1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-22 04:22:57 +02:00
llvm-mirror/test/Transforms/BBVectorize
Hal Finkel d1fc849359 BBVectorize: Choose pair ordering to minimize shuffles
BBVectorize would, except for loads and stores, always fuse instructions
so that the first instruction (in the current source order) would always
represent the low part of the input vectors and the second instruction
would always represent the high part. This lead to too many shuffles
being produced because sometimes the opposite order produces fewer of them.

With this change, BBVectorize tracks the kind of pair connections that form
the DAG of candidate pairs, and uses that information to reorder the pairs to
avoid excess shuffles. Using this information, a future commit will be able
to add VTTI-based shuffle costs to the pair selection procedure. Importantly,
the number of remaining shuffles can now be estimated during pair selection.

There are some trivial instruction reorderings in the test cases, and one
simple additional test where we certainly want to do a reordering to
avoid an unnecessary shuffle.

llvm-svn: 167122
2012-10-31 15:17:07 +00:00
..
X86 BBVectorize: Choose pair ordering to minimize shuffles 2012-10-31 15:17:07 +00:00
cycle.ll BBVectorize: Choose pair ordering to minimize shuffles 2012-10-31 15:17:07 +00:00
func-alias.ll Update BBVectorize to use aliasesUnknownInst. 2012-02-10 15:52:40 +00:00
ld1.ll
lit.local.cfg Move target-specific BBVectorize tests into a separate directory. 2012-10-26 19:38:09 +00:00
loop1.ll BBVectorize: Choose pair ordering to minimize shuffles 2012-10-31 15:17:07 +00:00
mem-op-depth.ll
metadata.ll Teach BBVectorize to combine, when possible, or discard metadata when fusing instructions. 2012-06-16 20:34:06 +00:00
no-ldstr-conn.ll Fix an error in BBVectorize important for vectorizing pointer types. 2012-04-14 07:32:50 +00:00
req-depth.ll
search-limit.ll BBVectorize: Choose pair ordering to minimize shuffles 2012-10-31 15:17:07 +00:00
simple3.ll Allow BBVectorize to form non-2^n-length vectors. 2012-06-28 05:42:42 +00:00
simple-int.ll BBVectorize: Choose pair ordering to minimize shuffles 2012-10-31 15:17:07 +00:00
simple-ldstr-ptrs.ll Disable generation of pointer vectors by BBVectorize. 2012-10-26 00:05:26 +00:00
simple-ldstr.ll BBVectorize: Choose pair ordering to minimize shuffles 2012-10-31 15:17:07 +00:00
simple-sel.ll BBVectorize: Choose pair ordering to minimize shuffles 2012-10-31 15:17:07 +00:00
simple-tst.ll Don't vectorize target-specific types (ppc_fp128, x86_fp80, etc.). 2012-04-27 19:34:00 +00:00
simple.ll BBVectorize: Choose pair ordering to minimize shuffles 2012-10-31 15:17:07 +00:00