Merge tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next

Martin KaFai Lau says:

====================
pull-request: bpf-next 2024-11-14

We've added 9 non-merge commits during the last 4 day(s) which contain
a total of 3 files changed, 226 insertions(+), 84 deletions(-).

The main changes are:

1) Fixes to bpf_msg_push/pop_data and test_sockmap. The changes has
   dependency on the other changes in the bpf-next/net branch,
   from Zijian Zhang.

2) Drop netns codes from mptcp test. Reuse the common helpers in
   test_progs, from Geliang Tang.

* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next:
  bpf, sockmap: Fix sk_msg_reset_curr
  bpf, sockmap: Several fixes to bpf_msg_pop_data
  bpf, sockmap: Several fixes to bpf_msg_push_data
  selftests/bpf: Add more tests for test_txmsg_push_pop in test_sockmap
  selftests/bpf: Add push/pop checking for msg_verify_data in test_sockmap
  selftests/bpf: Fix total_bytes in msg_loop_rx in test_sockmap
  selftests/bpf: Fix SENDPAGE data logic in test_sockmap
  selftests/bpf: Add txmsg_pass to pull/push/pop in test_sockmap
  selftests/bpf: Drop netns helpers in mptcp
====================

Link: https://patch.msgid.link/20241114202832.3187927-1-martin.lau@linux.dev
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

2 files changed, 175 insertions, 47 deletions

diff --git a/tools/testing/selftests/bpf/prog_tests/mptcp.c b/tools/testing/selftests/bpf/prog_tests/mptcp.c
index be3cad2aff77..f8eb7f9d4fd2 100644
--- a/tools/testing/selftests/bpf/prog_tests/mptcp.c
+++ b/tools/testing/selftests/bpf/prog_tests/mptcp.c
@@ -69,24 +69,6 @@ struct mptcp_storage {
 	char ca_name[TCP_CA_NAME_MAX];
 };
 
-static struct nstoken *create_netns(void)
-{
-	SYS(fail, "ip netns add %s", NS_TEST);
-	SYS(fail, "ip -net %s link set dev lo up", NS_TEST);
-
-	return open_netns(NS_TEST);
-fail:
-	return NULL;
-}
-
-static void cleanup_netns(struct nstoken *nstoken)
-{
-	if (nstoken)
-		close_netns(nstoken);
-
-	SYS_NOFAIL("ip netns del %s", NS_TEST);
-}
-
 static int start_mptcp_server(int family, const char *addr_str, __u16 port,
 			      int timeout_ms)
 {
@@ -206,15 +188,15 @@ out:
 
 static void test_base(void)
 {
-	struct nstoken *nstoken = NULL;
+	struct netns_obj *netns = NULL;
 	int server_fd, cgroup_fd;
 
 	cgroup_fd = test__join_cgroup("/mptcp");
 	if (!ASSERT_GE(cgroup_fd, 0, "test__join_cgroup"))
 		return;
 
-	nstoken = create_netns();
-	if (!ASSERT_OK_PTR(nstoken, "create_netns"))
+	netns = netns_new(NS_TEST, true);
+	if (!ASSERT_OK_PTR(netns, "netns_new"))
 		goto fail;
 
 	/* without MPTCP */
@@ -237,7 +219,7 @@ with_mptcp:
 	close(server_fd);
 
 fail:
-	cleanup_netns(nstoken);
+	netns_free(netns);
 	close(cgroup_fd);
 }
 
@@ -322,21 +304,21 @@ out:
 
 static void test_mptcpify(void)
 {
-	struct nstoken *nstoken = NULL;
+	struct netns_obj *netns = NULL;
 	int cgroup_fd;
 
 	cgroup_fd = test__join_cgroup("/mptcpify");
 	if (!ASSERT_GE(cgroup_fd, 0, "test__join_cgroup"))
 		return;
 
