@@ -39,82 +39,166 @@ struct scale_test_def {
bool fails;
};
-void test_bpf_verif_scale(void)
-{
- struct scale_test_def tests[] = {
- { "loop3.o", BPF_PROG_TYPE_RAW_TRACEPOINT, true /* fails */ },
-
- { "test_verif_scale1.o", BPF_PROG_TYPE_SCHED_CLS },
- { "test_verif_scale2.o", BPF_PROG_TYPE_SCHED_CLS },
- { "test_verif_scale3.o", BPF_PROG_TYPE_SCHED_CLS },
-
- { "pyperf_global.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
- { "pyperf_subprogs.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
-
- /* full unroll by llvm */
- { "pyperf50.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
- { "pyperf100.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
- { "pyperf180.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
-
- /* partial unroll. llvm will unroll loop ~150 times.
- * C loop count -> 600.
- * Asm loop count -> 4.
- * 16k insns in loop body.
- * Total of 5 such loops. Total program size ~82k insns.
- */
- { "pyperf600.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
-
- /* no unroll at all.
- * C loop count -> 600.
- * ASM loop count -> 600.
- * ~110 insns in loop body.
- * Total of 5 such loops. Total program size ~1500 insns.
- */
- { "pyperf600_nounroll.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
-
- { "loop1.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
- { "loop2.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
- { "loop4.o", BPF_PROG_TYPE_SCHED_CLS },
- { "loop5.o", BPF_PROG_TYPE_SCHED_CLS },
- { "loop6.o", BPF_PROG_TYPE_KPROBE },
-
- /* partial unroll. 19k insn in a loop.
- * Total program size 20.8k insn.
- * ~350k processed_insns
- */
- { "strobemeta.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
-
- /* no unroll, tiny loops */
- { "strobemeta_nounroll1.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
- { "strobemeta_nounroll2.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
-
- /* non-inlined subprogs */
- { "strobemeta_subprogs.o", BPF_PROG_TYPE_RAW_TRACEPOINT },
-
- { "test_sysctl_loop1.o", BPF_PROG_TYPE_CGROUP_SYSCTL },
- { "test_sysctl_loop2.o", BPF_PROG_TYPE_CGROUP_SYSCTL },
-
- { "test_xdp_loop.o", BPF_PROG_TYPE_XDP },
- { "test_seg6_loop.o", BPF_PROG_TYPE_LWT_SEG6LOCAL },
- };
+static void scale_test(const char *file,
+ enum bpf_prog_type attach_type,
+ bool should_fail)
+{
libbpf_print_fn_t old_print_fn = NULL;
- int err, i;
+ int err;
if (env.verifier_stats) {
test__force_log();
old_print_fn = libbpf_set_print(libbpf_debug_print);
}
- for (i = 0; i < ARRAY_SIZE(tests); i++) {
- const struct scale_test_def *test = &tests[i];
-
- if (!test__start_subtest(test->file))
- continue;
-
- err = check_load(test->file, test->attach_type);
- CHECK_FAIL(err && !test->fails);
- }
+ err = check_load(file, attach_type);
+ if (should_fail)
+ ASSERT_ERR(err, "expect_error");
+ else
+ ASSERT_OK(err, "expect_success");
if (env.verifier_stats)
libbpf_set_print(old_print_fn);
}
+
+void test_verif_scale1()
+{
+ scale_test("test_verif_scale1.o", BPF_PROG_TYPE_SCHED_CLS, false);
+}
+
+void test_verif_scale2()
+{
+ scale_test("test_verif_scale2.o", BPF_PROG_TYPE_SCHED_CLS, false);
+}
+
+void test_verif_scale3()
+{
+ scale_test("test_verif_scale3.o", BPF_PROG_TYPE_SCHED_CLS, false);
+}
+
+void test_verif_scale_pyperf_global()
+{
+ scale_test("pyperf_global.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_pyperf_subprogs()
+{
+ scale_test("pyperf_subprogs.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_pyperf50()
+{
+ /* full unroll by llvm */
+ scale_test("pyperf50.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_pyperf100()
+{
+ /* full unroll by llvm */
+ scale_test("pyperf100.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_pyperf180()
+{
+ /* full unroll by llvm */
+ scale_test("pyperf180.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_pyperf600()
+{
+ /* partial unroll. llvm will unroll loop ~150 times.
