Re: bin/50108 (fsck_ffs fails replaying wapbl journal on filesystem with 4k sectors)

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

The following reply was made to PR bin/50108; it has been noted by GNATS.

From: David Holland <dholland-bugs%netbsd.org@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: bin/50108 (fsck_ffs fails replaying wapbl journal on filesystem
 with 4k sectors)
Date: Sat, 5 Sep 2015 23:43:52 +0000

 On Mon, Aug 17, 2015 at 08:05:01PM +0000, Eugene wrote:
  >  [...]
  >  > problem is not understood unless the explanation is posted to
  >  > gnats :-)
  >  Why? Michael's explanation and fix look right: I've tested his proposed
  >  solution, and it works right for both 4k and regular 512 byte sectors. 
 
 It's not included in the PR. It *is* included in 45933, but I'd
 forgotten about that one and it wasn't referenced in here here until
 later.
 
 If the issue is "understood" but doesn't actually get fixed, what
 happens is that someone coming by months (or years) later has no idea
 what was understood and has to redo the work. (Or if they're lucky
 they can ask the original person, but things get forgotten.)
 
 so it was just a procedural complaint.
 
  >  Wouldn't it be better to modify fsck_ffs code (that is the
  >  wapbl_read()/wapbl_write() functions in sbin/fsck_ffs/wapbl.c -
  >  kern/vfs_wapbl.c calls them back for IO when working in user mode) 
  >  to use the same DEV_BSIZE unit as kernel when replaying the journal? 
 
 This is a thorny and ugly issue and there isn't any one good solution
 to it that doesn't involve digging up and chewing over all of ffs. :(
 
 -- 
 David A. Holland
 dholland%netbsd.org@localhost