Displaying 10 results from an estimated 10 matches for "anticonverg".
2018 Dec 19
5
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
Hi all,
LLVM needs a solution to the long-standing problem that the IR is unable
to express certain semantics expected by high-level programming
languages that target GPUs.
Solving this issue is necessary both for upstream use of LLVM as a
compiler backend for GPUs and for correctly supporting LLVM IR <->
SPIR-V roundtrip translation. It may also be useful for compilers
targeting
2018 Dec 29
1
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...functions like
> syncthreads(), to which we cannot add control-flow dependencies.
> That is, it's legal to hoist syncthreads out of an "if", but it's
> not legal to sink it into an "if". It's not clear to me why we
> can't have "anticonvergent" (terrible name) functions which cannot
> have control-flow dependencies removed from them? ballot() would be
> both convergent and anticonvergent.
>
> Would that solve your problem?
>
>
> I think it's important to note that we already have such an a...
2019 Jan 24
3
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...> > syncthreads(), to which we cannot add control-flow dependencies.
> > That is, it's legal to hoist syncthreads out of an "if", but it's
> > not legal to sink it into an "if". It's not clear to me why we
> > can't have "anticonvergent" (terrible name) functions which cannot
> > have control-flow dependencies removed from them? ballot() would be
> > both convergent and anticonvergent.
> >
> > Would that solve your problem?
> >
> >
> > I think it's important to note t...
2019 Jan 28
2
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...ncthreads(), to which we cannot add control-flow dependencies.
> > > That is, it's legal to hoist syncthreads out of an "if", but it's
> > > not legal to sink it into an "if". It's not clear to me why we
> > > can't have "anticonvergent" (terrible name) functions which cannot
> > > have control-flow dependencies removed from them? ballot() would
be
> > > both convergent and anticonvergent.
> > >
> > > Would that solve your problem?
> > >
> > >
> > > I...
2019 Jan 30
3
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...which we cannot add control-flow dependencies.
> > > > That is, it's legal to hoist syncthreads out of an "if", but it's
> > > > not legal to sink it into an "if". It's not clear to me why we
> > > > can't have "anticonvergent" (terrible name) functions which cannot
> > > > have control-flow dependencies removed from them? ballot() would be
> > > > both convergent and anticonvergent.
> > > >
> > > > Would that solve your problem?
> > > >
> &...
2019 Jan 31
2
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...control-flow dependencies.
> > > > > That is, it's legal to hoist syncthreads out of an "if", but
> it's
> > > > > not legal to sink it into an "if". It's not clear to me why we
> > > > > can't have "anticonvergent" (terrible name) functions which
> cannot
> > > > > have control-flow dependencies removed from them? ballot()
> would be
> > > > > both convergent and anticonvergent.
> > > > >
> > > > > Would that solve your probl...
2019 Jan 31
3
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...control-flow dependencies.
> > > > > That is, it's legal to hoist syncthreads out of an "if", but
> it's
> > > > > not legal to sink it into an "if". It's not clear to me why we
> > > > > can't have "anticonvergent" (terrible name) functions which
> cannot
> > > > > have control-flow dependencies removed from them? ballot()
> would be
> > > > > both convergent and anticonvergent.
> > > > >
> > > > > Would that solve your probl...
2019 Jan 30
2
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...control-flow dependencies.
> > > > > That is, it's legal to hoist syncthreads out of an "if", but
> it's
> > > > > not legal to sink it into an "if". It's not clear to me why we
> > > > > can't have "anticonvergent" (terrible name) functions which
> cannot
> > > > > have control-flow dependencies removed from them? ballot()
> would be
> > > > > both convergent and anticonvergent.
> > > > >
> > > > > Would that solve your probl...
2019 Feb 01
2
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...> > > > > That is, it's legal to hoist syncthreads out of an "if",
> but it's
> > > > > not legal to sink it into an "if". It's not clear to me
> why we
> > > > > can't have "anticonvergent" (terrible name) functions
> which cannot
> > > > > have control-flow dependencies removed from them?
> ballot() would be
> > > > > both convergent and anticonvergent.
> > > > >
> > > > > ...
2019 Feb 09
1
[RFC] Adding thread group semantics to LangRef (motivated by GPUs)
...; > > That is, it's legal to hoist syncthreads out of an "if",
> > but it's
> > > > > > not legal to sink it into an "if". It's not clear to me
> > why we
> > > > > > can't have "anticonvergent" (terrible name) functions
> > which cannot
> > > > > > have control-flow dependencies removed from them?
> > ballot() would be
> > > > > > both convergent and anticonvergent.
> > > > > >
> >...