Merge branch 'veth-flexible-channel-numbers'

Paolo Abeni says:

====================
veth: more flexible channels number configuration

XDP setups can benefit from multiple veth RX/TX queues. Currently
veth allow setting such number only at creation time via the
'numrxqueues' and 'numtxqueues' parameters.

This series introduces support for the ethtool set_channel operation
and allows configuring the queue number via a new module parameter.

The veth default configuration is not changed.

Finally self-tests are updated to check the new features, with both
valid and invalid arguments.

This iteration is a rebase of the most recent RFC, it does not provide
a module parameter to configure the default number of queues, but I
think could be worthy

RFC v1 -> RFC v2:
 - report more consistent 'combined' count
 - make set_channel as resilient as possible to errors
 - drop module parameter - but I would still consider it.
 - more self-tests
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

mm/memcontrol.c

@@ -968,7 +968,7 @@ static __always_inline bool memcg_kmem_bypass(void)
 		return false;
 
 	/* Memcg to charge can't be determined. */
-	if (in_interrupt() || !current->mm || (current->flags & PF_KTHREAD))
+	if (!in_task() || !current->mm || (current->flags & PF_KTHREAD))
 		return true;
 
 	return false;

net/8021q/vlan.c

@@ -67,7 +67,7 @@ static int vlan_group_prealloc_vid(struct vlan_group *vg,
 		return 0;
 
 	size = sizeof(struct net_device *) * VLAN_GROUP_ARRAY_PART_LEN;
-	array = kzalloc(size, GFP_KERNEL);
+	array = kzalloc(size, GFP_KERNEL_ACCOUNT);
 	if (array == NULL)
 		return -ENOBUFS;
 

net/core/dev.c

@@ -10119,7 +10119,7 @@ static int netif_alloc_rx_queues(struct net_device *dev)
 
 	BUG_ON(count < 1);
 
-	rx = kvzalloc(sz, GFP_KERNEL | __GFP_RETRY_MAYFAIL);
+	rx = kvzalloc(sz, GFP_KERNEL_ACCOUNT | __GFP_RETRY_MAYFAIL);
 	if (!rx)
 		return -ENOMEM;
 
@@ -10186,7 +10186,7 @@ static int netif_alloc_netdev_queues(struct net_device *dev)
 	if (count < 1 || count > 0xffff)
 		return -EINVAL;
 
-	tx = kvzalloc(sz, GFP_KERNEL | __GFP_RETRY_MAYFAIL);
+	tx = kvzalloc(sz, GFP_KERNEL_ACCOUNT | __GFP_RETRY_MAYFAIL);
 	if (!tx)
 		return -ENOMEM;
 
@@ -10826,7 +10826,7 @@ struct net_device *alloc_netdev_mqs(int sizeof_priv, const char *name,
 	/* ensure 32-byte alignment of whole construct */
 	alloc_size += NETDEV_ALIGN - 1;
 
-	p = kvzalloc(alloc_size, GFP_KERNEL | __GFP_RETRY_MAYFAIL);
+	p = kvzalloc(alloc_size, GFP_KERNEL_ACCOUNT | __GFP_RETRY_MAYFAIL);
 	if (!p)
 		return NULL;
 

net/core/fib_rules.c

@@ -57,7 +57,7 @@ int fib_default_rule_add(struct fib_rules_ops *ops,
 {
 	struct fib_rule *r;
 
-	r = kzalloc(ops->rule_size, GFP_KERNEL);
+	r = kzalloc(ops->rule_size, GFP_KERNEL_ACCOUNT);
 	if (r == NULL)
 		return -ENOMEM;
 
@@ -541,7 +541,7 @@ static int fib_nl2rule(struct sk_buff *skb, struct nlmsghdr *nlh,
 			goto errout;
 	}
 
