[Bug 1810235] Re: cryptsetup silently fails in live environment when given a raw "bcached" block device

2019-01-28 Thread Robie Basak
I hope my change of bug title clarifies the edge case when this bug occurs. If I have misunderstood I'd appreciate being corrected. Because this is a pretty uncommon use case, I'm marking the bug importance as Low. Volunteers welcome - though I suspect that any volunteering would best be done

[Bug 1810235] Re: cryptsetup silently fails in live environment when given a raw "bcached" block device

2019-01-28 Thread Robie Basak
There is no action possible in the bcache-tools package so I'm marking that task as Invalid - but the cryptsetup task remains open. ** Changed in: bcache-tools (Ubuntu) Status: New => Invalid -- You received this bug notification because you are a member of Ubuntu Bugs, which is

[Bug 1810235] Re: cryptsetup silently fails in live environment

2019-01-25 Thread TJ
It may be that cryptsetup needs to be taught to use offset= and skip= for LUKS as well as plain. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1810235 Title: cryptsetup silently fails in live

[Bug 1810235] Re: cryptsetup silently fails in live environment

2019-01-21 Thread Nathan
Possibly, if there is a way to ensure bcache only caches encrypted block data and not the cleartext blocks. There may be a way of setting the backing drive to work with a different ordering but I'll need to run some tests and I'm not the most knowledgeable in regards to bcache, its been a

[Bug 1810235] Re: cryptsetup silently fails in live environment

2019-01-14 Thread Andreas Hasenack
Could this be solved with different ordering? Luks first, then bcache? Luks definitely expects its header to be in a certain position: LUKS EXTENSION LUKS, the Linux Unified Key Setup, is a standard for disk encryption. It adds a standardized header at the start of the device, a key-slot

[Bug 1810235] Re: cryptsetup silently fails in live environment

2019-01-14 Thread Andreas Hasenack
Could this be solved with different ordering? Luks first, then bcache? Luks definitely expects its header to be in a certain position: LUKS EXTENSION LUKS, the Linux Unified Key Setup, is a standard for disk encryption. It adds a standardized header at the start of the device, a key-slot

[Bug 1810235] Re: cryptsetup silently fails in live environment

2019-01-12 Thread Nathan
Following up and to clarify my initial post a bit, I've found the main issue is with cryptsetup (or related system) expecting LUKS to be located within a certain boundary and bcache creates an offset on the backing drive that shifts this boundary outside where cryptsetup looks. The package should

[Bug 1810235] Re: cryptsetup silently fails in live environment

2019-01-01 Thread Nathan
Additional Information: gdisk shows the partition type as , the drive (sda3) boots and runs without issue. Cryptsetup however; cannot access it after boot but it does properly mount and run the contents of the encrypted partition (i.e. root drive). The drive was originally set up following