Re: kern/47778: Kernel panics when accessing file system on faulty medium

[David Holland <dholland-bugs%netbsd.org@localhost>]

The following reply was made to PR kern/47778; it has been noted by GNATS.

From: David Holland <dholland-bugs%netbsd.org@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: kern/47778: Kernel panics when accessing file system on faulty
 medium
Date: Sun, 28 Apr 2013 16:04:27 +0000

 On Sun, Apr 28, 2013 at 01:35:00PM +0000, Aleksej Saushev wrote:
  > When accessing partially faulty medium (HDD), kernel panics with
  > the following backtrace (recorded manually since either savecore
  > or core dump seems not to function correctly):
 
 Yes, corrupted filesystems will crash the kernel. It would be nice if
 this were not the case, but it is, and it's basically not fixable.
 Part of the point of pooka's stuff is to allow running filesystem code
 in a way such that these crashes are not fatal. (You might want to try
 that.)
 
 -- 
 David A. Holland
 dholland%netbsd.org@localhost