-	nlrule = kzalloc(ops->rule_size, GFP_KERNEL);
+	nlrule = kzalloc(ops->rule_size, GFP_KERNEL_ACCOUNT);
 	if (!nlrule) {
 		err = -ENOMEM;
 		goto errout;

net/core/scm.c

@@ -79,7 +79,7 @@ static int scm_fp_copy(struct cmsghdr *cmsg, struct scm_fp_list **fplp)
 
 	if (!fpl)
 	{
-		fpl = kmalloc(sizeof(struct scm_fp_list), GFP_KERNEL);
+		fpl = kmalloc(sizeof(struct scm_fp_list), GFP_KERNEL_ACCOUNT);
 		if (!fpl)
 			return -ENOMEM;
 		*fplp = fpl;
@@ -355,7 +355,7 @@ struct scm_fp_list *scm_fp_dup(struct scm_fp_list *fpl)
 		return NULL;
 
 	new_fpl = kmemdup(fpl, offsetof(struct scm_fp_list, fp[fpl->count]),
-			  GFP_KERNEL);
+			  GFP_KERNEL_ACCOUNT);
 	if (new_fpl) {
 		for (i = 0; i < fpl->count; i++)
 			get_file(fpl->fp[i]);

net/dccp/proto.c

@@ -1126,7 +1126,7 @@ static int __init dccp_init(void)
 	dccp_hashinfo.bind_bucket_cachep =
 		kmem_cache_create("dccp_bind_bucket",
 				  sizeof(struct inet_bind_bucket), 0,
-				  SLAB_HWCACHE_ALIGN, NULL);
+				  SLAB_HWCACHE_ALIGN | SLAB_ACCOUNT, NULL);
 	if (!dccp_hashinfo.bind_bucket_cachep)
 		goto out_free_hashinfo2;
 

net/ipv4/devinet.c

@@ -215,7 +215,7 @@ static void devinet_sysctl_unregister(struct in_device *idev)
 
 static struct in_ifaddr *inet_alloc_ifa(void)
 {
-	return kzalloc(sizeof(struct in_ifaddr), GFP_KERNEL);
+	return kzalloc(sizeof(struct in_ifaddr), GFP_KERNEL_ACCOUNT);
 }
 
 static void inet_rcu_free_ifa(struct rcu_head *head)

net/ipv4/fib_trie.c

@@ -2380,11 +2380,11 @@ void __init fib_trie_init(void)
 {
 	fn_alias_kmem = kmem_cache_create("ip_fib_alias",
 					  sizeof(struct fib_alias),
-					  0, SLAB_PANIC, NULL);
+					  0, SLAB_PANIC | SLAB_ACCOUNT, NULL);
 
 	trie_leaf_kmem = kmem_cache_create("ip_fib_trie",
 					   LEAF_SIZE,
-					   0, SLAB_PANIC, NULL);
+					   0, SLAB_PANIC | SLAB_ACCOUNT, NULL);
 }
 
 struct fib_table *fib_trie_table(u32 id, struct fib_table *alias)

net/ipv4/tcp.c

@@ -4512,7 +4512,9 @@ void __init tcp_init(void)
 	tcp_hashinfo.bind_bucket_cachep =
 		kmem_cache_create("tcp_bind_bucket",
 				  sizeof(struct inet_bind_bucket), 0,
-				  SLAB_HWCACHE_ALIGN|SLAB_PANIC, NULL);
+				  SLAB_HWCACHE_ALIGN | SLAB_PANIC |
+				  SLAB_ACCOUNT,
+				  NULL);
 
 	/* Size and allocate the main established and bind bucket
 	 * hash tables.

net/ipv6/addrconf.c

@@ -1080,7 +1080,7 @@ ipv6_add_addr(struct inet6_dev *idev, struct ifa6_config *cfg,
 			goto out;
 	}
 
