Locality Optimizations - | Study Material, Lecturing Notes, Assignment, Reference, Wiki description explanation, brief detail |

Chapter: Compilers - Principles, Techniques, & Tools : Optimizing for Parallelism and Locality

Locality Optimizations

1 Temporal Locality of Computed Data 2 Array Contraction 3 Partition Interleaving 4 Putting it All Together 5 Exercises for Section 11.10

Locality Optimizations

 

1 Temporal Locality of Computed Data

2 Array Contraction

3 Partition Interleaving

4 Putting it All Together

5 Exercises for Section 11.10

 

The performance of a processor, be it a part of a multiprocessor or not, is highly sensitive to its cache behavior. Misses in the cache can take tens of clock cycles, so high cache-miss rates can lead to poor processor performance. In the context of a multiprocessor with a common memory bus, contention on the bus can further add to the penalty of poor data locality.

 

As we shall see, even if we just wish to improve the locality of uniprocessors, the affine-partitioning algorithm for parallelization is useful as a means of iden-tifying opportunities for loop transformations. In this section, we describe three techniques for improving data locality in uniprocessors and multiprocessors.

 

1.          We improve the temporal locality of computed results by trying to use the results as soon as they are generated. We do so by dividing a computation into independent partitions and executing all the dependent operations in each partition close together.

 

 

2.      Array  contraction  reduces  the  dimensions  of  an  array  and  reduces  the number of memory locations accessed. We can apply array contraction if only one location of the array is used at a given time.

 

3. Besides improving temporal locality of computed results, we also need to optimize for the spatial locality of computed results, and for both the temporal and spatial locality of read-only data. Instead of executing each partition one after the other, we interleave a number of the partitions so that reuses among partitions occur close together.

 

1. Temporal Locality of Computed Data

 

The affine-partitioning algorithm pulls all the dependent operations together; by executing these partitions serially we improve temporal locality of computed data. Let us return to the multigrid example discussed in Section 11.7.1. Ap-plying Algorithm 11.43 to parallelize the code in Fig 11.23 finds two degrees of parallelism. The code in Fig 11.24 contains two outer loops that iterate through the independent partitions serially. This transformed code has im-proved temporal locality, since computed results are used immediately in the same iteration.

 

 

Thus, even if our goal is to optimize for sequential execution, it is profitable to use parallelization to find these related operations and place them together. The algorithm we use here is similar to that of Algorithm 11.64, which finds all the granularities of parallelism starting with the outermost loop. As discussed in Section 11.9.9, the algorithm parallelizes strongly connected components in-dividually, if we cannot find synchronization-free parallelism at each level. This parallelization tends to increase communication. Thus, we combine separately parallelized strongly connected components greedily, if they share reuse.

 

 

2. Array Contraction

 

The optimization of array contraction provides another illustration of the trade-off between storage and parallelism, which was first introduced in the context of instruction-level parallelism in Section 10.2.3. Just as using more registers al-lows for more instruction-level parallelism, using more memory allows for more loop-level parallelism. As shown in the multigrid example in Section 11.7.1, expanding a temporary scalar variable into an array allows different iterations to keep different instances of the temporary variables and to execute at the same time. Conversely, when we have a sequential execution that operates on one array element at a time serially, we can contract the array, replace it with a scalar, and have each iteration use the same location.

 

In the transformed multigrid program shown in Fig. 11.24, each iteration of the inner loop produces and consumes a different element of AP, AM, T, and a row of D. If these arrays are not used outside of the code excerpt, the iterations can serially reuse the same data storage instead of putting the values in different elements and rows, respectively. Figure 11.62 shows the result of reducing the dimensionality of the arrays. This code runs faster than the original, because it reads and writes less data. Especially in the case when an array is reduced to a scalar variable, we can allocate the variable to a register and eliminate the need to access memory altogether.

 

As less storage is used, less parallelism is available. Iterations in the trans-formed code in Fig. 11.62 now share data dependences and no longer can be executed in parallel. To parallelize the code on P processors, we can expand each of the  scalar variables by a factor of P and have each processor access its own private copy.  Thus, the amount by which the storage is expanded is


directly correlated to the amount of parallelism exploited.

 

There are three reasons it is common to find opportunities for array con-traction:

 

 

1. Higher-level programming languages for scientific applications, such as Matlab and Fortran 90, support array-level operations. Each subexpres-sion of array operations produces a temporary array. Because the arrays can be large, every array operation such as a multiply or add would require reading and writing many memory locations, while requiring relatively few arithmetic operations. It is important that we reorder operations so that data is consumed as it is produced and that we contract these arrays into scalar variables.

 

 

2.  Supercomputers built in the 80's and 90's are all vector machines, so many scientific applications developed then have been optimized for such machines. Even though vectorizing compilers exist, many programmers still write their code to operate on vectors at a time. The multigrid code example of this chapter is an example of this style.

 

3.  Opportunities for contraction are also introduced by the compiler. As illustrated by variable T in the multigrid example, a compiler would ex-pand arrays to improve parallelization. We have to contract them when the space expansion is not necessary.


