Re: [PATCH bpf-next v2 1/2] selftests/bpf: Convert CHECK macros to ASSERT_* macros in bpf_iter

From: Yonghong Song
Date: Fri Oct 27 2023 - 00:43:10 EST



On 10/25/23 9:33 PM, Kui-Feng Lee wrote:


On 10/25/23 19:03, Yuran Pereira wrote:
As it was pointed out by Yonghong Song [1], in the bpf selftests the use
of the ASSERT_* series of macros is preferred over the CHECK macro.
This patch replaces all CHECK calls in bpf_iter with the appropriate
ASSERT_* macros.

[1] https://lore.kernel.org/lkml/0a142924-633c-44e6-9a92-2dc019656bf2@xxxxxxxxx

Suggested-by: Yonghong Song <yonghong.song@xxxxxxxxx>
Signed-off-by: Yuran Pereira <yuran.pereira@xxxxxxxxxxx>
---
  .../selftests/bpf/prog_tests/bpf_iter.c       | 82 +++++++++----------
  1 file changed, 39 insertions(+), 43 deletions(-)

diff --git a/tools/testing/selftests/bpf/prog_tests/bpf_iter.c b/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
index 1f02168103dd..7db6972ed952 100644
--- a/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
+++ b/tools/testing/selftests/bpf/prog_tests/bpf_iter.c
@@ -34,8 +34,6 @@
  #include "bpf_iter_ksym.skel.h"
  #include "bpf_iter_sockmap.skel.h"
  -static int duration;
-
  static void test_btf_id_or_null(void)
  {
      struct bpf_iter_test_kern3 *skel;
@@ -64,7 +62,7 @@ static void do_dummy_read_opts(struct bpf_program *prog, struct bpf_iter_attach_
      /* not check contents, but ensure read() ends without error */
      while ((len = read(iter_fd, buf, sizeof(buf))) > 0)
          ;
-    CHECK(len < 0, "read", "read failed: %s\n", strerror(errno));
+    ASSERT_GE(len, 0, "read");
        close(iter_fd);
  @@ -413,7 +411,7 @@ static int do_btf_read(struct bpf_iter_task_btf *skel)
          goto free_link;
      }
  -    if (CHECK(err < 0, "read", "read failed: %s\n", strerror(errno)))
+    if (!ASSERT_GE(err, 0, "read"))
          goto free_link;
        ASSERT_HAS_SUBSTR(taskbuf, "(struct task_struct)",
@@ -526,11 +524,11 @@ static int do_read_with_fd(int iter_fd, const char *expected,
      start = 0;
      while ((len = read(iter_fd, buf + start, read_buf_len)) > 0) {
          start += len;
-        if (CHECK(start >= 16, "read", "read len %d\n", len))
+        if (!ASSERT_LT(start, 16, "read"))
              return -1;
          read_buf_len = read_one_char ? 1 : 16 - start;
      }
-    if (CHECK(len < 0, "read", "read failed: %s\n", strerror(errno)))
+    if (!ASSERT_GE(len, 0, "read"))
          return -1;
        if (!ASSERT_STREQ(buf, expected, "read"))
@@ -571,8 +569,7 @@ static int do_read(const char *path, const char *expected)
      int err, iter_fd;
        iter_fd = open(path, O_RDONLY);
-    if (CHECK(iter_fd < 0, "open", "open %s failed: %s\n",
-          path, strerror(errno)))
+    if (!ASSERT_GE(iter_fd, 0, "open"))
          return -1;
        err = do_read_with_fd(iter_fd, expected, false);
@@ -600,7 +597,7 @@ static void test_file_iter(void)
      unlink(path);
        err = bpf_link__pin(link, path);
-    if (CHECK(err, "pin_iter", "pin_iter to %s failed: %d\n", path, err))
+    if (!ASSERT_OK(err, "pin_iter"))
          goto free_link;
        err = do_read(path, "abcd");
@@ -651,12 +648,10 @@ static void test_overflow(bool test_e2big_overflow, bool ret1)
       * overflow and needs restart.
       */
      map1_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, 4, 8, 1, NULL);
-    if (CHECK(map1_fd < 0, "bpf_map_create",
-          "map_creation failed: %s\n", strerror(errno)))
+    if (!ASSERT_GE(map1_fd, 0, "bpf_map_create"))
          goto out;
      map2_fd = bpf_map_create(BPF_MAP_TYPE_ARRAY, NULL, 4, 8, 1, NULL);
-    if (CHECK(map2_fd < 0, "bpf_map_create",
-          "map_creation failed: %s\n", strerror(errno)))
+    if (!ASSERT_GE(map2_fd, 0, "bpf_map_create"))
          goto free_map1;
        /* bpf_seq_printf kernel buffer is 8 pages, so one map
@@ -685,14 +680,12 @@ static void test_overflow(bool test_e2big_overflow, bool ret1)
      /* setup filtering map_id in bpf program */
      map_info_len = sizeof(map_info);
      err = bpf_map_get_info_by_fd(map1_fd, &map_info, &map_info_len);
-    if (CHECK(err, "get_map_info", "get map info failed: %s\n",
-          strerror(errno)))
+    if (!ASSERT_OK(err, "get_map_info"))
          goto free_map2;
      skel->bss->map1_id = map_info.id;
        err = bpf_map_get_info_by_fd(map2_fd, &map_info, &map_info_len);
-    if (CHECK(err, "get_map_info", "get map info failed: %s\n",
-          strerror(errno)))
+    if (!ASSERT_OK(err, "get_map_info"))
          goto free_map2;
      skel->bss->map2_id = map_info.id;
  @@ -714,16 +707,15 @@ static void test_overflow(bool test_e2big_overflow, bool ret1)
          while ((len = read(iter_fd, buf, expected_read_len)) > 0)
              total_read_len += len;
  -        CHECK(len != -1 || errno != E2BIG, "read",
-              "expected ret -1, errno E2BIG, but get ret %d, error %s\n",
-              len, strerror(errno));
+        if (!ASSERT_EQ(len, -1, "read"))
+            goto free_buf;
+        ASSERT_EQ(errno, E2BIG, "read");

