Skip to contentSkip to conversation

scary messages: HSM violation during boot of 2.6.18/amd64

DISCUSSION Tomasz Chmielewski 4 participants 7 messages Nov 10, 2006 – Nov 11, 2006

lore
Share & tools
lore
#0 lore raw
Quoted 14 lines

Hi, I just upgraded my workstation to 2.6.18.2. It has four SATA drives in a RAID10, connected to the system in pairs on Promise and Via on-board controllers.

Now on every boot, I see several messages like this for the drives connected to the VIA controller (VT6420):

ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen ata2.00: tag 0 cmd 0xb0 Emask 0x2 stat 0x50 err 0x0 (HSM violation) ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen ata1.00: tag 0 cmd 0xb0 Emask 0x2 stat 0x50 err 0x0 (HSM violation)

What do these mean?

Quoted 8 lines

Also, for the first of the two drives on the Promise/FastTrak PDC20378 controller, I see messages like this:

ata3: translated ATA stat/err 0x50/00 to SCSI SK/ASC/ASCQ 0xb/00/00

What about those?

I saw none of that under 2.6.17.x. Should I be worried?

I saw similar when using smartctl / smartd with wrong options (without -d ata; in short, smartd tried to talk "IDE language" to SATA device...).

Signature · 2 lines
Tomasz Chmielewski
http://wpkg.org
#1 lore raw

also sprach Tomasz Chmielewski <[email protected]> [2006.11.10.1612 +0100]:

Quoted 2 lines

I saw similar when using smartctl / smartd with wrong options (without -d ata; in short, smartd tried to talk "IDE language" to SATA device...).

I am using smartd, and you are right: it triggers those messages after being started:

[smartd start]
kernel: ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen
kernel: ata2.00: tag 0 cmd 0xb0 Emask 0x2 stat 0x50 err 0x0 (HSM violation)
kernel: ata2: SATA link up 1.5 Gbps (SStatus 113 SControl 300)
kernel: ata2: soft resetting port
kernel: ata2.00: configured for UDMA/133
kernel: ata2: EH complete
[...]
kernel: ata3: no sense translation for status: 0x50
kernel: ata3: translated ATA stat/err 0x50/00 to SCSI SK/ASC/ASCQ 0xb/00/00
kernel: ata3: status=0x50 { DriveReady SeekComplete }

Signature · 7 lines
martin;              (greetings from the heart of the sun.)
  \____ echo mailto: !#^."<*>"|tr "<*> mailto:" net@madduck
 
spamtraps: [email protected]
 
"it is only the modern that ever becomes old-fashioned." 
                                                        -- oscar wilde
#2 lore raw
In reply to Tomasz Chmielewski
Quoted 5 lines · Tomasz Chmielewski
Quoted 2 lines

ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen ata2.00: tag 0 cmd 0xb0 Emask 0x2 stat 0x50 err 0x0 (HSM violation)

I saw similar when using smartctl / smartd with wrong options (without -d ata; in short, smartd tried to talk "IDE language" to SATA device...).

(Foo on me for not chasing this earlier - I've been seeing it ever since I started using the libata stuff several months ago - Alan Cox got things fixed so I had the ata_piix driver at UDMA/100, and I never followed up on this part)

[   21.734165] ata1.00: ATA-6, max UDMA/100, 117210240 sectors: LBA
[   21.734171] ata1.00: ata1: dev 0 multi count 8
[   21.884789] ata1.01: ATAPI, max UDMA/33
[   21.889464] ata1.00: configured for UDMA/100
[   22.041527] ata1.01: configured for UDMA/33
[   22.041536] scsi1 : ata_piix 
[   22.192756] scsi 0:0:0:0: Direct-Access     ATA      FUJITSU MHV2060A 0000 PQ: 0 ANSI:5

(ata1.01 is a CD)

when smartd starts up:

[   92.567000] ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen
[   92.568000] ata1.00: tag 0 cmd 0xb0 Emask 0x2 stat 0x50 err 0x0 (HSM violation)
[   92.569000] ata1: soft resetting port
[   92.879000] ata1.00: configured for UDMA/100
[   93.031000] ata1.01: configured for UDMA/33
[   93.031000] ata1: EH complete
(set of 6 messages repeated 5 more times)
[   95.346000] SCSI device sda: 117210240 512-byte hdwr sectors (60012 MB)
[   95.348000] sda: Write Protect is off
[   95.349000] sda: Mode Sense: 00 3a 00 00
[   95.349000] SCSI device sda: drive cache: write back
[   95.350000] SCSI device sda: 117210240 512-byte hdwr sectors (60012 MB)
[   95.351000] sda: Write Protect is off
[   95.352000] sda: Mode Sense: 00 3a 00 00
[   95.353000] SCSI device sda: drive cache: write back

and then the entire set repeats again (6 sets of 6 msgs, then final 8).

