Unexpected Bus Free Phase Error Condition; Message Out Phase Parity Error; Message In Phase Parity Error (Message Parity Error); Command Phase Parity Error - IBM DCAS-32160 - Ultrastar 2.1 GB Hard Drive Specifications

9.4.1 Unexpected BUS FREE Phase Error Condition

There are several error conditions that will cause the file to immediately change to the BUS F R E E phase,
regardless of the state of the ATN signal. The file will not attempt to reconnect to the initiator to complete
the operation that was in progress when the error condition was detected. The initiator should interpret this
as a catastrophic error condition.
If the L U N was identified by the file prior to the error condition, then the file will abort the active command
for this initiator/LUN and generate sense data for this initiator/LUN to describe the cause of the catastrophic
error. The initiator may retrieve this sense data by issuing a REQUEST SENSE command to this LUN.
Note however, that the REQUEST SENSE command may fail if the catastrophic error condition persists.
If the L U N was not identified by the file prior to the error condition, then the file will not affect the sense
data or the operation of any currently executing command for this initiator or any other initiator.

9.4.2 MESSAGE OUT Phase Parity Error

If the file detects a parity error during the MESSAGE O U T phase, the file will retry the MESSAGE O U T
phase one time as follows:
1. Continue the REQ/ACK handshakes until the initiator drops ATN.
remaining MESSAGE O U T phase bytes received after the parity error.
2. Assert the R E Q signal prior to changing to any other phase. After detecting this condition, the initiator
must resend all of the previous message byte(s) sent during this MESSAGE O U T phase.
3. Repeat the transfer of the MESSAGE O U T phase bytes(s). If the file receives all of the message byte(s)
successfully, the file will change to:
Any other information transfer phase and transfer at least one byte.
BUS F R E E phase if the message received was ABORT or BUS DEVICE RESET.
If a second parity error is detected, the target will abort the current command with CHECK CONDI-
TION status and sense data of ABORTED C O M M A N D / SCSI PARITY E R R O R .

9.4.3 MESSAGE IN Phase Parity Error (Message Parity Error)

If the file receives a MESSAGE PARITY E R R O R message, it is considered a retrievable error. The file will
do the following one time if no previous retrievable error and if a MESSAGE IN phase has just occurred:
1. Change phase to MESSAGE IN.
2. Send the last message again.
If this is the second retrievable error, the file will terminate the current command as follows:
1. Change to the BUS F R E E phase, regardless of the state of the ATN signal.
2. Abort the active command for this initiator/LUN and set the sense data to ABORTED C O M M A N D /
SCSI PARITY E R R O R .

9.4.4 COMMAND Phase Parity Error

1. Change phase to MESSAGE IN and send a RESTORE POINTERS message.
2. If RESTORE POINTERS message is accepted, then change phase to C O M M A N D O U T and receive
the command again.
If a second parity error is detected, the file will abort the current command with CHECK CONDITION
status and sense data of ABORTED C O M M A N D / SCSI PARITY E R R O R .
The file will ignore all the
SCSI MESSAGE SYSTEM
177