I think you can just do

  ASSERT_EQ(len, -1, "read");
  ASSERT_EQ(errno, E2BG, "read");

without a check here.
WDYT?

Many recent selftests have ASSERT_* similar to what Kui-Feng is suggested.
So I think it is okay to do adjustment like it. The same for some other suggestions
below.

But since this patch intends to convert CHECK to ASSERT_*, so other possible
'optimizations' like above ASSERT_EQ can stay as is since they do not really
affect functionality.

In the next revision, please put three patches 0/2, 1/2 and 2/2 together as a single
patch set. Thanks.


          goto free_buf;
      } else if (!ret1) {
          while ((len = read(iter_fd, buf, expected_read_len)) > 0)
              total_read_len += len;
  -        if (CHECK(len < 0, "read", "read failed: %s\n",
-              strerror(errno)))
+        if (!ASSERT_GE(len, 0, "read"))
              goto free_buf;
      } else {
          do {
@@ -732,8 +724,7 @@ static void test_overflow(bool test_e2big_overflow, bool ret1)
                  total_read_len += len;
          } while (len > 0 || len == -EAGAIN);
  -        if (CHECK(len < 0, "read", "read failed: %s\n",
-              strerror(errno)))
+        if (!ASSERT_GE(len, 0, "read"))
              goto free_buf;
      }
  @@ -836,7 +827,7 @@ static void test_bpf_hash_map(void)
      /* do some tests */
      while ((len = read(iter_fd, buf, sizeof(buf))) > 0)
          ;
-    if (CHECK(len < 0, "read", "read failed: %s\n", strerror(errno)))
+    if (!ASSERT_GE(len, 0, "read"))
          goto close_iter;
        /* test results */
@@ -917,7 +908,7 @@ static void test_bpf_percpu_hash_map(void)
      /* do some tests */
      while ((len = read(iter_fd, buf, sizeof(buf))) > 0)
          ;
-    if (CHECK(len < 0, "read", "read failed: %s\n", strerror(errno)))
+    if (!ASSERT_GE(len, 0, "read"))
          goto close_iter;
        /* test results */
@@ -983,17 +974,15 @@ static void test_bpf_array_map(void)
      start = 0;
      while ((len = read(iter_fd, buf + start, sizeof(buf) - start)) > 0)
          start += len;
-    if (CHECK(len < 0, "read", "read failed: %s\n", strerror(errno)))
+    if (!ASSERT_GE(len, 0, "read"))
          goto close_iter;
        /* test results */
      res_first_key = *(__u32 *)buf;
      res_first_val = *(__u64 *)(buf + sizeof(__u32));
-    if (CHECK(res_first_key != 0 || res_first_val != first_val,
-          "bpf_seq_write",
-          "seq_write failure: first key %u vs expected 0, "
-          " first value %llu vs expected %llu\n",
-          res_first_key, res_first_val, first_val))
+    if (!ASSERT_EQ(res_first_key, 0, "bpf_seq_write"))
+        goto close_iter;
+    else if (!ASSERT_EQ(res_first_val, first_val, "bpf_seq_write"))
          goto close_iter;

Similiar here!

 if (!ASSERT_EQ(...) ||
     !ASSERT_EQ(...))
      goto close_iter;

[...]