[05:03] infinity and apw do you want to be debugs? === ara is now known as Guest43853 [11:28] so when do we get new kernels without rebooting? [13:47] Does the ubuntu kernel in any way change the default mount options for ext4? IE barriers=1, data=ordered? [13:50] alexbligh1, not using fstab you mean? not that i know of, but not something i have wanted to do [13:51] apw, I mean fstab is specifying nothing fancy, yet I'm seeing a (postgres) crash behaviour that is similar to what I'd expect if barriers=0. I'm wondering if the (old) hdd doesn't support barriers (though the telltale message isn't in the log) or whether there's an 'optimisation' in the kernel (cough). (f)stab in the dark. [13:52] alexbligh1, not that i am aware of, no ... hrm [13:53] apw, dim recollection of different distros using different default mount options but I think that was for ext3 a hundred years ago. [13:56] alexbligh1, yep i'd agree with that feeling [13:58] ape, thx anyway [13:58] hi, you probably also want 808f80b46 in addition to 46e971cdcd (upstream 005efedf2c) in 3.19-queue [13:59] (btrfs read corruption issues) [14:00] henrix, ^ [14:00] kamal, ^ [14:04] jtaylor: ack, thanks. seems to be urgent, so we'll probably squash it into the stable kernels under review [14:05] not super urgent as you need to compress and deduplicate, but as you have one of the patches you might as well have both [14:07] jtaylor: yeah, since that extra commit is also tagged for stable, it will get into stable kernels (in fact, it's already in the 3.16 -queue!) [14:07] its already in 4.2 stable [15:16] jtaylor, yes, thanks for the heads-up. I'll apply this to 3.19-stable and 3.13-stable today and so it'll release with the other btrfs read corruption fix [15:16] 808f80b Btrfs: update fix for read corruption of compressed and shared extents [15:16] henrix, fyi ^ [15:17] kamal: yep, i was going to do exactly the same for 3.16 ;) === henrix_ is now known as henrix === th3s3_3y3s is now known as kingofanything === kingofanything is now known as th3s3_3y3s