Explicilty depend on HostCPUFeatures and CPUSummary, bump verion.

Author: chriselrod

Project.toml

@@ -5,8 +5,10 @@ version = "0.12.73"
 
 [deps]
 ArrayInterface = "4fba245c-0d91-5ea0-9b3e-6abc04ee57a9"
+CPUSummary = "2a0fbf3d-bb9c-48f3-b0a9-814d99fd7ab9"
 CloseOpenIntervals = "fb6a15b2-703c-40df-9091-08a04967cfa9"
 DocStringExtensions = "ffbed154-4ef7-542d-bbb7-c09d3a79fcae"
+HostCPUFeatures = "3e5b6fbb-0976-4d2c-9146-d79de83f2fb0"
 IfElse = "615f187c-cbe4-4ef1-ba3b-2fcf58d6d173"
 LayoutPointers = "10f19ff3-798f-405d-979b-55457f8fc047"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
@@ -21,7 +23,10 @@ VectorizationBase = "3d5dd08c-fd9d-11e8-17fa-ed2836048c2f"
 
 [compat]
 ArrayInterface = "3.1.32"
+CPUSummary = "0.1.3"
+CloseOpenIntervals = "0.1.2"
 DocStringExtensions = "0.8"
+HostCPUFeatures = "0.1.3"
 IfElse = "0.1"
 LayoutPointers = "0.1.2"
 OffsetArrays = "1.4.1"

src/LoopVectorization.jl

@@ -2,8 +2,8 @@ module LoopVectorization
 
 using Static: StaticInt, gt
 using VectorizationBase, SLEEFPirates, UnPack, OffsetArrays
-using VectorizationBase: register_size, register_count, cache_linesize, cache_size, has_opmask_registers,
-  mask, pick_vector_width, MM, AbstractMask, data, grouped_strided_pointer, AbstractSIMD,
+using VectorizationBase: 
+  mask, MM, AbstractMask, data, grouped_strided_pointer, AbstractSIMD,
   vzero, offsetprecalc, lazymul,
   vadd_nw, vadd_nsw, vadd_nuw, vsub_nw, vsub_nsw, vsub_nuw, vmul_nw, vmul_nsw, vmul_nuw,
   vfmaddsub, vfmsubadd, vpermilps177, vmovsldup, vmovshdup,
@@ -21,10 +21,10 @@ using VectorizationBase: register_size, register_count, cache_linesize, cache_si
     contract_min, collapse_min,
     contract_and, collapse_and,
     contract_or,  collapse_or,
-    num_threads, num_cores,
     max_mask, maybestaticsize#,zero_mask
 
-
+using HostCPUFeatures: pick_vector_width, register_size, register_count, has_opmask_registers
+using CPUSummary: num_threads, num_cores, cache_linesize, cache_size
 
 using LayoutPointers: stridedpointer_preserve, GroupedStridedPointers
 

src/simdfunctionals/map.jl

@@ -193,7 +193,7 @@ function vmap_multithread!(
     W, Wshift = VectorizationBase.pick_vector_width_shift(T)
     ptry, ptrargs, N = setup_vmap!(f, y, Val{NonTemporal}(), args...)
     # nt = min(Threads.nthreads(), VectorizationBase.SYS_CPU_THREADS, N >> (Wshift + 3))
-    nt = min(Threads.nthreads(), VectorizationBase.num_cores(), N >> (Wshift + 5))
+    nt = min(Threads.nthreads(), num_cores(), N >> (Wshift + 5))
 
     # if !((nt > 1) && iszero(ccall(:jl_in_threaded_region, Cint, ())))
     if nt < 2