Re: unable to read RDB block due to b9684a71fca793

[Christoph Hellwig]

On Wed, Jul 06, 2022 at 09:53:36AM +0700, Bagas Sanjaya wrote:
> Hi,
> 
> I first noticed this issue on v5.18.9, and still found on v5.19-rc5.
> 
> Looking at dmesg, I see error message on loop8 device:
> 
> [   41.319699] Dev loop8: unable to read RDB block 8
> [   41.320566]  loop8: unable to read partition table
> [   41.320597] loop8: partition table beyond EOD, truncated
> 
> My Debian 11 laptop is also run LXD (as development server).
> 
> Bisecting between v5.18 and v5.19-rc5, the culprit is commit b9684a71fca793
> ("block, loop: support partitions without scanning").

Which just restores the previous behavior of optionally allowing to
scan for partitions on all loop devices.  So that error had been
there before and just disappeared due to a regression.