[LTP] [RFC PATCH v5 0/3] Simplify network setup

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Petr Vorel]

Hi,

another version of tst_net_vars network setup helper.
v4->v5:
* Clenup unneded variables (reverse IP related).
* Don't require to have mask in IP address. Use netlink to detect prefix when
  address is presented on some of the NIC on the system or fallback to default
  ones (24 for IPv4, 64 for IPv6).
* Fix detecting wrong subnet with network not round up by 8 (IPv4 resp 16
  (IPv6). E.g.: problem IP addressed like 10.2.0.1/22 10.2.1.1/22.
* Add more environment variables (e.g. prefix, whole IP address)
* Rewritten (again) unused network generation to fix overlap with used IP addresses.
* Simplify tst_ipaddr() (in testcases/lib/test_net.sh).
* Use tst_brk() to exit on error.
* Cleanup warnings and formatting.

There is still requirement network prefix to be < 8 (IPv4) resp. 16 (IPv6) as
then IPV{4,6}_NETWORK variables would be empty, which is not expected by tests
(they join it with address separators with {L,R}HOST_IPV{4,6}_HOST. Even we got
rid of reverse related variables we still need IPV{4,6}_NETWORK variables for
some tests.

BTW: I probably have source lines > 80 chars. My Vim doesn't highlight
anything, but checkpatch.pl complain. I might have wrong vim configuration.

Petr Vorel (3):
  gitignore: Prefixing filenames
  network: Add tool for setup IP variables
  network: Use tst_net_vars to set up IPv4 and IPv6 related variables

 testcases/lib/.gitignore     |  13 +-
 testcases/lib/Makefile       |   2 +-
 testcases/lib/test_net.sh    |  43 +--
 testcases/lib/tst_net_vars.c | 760 +++++++++++++++++++++++++++++++++++++++++++
 4 files changed, 781 insertions(+), 37 deletions(-)
 create mode 100644 testcases/lib/tst_net_vars.c

-- 
2.12.2

[LTP] [RFC PATCH v5 1/3] gitignore: Prefixing filenames

[Petr Vorel]

Signed-off-by: Petr Vorel <pvorel@suse.cz>
---
 testcases/lib/.gitignore | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/testcases/lib/.gitignore b/testcases/lib/.gitignore
index 920817cac..522889bed 100644
--- a/testcases/lib/.gitignore
+++ b/testcases/lib/.gitignore
@@ -1,6 +1,6 @@
-tst_sleep
-tst_random
-tst_checkpoint
-tst_rod
-tst_kvcmp
-tst_device
+/tst_sleep
+/tst_random
+/tst_checkpoint
+/tst_rod
+/tst_kvcmp
+/tst_device
-- 
2.12.2

[LTP] [RFC PATCH v5 2/3] network: Add tool for setup IP variables

[Petr Vorel]

These environment variables are exported by setup_network_variables:
IPV4_LHOST: IPv4 address of the local host
IPV4_RHOST: IPv4 address of the local host
IPV4_NETWORK: IPv4 common part of IPV4_LHOST and IPV4_RHOST
LHOST_IPV4_HOST IPv4 unique part of IPV4_LHOST
RHOST_IPV4_HOST IPv4 unique part of IPV4_RHOST
IPV4_NET16_UNUSED: IPv4 16 bit unused subnet
IPV4_PREFIX: IPv4 prefix
IPV6_LHOST: IPv6 address of local host
IPV6_RHOST: IPv6 address of the remote host
IPV6_NETWORK: IPv6 common part of IPV6_LHOST and IPV6_RHOST
LHOST_IPV6_HOST: IPv6 unique part of IPV6_LHOST
RHOST_IPV6_HOST: IPv6 unique part of IPV6_RHOST
IPV6_NET32_UNUSED: IPv6 32 bit unused subnet
IPV6_PREFIX: IPv6 prefix

Signed-off-by: Petr Vorel <pvorel@suse.cz>
---
 testcases/lib/.gitignore     |   1 +
 testcases/lib/Makefile       |   2 +-
 testcases/lib/tst_net_vars.c | 760 +++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 762 insertions(+), 1 deletion(-)
 create mode 100644 testcases/lib/tst_net_vars.c

diff --git a/testcases/lib/.gitignore b/testcases/lib/.gitignore
index 522889bed..e1274df64 100644
--- a/testcases/lib/.gitignore
+++ b/testcases/lib/.gitignore
@@ -4,3 +4,4 @@
 /tst_rod
 /tst_kvcmp
 /tst_device
+/tst_net_vars
diff --git a/testcases/lib/Makefile b/testcases/lib/Makefile
index 1127a59fe..49ef2ec01 100644
--- a/testcases/lib/Makefile
+++ b/testcases/lib/Makefile
@@ -27,6 +27,6 @@ include $(top_srcdir)/include/mk/testcases.mk
 INSTALL_TARGETS		:= *.sh
 
 MAKE_TARGETS		:= tst_sleep tst_random tst_checkpoint tst_rod tst_kvcmp\
-			   tst_device
+			   tst_device tst_net_vars
 
