Displaying 7 results from an estimated 7 matches for "computepairsconnectedto".
2011 Dec 29
2
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
...ot have any strong opinion about the other top-level
maps. Just decide yourself what is easier to read.
> 2. I have not (yet) changed the types of the maps from holding Value* to
> Instruction*. Doing so would eliminate a few casts, but would cause
> even-more casts (and maybe checks) in computePairsConnectedTo. Even so,
> it might be worthwhile to make the change for clarity (conceptually,
> those maps do hold only pointers to instructions).
I believe if code relies on the assumption that some data structures
only objects of a certain type, we should use that type to define the
data structures....
2011 Dec 29
0
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
...n about the other top-level
> maps. Just decide yourself what is easier to read.
>
> > 2. I have not (yet) changed the types of the maps from holding Value* to
> > Instruction*. Doing so would eliminate a few casts, but would cause
> > even-more casts (and maybe checks) in computePairsConnectedTo. Even so,
> > it might be worthwhile to make the change for clarity (conceptually,
> > those maps do hold only pointers to instructions).
>
> I believe if code relies on the assumption that some data structures
> only objects of a certain type, we should use that type to defi...
2011 Dec 14
0
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
...on every call to runOnBasicBlock, then I
should probably do the same with all of the top-level maps, etc. Do you
agree?
2. I have not (yet) changed the types of the maps from holding Value* to
Instruction*. Doing so would eliminate a few casts, but would cause
even-more casts (and maybe checks) in computePairsConnectedTo. Even so,
it might be worthwhile to make the change for clarity (conceptually,
those maps do hold only pointers to instructions).
Please look it over and let me know what you think.
Thanks again,
Hal
On Fri, 2011-12-02 at 17:07 +0100, Tobias Grosser wrote:
> On 11/23/2011 05:52 PM, Hal Finkel...
2011 Dec 02
5
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
...eSet<Value *> &Users,
> + AliasSetTracker&WriteSet, Instruction *I,
> + Instruction *J, bool&UsesI, bool UpdateUsers = true,
> + std::multimap<Value *, Value *> *LoadMoveSet = 0);
> +
> + void computePairsConnectedTo(
> + std::multimap<Value *, Value *> &CandidatePairs,
> + std::vector<Value *> &PairableInsts,
> + std::multimap<ValuePair, ValuePair> &ConnectedPairs,
> + ValuePair P);...
2011 Nov 23
0
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
On Mon, 2011-11-21 at 21:22 -0600, Hal Finkel wrote:
> On Mon, 2011-11-21 at 11:55 -0600, Hal Finkel wrote:
> > Tobias,
> >
> > I've attached an updated patch. It contains a few bug fixes and many
> > (refactoring and coding-convention) changes inspired by your comments.
> >
> > I'm currently trying to fix the bug responsible for causing a compile
2011 Dec 02
0
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
...;Users,
> > + AliasSetTracker&WriteSet, Instruction *I,
> > + Instruction *J, bool&UsesI, bool UpdateUsers = true,
> > + std::multimap<Value *, Value *> *LoadMoveSet = 0);
> > +
> > + void computePairsConnectedTo(
> > + std::multimap<Value *, Value *> &CandidatePairs,
> > + std::vector<Value *> &PairableInsts,
> > + std::multimap<ValuePair, ValuePair> &ConnectedPairs,
> > +...
2011 Nov 22
5
[LLVMdev] [llvm-commits] [PATCH] BasicBlock Autovectorization Pass
On Mon, 2011-11-21 at 11:55 -0600, Hal Finkel wrote:
> Tobias,
>
> I've attached an updated patch. It contains a few bug fixes and many
> (refactoring and coding-convention) changes inspired by your comments.
>
> I'm currently trying to fix the bug responsible for causing a compile
> failure when compiling
>