Displaying 4 results from an estimated 4 matches for "value_pair".
2011 Nov 17
2
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
...er of instcombine and GVN passes. I don't see a strong reason
why we would not want to do this after vectorization (In case it is
known to be useful in general).
> +++ b/lib/Transforms/Vectorize/BBVectorize.cpp
> @@ -0,0 +1,1338 @@
> + typedef std::pair<Value *, Value *> value_pair;
> + typedef std::pair<value_pair, value_pair> vp_pair;
> + typedef std::pair<std::multimap<Value *, Value *>::iterator,
> + std::multimap<Value *, Value *>::iterator> vp_iterator_pair;
> + typedef std::pair<std::multimap<value_pair,...
2011 Nov 21
0
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
...#39;t see a strong reason
> why we would not want to do this after vectorization (In case it is
> known to be useful in general).
>
>
> > +++ b/lib/Transforms/Vectorize/BBVectorize.cpp
> > @@ -0,0 +1,1338 @@
>
> > + typedef std::pair<Value *, Value *> value_pair;
> > + typedef std::pair<value_pair, value_pair> vp_pair;
> > + typedef std::pair<std::multimap<Value *, Value *>::iterator,
> > + std::multimap<Value *, Value *>::iterator> vp_iterator_pair;
> > + typedef std::pair<std::mul...
2011 Nov 16
0
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
Tobias, et al.,
Attached is the my autovectorization pass. I've fixed a bug that appears
when using -bb-vectorize-aligned-only, fixed some 80-col violations,
etc., and at least on x86_64, all test cases pass except for a few; and
all of these failures look like instruction-selection bugs. For example:
MultiSource/Applications/ClamAV - fails to compile shared_sha256.c with
an error: error in
2011 Nov 15
3
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
Tobias,
I've attached the latest version of my autovectorization patch. I was
able to add support for using the ScalarEvolution analysis for
load/store pairing (thanks for your help!). This led to a modest
performance increase and a modest compile-time increase. This version
also has a cutoff as you suggested (although the default value is set
high (4000 instructions between pairs) because