 include $(top_srcdir)/include/mk/generic_leaf_target.mk
diff --git a/testcases/lib/tst_net_vars.c b/testcases/lib/tst_net_vars.c
new file mode 100644
index 000000000..a57f84f31
--- /dev/null
+++ b/testcases/lib/tst_net_vars.c
@@ -0,0 +1,760 @@
+/*
+ * Copyright (c) 2017 Petr Vorel <pvorel@suse.cz>
+ * Copyright (c) 1997-2015 Red Hat, Inc. All rights reserved.
+ * Copyright (c) 2011-2013 Rich Felker, et al.
+ *
+ * This program is free software: you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation, either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <arpa/inet.h>
+#include <arpa/nameser.h>
+#include <linux/rtnetlink.h>
+#include <errno.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <time.h>
+#include <stdarg.h>
+
+#define TST_NO_DEFAULT_MAIN
+#include "tst_test.h"
+
+#define FLAG_GET_NETWORK 1
+#define FLAG_GET_HOST 2
+
+#define DEFAULT_IPV4_PREFIX 24
+#define DEFAULT_IPV6_PREFIX 64
+
+#define BASE_IPV4_PREFIX 8
+#define BASE_IPV6_PREFIX 16
+
+#define DEFAULT_IPV4_UNUSED_PART1 10
+#define DEFAULT_IPV6_UNUSED_PART1 0xfd
+
+#define DEFAULT_IPV4_UNUSED_PART2 23
+#define DEFAULT_IPV6_UNUSED_PART2 0x23
+
+#define PROGRAM_NAME "tst_net_vars"
+
+typedef struct ltp_net_variables {
+	char *ipv4_lhost;
+	char *ipv4_rhost;
+	char *ipv4_network;
+	char *lhost_ipv4_host;
+	char *rhost_ipv4_host;
+	char *ipv6_lhost;
+	char *ipv6_rhost;
+	char *ipv6_network;
+	char *lhost_ipv6_host;
+	char *rhost_ipv6_host;
+	char *ipv4_net16_unused;
+	char *ipv6_net32_unused;
+	int ipv4_prefix;
+	int ipv6_prefix;
+} ltp_net_variables;
+
+static ltp_net_variables vars;
+
+static void usage(void)
+{
+	fprintf(stderr, "USAGE:\n"
+		"%s IPv4/PREFIX IPv4/PREFIX\n"
+		"%s IPv6/PREFIX IPv6/PREFIX\n\n"
+		"%s IPv4 IPv4\n"
+		"%s IPv6 IPv6\n\n"
+		"IP addresses must be different, within the same subnet.\n"
+		"Prefixes must be the same.\n"
+		"IPv4 prefixes must be in range <8, 31>.\n"
+		"IPv6 prefixes must be <16, 127>.\n"
+		"Default IPv4 prefix: %d.\n"
+		"Default IPv6 prefix: %d.\n\n"
+		"Exported variables:\n"
+		"IPV4_LHOST: IPv4 address of the local host\n"
+		"IPV4_RHOST: IPv4 address of the local host\n"
+		"IPV4_NETWORK: IPv4 common part of IPV4_LHOST and IPV4_RHOST\n"
+		"LHOST_IPV4_HOST IPv4 unique part of IPV4_LHOST\n"
+		"RHOST_IPV4_HOST IPv4 unique part of IPV4_RHOST\n"
+		"IPV4_NET16_UNUSED: IPv4 16 bit unused subnet\n"
+		"IPV4_PREFIX: IPv4 prefix\n"
+		"IPV6_LHOST: IPv6 address of local host\n"
+		"IPV6_RHOST: IPv6 address of the remote host\n"
+		"IPV6_NETWORK: IPv6 common part of IPV6_LHOST and IPV6_RHOST\n"
+		"LHOST_IPV6_HOST: IPv6 unique part of IPV6_LHOST\n"
+		"RHOST_IPV6_HOST: IPv6 unique part of IPV6_RHOST\n"
+		"IPV6_NET32_UNUSED: IPv6 32 bit unused subnet\n"
+		"IPV6_PREFIX: IPv6 prefix\n",
+		PROGRAM_NAME, PROGRAM_NAME, PROGRAM_NAME, PROGRAM_NAME,
+		DEFAULT_IPV4_PREFIX, DEFAULT_IPV6_PREFIX);
+}
+
+static int get_in_addr(const char *ip_str, struct in_addr *ip)
+{
+	if (inet_pton(AF_INET, ip_str, ip) <= 0) {
+		fprintf(stderr, "Bad IPv4 address: '%s'\n", ip_str);
+		return -1;
+	}
+
+	return 0;
+}
+
+static int get_in6_addr(const char *ip_str, struct in6_addr *ip6)
+{
+	if (inet_pton(AF_INET6, ip_str, ip6) <= 0) {
+		fprintf(stderr, "bad IPv6 address: '%s'\n", ip_str);
+		return -1;
+	}
+	return 0;
+}
+
+/*
+ * Function prefix2mask is from ipcalc project, ipcalc.c.
+ */
+struct in_addr prefix2mask(int prefix)
+{
+	struct in_addr mask;
+
+	memset(&mask, 0, sizeof(mask));
+	if (prefix)
+		mask.s_addr = htonl(~((1 << (32 - prefix)) - 1));
+	else
+		mask.s_addr = htonl(0);
+
+	return mask;
+}
+
+/*
+ * Function calc_network is based on ipcalc project,
+ * calc_network/ipcalc.c.
+ */
+struct in_addr calc_network(const struct in_addr *ip, struct in_addr *mask)
+{
+	struct in_addr network;
+
+	memset(&network, 0, sizeof(network));
+	network.s_addr = ip->s_addr & mask->s_addr;
+
+	return network;
+}
+
+static int is_in_subnet_ipv4(const struct in_addr *network,
+	const struct in_addr *mask, const struct in_addr *ip)
+{
+	return (ip->s_addr & mask->s_addr) ==
+		(network->s_addr & mask->s_addr);
+}
+
+int is_in_subnet_ipv6(const struct in6_addr *network,
+	const struct in6_addr *mask, const struct in6_addr *ip6)
+{
+	unsigned int i;
+
+	for (i = 0; i < sizeof(struct in6_addr) / sizeof(int); i++) {
+		if (((((int *) ip6)[i] & ((int *) mask)[i])) !=
+			(((int *) network)[i] & ((int *) mask)[i]))
+			return 0;
+	}
+
+	return 1;
+}
+
+/*
+ * For unused network we use
+ * DEFAULT_IPV4_UNUSED_PART1:DEFAULT_IPV4_UNUSED_PART2 or
+ * {DEFAULT_IPV4_UNUSED_PART1}.XY, when there is a collision with IP.
+ */
+static char *get_ipv4_net16_unused(const struct in_addr *ip,
+	unsigned int prefix)
+{
+	struct in_addr mask, network;
+	char buf[128], net_unused[128];
+
+	mask = prefix2mask(prefix);
+	network = calc_network(ip, &mask);
+