-	nstoken = create_netns();
-	if (!ASSERT_OK_PTR(nstoken, "create_netns"))
+	netns = netns_new(NS_TEST, true);
+	if (!ASSERT_OK_PTR(netns, "netns_new"))
 		goto fail;
 
 	ASSERT_OK(run_mptcpify(cgroup_fd), "run_mptcpify");
 
 fail:
-	cleanup_netns(nstoken);
+	netns_free(netns);
 	close(cgroup_fd);
 }
 
@@ -414,7 +396,7 @@ close_server:
 static void test_subflow(void)
 {
 	struct mptcp_subflow *skel;
-	struct nstoken *nstoken;
+	struct netns_obj *netns;
 	int cgroup_fd;
 
 	cgroup_fd = test__join_cgroup("/mptcp_subflow");
@@ -437,8 +419,8 @@ static void test_subflow(void)
 	if (!ASSERT_OK_PTR(skel->links._getsockopt_subflow, "attach _getsockopt_subflow"))
 		goto skel_destroy;
 
-	nstoken = create_netns();
-	if (!ASSERT_OK_PTR(nstoken, "create_netns: mptcp_subflow"))
+	netns = netns_new(NS_TEST, true);
+	if (!ASSERT_OK_PTR(netns, "netns_new: mptcp_subflow"))
 		goto skel_destroy;
 
 	if (endpoint_init("subflow") < 0)
@@ -447,7 +429,7 @@ static void test_subflow(void)
 	run_subflow();
 
 close_netns:
-	cleanup_netns(nstoken);
+	netns_free(netns);
 skel_destroy:
 	mptcp_subflow__destroy(skel);
 close_cgroup:
diff --git a/tools/testing/selftests/bpf/test_sockmap.c b/tools/testing/selftests/bpf/test_sockmap.c
index 075c93ed143e..e5c7ecbe57e3 100644
--- a/tools/testing/selftests/bpf/test_sockmap.c
+++ b/tools/testing/selftests/bpf/test_sockmap.c
@@ -88,6 +88,10 @@ int ktls;
 int peek_flag;
 int skb_use_parser;
 int txmsg_omit_skb_parser;
+int verify_push_start;
+int verify_push_len;
+int verify_pop_start;
+int verify_pop_len;
 
 static const struct option long_options[] = {
 	{"help",	no_argument,		NULL, 'h' },
@@ -420,16 +424,18 @@ static int msg_loop_sendpage(int fd, int iov_length, int cnt,
 {
 	bool drop = opt->drop_expected;
 	unsigned char k = 0;
+	int i, j, fp;
 	FILE *file;
-	int i, fp;
 
 	file = tmpfile();
 	if (!file) {
 		perror("create file for sendpage");
 		return 1;
 	}
-	for (i = 0; i < iov_length * cnt; i++, k++)
-		fwrite(&k, sizeof(char), 1, file);
+	for (i = 0; i < cnt; i++, k = 0) {
+		for (j = 0; j < iov_length; j++, k++)
+			fwrite(&k, sizeof(char), 1, file);
+	}
 	fflush(file);
 	fseek(file, 0, SEEK_SET);
 
