We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent a89c008 commit 4c7c11fCopy full SHA for 4c7c11f
examples/benchmark/benchmark.hpp
@@ -4,6 +4,18 @@
4
5
#include "format_bin.hpp"
6
7
+// It seems that there is a "threshold" to the amount of functions being
8
+// benchmarked. Having "too many" of them makes them become slow(er). This
9
+// appears to be due to code alignment. See similar issue and video referencing
10
+// that issue:
11
+//
12
+// * https://github.com/google/benchmark/issues/461
13
+// * https://www.youtube.com/watch?v=10MQW-aJU3g&t=197s
14
15
+// Having split up many of the benchmarks into different executables solved it
16
+// for a while, but the problem came back for bm_pico_kd_tree. For now,
17
+// benchmarks are just enabled 1-by-1.
18
+
19
namespace pico_tree {
20
21
class Benchmark : public benchmark::Fixture {
0 commit comments