+	sprintf(net_unused, "%d.%d", DEFAULT_IPV4_UNUSED_PART1,
+			DEFAULT_IPV4_UNUSED_PART2);
+	sprintf(buf, "%s.0.0", net_unused);
+
+	if (get_in_addr(buf, &network))
+		return NULL;
+
+	if (!is_in_subnet_ipv4(ip, &mask, &network))
+		return strdup(net_unused);
+
+	srand(time(NULL));
+
+	/* try to randomize second group */
+	sprintf(net_unused, "%d.%d", DEFAULT_IPV4_UNUSED_PART1,
+		(rand() % 128) + (((ip->s_addr >> 8) & 0xff) < 128 ? 128 : 0));
+	sprintf(buf, "%s.0.0", net_unused);
+
+	if (get_in_addr(buf, &network))
+		return NULL;
+
+	if (!is_in_subnet_ipv4(ip, &mask, &network))
+		return strdup(net_unused);
+
+	/* try to randomize first group */
+	sprintf(net_unused, "%d.%d", (rand() % 128) + (((ip->s_addr) & 0xff)
+			< 128 ? 128 : 0), DEFAULT_IPV4_UNUSED_PART2);
+	sprintf(buf, "%s.0.0", net_unused);
+
+	if (get_in_addr(buf, &network))
+		return NULL;
+
+	if (!is_in_subnet_ipv4(ip, &mask, &network))
+		return strdup(net_unused);
+
+	return NULL;
+}
+
+/*
+ * Function get_ipv6_net32_unused is inspired by ipcalc project,
+ * get_ipv6_info/ipcalc.c.
+ *
+ * For unused network we use DEFAULT_IPV6_UNUSED_PART1:DEFAULT_IPV6_UNUSED_PART2
+ * if no collision with existing IP range.
+ * Otherwise we try to use
+ * {DEFAULT_IPV6_UNUSED_PART1}XY:DEFAULT_IPV6_UNUSED_PART2 or
+ * XY:DEFAULT_IPV6_UNUSED_PART2.
+ */
+static char *get_ipv6_net32_unused(const struct in6_addr *ip6,
+	unsigned int prefix)
+{
+	int i, j;
+	struct in6_addr mask, network;
+	char buf[128], net_unused[128];
+
+	if (prefix > 128)
+		return NULL;
+
+	memset(&mask, 0x0, sizeof(mask));
+	for (i = prefix, j = 0; i > 0; i -= 8, j++) {
+		if (i >= 8)
+			mask.s6_addr[j] = 0xff;
+		else
+			mask.s6_addr[j] = (unsigned long)(0xffU << (8 - i));
+	}
+
+	sprintf(net_unused, "%x:%x", 256 * DEFAULT_IPV6_UNUSED_PART1,
+			DEFAULT_IPV6_UNUSED_PART2);
+	sprintf(buf, "%s::", net_unused);
+
+	if (get_in6_addr(buf, &network))
+		return NULL;
+
+	if (!is_in_subnet_ipv6(ip6, &mask, &network))
+		return strdup(net_unused);
+
+	srand(time(NULL));
+
+	/* try to randomize second group */
+	sprintf(net_unused, "%x:%x", 256 * DEFAULT_IPV6_UNUSED_PART1 +
+			(rand() % 128) + (ip6->s6_addr[1] < 128 ? 128 : 0),
+			DEFAULT_IPV6_UNUSED_PART2);
+	sprintf(buf, "%s::", net_unused);
+
+	if (get_in6_addr(buf, &network))
+		return NULL;
+
+	if (!is_in_subnet_ipv6(ip6, &mask, &network))
+		return strdup(net_unused);
+
+	/* try to randomize first group */
+	sprintf(net_unused, "%x:%x",
+			256 * (rand() % 128) + (256 * ip6->s6_addr[0] < 128 ?
+			128 : 0), DEFAULT_IPV6_UNUSED_PART2);
+	sprintf(buf, "%s::", net_unused);
+
+	if (get_in6_addr(buf, &network))
+		return NULL;
+
+	if (!is_in_subnet_ipv6(ip6, &mask, &network))
+		return strdup(net_unused);
+
+	return NULL;
+}
+
+/*
+ * Function inet_ntop6_impl is based on musl libc project,
+ * inet_ntop/inet_ntop.c.
+ */
+static char *inet_ntop6_impl(const unsigned char *a0, unsigned int prefix,
+	int flags)
+{
+	const unsigned char *a = a0;
+	unsigned int i, j, max, best, border = 0;
+	char buf[100];
+	char ret[100];
+	char tmp[100];
+	char *p_ret = ret;
+	char *p_tmp = tmp;
+	size_t offset;
+
+	int is_net = flags & FLAG_GET_NETWORK;
+
+	snprintf(buf, sizeof(buf),
+		"%x:%x:%x:%x:%x:%x:%x:%x",
+		256 * a[0] + a[1], 256 * a[2] + a[3],
+		256 * a[4] + a[5], 256 * a[6] + a[7],
+		256 * a[8] + a[9], 256 * a[10] + a[11],
+		256 * a[12] + a[13], 256 * a[14] + a[15]);
+
+	for (i = 0; i < 8; i++) {
+		if (i < prefix >> 4) {
+			border += sprintf(p_tmp, "%x", 256 * a[2 * i] + a[2 * i + 1]);
+			if (i > 0)
+				border++;
+		}
+
+		if (is_net && i >= prefix >> 4)
+			break;
+
+		if (!is_net && i < prefix >> 4)
+			continue;
+
+		/* ':' only if no leading in host or ending in net */
+		if ((is_net && i > 0)
+			 || (!is_net && i > prefix >> 4))
+			*p_ret++ = ':';
+
+		offset = sprintf(p_ret, "%x", 256 * a[2 * i] + a[2 * i + 1]);
+		p_ret += offset;
+	}
+
+	*p_ret = '\0';
+
+	/* Find longest /(^0|:)[:0]{2,}/ */
+	for (i = best = 0, max = 2; buf[i]; i++) {
+		if (i && buf[i] != ':')
+			continue;
+		j = strspn(buf + i, ":0");
+
+		if (j > max)
+			best = i, max = j;
+	}
+
+	size_t length = strlen(ret);
+	size_t best_end = best + max - 1;
+
+	if (max > 2 && ((is_net && best < border) ||
+					(!is_net && best_end + 2 > border))) {
+		p_ret = ret;
+		/* Replace longest /(^0|:)[:0]{2,}/ with "::" */
+		if (is_net) {
+			if (best == 0 && best_end >= border) {
+				/* zeros in whole net part or continue to host */
+				ret[0] = ':';
+				ret[1] = '\0';
+			} else if (best == 0 && best_end < border) {
+				/* zeros on beginning, not whole part */
+				ret[0] = ':';
+				memmove(p_ret + 1, p_ret + best_end, border - best_end + 1);
+			} else if (best > 0 && best_end >= border) {
+				/* zeros not from beginning to border or continue to host */
+				ret[best] = ':';
+				ret[best + 1] = '\0';
+			} else {
