post-3979

that’s flattering – one of the tests available for the OSDL’s STP (scalable test platform) is my old tiobench – they’re asking for some changes, and it’s a nice 2-way street – i finally get the “fleet” of test machines (CPU’s ranging from 1 to 8, memory ranging from 256MB to 8GB or something like that) to play around with and they’ll get a better tiobench out of it :)

I’m actually glad this happened, mainly because when I developed tiobench (esp. the mmap stuff, since that’s how I was processing our 128GB hardware traces, and hence the need for tiobench at all), madvise wasn’t in the mainline kernel and the impl. from the scability project wasn’t quite there yet, so it’ll be good to get back to that and finally have a chance to play with madvise.

This is one of those days that I really hate not taking that job in Portland to be a Linux kernel hacker for the IBM LTC (Linux Technology Center). I shoulda listened to :P

I forgot to mention – over the course of trying to pitch the OSDL (unsuccessfully) to fund the next-generate hardware trace tool, I ended up with some emails with Paul McKenney, which was nice since he was the main tech interviewer of me for the Portland job. That was a fun interview, but it turns out he got moved over to their storage area (which is a good sign for storage, a large chunk of the SCO-contested things like RCU are his doing) and he misses LTC work :)

I should probably look around again… if I could get an IBM LTC position in RTP, that’d be worth considering.

About these ads