Transcript SHARC programming model

Input and output
Supervisor mode, exceptions,
traps
KAIST 전산학과
맹 승 렬
[email protected]
I/O devices
CPU
status
reg
data
reg
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
mechanism
 Usually includes some non-digital component
 Typical digital interface to CPU:
2
Example : 8251 UART
 Universal asynchronous receiver transmitter
(UART) : provides serial communication
 8251 functions are integrated into standard PC
interface chip
 Allows many communication parameters to be
programmed
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
3
Serial communication
 Characters are transmitted separately:
no
char
start
bit 0
bit 1
...
bit n-1 stop
time
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
4
Serial communication parameters





Baud (bit) rate
Number of bits per character
Parity/no parity
Even/odd parity
Length of stop bit (1, 1.5, 2 bits)
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
5
8251 CPU interface
CPU
status
(8 bit)
8251
data
(8 bit)
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
xmit/
rcv
serial
port
6
How to communicate with the registers
in the interface
 Memory-Mapped I/O
 Isolated I/O (standard I/O)
 Advantages and disadvantages
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
7
Programming I/O
 Two types of instructions can support I/O
• special-purpose I/O instructions
• memory-mapped load/store instructions
 Intel x86 provides in, out instructions
• Most other CPUs use memory-mapped I/O
• I/O instructions do not preclude memory-mapped I/O
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
8
ARM memory-mapped I/O
 Define location for device:
DEV1 EQU 0x1000
 Read/write code:
LDR
LDR
LDR
STR
r1,#DEV1 ; set up device adrs
r0,[r1] ; read DEV1
r0,#8 ; set up value to write
r0,[r1] ; write value to device
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
9
Modes of Transfer
 Suppose a peripheral intermittently receives data,
which must be serviced by the processor
• The processor can poll the peripheral regularly to see if
data has arrived – wasteful
• The peripheral can interrupt the processor when it has
data
 Data transfer under program control
• Busy-waiting, Polling
 Interrupt
 DMA transfer
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
10
Busy/wait output
 Simplest way to program device
• Use instructions to test when device is ready
current_char = mystring;
while (*current_char != ‘\0’) {
poke(OUT_CHAR,*current_char);
while (peek(OUT_STATUS) != 0);
current_char++;
}
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
11
Interrupt-driven I/O
 Busy/wait is very inefficient (Why?)
• CPU can’t do other work while testing device
• Hard to do simultaneous I/O
 Interrupts allow a device to change the flow of
control in the CPU
• Causes subroutine call to handle device
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
12
CPU
PC
IR
intreq
intack
data/address
status
reg
data
reg
mechanism
Interrupt interface
 Requires extra pins: Intreq, Intack
• If Intreq is 1, processor suspends current program, jumps
to an Interrupt Service Routine, or ISR
• Known as interrupt-driven I/O
• Essentially, “polling” of the interrupt pin is built-into the
hardware, so no extra time!
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
13
Interrupt sequence





