Investigate block allocation in FS

[tixy] Write skeleton for simulation tool: DONE
[tixy] Implement algorithm "pure linear access": DONE
[tixy] Implement algorithm "block remapping within one erase-block": DONE
[tixy] Implement algorithm "data logging": DONE
[tixy] Implement algorithm "cache for small writes": DONE
[tixy] Create 'algorithm exposer' tool: DONE
[tixy] Validate simulations by comparing their results to real timings: DONE
[tixy] Automate creation of an aged file system: TODO

[tixy] Gather media write stats for all automated test cases on all filesystem types and options: DONE
[tixy] Repeat above on real ARM hardware, both uniprocessor and SMP: TODO
[tixy] Repeat selected test runs so we can determine noise from real differences: DONE
[tixy] Create 'desktop system monitor' script: TODO
Get volunteers to run 'desktop system monitor': TODO

Collect block trace data: TODO
- distribution installation using debootstrap: TODO
- parallel kernel compile: TODO
- git clone/checkout: TODO
- streaming data write: TODO
- regular desktop use: TODO
- others (to be added): TODO

Find a set of ext4 mount options to try (Surbi): TODO

Create new Linaro project for modeling tool: TODO
- Choose an appropriate license: TODO
- Create a source repository: TODO

Build a cross-reference of workloads and known devices: TODO

Discuss test results to mailing lists: TODO
Present test resutls to TSC: TODO
Follow up with test results on lwn.net: TODO

Based on results, need to decided whether to focus on block allocator or remmaper.

Work items:
[tixy] Create script for running test cases: DONE
[tixy] Modify blktrace to generate raw dump: DONE
[tixy] Run test script to gather media write stats for test cases on a subset of filesystems: DONE