[issue1158] Problems with a SATA DVD burner

Jeremy Chadwick jdc at parodius.com
Thu Dec 11 04:51:29 PST 2008


On Thu, Dec 11, 2008 at 12:24:08PM +0000, Steve O'Hara-Smith wrote:
> On Wed, 10 Dec 2008 23:38:08 -0000
> Matthew Dillon <bugs at lists.dragonflybsd.org> wrote:
> 
> > Matthew Dillon <dillon at apollo.backplane.com> added the comment:
> > 
> > I picked up a HP SATA DVD Writer today to play and so far it has worked
> >     flawlessly with the latest HEAD.  I was able to format and write
> 
> 	Seems I didn't try hard enough with the latest version of the code.
> At boot I get reams of:
> 
> Dec 11 12:05:03 df1 kernel: intr 10 at 40001/40000 hz, livelocked limit
> engaged! Dec 11 12:05:04 df1 kernel: intr 10 at 8604/20000 hz, livelock
> removed
> 
> 	With intermittent:
> 
> Dec 11 12:05:04 df1 kernel: unknown: FAILURE - INQUIRY timed out
> 
> 	Early on there are messages like:
> 
> Dec 11 11:47:05 df1 kernel: **WARNING** waiting for the following device to
> fini sh configuring:
> Dec 11 11:47:05 df1 kernel: xpt:        func=0xc0146dd9 arg=0
> 
> 	After a few of these there is:
> 
> Dec 11 11:47:05 df1 kernel: Giving up, interrupt routing is probably hosed
> 
> 	Eventually however (and this is the bit I didn't wait long enough
> for last time round) I get:
> 
> Dec 11 12:05:04 df1 kernel: cd0 at ata4 bus 0 target 0 lun 0
> Dec 11 12:05:04 df1 kernel: cd0: <TSSTcorp CDDVDW SH-S223F SB00> Removable
> CD-ROM SCSI-0 device Dec 11 12:05:04 df1 kernel: cd0: 3.300MB/s transfers
> Dec 11 12:05:04 df1 kernel: cd0: cd present [2295104 x 2048 byte records]
> Dec 11 12:05:04 df1 kernel: cd1 at ata6 bus 0 target 0 lun 0
> Dec 11 12:05:04 df1 kernel: cd1: <LITE-ON DVD SOHD-167T 9S19> Removable
> CD-ROM SCSI-0 device Dec 11 12:05:04 df1 kernel: cd1: 3.300MB/s transfers
> Dec 11 12:05:04 df1 kernel: cd1: Attempt to query device size failed: NOT
> READY, Medium not present
> 
> 	The livelock messages go away - and once I reenable DMA on atapi
> (it gets turned off somewhere along the line in the boot process)
> everything works fine. It just takes 15-20 minutes to get there.
> 
> 	I'm using an Asus A8V Deluxe motherboard - dmesg | grep ata
> 
> atapci0: <Promise PDC20378 SATA150 controller> port 0x8800-0x887f,0x9000-0x900f,0x9400-0x943f mem 0xfb700000-0xfb71ffff,0xfb800000-0xfb800fff irq 5 at device 8.0 on pci0
> ata2: <ATA channel 0> on atapci0
> ad4: 239372MB <Maxtor 6Y250M0 YAR51HW0> at ata2-master SATA150
> ata3: <ATA channel 1> on atapci0
> ata4: <ATA channel 2> on atapci0
> ad8: 117246MB <Maxtor 6Y120P0 YAR41BW0> at ata4-master UDMA133
> atapci1: <VIA 6420 SATA150 controller> port 0xa800-0xa8ff,0xb000-0xb00f,0xb400-0xb403,0xb800-0xb807,0xc000-0xc003,0xc400-0xc407 irq 10 at device 15.0 on pci0
> ata5: <ATA channel 0> on atapci1
> ad10: 286168MB <Seagate ST3300622AS 3.AAE> at ata5-master SATA150
> ata6: <ATA channel 1> on atapci1
> atapci2: <VIA 8237 UDMA133 controller> port 0xfc00-0xfc0f,0x376,0x170-0x177,0x3f6,0x1f0-0x1f7 at device 15.1 on pci0
> ata0: <ATA channel 0> on atapci2
> ata1: <ATA channel 1> on atapci2
> cd0 at ata4 bus 0 target 0 lun 0
> cd1 at ata6 bus 0 target 0 lun 0
> 
> 	The reuse of ata4 seems odd - related or irrelevant ?

I'm trying very hard to stay out of this discussion because I'm in no
way familiar with DFBSD's changes to ATA and ATAPI.  I am, however,
quite familiar with ATA/SATA on FreeBSD.  One thing stands out here like
a sore thumb:

Your SATA CD/DVD drive is connected to a Promise RAID controller.  I
have never seen this work, and vendors even go so far to state "DO NOT
HOOK UP CD/DVD DRIVES TO RAID CONTROLLERS".  The RAID controller is
probably freaking out because it doesn't know how to speak ATAPI
properly.

Do you see the same problem if you hook the SATA CD/DVD drive to your
VIA SATA controller?

-- 
| Jeremy Chadwick                                jdc at parodius.com |
| Parodius Networking                       http://www.parodius.com/ |
| UNIX Systems Administrator                  Mountain View, CA, USA |
| Making life hard for others since 1977.              PGP: 4BD6C0CB |






More information about the Bugs mailing list