Investigate Cyanogenmod memcpy/memset implementations

Registered by Bernhard Rosenkraenzer on 2012-06-26

Why?
The memcpy and memset implementations found in Cyanogenmod may be better than ours. If they are we should merge them, if they're not then we should document our experiment on a Wiki for reference.

Context?
General Android optimization work.

What gets produced?
Android builds with the changes or a Wiki page with the results of the experiment.

Where will the work get put?
In Android builds or on a Wiki.

Blueprint information

Status:
Started
Approver:
Zach Pfeffer
Priority:
Low
Drafter:
Bernhard Rosenkraenzer
Direction:
Approved
Assignee:
Bernhard Rosenkraenzer
Definition:
Approved
Series goal:
Accepted for future
Implementation:
Started
Milestone target:
milestone icon backlog
Started by
Bernhard Rosenkraenzer on 2012-07-04

Related branches

Sprints

Whiteboard

Notes:
[2012/5/31 pfefferz] Put notes here.

Meta:
Roadmap id: ANDROID-OPTIMIZATION
Headline: Cyanogenmod's memcpy/memset implementations have been investigated for Linaro Android
Acceptance: Cyanogenmod's memcpy/memset implementations have been investigated and either merged, or documented as not being faster than ours.

(?)

Work Items

Work items:
Get stringbench and friends into the tree: DONE
Check out CyanogenMod's memcpy: TODO
Check out AOKP's memcpy: TODO
Check out CyanogenMod's memset: TODO
Check out AOKP's memset: TODO
Benchmark them against ours (Cortex-string's): TODO
If they're faster, merge them, otherwise create a Wiki page documenting the lack of performance gains: TODO

This blueprint contains Public information 
Everyone can see this information.