CPU acknowledges request
Device sends vector
CPU calls ISR
Execution of ISR
CPU restores state to foreground program
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
14
Priorities and vectors
 Two mechanisms allow us to make interrupts more
specific:
• Priorities determine what interrupt gets CPU first
• Vectors determine what code is called for each type of
interrupt
 Mechanisms are orthogonal: most CPUs provide
both
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
15
Interrupts - Vectors
 What is the address (interrupt address vector) of
the ISR?
• Fixed interrupt
– Address built into microprocessor, cannot be changed
– Either ISR stored at address or a jump to actual ISR
stored if not enough bytes available
• Vectored interrupt
– Peripheral must provide the address
– Common when microprocessor has multiple
peripherals connected by a system bus
• Compromise: interrupt address table
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
16
Interrupt-driven I/O using vectored
interrupt
Time
1(a): μP is executing its main program.
3: After completing instruction at 100, μP sees Int
asserted, saves the PC’s value of 100, and asserts
Inta.
5(a): μP jumps to the address on the bus (16).
The ISR there reads data from 0x8000, modifies
the data, and writes the resulting data to 0x8001.
1(b): P1 receives input data in a
register with address 0x8000.
2: P1 asserts Int to request servicing
by the microprocessor.
4: P1 detects Inta and puts interrupt
address vector 16 on the data bus.
5(b): After being read, P1 deasserts
Int.
6: The ISR returns, thus restoring PC to
100+1=101, where μP resumes executing.
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
17
Interrupt-driven I/O using vectored
interrupt
1(a): P is executing its main program
1(b): P1 receives input data in a register
with address 0x8000.
Program memory
ISR
16: MOV R0, 0x8000
17: # modifies R0
18: MOV 0x8001, R0
19: RETI # ISR return
...
Main program
...
100: instruction
101: instruction
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
μP
Data memory
System bus
Inta
Int
PC
100
P1
P2
16
0x8000
0x8001
18
Interrupt-driven I/O using vectored
interrupt
2: P1 asserts Int to request servicing by the
microprocessor
Program memory
ISR
16: MOV R0, 0x8000
17: # modifies R0
18: MOV 0x8001, R0
19: RETI # ISR return
...
Main program
...
100: instruction
101: instruction
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
μP
Data memory
System bus
Inta
Int
PC
100
P1
1
P2
16
0x8000
0x8001
19
Interrupt-driven I/O using vectored
interrupt
3: After completing instruction at 100, μP
sees Int asserted, saves the PC’s value of
100, and asserts Inta
Program memory
ISR
16: MOV R0, 0x8000
17: # modifies R0
18: MOV 0x8001, R0
19: RETI # ISR return
...
Main program
...
100: instruction
101: instruction
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
μP
Data memory
System bus
Inta
Int
PC
100
1
P1
P2
16
0x8000
0x8001
20
Interrupt-driven I/O using vectored
interrupt
4: P1 detects Inta and puts interrupt
address vector 16 on the data bus
Program memory
ISR
16: MOV R0, 0x8000
17: # modifies R0
18: MOV 0x8001, R0
19: RETI # ISR return
...
Main program
...
100: instruction
101: instruction
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
μP
Data memory
System bus
16
Inta
Int
PC
100
P1
P2
16
0x8000
0x8001
21
Interrupt-driven I/O using vectored
interrupt
5(a): PC jumps to the address on the bus
(16). The ISR there reads data from
0x8000, modifies the data, and writes the
resulting data to 0x8001.
5(b): After being read, P1 deasserts Int.
Program memory
ISR
16: MOV R0, 0x8000
17: # modifies R0
18: MOV 0x8001, R0
19: RETI # ISR return
...
Main program
...
100: instruction
101: instruction
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
μP
Data memory
System bus
Inta
Int
PC
100
P1
0
P2
16
0x8000
0x8001
22
Interrupt-driven I/O using vectored
interrupt
6: The ISR returns, thus restoring the PC to
100+1=101, where the μP resumes
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
Program memory
ISR
16: MOV R0, 0x8000
17: # modifies R0
18: MOV 0x8001, R0
19: RETI # ISR return
...
Main program
...
100: instruction
101: instruction
μP
Data memory
System bus
Int
PC
100
+1
P1
P2
0x8000
0x8001
23
Interrupt address table
 Compromise between fixed and vectored interrupts
• One interrupt pin
• Table in memory holding ISR addresses (maybe 256
words)
• Peripheral doesn’t provide ISR address, but rather index
into table
– Fewer bits are sent by the peripheral
– Can move ISR location without changing peripheral
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
24
Additional interrupt issues
 Maskable vs. non-maskable interrupts
• Maskable: programmer can set bit that causes processor to
ignore interrupt
– Important when in the middle of time-critical code
• Non-maskable: a separate interrupt pin that can’t be masked
– Typically reserved for drastic situations, like power failure
requiring immediate backup of data to non-volatile memory
 Jump to ISR
• Some microprocessors treat jump same as call of any subroutine
– Complete state saved (PC, registers) – may take hundreds of
cycles
• Others only save partial state, like PC only
– Thus, ISR must not modify registers, or else must save them
first
– Assembly-language programmer must be aware of which
registers stored
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
25
Arbitration: Priority arbiter
 Consider the situation where multiple peripherals request service
from single resource (e.g., microprocessor, DMA controller)
simultaneously - which gets serviced first?
 Priority arbiter
• Single-purpose processor
• Peripherals make requests to arbiter, arbiter makes requests to
resource
• Arbiter connected to system bus for configuration only
Microprocessor
System bus
Inta
Int
5
3
Priority
arbiter
7
Peripheral1
Ireq1
Iack1 6
Ireq2
2
Peripheral2
2
Iack2
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
26
Arbitration using a priority arbiter
Microprocessor
System bus
Inta
Int
5
3
Priority
arbiter
7
Peripheral1
Ireq1
Iack1 6
Ireq2
2
Peripheral2
2
Iack2
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
1. Microprocessor is executing its program.
2. Peripheral1 needs servicing so asserts Ireq1. Peripheral2 also needs servicing so asserts Ireq2.
3. Priority arbiter sees at least one Ireq input asserted, so asserts Int.
4. Microprocessor stops executing its program and stores its state.
5. Microprocessor asserts Inta.
6. Priority arbiter asserts Iack1 to acknowledge Peripheral1.
7. Peripheral1 puts its interrupt address vector on the system bus
8. Microprocessor jumps to the address of ISR read from data bus, ISR executes and returns
(and completes handshake with arbiter).
9. Microprocessor resumes executing its program.
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
27
Arbitration: Priority arbiter
 Types of priority
– Fixed priority
» each peripheral has unique rank
» highest rank chosen first with simultaneous
requests
» preferred when clear difference in rank between
peripherals
– Rotating priority (round-robin)
» priority changed based on history of servicing
» better distribution of servicing especially among
peripherals with similar priority demands
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
28
Arbitration: Daisy-chain arbitration
 Arbitration done by peripherals
• Built into peripheral or external logic added
– req input and ack output added to each peripheral
 Peripherals connected to each other in daisy-chain manner
• One peripheral connected to resource, all others connected
“upstream”
• Peripheral’s req flows “downstream” to resource, resource’s ack
flows “upstream” to requesting peripheral
• Closest peripheral has highest priority
P
System bus
Inta
Int
Peripheral1
Peripheral2
Ack_in Ack_out
Req_out Req_in
Ack_in Ack_out
Req_out Req_in
0
Daisy-chain aware peripherals
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
29
Arbitration: Daisy-chain arbitration
 Pros/cons
• Easy to add/remove peripheral - no system redesign
needed
• Does not support rotating priority
• One broken peripheral can cause loss of access to other
peripherals
Microprocessor
P
System bus
System bus
Inta
Int
Priority
arbiter
Peripheral
1
Peripheral
2
Ireq1
Iack1
Inta
Int
Peripheral1
Peripheral2
Ack_in Ack_out
Req_out Req_in
Ack_in Ack_out
Req_out Req_in
0
Ireq2
Iack2
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
Daisy-chain aware peripherals
30
Intel 8259 programmable priority
controller
D[7..0]
A[0..0]
RD
WR
INT
INTA
CAS[2..0]
SP/EN
Intel 8259
IR0
IR1
IR2
IR3
IR4
IR5
IR6
IR7
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
Signal
D[7..0]
Description
These wires are connected to the system bus and are used by the microprocessor to
write or read the internal registers of the 8259.
A[0..0]
This pin actis in cunjunction with WR/RD signals. It is used by the 8259 to decipher
various command words the microprocessor writes and status the microprocessor
wishes to read.
WR
When this write signal is asserted, the 8259 accepts the command on the data line, i.e.,
the microprocessor writes to the 8259 by placing a command on the data lines and
asserting this signal.
RD
When this read signal is asserted, the 8259 provides on the data lines its status, i.e., the
microprocessor reads the status of the 8259 by asserting this signal and reading the data
lines.
INT
This signal is asserted whenever a valid interrupt request is received by the 8259, i.e., it
is used to interrupt the microprocessor.
INTA
This signal, is used to enable 8259 interrupt-vector data onto the data bus by a sequence
of interrupt acknowledge pulses issued by the microprocessor.
IR
0,1,2,3,4,5,6,7
An interrupt request is executed by a peripheral device when one of these signals is
asserted.
CAS[2..0]
These are cascade signals to enable multiple 8259 chips to be chained together.
SP/EN
This function is used in conjunction with the CAS signals for cascading purposes.
31
Sources of interrupt overhead