 

Example  1 T . 6 7 :  The array expression Z = W + X + Y translates to


can speed it up considerably. Of course at the level of C code, we would not even have to use the temporary T, but could write the assignment to Z[i] as a single statement. However, here we are trying to model the intermediate-code level at which a vector processor would deal with the operations. •

 

A l g o r i t h m 1 1 . 6 8 :  Array contraction.

 

INPUT: A program transformed by Algorithm 11.64.

 

OUTPUT: An equivalent program with reduced array dimensions.

 

METHOD: A dimension of an array can be contracted to a single element if

 

1. Each independent partition uses only one element of the array,

 

2.          The value of the element upon entry to the partition is not used by the partition, and

 

 

3.          The value of the element is not live on exit from the partition.

 

Identify the contractable dimensions — those that satisfy the three condi-tions above — and replace them with a single element. •

 

Algorithm 11.68 assumes that the program has first been transformed by Al-gorithm 11.64 to pull all the dependent operations into a partition and execute the partitions sequentially. It finds those array variables whose elements' live ranges in different iterations are disjoint. If these variables are not live after the loop, it contracts the array and has the processor operate on the same scalar location. After array contraction, it may be necessary to selectively expand arrays to accommodate for parallelism and other locality optimizations.

 

The liveness analysis required here is more complex than that described in Section 9.2.5. If the array is declared as a global variable, or if it is a parameter, interprocedural analysis is required to ensure that the value on exit is not used. Furthermore, we need to compute the liveness of individual array elements, conservatively treating the array as a scalar would be too imprecise.

 

Partition Interleaving

Different partitions in a loop often read the same data, or read and write the same cache lines. In this and the next two sections, we discuss how to optimize for locality when reuse is found across partitions.

R e u s e in Innermost Blocks

 

We adopt the simple model that data can be found in the cache if it is reused within a small number of iterations. If the innermost loop has a large or un-known bound, only reuse across iterations of the innermost loop translates into a locality benefit. Blocking creates inner loops with small known bounds, al-lowing reuse within and across entire blocks of computation to be exploited. Thus, blocking has the effect of capitalizing on more dimensions of reuse.

 

 

E x a m p l e 1 1 . 6 9 : Consider the matrix-multiply code shown in Fig. 11.5 and its blocked version in Fig. 11.7. Matrix multiplication has reuse along every dimension of its three-dimensional iteration space. In the original code, the in-nermost loop has n iterations, where n is unknown and can be large. Our simple model assumes that only the data reused across iterations in the innermost loop is found in the cache.

 

In the blocked version, the three innermost loops execute a three-dimension-al block of computation, with B iterations on each side. The block size B is chosen by the compiler to be small enough so that all the cache lines read and written within the block of computation fit into the cache. Thus reused data across iterations in the third outermost loop can be found in the cache. •

 

We refer to the innermost set of loops with small known bounds as the inner-most block. It is desirable that the innermost block include all the dimensions of the iteration space that carry reuse, if possible. Maximizing the lengths of each side of the block is not as important. For the matrix-multiply example, 3-dimensional blocking reduces the amount of data accessed for each matrix by a factor of B2. If reuse is present, it is better to accommodate higher-dimensional blocks with shorter sides than lower-dimensional blocks with longer sides.

 

We can optimize locality of the innermost fully permutable loop nest by blocking the subset of loops that share reuse. We can generalize the notion of blocking to exploit reuses found among iterations of outer parallel loops, also. Observe that blocking primarily interleaves the execution of a small number of instances of the innermost loop. In matrix multiplication, each instance of the innermost loop computes one element of the array answer; there are n2 of them. Blocking interleaves the execution of a block of instances, computing B iterations from each instance at a time. Similarly, we can interleave iterations in parallel loops to take advantage of reuses between them.

 

We define two primitives below that can reduce the distance between reuses across different iterations. We apply these primitives repeatedly, starting from the outermost loop until all the reuses are moved adjacent to each other in the innermost block.

 

 

Interleaving Inner Loops in a Parallel Loop

 

Consider the case where an outer parallelizable loop contains an inner loop. To exploit reuse across iterations of the outer loop, we interleave the executions of a fixed number of instances of the inner loop, as shown in Fig. 11.63. Creating two-dimensional inner blocks, this transformation reduces the distance between reuse of consecutive iterations of the outer loop.


is known as stripmining. In the case where the outer loop in Fig. 11.63 has a small known bound, we need not stripmine it, but can simply permute the two loops in the original program.

 

 

Interleaving S t a t e m e n t s in a Parallel Loop

 

Consider the case where a parallelizable loop contains a sequence of statements si, S2, • • • , sm. If some of these statements are loops themselves, statements from consecutive iterations may still be separated by many operations. We can exploit reuse between iterations by again interleaving their executions, as shown in Fig. 11.64. This transformation distributes a stripmined loop across the statements. Again, if the outer loop has a small fixed number of iterations, we need not stripmine the loop but simply distribute the original loop over all the statements.