+				/* zeros somewhere in the middle */
+				ret[best] = ':';
+				memmove(p_ret + best + 1, p_ret + best_end,
+						border - best + 1);
+			}
+		} else {
+			if (best <= border + 1 && best_end >= length + border) {
+				/* zeros in whole host part or continue to net */
+				ret[0] = '0';
+				ret[1] = '\0';
+			} else if (best <= border + 1 && best_end < length + border) {
+				if (best == border) {
+					/* zeros start in host, ends before end */
+					p_ret[0] = ':';
+					memmove(p_ret + 1, p_ret + best_end - border, length +
+						border - best_end + 2);
+				} else
+					/* zeros start in net, ends before end */
+					memmove(p_ret, p_ret + best_end - border, length +
+						border - best_end + 1);
+			} else if (best > border && best_end == border + length) {
+				/* zeros at the end */
+				ret[best - border] = ':';
+				ret[best - border + 1] = '\0';
+			} else {
+				/* zeros somewhere in the middle */
+				ret[best - border] = ':';
+				memmove(p_ret + best - border + 1, p_ret + best_end - border,
+					length + border - best_end + 1);
+			}
+		}
+	}
+
+	if (length < INET6_ADDRSTRLEN)
+		return strdup(ret);
+
+	return NULL;
+}
+
+/*
+ * Function bit_count is from ipcalc project, ipcalc.c.
+ */
+static int bit_count(uint32_t i)
+{
+	int c = 0;
+	unsigned int seen_one = 0;
+
+	while (i > 0) {
+		if (i & 1) {
+			seen_one = 1;
+			c++;
+		} else {
+			if (seen_one)
+				return -1;
+		}
+		i >>= 1;
+	}
+
+	return c;
+}
+
+/*
+ * Function mask2prefix is from ipcalc project, ipcalc.c.
+ */
+static int mask2prefix(struct in_addr mask)
+{
+	return bit_count(ntohl(mask.s_addr));
+}
+
+/*
+ * Function ipv4_mask_to_int is from ipcalc project, ipcalc.c.
+ */
+static int ipv4_mask_to_int(const char *prefix)
+{
+	int ret;
+	struct in_addr in;
+
+	ret = inet_pton(AF_INET, prefix, &in);
+	if (ret == 0)
+		return -1;
+
+	return mask2prefix(in);
+}
+
+/*
+ * Function safe_atoi is from ipcalc project, ipcalc.c.
+ */
+static int safe_atoi(const char *s, int *ret_i)
+{
+	char *x = NULL;
+	long l;
+
+	errno = 0;
+	l = strtol(s, &x, 0);
+
+	if (!x || x == s || *x || errno)
+		return errno > 0 ? -errno : -EINVAL;
+
+	if ((long)(int)l != l)
+		return -ERANGE;
+
+	*ret_i = (int)l;
+
+	return 0;
+}
+
+static int read_prefix(const char *ip_str, int is_ipv6)
+{
+	uint8_t family = is_ipv6 ? AF_INET6 : AF_INET;
+
+	char buf[16384];
+	int len;
+
+	struct {
+		struct nlmsghdr nlhdr;
+		struct ifaddrmsg addrmsg;
+	} msg;
+
+	struct nlmsghdr *retmsg;
+
+	int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE);
+
+	memset(&msg, 0, sizeof(msg));
+	msg.nlhdr.nlmsg_len = NLMSG_LENGTH(sizeof(struct ifaddrmsg));
+	msg.nlhdr.nlmsg_flags = NLM_F_REQUEST | NLM_F_ROOT;
+	msg.nlhdr.nlmsg_type = RTM_GETADDR;
+	msg.addrmsg.ifa_family = family;
+
+	send(sock, &msg, msg.nlhdr.nlmsg_len, 0);
+	len = recv(sock, buf, sizeof(buf), 0);
+	retmsg = (struct nlmsghdr *)buf;
+
+	while NLMSG_OK(retmsg, len) {
+		struct ifaddrmsg *retaddr;
+		struct rtattr *retrta;
+		char pradd[128];
+		int attlen;
+
+		retaddr = (struct ifaddrmsg *)NLMSG_DATA(retmsg);
+		retrta = (struct rtattr *)IFA_RTA(retaddr);
+		attlen = IFA_PAYLOAD(retmsg);
+
+		while RTA_OK(retrta, attlen) {
+			if (retrta->rta_type == IFA_ADDRESS) {
+				inet_ntop(family, RTA_DATA(retrta), pradd, sizeof(pradd));
+				if (!strcmp(pradd, ip_str))
+					return retaddr->ifa_prefixlen;
+			}
+			retrta = RTA_NEXT(retrta, attlen);
+
+		}
+		retmsg = NLMSG_NEXT(retmsg, len);
+	}
+
+	return -1;
+}
+
+static int validate_prefix(int prefix, int is_ipv6)
+{
+	int base_prefix = is_ipv6 ? BASE_IPV6_PREFIX : BASE_IPV4_PREFIX;
+
+	return (prefix < base_prefix || (is_ipv6 && prefix == 128)) ||
+		(!is_ipv6 && prefix == 32) ? -1 : 0;
+}
+
+/*
+ * Function get_prefix use code from ipcalc project, str_to_prefix/ipcalc.c.
+ */
+static int get_prefix(const char *ip_str, int is_ipv6)
+{
+	char *prefix_str = NULL;
+	int prefix = -1, r;
+
+	prefix_str = strchr(ip_str, '/');
+	if (!prefix_str)
+		return -1;
+
+	*(prefix_str++) = '\0';
+
+	if (!is_ipv6 && strchr(prefix_str, '.'))
+		prefix = ipv4_mask_to_int(prefix_str);
+	else {
+		r = safe_atoi(prefix_str, &prefix);
+		if (r != 0)
+			tst_brk(TBROK, "Conversion error");
+	}
+
+	if (prefix < 0 || ((is_ipv6 && prefix > 128) ||
+		(!is_ipv6 && prefix > 32)))
+		tst_brk(TBROK, "Bad %s prefix: %s", is_ipv6 ? "IPv6" : "IPv4",
+			prefix_str);
+
+	return prefix;
+}
+
+static char *get_ipv4_host(int ip, int prefix)
+{
+	char buf[INET_ADDRSTRLEN + 1];
+	char *p_buf = buf;
+	unsigned char byte;
+	int i;
+
+	if (prefix < 0 || prefix > 32)
+		return NULL;
+
+	prefix &= 0x18;
+
+	for (i = 0; i < 32; i += 8) {
+		if (i < prefix)
+			continue;
+
+		if (i > prefix) {
+			sprintf(p_buf, ".");
+			p_buf++;
+		}
+
+		if (i == 0)
+			byte = ip & 0xff;
+		else
+			byte = (ip >> i) & 0xff;
+
+		sprintf(p_buf, "%d", byte);