@@ -512,12 +518,41 @@ unwind_iov:
 	return -ENOMEM;
 }
 
-/* TODO: Add verification logic for push, pull and pop data */
+/* In push or pop test, we need to do some calculations for msg_verify_data */
+static void msg_verify_date_prep(void)
+{
+	int push_range_end = txmsg_start_push + txmsg_end_push - 1;
+	int pop_range_end = txmsg_start_pop + txmsg_pop - 1;
+
+	if (txmsg_end_push && txmsg_pop &&
+	    txmsg_start_push <= pop_range_end && txmsg_start_pop <= push_range_end) {
+		/* The push range and the pop range overlap */
+		int overlap_len;
+
+		verify_push_start = txmsg_start_push;
+		verify_pop_start = txmsg_start_pop;
+		if (txmsg_start_push < txmsg_start_pop)
+			overlap_len = min(push_range_end - txmsg_start_pop + 1, txmsg_pop);
+		else
+			overlap_len = min(pop_range_end - txmsg_start_push + 1, txmsg_end_push);
+		verify_push_len = max(txmsg_end_push - overlap_len, 0);
+		verify_pop_len = max(txmsg_pop - overlap_len, 0);
+	} else {
+		/* Otherwise */
+		verify_push_start = txmsg_start_push;
+		verify_pop_start = txmsg_start_pop;
+		verify_push_len = txmsg_end_push;
+		verify_pop_len = txmsg_pop;
+	}
+}
+
 static int msg_verify_data(struct msghdr *msg, int size, int chunk_sz,
-				 unsigned char *k_p, int *bytes_cnt_p)
+			   unsigned char *k_p, int *bytes_cnt_p,
+			   int *check_cnt_p, int *push_p)
 {
-	int i, j, bytes_cnt = *bytes_cnt_p;
+	int bytes_cnt = *bytes_cnt_p, check_cnt = *check_cnt_p, push = *push_p;
 	unsigned char k = *k_p;
+	int i, j;
 
 	for (i = 0, j = 0; i < msg->msg_iovlen && size; i++, j = 0) {
 		unsigned char *d = msg->msg_iov[i].iov_base;
@@ -536,6 +571,37 @@ static int msg_verify_data(struct msghdr *msg, int size, int chunk_sz,
 		}
 
 		for (; j < msg->msg_iov[i].iov_len && size; j++) {
+			if (push > 0 &&
+			    check_cnt == verify_push_start + verify_push_len - push) {
+				int skipped;
+revisit_push:
+				skipped = push;
+				if (j + push >= msg->msg_iov[i].iov_len)
+					skipped = msg->msg_iov[i].iov_len - j;
+				push -= skipped;
+				size -= skipped;
+				j += skipped - 1;
+				check_cnt += skipped;
+				continue;
+			}
+
+			if (verify_pop_len > 0 && check_cnt == verify_pop_start) {
+				bytes_cnt += verify_pop_len;
+				check_cnt += verify_pop_len;
+				k += verify_pop_len;
+
+				if (bytes_cnt == chunk_sz) {
+					k = 0;
+					bytes_cnt = 0;
+					check_cnt = 0;
+					push = verify_push_len;
+				}
+
+				if (push > 0 &&
+				    check_cnt == verify_push_start + verify_push_len - push)
+					goto revisit_push;
+			}
+
 			if (d[j] != k++) {
 				fprintf(stderr,
 					"detected data corruption @iov[%i]:%i %02x != %02x, %02x ?= %02x\n",
@@ -543,15 +609,20 @@ static int msg_verify_data(struct msghdr *msg, int size, int chunk_sz,
 				return -EDATAINTEGRITY;
 			}
 			bytes_cnt++;
+			check_cnt++;
 			if (bytes_cnt == chunk_sz) {
 				k = 0;
 				bytes_cnt = 0;
+				check_cnt = 0;
+				push = verify_push_len;
 			}
 			size--;
 		}
 	}
 	*k_p = k;
 	*bytes_cnt_p = bytes_cnt;
+	*check_cnt_p = check_cnt;
+	*push_p = push;
 	return 0;
 }
 
@@ -604,12 +675,14 @@ static int msg_loop(int fd, int iov_count, int iov_length, int cnt,
 		}
 		clock_gettime(CLOCK_MONOTONIC, &s->end);
 	} else {
+		float total_bytes, txmsg_pop_total, txmsg_push_total;
 		int slct, recvp = 0, recv, max_fd = fd;
-		float total_bytes, txmsg_pop_total;
 		int fd_flags = O_NONBLOCK;
 		struct timeval timeout;
 		unsigned char k = 0;
 		int bytes_cnt = 0;
+		int check_cnt = 0;
+		int push = 0;
 		fd_set w;
 