-	ifa = kzalloc(sizeof(*ifa), gfp_flags);
+	ifa = kzalloc(sizeof(*ifa), gfp_flags | __GFP_ACCOUNT);
 	if (!ifa) {
 		err = -ENOBUFS;
 		goto out;

net/ipv6/ip6_fib.c

@@ -2449,8 +2449,8 @@ int __init fib6_init(void)
 	int ret = -ENOMEM;
 
 	fib6_node_kmem = kmem_cache_create("fib6_nodes",
-					   sizeof(struct fib6_node),
-					   0, SLAB_HWCACHE_ALIGN,
+					   sizeof(struct fib6_node), 0,
+					   SLAB_HWCACHE_ALIGN | SLAB_ACCOUNT,
 					   NULL);
 	if (!fib6_node_kmem)
 		goto out;

net/ipv6/route.c

@@ -6638,7 +6638,7 @@ int __init ip6_route_init(void)
 	ret = -ENOMEM;
 	ip6_dst_ops_template.kmem_cachep =
 		kmem_cache_create("ip6_dst_cache", sizeof(struct rt6_info), 0,
-				  SLAB_HWCACHE_ALIGN, NULL);
+				  SLAB_HWCACHE_ALIGN | SLAB_ACCOUNT, NULL);
 	if (!ip6_dst_ops_template.kmem_cachep)
 		goto out;
 

net/ipv6/sit.c

@@ -321,7 +321,7 @@ static int ipip6_tunnel_get_prl(struct net_device *dev, struct ifreq *ifr)
 	 * we try harder to allocate.
 	 */
 	kp = (cmax <= 1 || capable(CAP_NET_ADMIN)) ?
-		kcalloc(cmax, sizeof(*kp), GFP_KERNEL | __GFP_NOWARN) :
+		kcalloc(cmax, sizeof(*kp), GFP_KERNEL_ACCOUNT | __GFP_NOWARN) :
 		NULL;
 
 	rcu_read_lock();
@@ -334,7 +334,8 @@ static int ipip6_tunnel_get_prl(struct net_device *dev, struct ifreq *ifr)
 		 * For root users, retry allocating enough memory for
 		 * the answer.
 		 */
-		kp = kcalloc(ca, sizeof(*kp), GFP_ATOMIC);
+		kp = kcalloc(ca, sizeof(*kp), GFP_ATOMIC | __GFP_ACCOUNT |
+					      __GFP_NOWARN);
 		if (!kp) {
 			ret = -ENOMEM;
 			goto out;

tools/testing/selftests/net/veth.sh

@@ -13,7 +13,7 @@ readonly NS_DST=$BASE$DST
 readonly BM_NET_V4=192.168.1.
 readonly BM_NET_V6=2001:db8::
 
-readonly NPROCS=`nproc`
+readonly CPUS=`nproc`
 ret=0
 
 cleanup() {
@@ -75,6 +75,29 @@ chk_tso_flag() {
 	__chk_flag "$1" $2 $3 tcp-segmentation-offload
 }
 
+chk_channels() {
+	local msg="$1"
+	local target=$2
+	local rx=$3
+	local tx=$4
+
+	local dev=veth$target
+
+	local cur_rx=`ip netns exec $BASE$target ethtool -l $dev |\
+		grep RX: | tail -n 1 | awk '{print $2}' `
+		local cur_tx=`ip netns exec $BASE$target ethtool -l $dev |\
+		grep TX: | tail -n 1 | awk '{print $2}'`
+	local cur_combined=`ip netns exec $BASE$target ethtool -l $dev |\
+		grep Combined: | tail -n 1 | awk '{print $2}'`
+
+	printf "%-60s" "$msg"
+	if [ "$cur_rx" = "$rx" -a "$cur_tx" = "$tx" -a "$cur_combined" = "n/a" ]; then
+		echo " ok "
+	else
+		echo " fail rx:$rx:$cur_rx tx:$tx:$cur_tx combined:n/a:$cur_combined"
+	fi
+}
+
 chk_gro() {
 	local msg="$1"
 	local expected=$2
@@ -107,11 +130,100 @@ chk_gro() {
 	fi
 }
 
+__change_channels()
+{
+	local cur_cpu
+	local end=$1
+	local cur
+	local i
+
+	while true; do
+		printf -v cur '%(%s)T'
+		[ $cur -le $end ] || break
+
+		for i in `seq 1 $CPUS`; do
+			ip netns exec $NS_SRC ethtool -L veth$SRC rx $i tx $i
+			ip netns exec $NS_DST ethtool -L veth$DST rx $i tx $i
+		done
+
+		for i in `seq 1 $((CPUS - 1))`; do
+			cur_cpu=$((CPUS - $i))
+			ip netns exec $NS_SRC ethtool -L veth$SRC rx $cur_cpu tx $cur_cpu
+			ip netns exec $NS_DST ethtool -L veth$DST rx $cur_cpu tx $cur_cpu
+		done
+	done
+}
+
+__send_data() {
+	local end=$1
+
+	while true; do
+		printf -v cur '%(%s)T'
+		[ $cur -le $end ] || break
+
+		ip netns exec $NS_SRC ./udpgso_bench_tx -4 -s 1000 -M 300 -D $BM_NET_V4$DST
+	done
+}
+
+do_stress() {
+	local end
+	printf -v end '%(%s)T'
+	end=$((end + $STRESS))
+
+	ip netns exec $NS_SRC ethtool -L veth$SRC rx 3 tx 3
+	ip netns exec $NS_DST ethtool -L veth$DST rx 3 tx 3
+
+	ip netns exec $NS_DST ./udpgso_bench_rx &
+	local rx_pid=$!
+
+	echo "Running stress test for $STRESS seconds..."
+	__change_channels $end &
+	local ch_pid=$!
+	__send_data $end &
+	local data_pid_1=$!
+	__send_data $end &
+	local data_pid_2=$!
+	__send_data $end &
+	local data_pid_3=$!
+	__send_data $end &
+	local data_pid_4=$!
+
+	wait $ch_pid $data_pid_1 $data_pid_2 $data_pid_3 $data_pid_4
+	kill -9 $rx_pid
+	echo "done"
+
+	# restore previous setting
+	ip netns exec $NS_SRC ethtool -L veth$SRC rx 2 tx 2
+	ip netns exec $NS_DST ethtool -L veth$DST rx 2 tx 1
+}
+
+usage() {
+	echo "Usage: $0 [-h] [-s <seconds>]"
+	echo -e "\t-h: show this help"
+	echo -e "\t-s: run optional stress tests for the given amount of seconds"
+}
+
+STRESS=0
+while getopts "hs:" option; do
+	case "$option" in
+	"h")
+		usage $0
+		exit 0
+		;;
+	"s")
+		STRESS=$OPTARG
+		;;
+	esac
+done
+
 if [ ! -f ../bpf/xdp_dummy.o ]; then
 	echo "Missing xdp_dummy helper. Build bpf selftest first"
 	exit 1
 fi
 
+[ $CPUS -lt 2 ] && echo "Only one CPU available, some tests will be skipped"
+[ $STRESS -gt 0 -a $CPUS -lt 3 ] && echo " stress test will be skipped, too"
+
 create_ns
 chk_gro_flag "default - gro flag" $SRC off
 chk_gro_flag "        - peer gro flag" $DST off
@@ -134,6 +246,8 @@ chk_gro "        - aggregation with TSO off" 1
 cleanup
 
 create_ns
+chk_channels "default channels" $DST 1 1
+
 ip -n $NS_DST link set dev veth$DST down
 ip netns exec $NS_DST ethtool -K veth$DST gro on
 chk_gro_flag "with gro enabled on link down - gro flag" $DST on
@@ -147,6 +261,56 @@ chk_gro "        - aggregation with TSO off" 1
 cleanup
 
