Technical article 112243

What does the warning message "ABRT set, non-existent device or HW failure" mean?

OpenServer 5.0.6

Written · Last revised

Problem

The following warning message is displayed in the /var/adm/messages and /var/adm/syslog files:

"WARNING: WD: ctlr 0 dev 1: ABRT set, non-existent device or hardware failure"

This message is not shown during installation.  There are no
problems installing from the ATAPI CD-ROM.  This message is not
displayed on the hardware screen during system boot-up.

Solution

The "wd" driver for the IDE devices in SCO OpenServer 5.0.6 has
been rewritten to include 'udma'(Ultra DMA) support.  The way
that this driver discovers what devices are attached to each of
the IDE ports on either controller is as follows:

The udma wd driver first searches for ATA devices and then for ATAPI devices. It does this by sending an ATA_IDENTIFY command(0xEC) to both channels of the IDE controller to see if there are any responses back from any ATA devices.

It will then look for ATAPI devices by sending an ATAPI_IDENTIFY command(0xA1) to both channels of the IDE controller and monitor the responses.

If only an ATAPI CD-ROM device (which does not 'shadow' commands sent to the slave) is attached to the master port of the controller, then the status reported back to the operating system from an ATA_IDENTIFY command is typically either:

               Status     Error
Ctrl 0, Dev0, 0x00000051 0x00000004
Ctrl 0, Dev1, 0x00000051 0x00000004

This shows the Error register has the ABRT bit set, while the Status Register has the ERR, DRDY and DSC bits set.

OR:
               Status     Error
Ctrl 0, Dev0, 0x00000051 0x00000004
Ctrl 0, Dev1, 0x00000001 0x00000004

This shows the Error register has the ABRT bit set, while the Status Register has the ERR bit set.

With these settings of the Status and Error registers, the 'udma' driver will classify these devices as not being ATA devices and will now attempt to identify any devices with the ATAPI_IDENTIFY(0xA1) command.

This is successful and the ATAPI CD-ROM device can be accessed without any problems.

However, when a CD-ROM device such as the LG CRD 8400 is attached as the master device, it not only responds to the ATA_IDENTIFY command sent to the master, but will respond to commands sent to the non-existent slave devices. This is the effect of 'shadowing'.

                Status     Error
Ctrl 0, Dev0, 0x00000051 0x00000004
Ctrl 0, Dev1, 0x00000050 0x00000004

The 'udma' driver ignores the device on the master port of the IDE chip as it is an ATAPI rather than an ATA device. It does report a warning message regarding a possible issue with the device on the slave port, as the ABRT bit is set in the Error register, without the ERR bit being set in the Status register. This is reported in /var/adm/messages and /var/adm/syslog.

"WARNING: WD: ctlr 0 dev 1: ABRT set, non-existent device or hardware failure"

The 'udma' driver will now attempt to identify any devices attached to the IDE controller with the ATAPI_IDENTIFY(0xA1) command.

This is successful and the ATAPI CD-ROM device can be accessed without any problems.

In summary, if the shadowing logic of the master device CD-ROM does not set the ERR bit in the Status register to slave device queries, and the ABRT bit is set in the Error register, then you will get a WARNING message.

Conversely, if the ERR bit is set in the Status register and the ABRT bit is set in the Error register, then there is no WARNING message.

Hence under the above circumstances, the WARNING message may be ignored, as the ATAPI CD-ROM (which 'shadows' slave commands) is attached to the master port of an IDE controller that reports an ABRT error without the corresponding ERR bit being set in the Status register, during the device discovery routines of system initialization.

To view the status and error registers during system boot, the following bootstring is required:

defbootstr wd.debug=udma
-----------------------------------------------------------
Further definition of the Error and Status registers follows:
-----------------------------------------------------------
Error Register
--------------

At the completion of any command, except EXECUTE DEVICE DIAGNOSTIC, the contents of the error register are valid when the ERR bit is equal to one in the Status register.

 --------------------------------------------------
| 7 |  6  | 5  |  4   |  3  |  2   |   1    |  0   |
 --------------------------------------------------
| r | UNC | MC | IDNF | MCR | ABRT | TKNONF | AMNF |
 --------------------------------------------------

Bit 7 is reserved UNC (Uncorrectable Data Error) MC (Media Changed) is reserved for use by removable media devices and indicates that new media is available to the operating system. IDNF(ID Not Found) indicates the requested sector's ID field could not be found. ABRT (Aborted Command) indicates the requested command has been aborted because the command code is invalid or another device error has occurred. MCR (Media Change Requested) is reserved for use by removable media devices and indicates that a request for media removal has been detected by the device. TKONF (Track 0 Not Found) indicates track 0 has not been found during a RECALIBRATE command. AMNF (Address Mark Not Found) indicates the data address mark has not been found after finding the correct ID field.

Status Register
---------------
------------------------------------------------
|  7  |  6   | 5  |  4  |  3  |  2   |  1  |  0  |
 ------------------------------------------------
| BSY | DRDY | DF | DSC | DRQ | CORR | IDX | ERR |
 ------------------------------------------------

BSY (Busy) is set whenever the device has control of the command Block registers. DRDY (Device Ready) is set to indicate that the device is capable of accepting all command codes. DF (Device Fault) indicates a device fault error has been detected. DSC (Device Seek Complete) indicates that the device heads are settled over a track. DRQ (Data Request) indicates that the device is ready to transfer a word or byte of data. CORR (Corrected Data) is used to indicate a correctable data error. IDX (Index) is vendor specific. ERR (Error) indicates that an error occurred during execution of the previous command. The bits in the error register have additional information regarding the cause of the error.

From the Xinuos technical article database, record 112243. Reproduced with permission. The original is at www.sco.com/ta/112243, which is reachable but blocked from search engines.

Next step

Need this fixed rather than explained?

Tell us the release, what the system runs and what it is doing. We will reply by email with what we would do next.

  • You deal with the people who do the work
  • No obligation, and no upgrade required before we will help
  • We tell you plainly if something is not worth doing

Send an enquiry

We reply by email. If a server is down, say so in the first line. You can also write to help@scoexpert.com.