| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
| |
Change-Id: Ifa04ea6f31cf0677bf95263188a137fd8dd721eb
|
| |
|
|
| |
Change-Id: I3f676929e1ce919fee254c2ac5320f2070b2a14f
|
| |
|
|
|
|
| |
Also a few more slight changes of emphasis.
Change-Id: If28f3389e7431f1ea37ffb118d8e0525fd2a00fd
|
| |
|
|
|
|
|
| |
(I'll start checking in all my benchmarks -- not just the ones relevant to this
document -- somewhere under dalvik/libcore/ tomorrow, and we can see about
linking to them for those who want something they can actually run, or just
want to see the exact code that was benchmarked.)
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
A lot of this documentation isn't even true of the G1, let alone Froyo running
on a Nexus One. Distinguish between truth and fiction, clarify where the JIT
affects things, and clarify certain confusions (such as the difference between
intrinsics and native methods).
I still need to include updated performance numbers in the final section. I
should also make the benchmark code available so that people don't have to
take our word for these things, and so it's easier for them to get an idea of
the performance of future devices and builds. (Though hopefully we can update
this every release in future.)
Anyway, just removing the untruths is a big step forward.
|
| |
|
|
|
|
|
| |
Clean up blockquote usage in other files.
Bug: 2160782
Change-Id: I537194ecddc53d4f0e8456ce958ae8db6f521593
|
| |
|
|
|
| |
Bug:2160782
Change-Id: Iaf5d2cc2e3c657700469e8b7394a95bc03fc26f3
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
