Considerably speed up dua interactive by allowing to use all (logical…

…) cores

src/aggregate.rs

@@ -37,12 +37,10 @@ pub fn aggregate(
                             if options.apparent_size {
                                 m.len()
                             } else {
-                                entry.path().size_on_disk_fast(m).unwrap_or_else(
-                                    |_| {
-                                        num_errors += 1;
-                                        0
-                                    },
-                                )
+                                entry.path().size_on_disk_fast(m).unwrap_or_else(|_| {
+                                    num_errors += 1;
+                                    0
+                                })
                             }
                         }
                         Some(Ok(_)) => 0,

src/traverse.rs

@@ -76,7 +76,7 @@ impl Traversal {
         if walk_options.threads == 0 {
             // avoid using the global rayon pool, as it will keep a lot of threads alive after we are done.
             // Also means that we will spin up a bunch of threads per root path, instead of reusing them.
-            walk_options.threads = num_cpus::get_physical();
+            walk_options.threads = num_cpus::get();
         }
         for path in input.into_iter() {
             let mut last_seen_eid = 0;