / 1788ns
Other calls Mako AOT:
Method: Original / New
cos: 635ns / 270ns
exp: 566ns / 324ns
floor: 604ns / 150ns
hypot: 631ns / 232ns
pow: 936ns / 643ns
sin: 641ns / 299ns
The advice to use Math directly, in preference to
FloatMath, is still good. FloatMath will be deprecated
separately.
Bug: https://code.google.com/p/android/issues/detail?id=36199
Change-Id: If07fcbd78543d13bc6d75f9743f999860e8d58d7
Every incoming Binder call ends up triggering a BlockGuardPolicy
update, which would thrash between new AndroidBlockGuardPolicy
instances and BlockGuard.LAX_POLICY, causing GC churn.
This change avoids the extra allocations by recycling a single
AndroidBlockGuardPolicy in a ThreadLocal. Worst-case thrashing case
is now 10% faster; from 7.46us to 6.65us.
Bug: 9424568
Change-Id: I9c3b1c097a2aecc9b1f109a824cf3ea319fb3393
Current timings on a mako listed below, which show minimal overhead
compared to writing directly. The increased code readability and
automatic indenting is worth the overhead.
Also worth noting is that writing concatenated strings is
substantially faster than printing components separately.
benchmark us linear runtime
ComplexDirect 10712.7 =========================
ComplexIndenting 12623.7 ==============================
PairIndenting 45.8 =
PairRaw 82.1 =
SimpleDirect 282.2 =
SimpleIndenting 294.4 =
Change-Id: I7c38690c14b017fa46248ebb0be69f6beff03371
Switch to parsing detailed network stats with native code, which
is 71% faster than ProcFileReader.
Change-Id: I2525aaee74d227ce187ba3a74dd08a2b06514deb
Deprecate transport layer statistics, leaving only the summarized
network layer statistics.
Improve documentation to be clear about layers where measurements
occur, and their behavior since boot. Under the hood, move to using
xt_qtaguid UID statistics.
Bug: 6818637, 7013662
Change-Id: I9f26992e5fcdebd88c671e5765bd91229e7b0016
