aospx

Investigate CodeAurora memcpy/memset implementations

Registered by Rob Beane on 2012-07-16

The memcpy and memset implementations found in CodeAurora may be better than ours. If they are we should merge them

Blueprint information

Status:: Complete

Approver:: Rob Beane

Priority:: Undefined

Drafter:: Rob Beane

Direction:: Approved

Assignee:: Rob Beane

Definition:: Obsolete

Series goal:: None

Implementation:: Blocked

Milestone target:: obsolete

Started by: Rob Beane on 2012-07-16

Completed by: Rob Beane on 2012-07-22

Related branches

Related bugs

Sprints

Whiteboard

[2012/07/22 robbeane] After some testing it appears that Linaro's mem* routines are faster than CodeAurora's version. So at this point we will continue to use Linaro's memcpy/memset routines.

0xbench VM benchmark showed an average increase of 150ms with CodeAurora mem* routines (lower is better).

(?)

Work Items

Work items:
Check out CodeAurora memcpy: DONE
Check out CodeAurora memset: DONE
Benchmark them against ours (Cortex-string's): DONE
If they're faster, merge them: DONE

This blueprint contains Public information

Everyone can see this information.

Subscribers

No subscribers.