Fix align and pass package to thread comp

omlins · Jul 30, 2023 · abc049c · abc049c
1 parent a98bb96
commit abc049c
Showing 1 changed file with 7 additions and 6 deletions.
diff --git a/src/ParallelKernel/parallel.jl b/src/ParallelKernel/parallel.jl
@@ -244,13 +244,13 @@ end
 
 function parallel_call_gpu(nblocks::Union{Symbol,Expr}, nthreads::Union{Symbol,Expr}, kernelcall::Expr, backend_kwargs_expr::Array, async::Bool, package::Symbol; stream::Union{Symbol,Expr}=default_stream(package), shmem::Union{Symbol,Expr,Nothing}=nothing, launch::Bool=true, configcall::Expr=kernelcall)
     maxsize = :( $nblocks .* $nthreads )
-    ranges  = :(ParallelStencil.ParallelKernel.compute_ranges($maxsize))
+    ranges  = :( ParallelStencil.ParallelKernel.compute_ranges($maxsize) )
     parallel_call_gpu(ranges, nblocks, nthreads, kernelcall, backend_kwargs_expr, async, package; stream=stream, shmem=shmem, launch=launch)
 end
 
 function parallel_call_gpu(ranges::Union{Symbol,Expr}, kernelcall::Expr, backend_kwargs_expr::Array, async::Bool, package::Symbol; stream::Union{Symbol,Expr}=default_stream(package), shmem::Union{Symbol,Expr,Nothing}=nothing, launch::Bool=true, configcall::Expr=kernelcall)
-    maxsize  = :(length.(ParallelStencil.ParallelKernel.promote_ranges($ranges)))
-    nthreads = :( ParallelStencil.ParallelKernel.compute_nthreads($maxsize, package) )
+    maxsize  = :( length.(ParallelStencil.ParallelKernel.promote_ranges($ranges)) )
+    nthreads = :( ParallelStencil.ParallelKernel.compute_nthreads($maxsize, $package) )
     nblocks  = :( ParallelStencil.ParallelKernel.compute_nblocks($maxsize, $nthreads) )
     parallel_call_gpu(ranges, nblocks, nthreads, kernelcall, backend_kwargs_expr, async, package; stream=stream, shmem=shmem, launch=launch)
 end
@@ -284,7 +284,7 @@ end
 
 function parallel_call_threads(nblocks::Union{Symbol,Expr}, nthreads::Union{Symbol,Expr}, kernelcall::Expr, async::Bool; launch::Bool=true, configcall::Expr=kernelcall)
     maxsize = :( $nblocks .* $nthreads )
-    ranges  = :(ParallelStencil.ParallelKernel.compute_ranges($maxsize))
+    ranges  = :( ParallelStencil.ParallelKernel.compute_ranges($maxsize) )
     parallel_call_threads(ranges, kernelcall, async; launch=launch)
 end
 
@@ -497,8 +497,9 @@ end
 
 function compute_nthreads(maxsize, package; nthreads_max=NTHREADS_MAX, flatdim=0) # This is a heuristic, which results in (32,8,1) threads, except if maxsize[1] < 32 or maxsize[2] < 8.
     maxsize = promote_maxsize(maxsize)
-    if     (package == PKG_CUDA)   nthreads_x = min(32,              (flatdim==1) ? 1 : maxsize[1])
-    elseif (package == PKG_AMDGPU) nthreads_x = min(128,             (flatdim==1) ? 1 : maxsize[1])
+    if     (Symbol(package) == PKG_CUDA)   nthreads_x = min(32,              (flatdim==1) ? 1 : maxsize[1])
+    elseif (Symbol(package) == PKG_AMDGPU) nthreads_x = min(128,             (flatdim==1) ? 1 : maxsize[1])
+    else                                   nthreads_x = min(32,              (flatdim==1) ? 1 : maxsize[1])
     end
     nthreads_y = min(ceil(Int,nthreads_max/nthreads_x),              (flatdim==2) ? 1 : maxsize[2])
     nthreads_z = min(ceil(Int,nthreads_max/(nthreads_x*nthreads_y)), (flatdim==3) ? 1 : maxsize[3])