Bug 11192

Summary: Since Core 104, 'hdparm' generates kernel errors.
Product: IPFire Reporter: Matthias Fischer <matthias.fischer>
Component: hdparmAssignee: Arne.F <arne.fitzenreiter>
Status: CLOSED CANTFIX QA Contact:
Severity: Balancing    
Priority: - Unknown - CC: mail, matthias.fischer, michael.tremer
Version: 2   
Hardware: all   
OS: Linux   

Description Matthias Fischer 2016-09-24 00:37:59 UTC 
Hi,

directly after upgrading to Core 104, every five minutes '/var/log/messages' gets filled with errors like this:

...
Sep 23 21:30:00 ipfire kernel: sd 0:0:0:0: [sda] Unhandled error code
Sep 23 21:30:00 ipfire kernel: sd 0:0:0:0: [sda]  
Sep 23 21:30:00 ipfire kernel: Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK
Sep 23 21:30:00 ipfire kernel: sd 0:0:0:0: [sda] CDB: 
Sep 23 21:30:00 ipfire kernel: ATA command pass through(16): 85 06 20 00 00 00 00 00 00 00 00 00 00 40 e5 00
Sep 23 21:30:00 ipfire kernel: sd 0:0:0:0: [sda] Unhandled error code
Sep 23 21:30:00 ipfire kernel: sd 0:0:0:0: [sda]  
Sep 23 21:30:00 ipfire kernel: Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK
Sep 23 21:30:00 ipfire kernel: sd 0:0:0:0: [sda] CDB: 
Sep 23 21:30:00 ipfire kernel: ATA command pass through(16): 85 06 20 00 00 00 00 00 00 00 00 00 00 40 98 00
...

First reported in IPFire-Forum:
https://forum.ipfire.org/viewtopic.php?f=17&t=17195&p=100914#p100635

Workaround:
Comment the 'hdparm'-call in '/usr/local/bin/makegraphs':

...
#   my $status = `hdparm -C /dev/$disk | tail -1 | cut -d: -f2`;
#   chomp $status;
...

Besides, starting 'hdparm' on console gives me nothing useful at all:

...
root@ipfire: / # hdparm -C /dev/sda

/dev/sda:
 drive state is:  unknown
...

Older 'hdparm'-version triggered this error too, it must have something to do with the core update.

Best,
Matthias
Comment 1 Michael Tremer 2016-09-24 13:34:45 UTC 
Did you try updating hdparm?
Comment 2 Matthias Fischer 2016-09-24 13:54:54 UTC 
Hi,

I used 9.48, that is the last version ("modified 2015-06-17"). ;-)

In fact, I tried *down*grading to 9.42 - same results.

Best,
Matthias
Comment 3 Matthias Fischer 2016-09-24 20:17:04 UTC 
I forgot:

I had to deactivate the cron job for '/usr/local/bin/hddshutdown' in 'root.orig'.

It contains a few 'hdparm' calls.

Best,
Matthias
Comment 4 Matthias Fischer 2016-10-23 18:24:30 UTC 
Hi,

upgraded to 'hdparm 9.50', no chane, same problems.

Best,
Matthias
Comment 5 Matthias Fischer 2016-10-23 18:25:33 UTC 
Hi,

upgraded to 'hdparm 9.50', no change, same problems.

Best,
Matthias
Comment 6 Matthias Fischer 2017-01-15 09:38:38 UTC 
Hi,

*Bump* and Update: ;-)


On "IPFire Duo Box" and "Core 108" I get:

...
root@ipfire: /var # hdparm -C /dev/sda

/dev/sda:
 drive state is:  active/idle
...

So it could be a problem with certain Controllers/Drives. The other one is pure SCSI (Adaptec/Hitachi 40GB Drive).

Best,
Matthias
Comment 7 Arne.F 2018-08-30 08:59:10 UTC 
is this still present with kernel 4.14 ?