Re: kern/46955 (process deadlock (tstile) runing amanda sendsize - rename bug in tmpfs ???)

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

The following reply was made to PR kern/46955; it has been noted by GNATS.

From: David Holland <dholland%netbsd.org@localhost>
To: Wolfgang Stukenbrock <wolfgang.stukenbrock%nagler-company.com@localhost>
Cc: gnats-bugs%NetBSD.org@localhost
Subject: Re: kern/46955 (process deadlock (tstile) runing amanda sendsize -
 rename bug in tmpfs ???)
Date: Sun, 27 Apr 2014 21:11:45 +0000

 On Sun, Apr 27, 2014 at 10:37:06AM +0200, Wolfgang Stukenbrock wrote:
  > I've had a try with tmpfs changes into 5.x, and it seems to fix most
  > issues, but accedently not all - I've had one dead lock in tmpfs again some
  > time ago.
 
 There may be other issues. There were a bunch of stability fixes to
 tmpfs last fall; I thought many of them had gotten into 6.x but I'm
 told not. (rmind@ says he doesn't have time but can help out...)
 
 However, none of them AFAIK have been merged into 5.x, so there are a
 bunch of things that could have bitten you.
 
  > Due to my limited time and the massive changes in the fs-code and the way
  > process notifications takes place between 5.x and 6.x, I've stopped working
  > on it and started migration of all of our systems to 6.x. (Currently
  > suspended due to the dead-lock-problem described in PR 48733.)
 
 That seems reasonable. We are hoping to get 7.x branched any month
 now; it's definitely getting to the point where working on 5.x isn't
 all that worthwhile.
 
  > My ufs patch to 5.x seems to be stable (of cause, it is ugly ...), but I
  > think I've already send that one long time ago.
 
 Yup.
 
 -- 
 David A. Holland
 dholland%netbsd.org@localhost