 

We use Si(j) to denote the execution of statement si in iteration j. Instead of the original sequential execution order shown in Fig. 11.65(a), the code executes in the order shown in Fig. 11.65(b).

 

E x a m p l e 1 1 . 7 0 : We now return to the multigrid example and show how we exploit reuse between iterations of outer parallel loops. We observe that references DW[1, k,j, i], DW[1, k-l,j,i], and DW[l,k+l,j,i]m the innermost loops of the code in Fig. 11.62 have rather poor spatial locality. From reuse analysis, as discussed in Section 11.5, the loop with index i carries spatial locality and the loop with index k carries group reuse. The loop with index k is already the innermost loop, so we are interested in interleaving operations on DW from a block of partitions with consecutive i values.

 

We apply the transform to interleave statements in the loop to obtain the code in Fig. 11.66, then apply the transform to interleave inner loops to obtain the code in Fig. 11.67. Notice that as we interleave B iterations from loop with

 

index i, we need to expand variables AP, AM,T into arrays that hold B results at a time.    •

 

 

4. Putting it All Together

 

Algorithm 11.71 optimizes locality for a uniprocessor, and Algorithm 11.72 optimizes both parallelism and locality for a multiprocessor.

 

A l g o r i t h m 1 1 . 7 1 :  Optimize data locality on a uniprocessor.

 

INPUT: A program with affine array accesses.

 

OUTPUT: An equivalent program that maximizes data locality.

 

METHOD: Do the following steps:

 

1.         Apply Algorithm 11.64 to optimize the temporal locality of computed results.

 

 

2.         Apply Algorithm 11.68 to contract arrays where possible.

 

3.             Determine the iteration subspace that may share the same data or cache lines using the technique described in Section 11.5. For each statement, identify those outer parallel loop dimensions that have data reuse.

4.             For each outer parallel loop carrying reuse, move a block of the iterations into the innermost block by applying the interleaving primitives repeat-edly.


 

5.           Apply blocking to the subset of dimensions in the innermost fully per-mutable loop nest that carries reuse.

 

6.           Block outer fully permutable loop nest for higher levels of memory hier-archies, such as the third-level cache or the physical memory.

 

7.           Expand scalars and arrays where necessary by the lengths of the blocks.

 

 

A l g o r i t h m  1 1 . 7 2 :  Optimize parallelism and data locality for multiprocessors.

 

INPUT: A program with affine array accesses.

 

OUTPUT: An equivalent program that maximizes parallelism and data locality.

 

METHOD: Do the following:

 

1. Use the Algorithm 11.64 to parallelize the program and create an SPMD program.


 

2. Apply Algorithm 11.71 to the SPMD program produced in Step 1 to optimize its locality.

 

 

5. Exercises for Section 11.10

 

Exercise 11 . 10 . 1 :  Perform array contraction on the following vector operations:

 

for (i=0; i<n; i++) T[i] = A[i] * B[i]; for (i=0; i<n; i++) D[i] = T[i] + C[i];

 

Exercise 11 . 10 . 2:  Perform array contraction on the following vector operations:

 

for (j = 2, j <= jl, j++)

 

for  (ii = 2,  ii  <= il,  ii+=b)  {

 

for  (i = ii;  i <= min(ii+b-l,il);  i++)  {

 

ib                             =  i-ii+1;

 

AP[ib]                     =  ...;

 

T                                =1 . 0/(1 . 0 +AP[ib]);

 

D[2,ib]                   = T*AP[ib] ;

 

DW[l,2,j,i]  = T*DW[l,2,j,i] ;

 

}

 

for (k=3, k <= kl-1, k++)

 

for  (i = ii;  i <= min(ii+b-l,il);  i++)  {

 

ib                                   =  i-ii+1;

 

AM                                  = AP [ib];

 

AP[ib]                          =  ...;

 

T                                      = . . .AP[ib]-AM*D[ib,k-l] . . .;

 

D[ib,k]                        = T*AP;

 

DW[l,k,j,i] = T*(DW[l,k,j,i]+DW[l,k-l,j,i])...;

 

}

 

for  (k=kl-l,  k>=2,  k — )  {

 

for  (i = ii;  i <= min(ii+b-l,il);  i++)

 

DW[l,k,j,i] = DW[l,k,j,i] +D[iw,k]*DW[l,k+l,j,i] ;

 

/*  Ends  code to be executed by processor  (j,i)  */

 

}

 

}

 

 

Figure 11.67: Excerpt of Fig. 11.23 after partitioning, array contraction, and blocking

 

 

for (i=0; i<n; i++) T[i] = A[i] + B[i]; for (i=0; i<n; i++) S[i] = C[i] + D[i]; for (i=0; i<n; i++) E[i] = T[i] * S[i];

 

Exercise 1 1 . 1 0 . 3 :  Stripmine the outer loop

 

for (i=n-l; i>=0; i — ) for (j=0; j<n; j++)

 

into strips of width 10.


Study Material, Lecturing Notes, Assignment, Reference, Wiki description explanation, brief detail


Copyright © 2018-2021 BrainKart.com; All Rights Reserved. (BS) Developed by Therithal info, Chennai.