ISR execution time
Interrupt mechanism overhead
Register save/restore
Pipeline-related penalties
Cache-related penalties
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
32
ARM interrupts
 ARM7 supports two types of interrupts:
• Fast interrupt requests (FIQs)
• Interrupt requests (IRQs)
 Exception table starts at location 0
• 0xFFFF0000 if high vector address is defined
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
33
Exception processing modes
Exception type
Mode
Normal address
High vector
address
Reset
Supervisor
0x00000000
0xFFFF0000
Undefined
instructions
Undefined
0x00000004
0xFFFF0004
SWI
Supervisor
0x00000008
0xFFFF0008
Prefetch Abort
Abort
0x0000000C
0xFFFF000C
Data Abort
Abort
0x00000010
0xFFFF0010
IRQ
IRQ
0x00000018
0xFFFF0018
FIQ
FIQ
0x0000001C
0xFFFF001C
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
34
Priorities
priority
exception
highest 1
Reset
2
Data Abort
3
FIQ
4
IRQ
5
Prefetch Abort
lowest 6
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
Undefined, SWI
35
ARM interrupt procedure
 CPU actions:
• Save PC. Copy CPSR to SPSR
• Force bits in CPSR to record interrupt
• Force PC to vector
 Handler responsibilities:
• Restore proper PC
• Restore CPSR from SPSR
• Clear interrupt disable flags
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
36
Interrupt request exception
 IRQ is generated externally by asserting the IRQ
input on the processor
 Lower priority than FIQ
7
N Z C V Q