+		p_buf += strlen(p_buf);
+	}
+
+	return strdup(buf);
+}
+
+static char *get_ipv4_network(int ip, int prefix)
+{
+	char buf[INET_ADDRSTRLEN + 1];
+	char *p_buf = buf;
+	unsigned char byte;
+	int i;
+
+	if (prefix < 0 || prefix > 32)
+		return NULL;
+
+	prefix &= 0x18;
+
+	for (i = 0; i < 32 && i < prefix; i += 8) {
+		if (i == 0) {
+			byte = ip & 0xff;
+			sprintf(p_buf, "%d", byte);
+		} else {
+			byte = (ip >> i) & 0xff;
+			sprintf(p_buf, ".%d", byte);
+		}
+		p_buf += strlen(p_buf);
+	}
+
+	return strdup(buf);
+}
+
+static void get_ipv4_info(struct in_addr *lip, struct in_addr *rip, int prefix)
+{
+	vars.ipv4_network = get_ipv4_network(lip->s_addr, prefix);
+	if (strcmp(vars.ipv4_network, get_ipv4_network(rip->s_addr, prefix)))
+		tst_brk(TBROK, "Please use the same network for both IP addresses");
+
+	vars.lhost_ipv4_host = get_ipv4_host(lip->s_addr, prefix);
+	vars.rhost_ipv4_host = get_ipv4_host(rip->s_addr, prefix);
+	vars.ipv4_net16_unused = get_ipv4_net16_unused(lip, prefix);
+
+}
+
+static void get_ipv6_info(struct in6_addr *lip, struct in6_addr *rip,
+	int prefix)
+{
+	vars.ipv6_network = inet_ntop6_impl(lip->s6_addr, prefix, FLAG_GET_NETWORK);
+	if (strcmp(vars.ipv6_network,
+		inet_ntop6_impl(rip->s6_addr, prefix, FLAG_GET_NETWORK)))
+		tst_brk(TBROK, "Please use the same network for both IP addresses");
+
+	vars.lhost_ipv6_host = inet_ntop6_impl(lip->s6_addr, prefix, FLAG_GET_HOST);
+	vars.rhost_ipv6_host = inet_ntop6_impl(rip->s6_addr, prefix, FLAG_GET_HOST);
+	vars.ipv6_net32_unused = get_ipv6_net32_unused(lip, prefix);
+}
+
+static void print_svar(const char *name, const char *val)
+{
+	if (name && val)
+		printf("export %s='%s'\n", name, val);
+}
+static void print_ivar(const char *name, const int val)
+{
+	if (name)
+		printf("export %s=%d\n", name, val);
+}
+
+static void print_vars(int is_ipv6)
+{
+	if (is_ipv6) {
+		print_svar("IPV6_LHOST", vars.ipv6_lhost);
+		print_svar("IPV6_RHOST", vars.ipv6_rhost);
+		print_svar("IPV6_NETWORK", vars.ipv6_network);
+		print_svar("LHOST_IPV6_HOST", vars.lhost_ipv6_host);
+		print_svar("RHOST_IPV6_HOST", vars.rhost_ipv6_host);
+		print_svar("IPV6_NET32_UNUSED", vars.ipv6_net32_unused);
+		print_ivar("IPV6_PREFIX", vars.ipv6_prefix);
+	} else {
+		print_svar("IPV4_LHOST", vars.ipv4_lhost);
+		print_svar("IPV4_RHOST", vars.ipv4_rhost);
+		print_svar("IPV4_NETWORK", vars.ipv4_network);
+		print_svar("LHOST_IPV4_HOST", vars.lhost_ipv4_host);
+		print_svar("RHOST_IPV4_HOST", vars.rhost_ipv4_host);
+		print_svar("IPV4_NET16_UNUSED", vars.ipv4_net16_unused);
+		print_ivar("IPV4_PREFIX", vars.ipv4_prefix);
+	}
+}
+
+int main(int argc, char *argv[])
+{
+	char *lip_str = NULL, *rip_str = NULL;
+	struct in_addr lip, rip;
+	struct in6_addr lip6, rip6;
+	int is_ipv6, lprefix, rprefix, prefix, base_prefix, is_read_prefix = 0;
+
+	int is_usage = argc > 1 && (!strcmp(argv[1], "-h") ||
+		!strcmp(argv[1], "--help"));
+	if (argc < 3 || is_usage) {
+		usage();
+		exit(is_usage ? 0 : 1);
+	}
+
+	lip_str = argv[1];
+	rip_str = argv[2];
+
+	is_ipv6 = !!strchr(lip_str, ':');
+	if (is_ipv6 != !(strchr(rip_str, ':') == NULL))
+		tst_brk(TBROK, "Mixed IPv4 and IPv6 addresses");
+
+	lprefix = get_prefix(lip_str, is_ipv6);
+	rprefix = get_prefix(rip_str, is_ipv6);
+
+	if (lprefix < 0 && rprefix < 0) {
+		is_read_prefix = 1;
+		lprefix = read_prefix(lip_str, is_ipv6);
+		if (lprefix < 0)
+			lprefix = read_prefix(rip_str, is_ipv6);
+		if (lprefix < 0) {
+			lprefix = is_ipv6 ? DEFAULT_IPV6_PREFIX : DEFAULT_IPV4_PREFIX;
+			fprintf(stderr, "Missing prefix, using default: %d\n", lprefix);
+		}
+		rprefix = lprefix;
+	} else {
+		if (lprefix < 0 && rprefix > 0)
+			lprefix = rprefix;
+		else if (rprefix < 0 && lprefix > 0)
+			rprefix = lprefix;
+	}
+
+	if (validate_prefix(lprefix, is_ipv6))
+		tst_brk(TBROK, is_read_prefix ?
+					 "Please don't use interface with prefix: %d for %s" :
+					 "Please don't use prefix: %d for %s",
+				 lprefix, is_ipv6 ?
+				 "IPv6" : "IPv4");
+
+	if (lprefix != rprefix)
+		tst_brk(TBROK, "Prefixes must be the same");
+
+	if (is_ipv6) {
+		get_in6_addr(lip_str, &lip6);
+		get_in6_addr(rip_str, &rip6);
+		vars.ipv6_lhost = strdup(lip_str);
+		vars.ipv6_rhost = strdup(rip_str);
+		vars.ipv6_prefix = lprefix;
+	} else {
+		get_in_addr(lip_str, &lip);
+		get_in_addr(rip_str, &rip);
+		vars.ipv4_lhost = strdup(lip_str);
+		vars.ipv4_rhost = strdup(rip_str);
+		vars.ipv4_prefix = lprefix;
+	}
+
+	if (!strcmp(lip_str, rip_str))
+		tst_brk(TBROK, "IP addresses cannot be the same");
+
+	/* Round down prefix */
+	base_prefix = is_ipv6 ? BASE_IPV6_PREFIX : BASE_IPV4_PREFIX;
+	prefix = lprefix / base_prefix * base_prefix;
+
+	if (is_ipv6)
+		get_ipv6_info(&lip6, &rip6, prefix);
+	else
+		get_ipv4_info(&lip, &rip, prefix);
+
+	print_vars(is_ipv6);
+
+	exit(0);
+}
-- 
2.12.2

