BLI_task: change BLI_task_parallel_range_ex() to just take a bool whether to use threading or not, instead of threshold.

From recent experience, turns out we often do want to use something else than basic
range of parallelized forloop as control parameter over threads usage, so now BLI func
only takes a boolean, and caller defines best check for its own case.

source/blender/modifiers/intern/MOD_uvwarp.c

@@ -210,7 +210,7 @@ static DerivedMesh *applyModifier(ModifierData *md, Object *ob,
 	UVWarpData data = {.mpoly = mpoly, .mloop = mloop, .mloopuv = mloopuv,
 	                   .dvert = dvert, .defgrp_index = defgrp_index,
 	                   .warp_mat = warp_mat, .axis_u = axis_u, .axis_v = axis_v};
-	BLI_task_parallel_range_ex(0, numPolys, &data, NULL, 0, uv_warp_compute, 1000, false);
+	BLI_task_parallel_range_ex(0, numPolys, &data, NULL, 0, uv_warp_compute, numPolys > 1000, false);
 
 	dm->dirty |= DM_DIRTY_TESS_CDLAYERS;
 

source/blender/modifiers/intern/MOD_weightvgproximity.c

@@ -176,7 +176,8 @@ static void get_vert2geom_distance(int numVerts, float (*v_cos)[3],
 	data.dist[1] = dist_e;
 	data.dist[2] = dist_f;
 
-	BLI_task_parallel_range_ex(0, numVerts, &data, &data_chunk, sizeof(data_chunk), vert2geom_task_cb, 10000, false);
+	BLI_task_parallel_range_ex(
+	            0, numVerts, &data, &data_chunk, sizeof(data_chunk), vert2geom_task_cb, numVerts > 10000, false);
 
 	if (dist_v)
 		free_bvhtree_from_mesh(&treeData_v);