6
M M M MM
I F T
4 3 2 1 0
I bit : disable IRQ when it is set
F bit: disable FIQ when it is set
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
37
IRQ
R14_irq = address of next instruction to be executed + 4
SPSR_irq = CPSR
CPSR[4:0] = 0b10010
/* enter IRQ mode */
CPSR[5] = 0
/* execute in ARM state */
/*CPSR[6] is unchanged */
CPSR[7] = 1
/* disable normal interrupt */
If high vectors configured then PC = 0xFFFF0018
else PC = 0x00000018
To return
SUBS PC, R14, #4 /* PC from R14_irq, CPSR from SPSR_irq */
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
38
FIQ
R14_fiq = address of next instruction to be executed + 4
SPSR_fiq = CPSR
CPSR[4:0] = 0b10001
/* enter FIQ mode */
CPSR[5] = 0
/* execute in ARM state */
CPSR[6] = 1
/* disable fast interrupt */
CPSR[7] = 1
/* disable normal interrupt */
If high vectors configured then PC = 0xFFFF001C
else PC = 0x0000001C
To return
SUBS PC, R14, #4 /* PC from R14_fiq, CPSR from SPSR_fiq*/
* The FIQ handler can be placed directly at address 0x0000001C or 0xFFFF001C,
without requiring a branch instruction from the vector.
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
39
ARM interrupt latency
 Worst-case latency to respond to interrupt is 27
cycles:
•
•
•
•
Two cycles to synchronize external request
Up to 20 cycles to complete current instruction
Three cycles for data abort
Two cycles to enter interrupt handling state
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
40
Interrupt
MPC 850 interrupt
Supervisor mode, exceptions, traps
Co-processors
KAIST 전산학과
맹 승 렬
[email protected]
Exceptions
back
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
42
Exceptions (cont’d)
back
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
43
Exception Handling
 Vector Table Starting Address
 Registers: SRR0, SRR1
• SRR0: return address
• SRR1: Machine State (from MSR)
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
44
External Interrupt Exception
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
45
Priority of Interrupt Sources
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
46
SIU interrupt Processing
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
47
Maskable, Nonmaskable Interrupts
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
48
Programming the SIU interrupt
controller




SIU interrupt pending register: SIPEND
SIU interrupt mask register: SIMASK
SIU interrupt edge/level register: SIEL
SIU interrupt vector register: SIVEC
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
49
Internal Memory Map Register (IMMR)

ISB : internal space base; all 16 bits can be programmed
• MPC 850 memory resource: a contiguous blocks of 16 KB
• Can be mapped on 64-KB resolution through IMMR
 IMMR: MPC850-specific SPR #638
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
50
Memory Map
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
51
SIMASK
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
52
SIEL
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
53
SIPEND
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
54
SIVEC
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
55
Interrupt table handling example
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
56
Supervisor mode
 May want to provide protective barriers between
programs
• Avoid memory corruption
 Need supervisor mode to manage the various
programs
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
57
ARM supervisor mode
 Use SWI instruction to enter supervisor mode,
similar to subroutine:
SWI CODE_1
• Sets PC to 0x08
• Argument to SWI is passed to supervisor mode code
• Saves CPSR in SPSR
 RESET
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
58
Exception
 Exception: internally detected error
 Exceptions are synchronous with instructions but
unpredictable
 Build exception mechanism on top of interrupt
mechanism
 Exceptions are usually prioritized and vectorized
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
59
ARM Exceptions
 Undefined
• Can be used for SW emulation of a coprocessor
 Prefetch Abort
• Instruction fetch memory abort
– ex) translation faults, permission faults, parity error
 Data Abort
• Data access memory abort
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
60
Trap
 Trap (software interrupt): an exception generated
by an instruction
• Call supervisor mode
 ARM uses SWI instruction for traps
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
61
MPC850 Exceptions
 PowerPC-defined exceptions
 Implementation-specific exceptions
2004
전문대교수연수
([email protected])
2004년CS310
Microprocessors
& Lab
62