[LTP] [RFC PATCH v5 3/3] network: Use tst_net_vars to set up IPv4 and IPv6 related variables

[Petr Vorel]

tst_net_vars simplifies setup as it only needs to have passed 4 IP
addresses (2 IPv4 and 2 IPv6).

Signed-off-by: Petr Vorel <pvorel@suse.cz>
---
 testcases/lib/test_net.sh | 43 +++++++++++++------------------------------
 1 file changed, 13 insertions(+), 30 deletions(-)

diff --git a/testcases/lib/test_net.sh b/testcases/lib/test_net.sh
index 08493558b..42bfc8dbc 100644
--- a/testcases/lib/test_net.sh
+++ b/testcases/lib/test_net.sh
@@ -191,18 +191,11 @@ tst_ipaddr()
 {
 	local type="${1:-lhost}"
 	local ipv="${TST_IPV6:-4}"
-	local tst_host=
-
-	if [ "$type" = "lhost" ]; then
-		eval "tst_host=\$LHOST_IPV${ipv}_HOST"
-	else
-		eval "tst_host=\$RHOST_IPV${ipv}_HOST"
-	fi
 
 	if [ "$TST_IPV6" ]; then
-		echo "${IPV6_NETWORK}:${tst_host}"
+		[ "$type" = "lhost" ] && echo "$IPV6_LHOST" || echo "$IPV6_RHOST"
 	else
-		echo "${IPV4_NETWORK}.${tst_host}"
+		[ "$type" = "lhost" ] && echo "$IPV4_LHOST" || echo "$IPV4_RHOST"
 	fi
 }
 
@@ -242,8 +235,8 @@ tst_add_ipaddr()
 	local type="${1:-lhost}"
 	local link_num="${2:-0}"
 
-	local mask=24
-	[ "$TST_IPV6" ] && mask=64
+	local mask=$IPV4_PREFIX
+	[ "$TST_IPV6" ] && mask=$IPV6_PREFIX
 
 	local iface=$(tst_iface $type $link_num)
 
@@ -487,25 +480,15 @@ export PASSWD="${PASSWD:-}"
 # Don't use it in new tests, use tst_rhost_run() from test_net.sh instead.
 export LTP_RSH="${LTP_RSH:-rsh -n}"
 
-# Test Links
-# Set first three octets of the network address, default is '10.0.0'
-export IPV4_NETWORK="${IPV4_NETWORK:-10.0.0}"
-# Set local host last octet, default is '2'
-export LHOST_IPV4_HOST="${LHOST_IPV4_HOST:-2}"
-# Set remote host last octet, default is '1'
-export RHOST_IPV4_HOST="${RHOST_IPV4_HOST:-1}"
-# Set the reverse of IPV4_NETWORK
-export IPV4_NET_REV="${IPV4_NET_REV:-0.0.10}"
-# Set first three octets of the network address, default is 'fd00:1:1:1'
-export IPV6_NETWORK="${IPV6_NETWORK:-fd00:1:1:1}"
-# Set local host last octet, default is '2'
-export LHOST_IPV6_HOST="${LHOST_IPV6_HOST:-:2}"
-# Set remote host last octet, default is '1'
-export RHOST_IPV6_HOST="${RHOST_IPV6_HOST:-:1}"
-
-# Networks that aren't reachable through the test links
-export IPV4_NET16_UNUSED="${IPV4_NET16_UNUSED:-10.23}"
-export IPV6_NET32_UNUSED="${IPV6_NET32_UNUSED:-fd00:23}"
+# tst_net_vars exports environment variables related to test links + networks
+# that aren't reachable through the test links.
+# See tst_net_vars -h for a list of exported variables.
+IPV4_LHOST="${IPV4_LHOST:-10.0.0.2}"
+IPV4_RHOST="${IPV4_RHOST:-10.0.0.1}"
+eval `tst_net_vars $IPV4_LHOST $IPV4_RHOST`
+IPV6_LHOST="${IPV6_LHOST:-fd00:1:1:1::2}"
+IPV6_RHOST="${IPV6_RHOST:-fd00:1:1:1::1}"
+eval `tst_net_vars $IPV6_LHOST $IPV6_RHOST`
 
 export HTTP_DOWNLOAD_DIR="${HTTP_DOWNLOAD_DIR:-/var/www/html}"
 export FTP_DOWNLOAD_DIR="${FTP_DOWNLOAD_DIR:-/var/ftp}"
-- 
2.12.2

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Alexey Kodanev]

On 03.05.2017 19:07, Petr Vorel wrote:
> Hi,
>
> another version of tst_net_vars network setup helper.
> v4->v5:
> * Clenup unneded variables (reverse IP related).
> * Don't require to have mask in IP address. Use netlink to detect prefix when
>   address is presented on some of the NIC on the system or fallback to default
>   ones (24 for IPv4, 64 for IPv6).
> * Fix detecting wrong subnet with network not round up by 8 (IPv4 resp 16
>   (IPv6). E.g.: problem IP addressed like 10.2.0.1/22 10.2.1.1/22.
> * Add more environment variables (e.g. prefix, whole IP address)
> * Rewritten (again) unused network generation to fix overlap with used IP addresses.
> * Simplify tst_ipaddr() (in testcases/lib/test_net.sh).
> * Use tst_brk() to exit on error.
> * Cleanup warnings and formatting.
>
> There is still requirement network prefix to be < 8 (IPv4) resp. 16 (IPv6) as
> then IPV{4,6}_NETWORK variables would be empty, which is not expected by tests
> (they join it with address separators with {L,R}HOST_IPV{4,6}_HOST. Even we got
> rid of reverse related variables we still need IPV{4,6}_NETWORK variables for
> some tests.

Hi Petr,

But you removed such dependency on the prefix from test_net.sh  in
the third patch... so we could just print the warning that configuration
is not appropriate for the old tests, that is for stress/multicast,
stress/route groups only.

Thanks,
Alexey

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Richard Palethorpe]

Hi,

On Wed,  3 May 2017 18:07:48 +0200
"Petr Vorel" <pvorel@suse.cz> wrote:

> BTW: I probably have source lines > 80 chars. My Vim doesn't highlight
> anything, but checkpatch.pl complain. I might have wrong vim configuration.

Are you using checkpatch.pl from kernel 4.11? It was recently changed to
follow the actual style guidelines which say lines over 80 chars are
preferable to strings being split over multiple lines. Maybe Vim is aware of
that?

Thank you,
Richard.

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Alexey Kodanev]

On 04.05.2017 14:27, Richard Palethorpe wrote:
> Hi,
>
> On Wed,  3 May 2017 18:07:48 +0200
> "Petr Vorel" <pvorel@suse.cz> wrote:
>
>> BTW: I probably have source lines > 80 chars. My Vim doesn't highlight
>> anything, but checkpatch.pl complain. I might have wrong vim configuration.
> Are you using checkpatch.pl from kernel 4.11? It was recently changed to
> follow the actual style guidelines which say lines over 80 chars are
> preferable to strings being split over multiple lines. Maybe Vim is aware of
> that?