+ * C loop count -> 600.
+ * Asm loop count -> 4.
+ * 16k insns in loop body.
+ * Total of 5 such loops. Total program size ~82k insns.
+ */
+ scale_test("pyperf600.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_pyperf600_nounroll()
+{
+ /* no unroll at all.
+ * C loop count -> 600.
+ * ASM loop count -> 600.
+ * ~110 insns in loop body.
+ * Total of 5 such loops. Total program size ~1500 insns.
+ */
+ scale_test("pyperf600_nounroll.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_loop1()
+{
+ scale_test("loop1.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_loop2()
+{
+ scale_test("loop2.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_loop3_fail()
+{
+ scale_test("loop3.o", BPF_PROG_TYPE_RAW_TRACEPOINT, true /* fails */);
+}
+
+void test_verif_scale_loop4()
+{
+ scale_test("loop4.o", BPF_PROG_TYPE_SCHED_CLS, false);
+}
+
+void test_verif_scale_loop5()
+{
+ scale_test("loop5.o", BPF_PROG_TYPE_SCHED_CLS, false);
+}
+
+void test_verif_scale_loop6()
+{
+ scale_test("loop6.o", BPF_PROG_TYPE_KPROBE, false);
+}
+
+void test_verif_scale_strobemeta()
+{
+ /* partial unroll. 19k insn in a loop.
+ * Total program size 20.8k insn.
+ * ~350k processed_insns
+ */
+ scale_test("strobemeta.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_strobemeta_nounroll1()
+{
+ /* no unroll, tiny loops */
+ scale_test("strobemeta_nounroll1.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_strobemeta_nounroll2()
+{
+ /* no unroll, tiny loops */
+ scale_test("strobemeta_nounroll2.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_strobemeta_subprogs()
+{
+ /* non-inlined subprogs */
+ scale_test("strobemeta_subprogs.o", BPF_PROG_TYPE_RAW_TRACEPOINT, false);
+}
+
+void test_verif_scale_sysctl_loop1()
+{
+ scale_test("test_sysctl_loop1.o", BPF_PROG_TYPE_CGROUP_SYSCTL, false);
+}
+
+void test_verif_scale_sysctl_loop2()
+{
+ scale_test("test_sysctl_loop2.o", BPF_PROG_TYPE_CGROUP_SYSCTL, false);
+}
+
+void test_verif_scale_xdp_loop()
+{
+ scale_test("test_xdp_loop.o", BPF_PROG_TYPE_XDP, false);
+}
+
+void test_verif_scale_seg6_loop()
+{
+ scale_test("test_seg6_loop.o", BPF_PROG_TYPE_LWT_SEG6LOCAL, false);
+}
Instead of using subtests in bpf_verif_scale selftest, turn each scale sub-test into its own test. Each subtest is compltely independent and just reuses a bit of common test running logic, so the conversion is trivial. For convenience, keep all of BPF verifier scale tests in one file. This conversion shaves off a significant amount of time when running test_progs in parallel mode. E.g., just running scale tests (-t verif_scale): BEFORE ====== Summary: 24/0 PASSED, 0 SKIPPED, 0 FAILED real 0m22.894s user 0m0.012s sys 0m22.797s AFTER ===== Summary: 24/0 PASSED, 0 SKIPPED, 0 FAILED real 0m12.044s user 0m0.024s sys 0m27.869s Ten second saving right there. test_progs -j is not yet ready to be turned on by default, unfortunately, and some tests fail almost every time, but this is a good improvement nevertheless. Ignoring few failures, here is sequential vs parallel run times when running all tests now: SEQUENTIAL ========== Summary: 206/953 PASSED, 4 SKIPPED, 0 FAILED real 1m5.625s user 0m4.211s sys 0m31.650s PARALLEL ======== Summary: 204/952 PASSED, 4 SKIPPED, 2 FAILED real 0m35.550s user 0m4.998s sys 0m39.890s Cc: Yucong Sun <sunyucong@gmail.com> Signed-off-by: Andrii Nakryiko <andrii@kernel.org> --- .../bpf/prog_tests/bpf_verif_scale.c | 220 ++++++++++++------ 1 file changed, 152 insertions(+), 68 deletions(-)