Displaying 3 results from an estimated 3 matches for "sync_region".
2012 Sep 26
0
[LLVMdev] [RFC] Parallelization metadata and intrinsics in LLVM (for OpenMP, etc.)
...(i32, ranging from 0 to (limit - 1) both inclusive) and `priv` are
passed to each of the invocations. The only invariant is that the 0th
iteration is always executed in the invoking thread.
It is legal to have calls to parallel_map inside a function being
parallel_map'ed over.
2. void @llvm.sync_region(void (i32, i8 *) fn, i8 type)
Semantics:
It is only legal to call sync_region from within the dynamic extent of
a parallel_map. It ensures that `limit` copies (the `limit` is the
limit from the parallel_map) of fn are executed with mutual exclusion.
`type` can either be 0 (`Any`) signifying tha...
2012 Aug 14
4
[LLVMdev] [RFC] Parallelization metadata and intrinsics in LLVM (for OpenMP, etc.)
On Tue, 14 Aug 2012 10:22:35 +0300
Pekka Jääskeläinen <pekka.jaaskelainen at tut.fi> wrote:
> On 08/13/2012 10:54 PM, Hal Finkel wrote:
> > I had thought about uses for shared-memory OpenCL implementations,
> > but I don't know enough about the use cases to make a specific
> > proposal. Is your metadata documented anywhere?
>
> It is now a quick "brute
2012 Sep 26
1
[LLVMdev] [RFC] Parallelization metadata and intrinsics in LLVM (for OpenMP, etc.)
...(i32, ranging from 0 to (limit - 1) both inclusive) and `priv` are
passed to each of the invocations. The only invariant is that the 0th
iteration is always executed in the invoking thread.
It is legal to have calls to parallel_map inside a function being
parallel_map'ed over.
2. void @llvm.sync_region(void (i32, i8 *) fn, i8 type)
Semantics:
It is only legal to call sync_region from within the dynamic extent of
a parallel_map. It ensures that `limit` copies (the `limit` is the
limit from the parallel_map) of fn are executed with mutual exclusion.
`type` can either be 0 (`Any`) signifying tha...