found by comparing the priority numbers,
as found inthe figure, from left to right until a mismatch is found. If the first inequality is between numeric
characters, either the two conditions
are mutually exclusive or, if both can
occur, the condition with the smaller
number is indicated. If the first
inequality is between alphabetic charac
ters, then the two conditions are not
exclusive, and it is unpredictable which
is indicated when both occur.
To understand the use ofthe table, consider an example involving the
instruction ADD DECIMAL, which is a
six-byte instruction. Assume that the
first four bytes of the instruction can
be accessed but that the instruction
crosses a boundary so thatan addressing
exception exists for the last two bytes.
Additionally, assume that the first
operand addressed by the instruction
contains invalid decimal digits and is
ina location that can be fetched from,
but not stored into, because of key
controlled protection.The three
exceptions which could result from
attempted execution of the ADDDECIMAL are: Priority
Number Exception
7.B Access exceptions for third
instruction halfword.
8.B Access exceptions (operand
1)• 8.D Data exception. Since the first inequality (7*8) is
between numeric characters, the address
ing exception would be indicated. If,
however, the entire ADD DECIMAL instruc
tion can be fetched, and onlythe second
two exceptions listed above exist, thenthe inequality (B*D) is between alpha
bet i c characters, and it. is
unpredictable whether the protection
exception or the data exception would be
indicated.Chapter 6. Interruptions 6-31
as found in
characters, either the two conditions
are mutually exclusive or, if both can
occur, the condition with the smaller
number is indicated. If the first
inequality is between alphabetic charac
ters, then the two conditions are not
exclusive, and it is unpredictable which
is indicated when both occur.
To understand the use of
instruction ADD DECIMAL, which is a
six-byte instruction. Assume that the
first four bytes of the instruction can
be accessed but that the instruction
crosses a boundary so that
exception exists for the last two bytes.
Additionally, assume that the first
operand addressed by the instruction
contains invalid decimal digits and is
in
but not stored into, because of key
controlled protection.
exceptions which could result from
attempted execution of the ADD
Number Exception
7.B Access exceptions for third
instruction halfword.
8.B Access exceptions (operand
1)
between numeric characters, the address
ing exception would be indicated. If,
however, the entire ADD DECIMAL instruc
tion can be fetched, and only
two exceptions listed above exist, then
bet i c characters, and it. is
unpredictable whether the protection
exception or the data exception would be
indicated.