And I *am* passing '-d ata' - /etc/smartd.conf contains:

/dev/sda -a -d ata -o on -S on -m root -l error -l selftest -s (S/../.././(00|06|12|18)|L/../.././03|O/../.././.[02468])

(Testing with removing '-d ata' results in smartd saying it can't talk to the scsi device at /dev/sda).

Any ideas/suggestions?

#3 lore raw

[email protected] wrote:

(...)

Quoted 8 lines

And I *am* passing '-d ata' - /etc/smartd.conf contains:

/dev/sda -a -d ata -o on -S on -m root -l error -l selftest -s (S/../.././(00|06|12|18)|L/../.././03|O/../.././.[02468])

(Testing with removing '-d ata' results in smartd saying it can't talk to the scsi device at /dev/sda).

Any ideas/suggestions?

You use old smartmontools :)

-o on / -S on is not supported for sata, unless you use a CVS version of smartmontools.

For more info, check smartmontools-support mailing list.

Signature · 3 lines
Tomasz Chmielewski

http://wpkg.org
#4 lore raw
Quoted 4 lines · Tomasz Chmielewski

You use old smartmontools :)

-o on / -S on is not supported for sata, unless you use a CVS version of smartmontools.

OK, thanks - the Fedora RPM from several days ago is still 5.36 based, I just pulled the CVS version and it starts up much more nicely. Only quirk:

Nov 10 18:04:32 turing-police smartd[18988]: Device: /dev/sda, opened 
Nov 10 18:04:32 turing-police smartd[18988]: Device /dev/sda: SATA disks accessed via libata are supported by Linux kernel versions 2.6.15-rc1 and above. Try adding '-d ata' or '-d sat' to the smartd.conf config file line. 

Well, yeah. I'm on 2.6.19-rc5-mm1 and the config file *has* '-d ata' already. ;) I'd file a bug report against that, except it's time for me to leave the office and forage for some dinner. :)

#5 lore raw
Quoted 7 lines · [email protected] wrote:
Quoted 4 lines · Tomasz Chmielewski

You use old smartmontools :)

-o on / -S on is not supported for sata, unless you use a CVS version of smartmontools.

Quoted 8 lines

OK, thanks - the Fedora RPM from several days ago is still 5.36 based, I just pulled the CVS version and it starts up much more nicely. Only quirk:

Nov 10 18:04:32 turing-police smartd[18988]: Device: /dev/sda, opened Nov 10 18:04:32 turing-police smartd[18988]: Device /dev/sda: SATA disks accessed via libata are supported by Linux kernel versions 2.6.15-rc1 and above. Try adding '-d ata' or '-d sat' to the smartd.conf config file line. 

Well, yeah. I'm on 2.6.19-rc5-mm1 and the config file *has* '-d ata' already. ;) I'd file a bug report against that, except it's time for me to leave the office and forage for some dinner. :)

For detailed info,

http://thread.gmane.org/gmane.linux.ide/13222/focus=13235

Signature · 1 line
tejun
#6 lore raw

also sprach Tomasz Chmielewski <[email protected]> [2006.11.10.2354 +0100]:

Quoted 1 line

You use old smartmontools :)

I just tried a CVS snapshot of smartmontools from today because you

Quoted 2 lines · said:

-o on / -S on is not supported for sata, unless you use a CVS version of smartmontools.

However, the messages are still there (and neither -o nor -S worked, smartd said it failed for both).

Quoted 1 line

For more info, check smartmontools-support mailing list.

I'll turn there. Thanks.

Signature · 7 lines
martin;              (greetings from the heart of the sun.)
  \____ echo mailto: !#^."<*>"|tr "<*> mailto:" net@madduck
 
spamtraps: [email protected]
 
warning at the gates of bill:
abandon hope, all ye who press <enter> here...