 		fcntl(fd, fd_flags);
@@ -623,12 +696,22 @@ static int msg_loop(int fd, int iov_count, int iov_length, int cnt,
 		 * This is really only useful for testing edge cases in code
 		 * paths.
 		 */
-		total_bytes = (float)iov_count * (float)iov_length * (float)cnt;
-		if (txmsg_apply)
+		total_bytes = (float)iov_length * (float)cnt;
+		if (!opt->sendpage)
+			total_bytes *= (float)iov_count;
+		if (txmsg_apply) {
+			txmsg_push_total = txmsg_end_push * (total_bytes / txmsg_apply);
 			txmsg_pop_total = txmsg_pop * (total_bytes / txmsg_apply);
-		else
+		} else {
+			txmsg_push_total = txmsg_end_push * cnt;
 			txmsg_pop_total = txmsg_pop * cnt;
+		}
+		total_bytes += txmsg_push_total;
 		total_bytes -= txmsg_pop_total;
+		if (data) {
+			msg_verify_date_prep();
+			push = verify_push_len;
+		}
 		err = clock_gettime(CLOCK_MONOTONIC, &s->start);
 		if (err < 0)
 			perror("recv start time");
@@ -701,10 +784,11 @@ static int msg_loop(int fd, int iov_count, int iov_length, int cnt,
 
 			if (data) {
 				int chunk_sz = opt->sendpage ?
-						iov_length * cnt :
+						iov_length :
 						iov_length * iov_count;
 
-				errno = msg_verify_data(&msg, recv, chunk_sz, &k, &bytes_cnt);
+				errno = msg_verify_data(&msg, recv, chunk_sz, &k, &bytes_cnt,
+							&check_cnt, &push);
 				if (errno) {
 					perror("data verify msg failed");
 					goto out_errno;
@@ -714,7 +798,9 @@ static int msg_loop(int fd, int iov_count, int iov_length, int cnt,
 								recvp,
 								chunk_sz,
 								&k,
-								&bytes_cnt);
+								&bytes_cnt,
+								&check_cnt,
+								&push);
 					if (errno) {
 						perror("data verify msg_peek failed");
 						goto out_errno;
@@ -796,8 +882,6 @@ static int sendmsg_test(struct sockmap_options *opt)
 
 	rxpid = fork();
 	if (rxpid == 0) {
-		if (txmsg_pop || txmsg_start_pop)
-			iov_buf -= (txmsg_pop - txmsg_start_pop + 1);
 		if (opt->drop_expected || txmsg_ktls_skb_drop)
 			_exit(0);
 
@@ -1466,8 +1550,8 @@ static void test_send_many(struct sockmap_options *opt, int cgrp)
 
 static void test_send_large(struct sockmap_options *opt, int cgrp)
 {
-	opt->iov_length = 256;
-	opt->iov_count = 1024;
+	opt->iov_length = 8192;
+	opt->iov_count = 32;
 	opt->rate = 2;
 	test_exec(cgrp, opt);
 }
@@ -1596,11 +1680,13 @@ static void test_txmsg_cork_hangs(int cgrp, struct sockmap_options *opt)
 static void test_txmsg_pull(int cgrp, struct sockmap_options *opt)
 {
 	/* Test basic start/end */
+	txmsg_pass = 1;
 	txmsg_start = 1;
 	txmsg_end = 2;
 	test_send(opt, cgrp);
 
 	/* Test >4k pull */
+	txmsg_pass = 1;
 	txmsg_start = 4096;
 	txmsg_end = 9182;
 	test_send_large(opt, cgrp);
@@ -1628,12 +1714,16 @@ static void test_txmsg_pull(int cgrp, struct sockmap_options *opt)
 
 static void test_txmsg_pop(int cgrp, struct sockmap_options *opt)
 {
+	bool data = opt->data_test;
+
 	/* Test basic pop */
+	txmsg_pass = 1;
 	txmsg_start_pop = 1;
 	txmsg_pop = 2;
 	test_send_many(opt, cgrp);
 
