Thanks Rob.  That’s an interesting thread.  I’ll look into that param and google around for similar.

 

 

From: tclug-list-bounces at mn-linux.org [mailto:tclug-list-bounces at mn-linux.org] On Behalf Of Robert Nesius
Sent: Friday, May 28, 2010 9:39 AM
To: TCLUG Mailing List
Subject: Re: [tclug-list] ata "failed command: WRITE DMA", "ATA bus error" messages

 

I don't have a silver-bullet solution/root-cause for you, but to me this looks like an issue with your newer kernel and drivers running against your older chipset and drives.  It seems the kernel is not wanting to use Direct Memory Access when talking to ATA (IDE) drives.  

Here is a link to a thread that has a suggestion for configuring the kernel to use a legacy mode when interacting with them, if I read it right.  And user Kane indicated specifying an irqpoll parameter also did the trick for him.   

http://bbs.archlinux.org/viewtopic.php?pid=332907

-Rob



On Fri, May 28, 2010 at 8:19 AM, Jeff Jensen <jjensen at apache.org> wrote:

Searching through the last boot log (trying to determine what is ata2), I
see:

For the old 40G:
ata2.00: ATAPI: CRD-8400B, 1.04, max UDMA/33

ata2.00: configured for PIO4

ata2.00: device is on DMA blacklist, disabling DMA

ata2.01: 78165360 sectors, multi 16: LBA
ata2.01: ATA-5: WDC WD400BB-00CLB0, 05.04E05, max UDMA/100

ata2.01: configured for UDMA/33

For the new 1TB:
ata6.00: ATA-8: WDC WD1001FALS-00J7B1, 05.00K05, max UDMA/133
ata6.00: configured for UDMA/133
ata6: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
ata6: SATA max UDMA/133 mmio m1048576 at 0xf4100000 port 0xf4128000 irq 5


It looks like ata1 & ata2 are the IDEs and ata3-ata6 are the new SATA card
(?).  So perhaps these messages are potential drive failure ones?



-----Original Message-----
From: Jeff Jensen [mailto:jjensen at apache.org]
Sent: Friday, May 28, 2010 8:02 AM
To: 'tclug-list at mn-linux.org'
Subject: ata "failed command: WRITE DMA", "ATA bus error" messages

To my old backup server (running BackupPC), I recently added a PCI SATA card
and 1TB drive, and installed Fedora 13 (and removed 2 IDE smaller drives;
the boot drive is still an older IDE 40G; was running Fedora 11).  Messages
log regularly has this set of messages, much more frequent when the backup
is running:

May 28 06:49:38 nacho kernel: ata2.01: exception Emask 0x0 SAct 0x0 SErr 0x0
action 0x6
May 28 06:49:38 nacho kernel: ata2.01: BMDMA stat 0x44
May 28 06:49:38 nacho kernel: ata2.01: failed command: WRITE DMA
May 28 06:49:38 nacho kernel: ata2.01: cmd
ca/00:c0:4f:8c:1a/00:00:00:00:00/f0 tag 0 dma 98304 out
May 28 06:49:38 nacho kernel:         res
51/84:00:0e:8d:1a/00:00:00:00:00/f0 Emask 0x10 (ATA bus error)
May 28 06:49:38 nacho kernel: ata2.01: status: { DRDY ERR }
May 28 06:49:38 nacho kernel: ata2.01: error: { ICRC ABRT }
May 28 06:49:38 nacho kernel: ata2: soft resetting link
May 28 06:49:38 nacho kernel: ata2.00: device is on DMA blacklist, disabling
DMA
May 28 06:49:38 nacho kernel: ata2.00: configured for PIO4
May 28 06:49:38 nacho kernel: ata2.01: configured for UDMA/33
May 28 06:49:38 nacho kernel: ata2: EH complete

I've googled various words from the messages, but what I find are CD drive
related messages and "now it takes longer to boot, so change modprobe" type
things.  I think mine is related to the new drive (just a hint from the "ATA
bus error" and 'ata' all over the messages ;-), possibly harmless messages
or maybe a SATA card config or compatibility problem(?).  I'm surprised they
keep repeating - if it was just a config error, I would think it would
adjust once (maybe at boot) and then be done.

Can anyone point me to an RTFM or hints how to research/what is the cause
pretty-please?!






_______________________________________________
TCLUG Mailing List - Minneapolis/St. Paul, Minnesota
tclug-list at mn-linux.org
http://mailman.mn-linux.org/mailman/listinfo/tclug-list

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.mn-linux.org/pipermail/tclug-list/attachments/20100530/68aba04e/attachment.htm