Re: panic with devclass_get_name (dc=0xdeadc0dedeadc0de)

From: Alexander Motin <mav_at_FreeBSD.org>
Date: Wed, 24 Nov 2010 19:57:36 +0200
Roger Hammerstein wrote:
> Using ATA_CAM it doesn't panic, it removes the bad disk from /dev after
> the problem.
> 
> Nov 24 12:36:15 butter kernel: ata4: SATA connect time=0ms status=00000113
> Nov 24 12:36:15 butter kernel: ata4: reset tp1 mask=01 ostat0=d0 ostat1=00
> Nov 24 12:36:15 butter kernel: ata4: stat0=0xd0 err=0x00 lsb=0x54 msb=0x02
> Nov 24 12:36:15 butter last message repeated 309 times
> Nov 24 12:36:15 butter kernel: ata4: reset tp2 stat0=d0 stat1=00 devices=0x0

It looks like controller or device stuck in some irresponsible state,
where soft reset unable to get device back. You can make driver execute
also hard reset before soft reset by changing in file ata-serverworks.c:
	if (ata_sata_phy_reset(dev, -1, 1))
to
	if (ata_sata_phy_reset(dev, -1, 0))

Please report what you get.

-- 
Alexander Motin
Received on Wed Nov 24 2010 - 16:57:51 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:09 UTC