Or may be vim's configuration has wrong tab size, should be 8 characters
for Linux coding style.

Thanks,
Alexey

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Alexey Kodanev]

On 04.05.2017 14:09, Alexey Kodanev wrote:
> On 03.05.2017 19:07, Petr Vorel wrote:
>> Hi,
>>
>> another version of tst_net_vars network setup helper.
>> v4->v5:
>> * Clenup unneded variables (reverse IP related).
>> * Don't require to have mask in IP address. Use netlink to detect prefix when
>>   address is presented on some of the NIC on the system or fallback to default
>>   ones (24 for IPv4, 64 for IPv6).
>> * Fix detecting wrong subnet with network not round up by 8 (IPv4 resp 16
>>   (IPv6). E.g.: problem IP addressed like 10.2.0.1/22 10.2.1.1/22.
>> * Add more environment variables (e.g. prefix, whole IP address)
>> * Rewritten (again) unused network generation to fix overlap with used IP addresses.
>> * Simplify tst_ipaddr() (in testcases/lib/test_net.sh).
>> * Use tst_brk() to exit on error.
>> * Cleanup warnings and formatting.
>>
>> There is still requirement network prefix to be < 8 (IPv4) resp. 16 (IPv6) as
>> then IPV{4,6}_NETWORK variables would be empty, which is not expected by tests
>> (they join it with address separators with {L,R}HOST_IPV{4,6}_HOST. Even we got
>> rid of reverse related variables we still need IPV{4,6}_NETWORK variables for
>> some tests.
> Hi Petr,
>
> But you removed such dependency on the prefix from test_net.sh  in
> the third patch... so we could just print the warning that configuration
> is not appropriate for the old tests, that is for stress/multicast,
> stress/route groups only.

Regarding IPV{4,6}_NETWORK calculation, I believe it would much
simpler ifwe saved the same part of address from IPV{4,6}_HOST
variables to_NETWORK variable and what is differ to the other
variables.There is no need to any additional checks for these
environment variables or I'm missing something?

Hmm, I didn't find how the patch is getting network prefix
on the remote host, is it because of the assumption that they
should be the same on both hosts? It is fine for now as it's
not so common for a test environment to have different ones,
so the better approach in the future is to make a separate tool
(installed on both hosts with LTP) and get prefixes from
local and remote host similar to tst_get_hwaddrs() in test_net.sh.

Thanks,
Alexey

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Petr Vorel]

Hi Alexey, Richard,

> >> BTW: I probably have source lines > 80 chars. My Vim doesn't highlight
> >> anything, but checkpatch.pl complain. I might have wrong vim configuration.
> > Are you using checkpatch.pl from kernel 4.11? It was recently changed to
> > follow the actual style guidelines which say lines over 80 chars are
> > preferable to strings being split over multiple lines. Maybe Vim is aware of
> > that?

> Or may be vim's configuration has wrong tab size, should be 8 characters
> for Linux coding style.
I'm using recent version of checkpatch.pl.
My fault, I've installed linuxsty plugin [1], but default is off, so I had these options:
tabstop=4 softtabstop=4 shiftwidth=4

[1]: https://github.com/vivien/vim-linux-coding-style/blob/master/plugin/linuxsty.vim

Thank you both for tips!

Petr

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Petr Vorel]

Hi Alexey,

> > There is still requirement network prefix to be < 8 (IPv4) resp. 16 (IPv6) as
> > then IPV{4,6}_NETWORK variables would be empty, which is not expected by tests
> > (they join it with address separators with {L,R}HOST_IPV{4,6}_HOST. Even we got
> > rid of reverse related variables we still need IPV{4,6}_NETWORK variables for
> > some tests.

> Hi Petr,

> But you removed such dependency on the prefix from test_net.sh  in
> the third patch... so we could just print the warning that configuration
> is not appropriate for the old tests, that is for stress/multicast,
> stress/route groups only.
Right, I'll remove requirement network prefix to be < 8 (IPv4) resp. 16 (IPv6).
There are other restrictions you probably don't like:
* Don't use 32 (IPv4) and 128 (IPv6): I'd keep this restriction. What do you think?
* Require to use the same prefix: I'd keep it for now.


Kind regards,
Petr

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Petr Vorel]

Hi Alexey,

> Regarding IPV{4,6}_NETWORK calculation, I believe it would much
> simpler ifwe saved the same part of address from IPV{4,6}_HOST
> variables to_NETWORK variable and what is differ to the other
> variables.There is no need to any additional checks for these
> environment variables or I'm missing something?
You're right - calculating just network part and then get host part by substring operation
will be probably less error-prone. I planned to write some tests latter (sorry for not
including it now), but this is really better solution.

> Hmm, I didn't find how the patch is getting network prefix
> on the remote host, is it because of the assumption that they
> should be the same on both hosts?
Yes. So far we have just one IPV{4,6}_NETWORK variable for both, so I guess this makes us to use
the same network for both.

> It is fine for now as it's
> not so common for a test environment to have different ones,
> so the better approach in the future is to make a separate tool
> (installed on both hosts with LTP) and get prefixes from
> local and remote host similar to tst_get_hwaddrs() in test_net.sh.
I'm ok to make C binary which reads prefix, which will be run on both local and remote
hosts. But to be actually used, we'd need to change all tests using IPV{4,6}_NETWORK to
use different variables for local and remote hosts. I'd like to do it after having this
merged. I have also in my TODO list creating tst_net_reverse_ip binary for
stress/multicast and stress/route tests groups.

BTW is there a better/simpler way to get prefix than what I wrote?


Kind regards,
Petr

[LTP] [RFC PATCH v5 2/3] network: Add tool for setup IP variables

[Richard Palethorpe]

Hi,

Just a couple of minor comments, I haven't properly read it.

On Wed,  3 May 2017 18:07:50 +0200
"Petr Vorel" <pvorel@suse.cz> wrote:

> +/*
> + * Function prefix2mask is from ipcalc project, ipcalc.c.
> + */
> +struct in_addr prefix2mask(int prefix)
> +{
> +	struct in_addr mask;
> +
> +	memset(&mask, 0, sizeof(mask));

FYI for these memsets you could use 'struct foo bar = { 0 }' instead. Although
it generates warnings on older compilers if the struct starts with another
composite data type.

