der Mouse wrote:
If so, then raidframe won't work with such large disks. You need [...] rf_reconstruct.c 1.95.0.2 or later.4.0.1 seems to use 1.95.2.something, so I'll probably just push the machine to 4.0.1. Thank you very much!Now done. Doesn't help. And yes, I've gone to some lengths to verify that the kernel that just now crashed is indeed the one I built an hour or two ago from the 4.0.1 source tree (which means, with rf_reconstruct.c,v 1.95.2.1). Or is 1.95.2.1 _not_ later than 1.95.0.2? I don't totally understand CVS numbering.
CVS version numbering isn't that strange really. You just need to read up on the documentation.
1.95.2.1 might not be later than 1.95.0.2.What you need to understand here is branch versions. Any CVS version number in which the next to last number is zero is a branch point. And the last number is then the number of that branch point (you can obviously have several different branches from the same point).
So, 1.95.0.2 is a branch, and the base point for 1.95.2.x. If you check out the branch, you will actually get the highest version that exists from that branch, and not the version where the branch occurred, even though if you look at the tag, that might be what you think you get.
So, in your case, you might get the same version, or you might get a newer version. Since 1.95.2.1 exists, you should never get 1.95 (which is what it branched from).
If you read the CVS documentation, you should look for magic versions (or something like that, I can't remember the exact title, and I don't have the CVS documentation in front of me).
Johnny