Chris Murphy posted on Sat, 10 May 2014 10:49:07 -0600 as excerpted: > On May 8, 2014, at 7:06 AM, Chris Korent <[email protected]> wrote: > >> Linux sysresccd 3.10.35-std420-amd64 #2 SMP Wed Apr 2 18:31:51 UTC >> >> btrfs --version Btrfs v3.14.1 > > I'm uncertain if this is an OK combination. I'm under the impression > that the kernel should be the same or newer version than progs. Also > kernel 3.10 is old in Btrfs terms. Before altering the file system > further, I'd try a normal mount with the newest kernel you can get your > hands on.
In theory it's OK as the progs are supposed to be backward compatible at least some way (I wouldn't try anything pre-2.6.32 for sure as that was the last big format change, IIRC, and anything pre-3.5 or so may work but I'd not suggest it), but... there are of course bugs, and btrfs is still under intense enough development that there are bugs being fixed every kernel... So while yes, in theory the above version spread should work, the recommendation to try the latest stable-kernel from the latest stable series 3.14.x is an absolutely solid one, and in fact we're late enough in the 3.15-rc cycle now that trying with that is ultimately the best bet. In fact, as the output from mkfs.btrfs suggests, running the latest stable at the oldest really is recommended, and kernel is a bit more critical in that regard than btrfs-progs, since the kernel is what actually handles the filesystem and thus running an old kernel really is basically uselessly risking your btrfs and the data on it to known and now fixed bugs. -- Duncan - List replies preferred. No HTML msgs. "Every nonfree program has a lord, a master -- and if you use the program, he is your master." Richard Stallman -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
