Fujitsu MAX3073NC - SCSI,80, Dell DC961,Hard Drive Scsi Logical Interface Specifications page 192

Parameter Data Formats
Table 4.1
EER PER DTE
0 1 1
0 1 1
1 0 0
1 0 0
1 0 1
1 0 1
4-8
Combinations of error recovery flags (2 of 3)
DCR
0 1. Rereading is tried repeatedly up to the number of times specified in the "Read
Retry Count", "Write Retry Count" or "Verify Retry Count" parameter. If
possible, data correction is executed afterward according to the ECC.
2. When error recovery has succeeded, or even when recovery is impossible,
execution of the command is terminated with a CHECK CONDITION status at
the point when error recovery processing is completed, and the sense data
indicate the data block address where that error occurred.
3. Transfer to the INIT of data in blocks with recovered errors is done, but transfer
to the INIT of data blocks in which the errors were unrecoverable is done in
accordance with the specification in the TB bit (read commands).
1 1. Rereading is tried repeatedly up to the number of times specified in the "Read
Retry Count", "Write Retry Count" or "Verify Retry Count" parameter. Data
correction is not executed according to the ECC.
2. When error recovery has succeeded, or even when recovery is impossible,
execution of the command is terminated with a CHECK CONDITION status at
the point when error recovery processing is completed, and the sense data
indicate the data block address where that error occurred.
3. Transfer to the INIT of data in blocks with recovered errors is done, but transfer
to the INIT of data blocks in which the errors were unrecoverable is done in
accordance with the specification in the TB bit (read commands).
0 1. When a correctable data check is detected, immediately, correction is executed
according to the ECC. For data checks which cannot be corrected, rereading is
tried repeatedly up to the number of times specified in the "Read Retry Count",
"Write Retry Count" or "Verify Retry Count" parameter. However, if a
correctable error is detected during reading, data correction is executed at that
point according to the ECC.
2. When error recovery has succeeded, processing of the command is continued.
3. The contents of recovered errors are not reported.
4. If unrecoverable errors have been detected, execution of that command is
terminated at that point.
5. Transfer of data in blocks which include unrecoverable errors to the INIT is
done in accordance with the specification in the TB bit (read commands).
1 (Setting prohibited) (See *1 at the end of this table.)
0 (Setting prohibited) (See *1 at the end of this table.)
1 (Setting prohibited) (See *1 at the end of this table.)
Error Recovery Procedure
C141-C010