On Wed, 10/8/14, Brian C. Lane <b...@redhat.com> wrote: > Subject: bug#18660: Partprobe and advanced format disks > To: 18...@debbugs.gnu.org > Date: Wednesday, October 8, 2014, 6:25 PM > > On Mon, Oct 06, 2014 at 10:06:45PM -0700, Nuzhna Pomoshch wrote: > > Partprobe is a fantastically useful too. It is absolutely > > vital to access new partitions without rebooting, as > > well as easily get to partitions on a mapped device. > > > > However, I am having a large problem with it now > > on a 4096 byte sector hard drive. > > You left out the most important tool :) What does the output from: > > parted -s /dev/mapper/disk u s p > > show?
Replied to you a month ago, but (following up) it seems that I am supposed to reply to the specific bug number (that was not clear to me before - this is the first time I have filed a bug here). Anyway, # parted -s /dev/mapper/disk u s p Model: Linux device-mapper (crypt) (dm) Disk /dev/mapper/disk: 732566133s Sector size (logical/physical): 4096B/4096B Partition Table: gpt Disk Flags: Number Start End Size File system Name Flags 1 256s 2097407s 2097152s Linux filesystem As I think I mentioned before (via command output), parted does see the disk having 4096 byte sectors. My guess (I haven't looked at the code, and I don't even know how much I could learn from doing so - I am not a programmer at all, and probably would not understand much, if anything) is that partprobe (and the libraries behind it) is just passing the sectors above to the device mapper (which uses 512 byte sectors exclusively) without making the necessary adjustment for the larger sector size. This would confirm that (again, these are 512 byte sectors): # dmsetup table disk1 0 2097152 linear 253:2 256 Nuzhna