 	/* Test pop with >4k */
+	txmsg_pass = 1;
 	txmsg_start_pop = 4096;
 	txmsg_pop = 4096;
 	test_send_large(opt, cgrp);
@@ -1644,6 +1734,12 @@ static void test_txmsg_pop(int cgrp, struct sockmap_options *opt)
 	txmsg_pop = 2;
 	test_send_many(opt, cgrp);
 
+	/* TODO: Test for pop + cork should be different,
+	 * - It makes the layout of the received data difficult
+	 * - It makes it hard to calculate the total_bytes in the recvmsg
+	 * Temporarily skip the data integrity test for this case now.
+	 */
+	opt->data_test = false;
 	/* Test pop + cork */
 	txmsg_redir = 0;
 	txmsg_cork = 512;
@@ -1657,16 +1753,21 @@ static void test_txmsg_pop(int cgrp, struct sockmap_options *opt)
 	txmsg_start_pop = 1;
 	txmsg_pop = 2;
 	test_send_many(opt, cgrp);
+	opt->data_test = data;
 }
 
 static void test_txmsg_push(int cgrp, struct sockmap_options *opt)
 {
+	bool data = opt->data_test;
+
 	/* Test basic push */
+	txmsg_pass = 1;
 	txmsg_start_push = 1;
 	txmsg_end_push = 1;
 	test_send(opt, cgrp);
 
 	/* Test push 4kB >4k */
+	txmsg_pass = 1;
 	txmsg_start_push = 4096;
 	txmsg_end_push = 4096;
 	test_send_large(opt, cgrp);
@@ -1677,19 +1778,64 @@ static void test_txmsg_push(int cgrp, struct sockmap_options *opt)
 	txmsg_end_push = 2;
 	test_send_many(opt, cgrp);
 
+	/* TODO: Test for push + cork should be different,
+	 * - It makes the layout of the received data difficult
+	 * - It makes it hard to calculate the total_bytes in the recvmsg
+	 * Temporarily skip the data integrity test for this case now.
+	 */
+	opt->data_test = false;
 	/* Test push + cork */
 	txmsg_redir = 0;
 	txmsg_cork = 512;
 	txmsg_start_push = 1;
 	txmsg_end_push = 2;
 	test_send_many(opt, cgrp);
+	opt->data_test = data;
 }
 
 static void test_txmsg_push_pop(int cgrp, struct sockmap_options *opt)
 {
+	/* Test push/pop range overlapping */
+	txmsg_pass = 1;
+	txmsg_start_push = 1;
+	txmsg_end_push = 10;
+	txmsg_start_pop = 5;
+	txmsg_pop = 4;
+	test_send_large(opt, cgrp);
+
+	txmsg_pass = 1;
 	txmsg_start_push = 1;
 	txmsg_end_push = 10;
 	txmsg_start_pop = 5;
+	txmsg_pop = 16;
+	test_send_large(opt, cgrp);
+
+	txmsg_pass = 1;
+	txmsg_start_push = 5;
+	txmsg_end_push = 4;
+	txmsg_start_pop = 1;
+	txmsg_pop = 10;
+	test_send_large(opt, cgrp);
+
+	txmsg_pass = 1;
+	txmsg_start_push = 5;
+	txmsg_end_push = 16;
+	txmsg_start_pop = 1;
+	txmsg_pop = 10;
+	test_send_large(opt, cgrp);
+
+	/* Test push/pop range non-overlapping */
+	txmsg_pass = 1;
+	txmsg_start_push = 1;
+	txmsg_end_push = 10;
+	txmsg_start_pop = 16;
+	txmsg_pop = 4;
+	test_send_large(opt, cgrp);
+
+	txmsg_pass = 1;
+	txmsg_start_push = 16;
+	txmsg_end_push = 10;
+	txmsg_start_pop = 5;
 	txmsg_pop = 4;
 	test_send_large(opt, cgrp);
 }