Digital Equipment AlphaServer 1200, Руководство по обслуживанию

Страница: 110 / 278

4-20 Service Manual PRELIMINARY
4.3.3 MCHK 670 Read Dirty CPU-Detected Failure
The error log in Example 4–1 shows the following:
ž
CPU0 logged the error in a system with two CPUs.
Ÿ
The External Interface Status Register records an uncorrectable ECC error
from the system (bit <30> set).
Both IOD CAP Error Registers logged an error.
?¡
The MC Error Info Registers 0 and 1 have captured the error information.
?¢
The commander at the time of the error was CPU0 (known from MC_ERR1)
?£
The command on the bus at the time was a read memory command.
?¤
The address read was a memory address, not an I/O address.
?¥
The data associated with the read was dirty.
From this information you know CPU0 requested data that was dirty; therefore,
memory did not provide it, nor did an I/O device. Only another CPU could have
provided the data from its cache. There is only one other CPU in this system, and it
is faulty. Had there been more than two CPUs you could not have identified the
error to a particular CPU. See Section 4.4 for a procedure designed to help with
IOD-detected errors.
NOTE: The error log example has been edited to decrease its size; registers of
interest are in bold type. The “MC” bus is the system bus.
Refer to Table 4-9 for information on decoding commands, and refer to Table 4-10
for information on node IDs.