Jack W.Davidson and Sanjay Jinturkar. Aggressive Loop Unrolling in a Retargetable, Optimizing Compiler. In Compiler Construction, volume 1060 of Lecture Notes in Computer Science, pages 59--73. Springer, April 1996. Home/Search Document Not in Database Summary Related Articles Check |
This paper is cited in the following contexts:
Effectively Exploiting Indirect Jumps - Uh (1997) (2 citations) (Correct)
....with the less expensive sequence. This technique has been used to eliminate conditional branches over short instruction sequences in many instances on the IBM RS 6000 [14] Loop unrolling has been used to avoid executions of the conditional branch associated with a loop termination condition [11]. Loop unswitching moves a conditional branch with a loop invariant condition before 12 13 the loop and duplicates the loop in each of the two destinations of the branch [1] Conditional branches have also been avoided by code duplication [20] This method determines if there are paths where the ....
J. W. Davidson and S. Jinturkar. Aggressive loop unrolling in a retargetable, optimizing compiler. In Proceedings of Compiler Construction Conference, pages 59--73, April 1996.
Implications of Memory Mappings on Cache Misses - Genius, Eisenbiegler (1998) (Correct)
....class of oblivious algorithms (defined by Lowe and Zimmermann [ZL94] Exclusively, reuse wrt. the innermost loop is considered, as the greatest effects can be achieved here. We concentrate on data caches. Instruction caches are not considered. Unless loops are unrolled extensively as e.g. in [DJ96], it is legitimate to leave instruction cache behaviour out of the focus. As noted above, the scientific applications for which the optimizations apply are oblivious and perfectly nested.References in the matrix multiplication algorithm (figure 2) are not uniformly generated.There are three ....
Jack W. Davidson and Sanjay Jinturkar. Aggressive loop unrolling in a retargetable, optimizing compiler. In Compiler Construction, volume 1060 of LNCS, pages 59--73, April 1996.
Implications of Memory Mapping on Cache Misses - Genius, Eisenbiegler (1997) (Correct)
....class of oblivious algorithms (defined by Lowe and Zimmermann [ZL94] Exclusively, reuse wrt. the innermost loop is considered, as the greatest effects can be achieved here. We concentrate on data caches. Instruction caches are not considered. Unless loops are unrolled extensively as e.g. in [DJ96], it is legitimate to leave instruction cache behaviour out of the focus. As noted above, the scientific applications for which the optimizations apply are oblivious, perfectly nested and free of loop carried dependencies. Indices are affine. The following are the criteria according to which loop ....
Jack W. Davidson and Sanjay Jinturkar. Aggressive loop unrolling in a retargetable, optimizing compiler. In Compiler Construction, volume 1060 of LNCS, pages 59--73, April 1996.
Techniques for Effectively Exploiting a Zero.. - Uh, Wang, Whalley.. Self-citation (Jinturkar) (Correct)
....Many code improving transformations and architectural features used to improve execution time for applications in general purpose computers do so at the expense of substantial code growth and more power consumption. For instance, loop unrolling is a popular technique to decrease loop overhead [2]. Yet, this approach often requires a signi cant increase in code size. Likewise, VLIW instructions can be used to reduce loop overhead at the expense of more power. Space increasing transformations and power inecient architectures are often unacceptable options for many DSP applications due to ....
....of iterations. After applying basic induction variable elimination and dead store elimination, the increment and initialization of a1 are removed. Thus, the loop overhead has been eliminated. a) Source Code of Loop a[i] 0; for (i = 0; i 10000; i ) r0 = a # r[0] ADDR( a) a2 = 0 # a[2]=0; L5: r0 = a2 # M[r[0] a[2] r[0] r[0] 2; r0 = a a2 = 0 r0 = a2 (c) After Using the ZOLB (b) DSP16000 Assembly and Corresponding RTLs without Using the ZOLB if le goto L5 # PC=IC =0 L5:PC; a1 = a1 1 # a[1] a[1] 1; IC=a[1] 1 0; a1 = 9999 # a[1] 9999; cloop = 10000 do cloop ....
[Article contains additional citation context not shown here]
Davidson, J.W., Jinturkar, S.: Aggressive Loop Unrolling in a Retargetable, Optimizing Compiler. Proceedings of Compiler Construction Conference. 59-73 (April 1996).
A Machine Learning Approach to Automatic Production of.. - Monsifrot, Bodin.. (2002) (4 citations) (Correct)
No context found.
Jack W.Davidson and Sanjay Jinturkar. Aggressive Loop Unrolling in a Retargetable, Optimizing Compiler. In Compiler Construction, volume 1060 of Lecture Notes in Computer Science, pages 59--73. Springer, April 1996.
Coalescing Conditional Branches into Efficient Indirect Jumps - Uh, Whalley (1997) (1 citation) (Correct)
No context found.
J.W. Davidson and S. Jinturkar. Aggressive loop unrolling in a retargetable, optimizing compiler. In Proceedings of Compiler Construction Conference, pages 59--73, April 1996.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC