I installed 6.0 on a system with two SATA drives in RAID-1. 5.4 is
installed on a standard IDE channel. All appeared to work well until
sometime after updating the system to 6-STABLE, installing the Nvidia
driver and copying a lot of data (mainly ogg files) from my old hard
drive onto it. I started seeing a multitude of READ_DMA (or maybe
WRITE_DMA) errors in /var/log/messages. The system was beginning to
slow down a lot, so I rebooted. Since then I have been unable to keep
it up without a panic after an fsck in single-user mode.
I realize this does not tell much, but just when ld.so.hints is being
removed due to being corrupt the follow panic appears: Panic:
ffs_valloc: dup alloc The panic may just be from a system that is too
corrupted. I am amazed it still boots.
I am not sure if it is -STABLE or -RELEASE where the problem really
started since I did not copy anything over until everything look good.
I did not lose anything important, so I do not need to save anything.
How should I proceed or does anyone know if this is a known problem? I
am able to mount the partitions within the 5.4 system without issue.
Possible related PR's:
i386/75531 (http://www.freebsd.org/cgi/query-pr.cgi?pr=i386/75531)
kern/89069 (http://www.freebsd.org/cgi/query-pr.cgi?pr=kern/89069)
Other oddities:
1. /var/log/messages showed some messages from my 5.4 system at first.
Are those stored in the BIOS?
2. While copying the data over, I noticed the system running slower and
slower. This may have been from all the messages being saved into
messages about READ_DMA.
Configuration:
ASUS A7V880
atapci0: <VIA 6420 SATA150 controller> port
0xeff0-0xeff7,0xefe4-0xefe7,0xefa8-0xefaf,0xefe0-0xefe3,0xef90-0xef9f,
0xe800-0xe8ff irq 20 at device 15.0 on pci0
ad4: 152627MB <Seagate ST3160827AS 3.42> at ata2-master SATA150
ad6: 152627MB <Seagate ST3160827AS 3.42> at ata3-master SATA150
ar0: 152627MB <VIA Tech V-RAID RAID1> status: READY
ar0: disk0 READY (master) using ad4 at ata2-master
ar0: disk1 READY (mirror) using ad6 at ata3-master
Se?n
--
sean-freebsd@farley.org
Brian Fundakowski Feldman
2005-Dec-21 14:07 UTC
File system corruption with ATA RAID-1 on 6-STABLE
On Sat, Nov 19, 2005 at 12:06:00PM -0600, Se?n C. Farley wrote:> I installed 6.0 on a system with two SATA drives in RAID-1. 5.4 is > installed on a standard IDE channel. All appeared to work well until > sometime after updating the system to 6-STABLE, installing the Nvidia > driver and copying a lot of data (mainly ogg files) from my old hard > drive onto it. I started seeing a multitude of READ_DMA (or maybe > WRITE_DMA) errors in /var/log/messages. The system was beginning to > slow down a lot, so I rebooted. Since then I have been unable to keep > it up without a panic after an fsck in single-user mode. > > I realize this does not tell much, but just when ld.so.hints is being > removed due to being corrupt the follow panic appears: Panic: > ffs_valloc: dup alloc The panic may just be from a system that is too > corrupted. I am amazed it still boots.That's how it appears, yeah. What does the drive health look like? Seems most like a chipset bug or nvidia driver bug, but could be things are just coinciding with drive or drive controller failure. -- Brian Fundakowski Feldman \'[ FreeBSD ]''''''''''\ <> green@FreeBSD.org \ The Power to Serve! \ Opinions expressed are my own. \,,,,,,,,,,,,,,,,,,,,,,\