> +
> +static int read_prefix(const char *ip_str, int is_ipv6)
> +{
> +	uint8_t family = is_ipv6 ? AF_INET6 : AF_INET;
> +
> +	char buf[16384];
> +	int len;
> +
> +	struct {
> +		struct nlmsghdr nlhdr;
> +		struct ifaddrmsg addrmsg;
> +	} msg;
> +
> +	struct nlmsghdr *retmsg;
> +
> +	int sock = socket(AF_NETLINK, SOCK_RAW, NETLINK_ROUTE);
> +
> +	memset(&msg, 0, sizeof(msg));
> +	msg.nlhdr.nlmsg_len = NLMSG_LENGTH(sizeof(struct ifaddrmsg));
> +	msg.nlhdr.nlmsg_flags = NLM_F_REQUEST | NLM_F_ROOT;
> +	msg.nlhdr.nlmsg_type = RTM_GETADDR;
> +	msg.addrmsg.ifa_family = family;
> +
> +	send(sock, &msg, msg.nlhdr.nlmsg_len, 0);
> +	len = recv(sock, buf, sizeof(buf), 0);
> +	retmsg = (struct nlmsghdr *)buf;

No error checking? I suppose it is unlikely to fail, but it can't hurt.

> +
> +	while NLMSG_OK(retmsg, len) {
> +		struct ifaddrmsg *retaddr;
> +		struct rtattr *retrta;
> +		char pradd[128];
> +		int attlen;
> +
> +		retaddr = (struct ifaddrmsg *)NLMSG_DATA(retmsg);
> +		retrta = (struct rtattr *)IFA_RTA(retaddr);
> +		attlen = IFA_PAYLOAD(retmsg);
> +
> +		while RTA_OK(retrta, attlen) {
> +			if (retrta->rta_type == IFA_ADDRESS) {
> +				inet_ntop(family, RTA_DATA(retrta), pradd, sizeof(pradd));
> +				if (!strcmp(pradd, ip_str))
> +					return retaddr->ifa_prefixlen;
> +			}
> +			retrta = RTA_NEXT(retrta, attlen);
> +
> +		}
> +		retmsg = NLMSG_NEXT(retmsg, len);
> +	}
> +
> +	return -1;
> +}
> +

Thank you,
Richard.

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Alexey Kodanev]

On 04.05.2017 16:46, Petr Vorel wrote:
> Hi Alexey,
>
>>> There is still requirement network prefix to be < 8 (IPv4) resp. 16 (IPv6) as
>>> then IPV{4,6}_NETWORK variables would be empty, which is not expected by tests
>>> (they join it with address separators with {L,R}HOST_IPV{4,6}_HOST. Even we got
>>> rid of reverse related variables we still need IPV{4,6}_NETWORK variables for
>>> some tests.
>> Hi Petr,
>> But you removed such dependency on the prefix from test_net.sh  in
>> the third patch... so we could just print the warning that configuration
>> is not appropriate for the old tests, that is for stress/multicast,
>> stress/route groups only.
> Right, I'll remove requirement network prefix to be < 8 (IPv4) resp. 16 (IPv6).
> There are other restrictions you probably don't like:
> * Don't use 32 (IPv4) and 128 (IPv6): I'd keep this restriction. What do you think?
> * Require to use the same prefix: I'd keep it for now.

I'm for a warning message that some network/stress tests not supporting
such config.

Thanks,
Alexey

[LTP] [RFC PATCH v5 0/3] Simplify network setup

[Petr Vorel]

Hi Alexey,

> > Right, I'll remove requirement network prefix to be < 8 (IPv4) resp. 16 (IPv6).
> > There are other restrictions you probably don't like:
> > * Don't use 32 (IPv4) and 128 (IPv6): I'd keep this restriction. What do you think?
> > * Require to use the same prefix: I'd keep it for now.

> I'm for a warning message that some network/stress tests not supporting
> such config.
Although I think prefix (mask) 32 resp. 128 doesn't make much sense, I'll remove both
restrictions. If I have enough time I'll implement use of different network variables for
local and remote hosts.

Thanks for your time and comments!


Kind regards,
Petr

[LTP] [RFC PATCH v5 2/3] network: Add tool for setup IP variables

[Petr Vorel]

Hi Richard,

> > +	struct in_addr mask;
> > +
> > +	memset(&mask, 0, sizeof(mask));

> FYI for these memsets you could use 'struct foo bar = { 0 }' instead. Although
> it generates warnings on older compilers if the struct starts with another
> composite data type.

Right, I used it for struct in_addr and in6_addr. Not used it for struct msg as it
contains another struct (even it does not produce a warning when compiled with
--std=gnu89).

> > +	struct {
> > +		struct nlmsghdr nlhdr;
> > +		struct ifaddrmsg addrmsg;
> > +	} msg;



> > +	msg.nlhdr.nlmsg_len = NLMSG_LENGTH(sizeof(struct ifaddrmsg));
> > +	msg.nlhdr.nlmsg_flags = NLM_F_REQUEST | NLM_F_ROOT;
> > +	msg.nlhdr.nlmsg_type = RTM_GETADDR;
> > +	msg.addrmsg.ifa_family = family;
> > +
> > +	send(sock, &msg, msg.nlhdr.nlmsg_len, 0);
> > +	len = recv(sock, buf, sizeof(buf), 0);
> > +	retmsg = (struct nlmsghdr *)buf;

> No error checking? I suppose it is unlikely to fail, but it can't hurt.
IMHO NLMSG_OK RTA_NEXT macros do some sort of checks [1], [2]. Sample code [3] checks for
more, but that's IMHO irrelevant.

> > +
> > +	while NLMSG_OK(retmsg, len) {
> > +		struct ifaddrmsg *retaddr;
> > +		struct rtattr *retrta;
> > +		char pradd[128];
> > +		int attlen;
> > +
> > +		retaddr = (struct ifaddrmsg *)NLMSG_DATA(retmsg);
> > +		retrta = (struct rtattr *)IFA_RTA(retaddr);
> > +		attlen = IFA_PAYLOAD(retmsg);
> > +
> > +		while RTA_OK(retrta, attlen) {
> > +			if (retrta->rta_type == IFA_ADDRESS) {
> > +				inet_ntop(family, RTA_DATA(retrta), pradd, sizeof(pradd));
> > +				if (!strcmp(pradd, ip_str))
> > +					return retaddr->ifa_prefixlen;
> > +			}
> > +			retrta = RTA_NEXT(retrta, attlen);
> > +
> > +		}
> > +		retmsg = NLMSG_NEXT(retmsg, len);
> > +	}
> > +
> > +	return -1;
> > +}
> > +

> Thank you,
> Richard.
Thanks for your comments and time!
Petr


[1] http://man7.org/linux/man-pages/man3/netlink.3.html
[2] http://man7.org/linux/man-pages/man3/rtnetlink.3.html
[3] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/samples/bpf/bpf_load.c?id=refs/tags/v4.11#n535