summaryrefslogtreecommitdiff
path: root/AbstractFreenix2004.mdwn
blob: 2575c35e1400db9cfc1d75dfa081a9c6240e3462 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
# <a name="Abstract for Freenix 2004"></a> Abstract for Freenix 2004

Bart proposed that the team submit a 3-5 page abstract to [Freenix 2004](http://www.usenix.org/events/usenix04/freenix.html), of which he and Keith are chairs. Possible topics:

- work rewriting the [[CanBusLinuxDriver]] (including CANtelope)
- overall progress on the [[FlightComputerSoftware]] since the Usenix paper.

Points to consider and areas of interest (from the above page)

- open source
- available online: Twiki, [[WebCVS]]
- project reports
- Usage and development experiences, good and bad
- Free and Open Source Software Engineering is a special focus this year
  - spec and design methods [eh, whats that? :-) ]
  - testing
  - reliability
  - performance
  - Team governance, admin, management [about PSAS, Twiki]
  - Planning and forecasting [eh, whats that? :-) ]
  - Measuring progress and assessing quality
- Embedded systems

**Submission deadline is 16-Dec-2003**

## <a name="CanBusLinuxDriver work"></a> [[CanBusLinuxDriver]] work

_Brainstorm. Feel free to comment and edit._

Motivation

- brief description of FC/LTC architecture
- brief description of CAN bus
- Goal: all aspects of rocket open source (CAN King -&gt; CANtalope)

Starting point

- MOPS board, Intel chip, Debian Linux 2.4.18
- existing open source CAN driver
- mods required
- device model, capabilities and limitations
  - no support for RTR messages
  - use of `read` for RTR breaks normal driver model
  - 1 device per message object doesn't make full use of the chip
  - no support for `select`
  - may not support telemetry bandwidth

Driver rewrite

- learning to write a driver
  - existing CAN driver
  - prior Linux experience
  - pointers to material on web
  - Intel chip docs vs. experimentation with `sucan`
  - interrupts
- methodology
  - borrow from existing driver
  - build environment, system
  - `ssh/scp` to FC/LTC
  - serial port, `minicom` for driver `kprintfs`
- general vs. specific use and requirements
  - simplify simplify simplify

Tools

- CAN King, motivation for CANtalope
  - [[CantalopeHardware]]
  - [[CantalopeSoftware]]
- `sucan` interactive chip driver
  - thanks to `iopl`
- [[RocketView]], unit testers, LTC, FC

Philosopy/Conclusions

- specific mods still useful to open source community? should we care?
- credit to previous driver
- make best use of what you got; no compromises for generality
- small steps
- learning experience

-- [[IanOsgood]] - 03 Dec 2003