Searched refs:unrolling (Results 1 – 25 of 62) sorted by relevance
123
| /openbsd/src/gnu/llvm/llvm/docs/ |
| D | TransformMetadata.rst | 53 Some attributes describe code transformations (unrolling, vectorizing,
91 specified at the same time, unrolling may occur either before or after
182 is present. Unrolling can be full unrolling, partial unrolling of a loop
183 with constant trip count or runtime unrolling of a loop with a trip
187 partial/runtime unrolling, the original loop of
287 as well as an additional inner loop unrolling. If no follow-up
394 - SimpleLoopUnroll/LoopFullUnroll (only performs full unrolling)
399 - LoopUnroll (partial and runtime unrolling)
404 - SimpleLoopUnroll/LoopFullUnroll (only performs full unrolling)
406 - LoopUnroll (partial and runtime unrolling)
[all …]
|
| D | CompileCudaWithLLVM.rst | 532 * Aggressive loop unrolling and function inlining -- Loop unrolling and
534 control flow transfer in GPU is more expensive. More aggressive unrolling and
538 (Programmers can force unrolling and inline using clang's `loop unrolling pragmas
|
| D | Vectorizers.rst | 348 Partial unrolling during vectorization
354 performing partial-unrolling of loops.
358 By unrolling the code the Loop Vectorizer allows two or more execution ports
|
| /openbsd/src/sys/lib/libkern/arch/mips64/ |
| D | bzero.S | 58 bne a0, a3, 2b # unrolling loop does not help
75 bne a0, a3, 2b # unrolling loop does not help
|
| /openbsd/src/gnu/llvm/llvm/docs/HistoricalNotes/ |
| D | 2001-06-01-GCCOptimizations.txt | 26 13. Loop unrolling
34 unrolling... these would be the major things we would miss out on if we
|
| /openbsd/src/lib/libc/arch/mips64/string/ |
| D | bzero.S | 54 bne a0, a3, 2b # unrolling loop does not help
|
| /openbsd/src/gnu/gcc/gcc/ |
| D | loop-unroll.c | 1793 determine_split_iv_delta (unsigned n_copy, unsigned n_copies, bool unrolling) in determine_split_iv_delta() argument
1795 if (unrolling) in determine_split_iv_delta()
2075 unsigned n_copies, bool unrolling, in apply_opt_in_copies() argument
2086 gcc_assert (!unrolling || rewrite_original_loop); in apply_opt_in_copies()
2100 unrolling); in apply_opt_in_copies()
2131 if (unrolling && opt_info->insns_with_var_to_expand) in apply_opt_in_copies()
2170 delta = determine_split_iv_delta (0, n_copies, unrolling); in apply_opt_in_copies()
|
| D | timevar.def | 108 DEFTIMEVAR (TV_COMPLETE_UNROLL , "complete unrolling")
|
| D | common.opt | 1034 Perform loop unrolling when iteration count is known
1038 Perform loop unrolling for all loops
|
| D | params.def | 146 …d, the maximum number of times that an individual variable will be expanded during loop unrolling",
|
| /openbsd/src/gnu/llvm/llvm/lib/Target/AArch64/ |
| D | AArch64SchedKryo.td | 25 // Enable partial & runtime unrolling. The magic number is chosen based on
|
| D | AArch64SchedA57.td | 29 // Enable partial & runtime unrolling. The magic number is chosen based on
|
| /openbsd/src/sys/lib/libkern/arch/alpha/ |
| D | divrem.m4 | 37 * Note that the loops could probably benefit from unrolling.
|
| /openbsd/src/lib/libc/arch/alpha/gen/ |
| D | divrem.m4 | 37 * Note that the loops could probably benefit from unrolling.
|
| /openbsd/src/gnu/gcc/gcc/config/sh/ |
| D | sh.opt | 175 Throttle unrolling to avoid thrashing target registers unless the unroll benefit outweighs this
|
| /openbsd/src/gnu/usr.bin/gcc/ |
| D | MAINTAINERS | 117 loop unrolling Jim Wilson wilson@tuliptree.org
|
| /openbsd/src/gnu/lib/libstdc++/libstdc++/include/ext/ |
| D | algorithm | 116 * optimizations such as unrolling).
|
| /openbsd/src/gnu/gcc/libstdc++-v3/include/ext/ |
| D | algorithm | 118 * optimizations such as unrolling).
|
| /openbsd/src/gnu/usr.bin/gcc/gcc/f/ |
| D | bugs.texi | 97 from unrolling loops.
|
| D | invoke.texi | 1604 @cindex loops, unrolling
1605 @cindex unrolling loops
1615 unrolling them and is probably generally appropriate for Fortran, though
1617 Note that outer loop unrolling isn't done specifically; decisions about
1634 to---unrolling. Loops written with @code{IF} and @code{GOTO} are not
1644 unrolling them in addition to iterative @code{DO} loops. In the absence
|
| /openbsd/src/gnu/llvm/clang/docs/ |
| D | OpenMPSupport.rst | 306 | loop | Loop unrolling transformation | :go…
|
| /openbsd/src/gnu/usr.bin/gcc/gcc/doc/ |
| D | passes.texi | 362 and optionally does strength-reduction and loop unrolling as well.
364 @file{loop.h} used for communication between them. Loop unrolling uses
|
| /openbsd/src/gnu/llvm/llvm/docs/CommandGuide/ |
| D | llvm-exegesis.rst | 237 unrolling tends to better hide the effects of the CPU frontend on architectures
|
| /openbsd/src/gnu/llvm/llvm/ |
| D | CREDITS.TXT | 70 D: Loop unrolling with run-time trip counts.
|
| /openbsd/src/gnu/gcc/gcc/doc/ |
| D | loop.texi | 229 unrolling or loop peeling. @code{can_duplicate_loop_p}
269 Without LCSSA, operations like loop unrolling may force creation of PHI
|
123