Test-and-set

From Wikipedia, the free encyclopedia
Jump to: navigation, search

In computer science, the test-and-set instruction is an instruction used to write to a memory location and return its old value as a single atomic (i.e., non-interruptible) operation. Typically, the value 1 is written to the memory location. If multiple processes may access the same memory location, and if a process is currently performing a test-and-set, no other process may begin another test-and-set until the first process is done. CPUs may use test-and-set instructions offered by other electronic components, such as dual-port RAM; CPUs may also offer a test-and-set instruction themselves.

A lock can be built using an atomic test-and-set instruction as follows:

function Lock(boolean *lock)
{
    while (test_and_set(lock) == 1);
}

The calling process obtains the lock if the old value was 0. It spins writing 1 to the variable until this occurs.

Maurice Herlihy (1991) proved that test-and-set has a finite consensus number, in contrast to the compare-and-swap operation. The test-and-set operation can solve the wait-free consensus problem for no more than two concurrent processes.[1] However, more than two decades before Herlihy's proof, IBM had already replaced Test-and-set by Compare-and-swap, which is a more general solution to this problem.

Hardware implementation of test-and-set[edit]

DPRAM test-and-set instructions can work in many ways. Here are two variations, both of which describe a DPRAM which provides exactly 2 ports, allowing 2 separate electronic components (such as 2 CPUs) access to every memory location on the DPRAM.

Variation 1[edit]

When CPU 1 issues a test-and-set instruction, the DPRAM first makes an "internal note" of this by storing the address of the memory location in a special place. If at this point, CPU 2 happens to issue a test-and-set instruction for the same memory location, the DPRAM first checks its "internal note", recognizes the situation, and issues a BUSY interrupt, which tells CPU 2 that it must wait and retry. This is an implementation of a busy waiting or spinlock using the interrupt mechanism. Since this all happens at hardware speeds, CPU 2's wait to get out of the spin-lock is very short.

Whether or not CPU 2 was trying to access the memory location, the DPRAM performs the test given by CPU 1. If the test succeeds, the DPRAM sets the memory location to the value given by CPU 1. Then the DPRAM wipes out its "internal note" that CPU 1 was writing there. At this point, CPU 2 could issue a test-and-set, which would succeed.

Variation 2[edit]

CPU 1 issues a test-and-set instruction to write to "memory location A". The DPRAM does not immediately store the value in memory location A, but instead simultaneously moves the current value to a special register, while setting the contents of memory location A to a special "flag value". If at this point, CPU 2 issues a test-and-set to memory location A, the DPRAM detects the special flag value, and as in Variation 1, issues a BUSY interrupt.

Whether or not CPU 2 was trying to access the memory location, the DPRAM now performs CPU 1's test. If the test succeeds, the DPRAM sets memory location A to the value specified by CPU 1. If the test fails, the DPRAM copies the value back from the special register to memory location A. Either operation wipes out the special flag value. If CPU 2 now issues a test-and-set, it will succeed.

Software implementation of test-and-set[edit]

Many processors have an atomic test-and-set machine language instruction. Those that do not can still implement an atomic test-and-set using a read-modify-write instruction.

The test and set instruction when used with boolean values behaves like the following function. Crucially the entire function is executed atomically: no process can interrupt the function mid-execution and hence see a state that only exists during the execution of the function. This code only serves to help explain the behaviour of test-and-set; atomicity requires explicit hardware support and hence can't be implemented as a simple function. NOTE: In this example, 'lock' is assumed to be passed by reference (or by name) but the assignment to 'initial' creates a new value (not just copying a reference).

function TestAndSet(boolean_ref lock) {
    boolean initial = lock
    lock = true
    return initial
}

The above code segment is not atomic in the sense of the test-and-set instruction. It also differs from the descriptions of DPRAM hardware test-and-set above in that here the "set" value and the test are fixed and invariant, and the "set" part of the operation is done regardless of the outcome of the test, whereas in the description of DPRAM test-and-set, the memory is set only upon passage of the test, and the value to set and the test condition are specified by the CPU. Here, the value to set can only be 1, but if 0 and 1 are considered the only valid values for the memory location, and "value is nonzero" is the only allowed test, then this equates to the case described for DPRAM hardware (or, more specifically, the DPRAM case reduces to this under these constraints). From that viewpoint, this can correctly be called "test-and-set" in the full conventional sense of the term. The essential point to note, which this software function does embody, is the general intent and principle of test-and-set: that a value both is tested and is set in one atomic operation such that no other program thread might cause the target memory location to change after it is tested but before it is set, which would violate the logic requiring that the location will only be set when it has a certain value. (That is, critically, as opposed to merely when it very recently had that value.)

In the C programming language, the implementation would be like:

 #define LOCKED 1
 
int TestAndSet(int* lockPtr) {
    int oldValue;
 
    // Start of atomic segment
    // The following statements should be interpreted as pseudocode for
    // illustrative purposes only.
    // Traditional compilation of this code will not guarantee atomicity, the
    // use of shared memory (i.e. not-cached values), protection from compiler
    // optimization, or other required properties.
    oldValue = *lockPtr;
    *lockPtr = LOCKED;
    // End of atomic segment
 
    return oldValue;
}

The code also shows that TestAndSet is really two operations: an atomic read-modify-write and a test. Only the read-modify-write needs to be atomic. (This is true because delaying the value comparison itself by any amount of time will not change the result of the test, once the value to test has been obtained. Once the code writes the initial value, the result of the test has been determined, even if it has not been computed yet—e.g. in the C language example, by the == operator.)

Implementing mutual exclusion with test-and-set[edit]

Main article: Test and Test-and-set

One way to implement mutual exclusion uses test-and-set in the following way:

boolean lock = false
function Critical(){
    while TestAndSet(lock)
        skip // spin until lock is acquired
    critical section // only one process can be in this section at a time
    lock = false // release lock when finished with the critical section
}

In pseudo C it would be like:

volatile int lock = 0;

void Critical() {
    while (TestAndSet(&lock) == 1);
    critical section // only one process can be in this section at a time
    lock = 0 // release lock when finished with the critical section
}

Note the volatile keyword. In absence of volatile, the compiler and/or the CPU(s) may optimize access to lock and/or use cached values, thus rendering the above code erroneous.

Conversely, and unfortunately, the presence of volatile does not guarantee that reads and writes are committed to memory. Some compilers issue memory barriers to ensure that operations are committed to memory, but since the semantics of volatile in C/C++ is quite vague, not all compilers will do that. Consult your compiler's documentation to determine if it does.

This function can be called by multiple processes, but it is guaranteed that only one process will be in the critical section at a time.

Another way to implement Mutual exclusion, known as Test and Test-and-set, is more efficient than the above technique on multiprocessor machines. The "Test and Test-and-set" technique uses the same "test-and-set" instruction as the above technique, but has better cache coherence.

Both the above technique and "Test and Test-and-set" are examples of a spinlock: the while-loop spins waiting to acquire the lock.

Assembly language implementation of critical sections[edit]

enter_region:        ; A "jump to" tag; function entry point.
 
  tsl reg, flag      ; Test and Set Lock; flag is the
                     ; shared variable; it is copied
                     ; into the register reg and flag
                     ; then atomically set to 1.
 
  cmp reg, #0        ; Was flag zero on entry_region?
 
  jnz enter_region   ; Jump to enter_region if
                     ; reg is non-zero; i.e.,
                     ; flag was non-zero on entry.
 
  ret                ; Exit; i.e., flag was zero on
                     ; entry. If we get here, tsl
                     ; will have set it non-zero; thus,
                     ; we have claimed the resource
                     ; associated with flag.
 
leave_region:
  move flag, #0      ; store 0 in flag
  ret                ; return to caller

We see that TSL tests a shared memory location (flag): It copies it to register reg, then it sets it to 1. TSL is atomic – it can't be interrupted. That is, all of its actions can be regarded as taking place at once. By using it with a spin lock (jnz enter_region), we repeatedly test if the resource is busy; when it is released (flag becomes zero) TSL will have reset it non-zero for us, and we will have claimed the resource because the spin lock testing reg will release, since reg has the previous value of flag, which was zero.

Implementing semaphores using test-and-set[edit]

It's possible to use the test-and-set instruction to implement semaphores. In uniprocessor systems this technique isn't needed (unless using multiple processes to access the same data); to use semaphores, it is sufficient to disable interrupts before accessing a semaphore. However, in multiprocessor systems, it is undesirable, if not impossible, to disable interrupts on all processors at the same time. Even with interrupts disabled, two or more processors could be attempting to access the same semaphore's memory at the same time. In this case, the test-and-set instruction may be used.

See also[edit]

References[edit]

  1. ^ Herlihy, Maurice (January 1991). "Wait-free synchronization". ACM Trans. Program. Lang. Syst. 13 (1): 124–149. doi:10.1145/114005.102808. Retrieved 2007-05-20. 

External links[edit]