Finally found some time[0] today to work on syscalltrack. Worked on the rules module, getting it to compile and link with kernel 2.6. Next up is seeing whether it loads, and then whether it actually runs. If it does, the hard part of the port to 2.6 is done, and everything else is just cleanup work. Once I get home[1], I’ll run it, and then we’ll see.
[0] Ugh, who am I kidding. It’s not time I was missing, it’s inclination. Thankfully, I seem to have found some. [1] I’m in Compilation class now, and afterwards I’ll head for a quicky workout at the gym.
Muli Ben-Yehuda's journal 
What did you learn in your compilation class? I am very envy of you. Over here, I am more worried about my grades, than to have the opportunity to learn. Guess I won’t have the chance to learn to write compilers here. I will be filing for an early graduation, get a degree with merit, and then return to the army. Hopefully, I get to do a masters programme later when I started working.
Do you know anything about Sun GridEngine, PVFS1/2, NUMA implementation, Rocks Clusters, etc? I am trying to learn… Might be useful for work.
Can’t wait to touch syscalltrack since my health is better now, school ending soon, holidays coming ๐
Comment by ideawerkz — November 2, 2003 @ 4:13 PM |
In compilation yesterday we went over several techniques related to LL(0), LL(1), SLR and LALR parsing. I missed the last couple of classes, so I spent most of the class with my nose buried in the book reading the material that I missed. We are working very closely with the dragon book, and have just finished chapter 4.
I care about grades, but I care more about studying interesting things. I try to balance work, life, free software and studying, and it’s never easy. I do need a high enough grade average to get into a masters program when I graduate.
I know a fair bit of stuff about various clustering types, yes. I’ve been involved with various clustering projects in the past and I think it’s the direction to go in the future. Running a single kernel on N processors where N is large enough is losing battle against complexity, whereas running N kernels that communicate amongst themsleves makes the complexity manageable.
Glad to hear your health is better, and there’s tons of work to do on syscalltrack. What would you like to work on?
Comment by mulix — November 3, 2003 @ 1:44 AM |
Seems like you are doing senior modules. Very interesting. I don’t know. I might be doing this module too if my supervisor refuses to let me drop honours. hehe.
Over here… to get into a good masters program, we can’t care about studying interesting things. we are required to have a flawless grade average. too bad for me.
if you have a choice to work on cluster-related project, would you choose to hack NUMA or pvfs to implement pvfs2 features, or play with sun gridengine?
i have not concentrate on syscalltrack because of my exams, but i will let you know which work i would like to work on soon! can you direct me to a list of TODOs if you have please?
Comment by ideawerkz — November 4, 2003 @ 2:29 PM
I will probably hack NUMA, or even better, Single System Image clusters (i.e. MOSIX).
There’s a list of TODOs in the source distribution, just look at the TODO file.
Comment by mulix — November 5, 2003 @ 5:21 AM
i have not tried mosix. i have decided on numa too, but i got to check the TODO. that is for my new job ๐
Comment by ideawerkz — November 5, 2003 @ 5:57 AM