Bug 2140537 - [RFE] dm-thin: allow supplying an offset to the table
Summary: [RFE] dm-thin: allow supplying an offset to the table
Keywords:
Status: NEW
Alias: None
Product: LVM and device-mapper
Classification: Community
Component: device-mapper
Version: unspecified
Hardware: All
OS: Linux
unspecified
unspecified
Target Milestone: ---
: ---
Assignee: Mikuláš Patočka
QA Contact: cluster-qe
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2022-11-07 05:26 UTC by Demi Marie Obenour
Modified: 2023-08-10 15:40 UTC (History)
7 users (show)

Fixed In Version:
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed:
Embargoed:
pm-rhel: lvm-technical-solution?


Attachments (Terms of Use)

Description Demi Marie Obenour 2022-11-07 05:26:30 UTC
Description of problem:
Unlike the linear and crypt targets, the thin target does not take an offset.

Version-Release number of selected component (if applicable):
6.0.7

How reproducible:
100%

Steps to Reproduce:
Try to overwrite the first part of a device (a partition table, say) with the following table

> 0 1 linear somedevice 0
> 1 100 thin 0 somepool / 1

Actual results:
There is no way to do this without a separate linear layer, which hurts performance.

Expected results:
It is possible.

Additional info:
“/” is used as a placeholder to indicate “no external origin device”.

Comment 1 Zdenek Kabelac 2023-02-07 16:35:17 UTC
Related to proposed patch: https://listman.redhat.com/archives/dm-devel/2023-February/053187.html
and developer's  NACK:  https://listman.redhat.com/archives/dm-devel/2023-February/053197.html

we really can't be adding 'linear' capabilities to every target.

Focus should go to optimize performance drop introduced by using 'extra' dm-layer on top of thin device.

Passing to Mikulas to consider what we can the to improve dm core logic here.


Note You need to log in before you can comment on or make changes to this bug.