Re: kern/50707: rm: fts_read: No such file or directory

[BERTRAND Joël <joel.bertrand%systella.fr@localhost>]

The following reply was made to PR kern/50707; it has been noted by GNATS.

From: =?UTF-8?Q?BERTRAND_Jo=c3=abl?= <joel.bertrand%systella.fr@localhost>
To: gnats-bugs%NetBSD.org@localhost, kern-bug-people%netbsd.org@localhost, gnats-admin%netbsd.org@localhost,
        netbsd-bugs%netbsd.org@localhost
Cc: 
Subject: Re: kern/50707: rm: fts_read: No such file or directory
Date: Tue, 5 Jul 2016 09:59:25 +0200

 BERTRAND Joël a écrit :
 > The following reply was made to PR kern/50707; it has been noted by GNATS.
 >
 > From: =?UTF-8?Q?BERTRAND_Jo=c3=abl?= <joel.bertrand%systella.fr@localhost>
 > To: gnats-bugs%NetBSD.org@localhost, kern-bug-people%netbsd.org@localhost, gnats-admin%netbsd.org@localhost,
 >          netbsd-bugs%netbsd.org@localhost
 > Cc:
 > Subject: Re: kern/50707: rm: fts_read: No such file or directory
 > Date: Mon, 27 Jun 2016 09:10:28 +0200
 >
 >   	Some news.
 >
 >   	Last saturday, I have tried to bissect fts code without any success,
 >   but I have seen that disk corruption only occurs on high disk I/O (for
 >   example make clean in ports tree, I use MAKE_JOBS=8 as this server
 >   contains an i7 CPU with 8 threads). All disk corruptions are directories
 >   without '.'. fsck returns that directories were 'half allocated'.
 >
 >   	Best regards,
 >
 >   	JKB
 >
 >
 
 	Hello,
 
 	Another one, but without any 'rm'. This morning, I have received a mail 
 from Charlie root :
 
 daily insecurity output for Tue Jul  5 03:15:01 CEST 2016
 Checking setuid files and devices:
 Setuid/device find errors:
 find: fts_read: No such file or directory
 
 	I wonder if this issue could be related to disk internal cache 
 strategy. Now, both disks return :
 legendre# legendre# dkctl /dev/wd1 getcache
 /dev/wd1d: read cache enabled
 /dev/wd1d: write-back cache enabled
 /dev/wd1d: read cache enable is not changeable
 /dev/wd1d: write cache enable is changeable
 /dev/wd1d: cache parameters are not savable
 legendre# dkctl /dev/wd0 getcache
 /dev/wd0d: read cache enabled
 /dev/wd0d: write-back cache enabled
 /dev/wd0d: read cache enable is not changeable
 /dev/wd0d: write cache enable is changeable
 /dev/wd0d: cache parameters are not savable
 legendre#
 
 	Is there a hope to fix this particularly bad issue ?
 
 	Best regards,
 
 	JKB