Re: kern/55362: bad inodes created after dump|restore

[Patrick Welche <prlw1%cam.ac.uk@localhost>]

The following reply was made to PR kern/55362; it has been noted by GNATS.

From: Patrick Welche <prlw1%cam.ac.uk@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: kern/55362: bad inodes created after dump|restore
Date: Fri, 17 Jul 2020 15:01:38 +0100

 Summary (not necessarily in time order):
 
 - all with a kernel without even DIAGNOSTIC
   (as already takes ca 17hours, and LOCKDEBUG panicked)
 - 2 rounds of SeaTools tests (including long) on all disks passed
 - dd zeros to raid device and cmp components OK
 - dd ONES  to raid device and cmp components OK
 - dump & restore to cgd on raid of zeros (as decribed above)
   FAILs cmp of components (see blocks of zeros on one component)
 - dump & restore to cgd on raid of ONES (as decribed above)
   FAILs cmp of components (see blocks of ones on one component)
 - moved dump to other computer, so dump & restore went across network
   to change timings - same results with ones & zeros
 - dump & restore only 1.5T data to raid rather than cgd - cmp OK! (chance?)
 - no difference with cgd before & after aes changes
 
 Issue with layered filesystem?
 
 Now trying with DIAGNOSTIC/DEBUG/LOCKDEBUG kernel and hit PR kern/55493