 create_ns
+
+CUR_TX=1
+CUR_RX=1
+if [ $CPUS -gt 1 ]; then
+	ip netns exec $NS_DST ethtool -L veth$DST tx 2
+	chk_channels "setting tx channels" $DST 1 2
+	CUR_TX=2
+fi
+
+if [ $CPUS -gt 2 ]; then
+	ip netns exec $NS_DST ethtool -L veth$DST rx 3 tx 3
+	chk_channels "setting both rx and tx channels" $DST 3 3
+	CUR_RX=3
+	CUR_TX=3
+fi
+
+ip netns exec $NS_DST ethtool -L veth$DST combined 2 2>/dev/null
+chk_channels "bad setting: combined channels" $DST $CUR_RX $CUR_TX
+
+ip netns exec $NS_DST ethtool -L veth$DST tx $((CPUS + 1)) 2>/dev/null
+chk_channels "setting invalid channels nr" $DST $CUR_RX $CUR_TX
+
+if [ $CPUS -gt 1 ]; then
+	# this also tests queues nr reduction
+	ip netns exec $NS_DST ethtool -L veth$DST rx 1 tx 2 2>/dev/null
+	ip netns exec $NS_SRC ethtool -L veth$SRC rx 1 tx 2 2>/dev/null
+	printf "%-60s" "bad setting: XDP with RX nr less than TX"
+	ip -n $NS_DST link set dev veth$DST xdp object ../bpf/xdp_dummy.o \
+		section xdp_dummy 2>/dev/null &&\
+		echo "fail - set operation successful ?!?" || echo " ok "
+
+	# the following tests will run with multiple channels active
+	ip netns exec $NS_SRC ethtool -L veth$SRC rx 2
+	ip netns exec $NS_DST ethtool -L veth$DST rx 2
+	ip -n $NS_DST link set dev veth$DST xdp object ../bpf/xdp_dummy.o \
+		section xdp_dummy 2>/dev/null
+	printf "%-60s" "bad setting: reducing RX nr below peer TX with XDP set"
+	ip netns exec $NS_DST ethtool -L veth$DST rx 1 2>/dev/null &&\
+		echo "fail - set operation successful ?!?" || echo " ok "
+	CUR_RX=2
+	CUR_TX=2
+fi
+
+if [ $CPUS -gt 2 ]; then
+	printf "%-60s" "bad setting: increasing peer TX nr above RX with XDP set"
+	ip netns exec $NS_SRC ethtool -L veth$SRC tx 3 2>/dev/null &&\
+		echo "fail - set operation successful ?!?" || echo " ok "
+	chk_channels "setting invalid channels nr" $DST 2 2
+fi
+
 ip -n $NS_DST link set dev veth$DST xdp object ../bpf/xdp_dummy.o section xdp_dummy 2>/dev/null
 chk_gro_flag "with xdp attached - gro flag" $DST on
 chk_gro_flag "        - peer gro flag" $SRC off
@@ -167,10 +331,27 @@ chk_gro_flag "        - after gro on xdp off, gro flag" $DST on
 chk_gro_flag "        - peer gro flag" $SRC off
 chk_tso_flag "        - tso flag" $SRC on
 chk_tso_flag "        - peer tso flag" $DST on
+
+if [ $CPUS -gt 1 ]; then
+	ip netns exec $NS_DST ethtool -L veth$DST tx 1
+	chk_channels "decreasing tx channels with device down" $DST 2 1
+fi
+
 ip -n $NS_DST link set dev veth$DST up
 ip -n $NS_SRC link set dev veth$SRC up
 chk_gro "        - aggregation" 1
 
+if [ $CPUS -gt 1 ]; then
+	[ $STRESS -gt 0 -a $CPUS -gt 2 ] && do_stress
+
+	ip -n $NS_DST link set dev veth$DST down
+	ip -n $NS_SRC link set dev veth$SRC down
+	ip netns exec $NS_DST ethtool -L veth$DST tx 2
+	chk_channels "increasing tx channels with device down" $DST 2 2
+	ip -n $NS_DST link set dev veth$DST up
+	ip -n $NS_SRC link set dev veth$SRC up
+fi
+
 ip netns exec $NS_DST ethtool -K veth$DST gro off
 ip netns exec $NS_SRC ethtool -K veth$SRC tx-udp-segmentation off
 chk_gro "aggregation again with default and TSO off" 10