Merge pull request #36 from gaelforget/fix_cuda_atomic_julia1p7

for Julia>=1.7 CUDA no longer exports @atomic
JuliaOcean · Sep 9, 2021 · 6a9fe59 · 6a9fe59
2 parents 712b6b1 + d68cfa0
commit 6a9fe59
Show file tree

Hide file tree

Showing 2 changed files with 16 additions and 16 deletions.
diff --git a/src/output/diagnostics_plankton.jl b/src/output/diagnostics_plankton.jl
@@ -6,7 +6,7 @@ function gpu_diags_proc_kernel!(diags_proc, proc, ac, x, y, z)
     index = (blockIdx().x - 1) * blockDim().x + threadIdx().x
     stride = blockDim().x * gridDim().x
     for i = index:stride:size(ac,1)
-        @inbounds @atomic diags_proc[x[i], y[i], z[i]] += proc[i] * ac[i]
+        @inbounds CUDA.@atomic diags_proc[x[i], y[i], z[i]] += proc[i] * ac[i]
     end
 end
 function diags_proc!(diags_proc, proc, ac, x, y, z, ::GPU)
@@ -53,4 +53,4 @@ function diags_spcs!(diags_sp, proc, plank, ac, x, y, z, arch::Architecture)
             throw(ArgumentError("$(diag) is not one of the diagnostics"))
         end
     end
-end
+end
diff --git a/src/plankton/physiology/counts.jl b/src/plankton/physiology/counts.jl
@@ -3,8 +3,8 @@
     index = (blockIdx().x - 1) * blockDim().x + threadIdx().x
     stride = blockDim().x * gridDim().x
     for i = index:stride:size(ac,1)
-        @inbounds @atomic ctschl[x[i], y[i], z[i]] += chl[i] * ac[i]
-        @inbounds @atomic ctspop[x[i], y[i], z[i]] += ac[i]
+        @inbounds CUDA.@atomic ctschl[x[i], y[i], z[i]] += chl[i] * ac[i]
+        @inbounds CUDA.@atomic ctspop[x[i], y[i], z[i]] += ac[i]
     end
     return nothing
 end
@@ -25,11 +25,11 @@ function gpu_calc_consume_kernel!(ctsdic, ctsdoc, ctsnh4, ctsno3, ctspo4, proc,
     index = (blockIdx().x - 1) * blockDim().x + threadIdx().x
     stride = blockDim().x * gridDim().x
     for i = index:stride:size(ac,1)
-        @inbounds @atomic ctsdic[x[i], y[i], z[i]] += (proc.resp[i] - proc.PS[i]) * ΔT * ac[i]
-        @inbounds @atomic ctsdoc[x[i], y[i], z[i]] += (proc.exu[i] - proc.VDOC[i]) * ΔT * ac[i]
-        @inbounds @atomic ctsnh4[x[i], y[i], z[i]] -= proc.VNH4[i] * ΔT * ac[i]
-        @inbounds @atomic ctsno3[x[i], y[i], z[i]] -= proc.VNO3[i] * ΔT * ac[i]
-        @inbounds @atomic ctspo4[x[i], y[i], z[i]] -= proc.VPO4[i] * ΔT * ac[i]
+        @inbounds CUDA.@atomic ctsdic[x[i], y[i], z[i]] += (proc.resp[i] - proc.PS[i]) * ΔT * ac[i]
+        @inbounds CUDA.@atomic ctsdoc[x[i], y[i], z[i]] += (proc.exu[i] - proc.VDOC[i]) * ΔT * ac[i]
+        @inbounds CUDA.@atomic ctsnh4[x[i], y[i], z[i]] -= proc.VNH4[i] * ΔT * ac[i]
+        @inbounds CUDA.@atomic ctsno3[x[i], y[i], z[i]] -= proc.VNO3[i] * ΔT * ac[i]
+        @inbounds CUDA.@atomic ctspo4[x[i], y[i], z[i]] -= proc.VPO4[i] * ΔT * ac[i]
     end
     return nothing
 end
@@ -55,12 +55,12 @@ function gpu_calc_loss_kernel!(ctsdoc, ctspoc, ctsdon, ctspon, ctsdop, ctspop, p
     index = (blockIdx().x - 1) * blockDim().x + threadIdx().x
     stride = blockDim().x * gridDim().x
     for i = index:stride:size(ac,1)
-        @inbounds @atomic ctsdoc[x[i], y[i], z[i]] += (plank.Bm[i] + plank.Cq[i]) * lossFracC * ac[i] * loss[i]
-        @inbounds @atomic ctsdon[x[i], y[i], z[i]] += (plank.Bm[i]*R_NC + plank.Nq[i]) * lossFracN * ac[i] * loss[i]
-        @inbounds @atomic ctsdop[x[i], y[i], z[i]] += (plank.Bm[i]*R_PC + plank.Pq[i]) * lossFracP * ac[i] * loss[i]
-        @inbounds @atomic ctspoc[x[i], y[i], z[i]] += (plank.Bm[i] + plank.Cq[i]) * (1.0-lossFracC) * ac[i] * loss[i]
-        @inbounds @atomic ctspon[x[i], y[i], z[i]] += (plank.Bm[i]*R_NC + plank.Nq[i]) * (1.0-lossFracN) * ac[i] * loss[i]
-        @inbounds @atomic ctspop[x[i], y[i], z[i]] += (plank.Bm[i]*R_PC + plank.Pq[i]) * (1.0-lossFracP) * ac[i] * loss[i]
+        @inbounds CUDA.@atomic ctsdoc[x[i], y[i], z[i]] += (plank.Bm[i] + plank.Cq[i]) * lossFracC * ac[i] * loss[i]
+        @inbounds CUDA.@atomic ctsdon[x[i], y[i], z[i]] += (plank.Bm[i]*R_NC + plank.Nq[i]) * lossFracN * ac[i] * loss[i]
+        @inbounds CUDA.@atomic ctsdop[x[i], y[i], z[i]] += (plank.Bm[i]*R_PC + plank.Pq[i]) * lossFracP * ac[i] * loss[i]
+        @inbounds CUDA.@atomic ctspoc[x[i], y[i], z[i]] += (plank.Bm[i] + plank.Cq[i]) * (1.0-lossFracC) * ac[i] * loss[i]
+        @inbounds CUDA.@atomic ctspon[x[i], y[i], z[i]] += (plank.Bm[i]*R_NC + plank.Nq[i]) * (1.0-lossFracN) * ac[i] * loss[i]
+        @inbounds CUDA.@atomic ctspop[x[i], y[i], z[i]] += (plank.Bm[i]*R_PC + plank.Pq[i]) * (1.0-lossFracP) * ac[i] * loss[i]
     end
     return nothing
 end
@@ -102,4 +102,4 @@ function calc_par!(par, arch::Architecture, chl, PARF, g::AbstractGrid, kc, kw)
     event = kernel!(par, chl, PARF, g, kc, kw)
     wait(device(arch), event)
     return nothing
-end
+end