On Thu, Jul 17, 2003 at 06:12:42PM +0200, Christophe Saout wrote: > Would it be possible to do development of these things (especially the > dm part) to an open place? I'm very interested in the inner workings of > device-mapper and the new 2.5/2.6 kernel block io code. Post your patches often, that way you'll get a lot more feedback. I tend to develop incremental patch sets. (eg. http://people.sistina.com/~thornber/internal-patches/raid1/) this allows people to see quickly what has changed. I have a little perl program that helps develop with such patchsets if you're interested. > If I had more time and would be able to understand these things better I > would have written some code by now (perhaps porting some of the missing > dm targets from 2.4 to 2.5 or so), but if I can see progress step by > step I'll perhaps be able to help. Or at least test it. I'm working on the missing targets (snapshots and mirroring) already, also these two are *far* more complex than your crypto target will be. Spend a day playing with dmsetup, and make sure you really understand how to write target types. The best way to learn is to try and write a trivial target type, eg. a 'reverse' target that divides the device into chunks and then reverses their order, ie. the first logical chunk is stored in the last logical chunk of the device. This should only take you ~150 lines of code (start with dm-linear.c and modify). > And, I was just wondering, would it be theoretically possible to write a > dm target that stores data in a file, so that it replaces what the loop > device does? Or was dm not designed for this? You could write a loop target, and people have suggested this already. It's not something that I'm interested in doing however. - Joe