[PATCH V1 net-next 0/4] Add destination MAC address to ethtool flow steering

STALE4995d

Revision v1 of 2 in this series.

8 messages, 3 authors, 2012-12-12 · open the first message on its own page

[PATCH V1 net-next 0/4] Add destination MAC address to ethtool flow steering

From: Amir Vadai <hidden>
Date: 2012-12-12 12:13:37

From: Yan Burman <redacted>

In vSwitch configuration it is often beneficial to create flow steering
rules for L3/L4 traffic based on VM port. This requires destination MAC
address of that port to be present. Note that today the mlx4_en driver 
adds the mac address of itself to the flow spec, where under the new
ethtool flag suggested here it doesn't.

It may also be useful in macvlan devices.

These patches add kernel support for the new field (does not break old
userspace compatibility, so new ethtool will work on old kernels and
old ethtool will work with new kernels).

Also present here is the ethtool userspace patch.

See more details here http ://marc.info/?t=134977576500003

Changes from V0:
- Get rid of full_mac, zero_mac in favour of
    is_zero_ether_addr and is_broadcast_ether_addr

Yan Burman (3):
  net: ethtool: Add destination MAC address to flow steering API
  net/mlx4_en: Use generic etherdevice.h functions.
  net/mlx4_en: Add support for destination MAC in steering rules

 drivers/net/ethernet/mellanox/mlx4/en_ethtool.c | 27 ++++++++++++++++---------
 include/uapi/linux/ethtool.h                    | 11 ++++++----
 2 files changed, 24 insertions(+), 14 deletions(-)

-- 
1.7.11.3

[PATCH V1 net-next 2/3] net/mlx4_en: Use generic etherdevice.h functions.

From: Amir Vadai <hidden>
Date: 2012-12-12 12:13:37

From: Yan Burman <redacted>

Get rid of full_mac, zero_mac in favour of
is_zero_ether_addr and is_broadcast_ether_addr.

Signed-off-by: Yan Burman <redacted>
Signed-off-by: Amir Vadai <redacted>
---
 drivers/net/ethernet/mellanox/mlx4/en_ethtool.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
index 4aaa7c3..cc7bb25 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c

@@ -613,8 +613,6 @@ static int mlx4_en_validate_flow(struct net_device *dev,
 	struct ethtool_usrip4_spec *l3_mask;
 	struct ethtool_tcpip4_spec *l4_mask;
 	struct ethhdr *eth_mask;
-	u64 full_mac = ~0ull;
-	u64 zero_mac = 0;
 
 	if (cmd->fs.location >= MAX_NUM_OF_FS_RULES)
 		return -EINVAL;

@@ -644,11 +642,11 @@ static int mlx4_en_validate_flow(struct net_device *dev,
 	case ETHER_FLOW:
 		eth_mask = &cmd->fs.m_u.ether_spec;
 		/* source mac mask must not be set */
-		if (memcmp(eth_mask->h_source, &zero_mac, ETH_ALEN))
+		if (!is_zero_ether_addr(eth_mask->h_source))
 			return -EINVAL;
 
 		/* dest mac mask must be ff:ff:ff:ff:ff:ff */
-		if (memcmp(eth_mask->h_dest, &full_mac, ETH_ALEN))
+		if (!is_broadcast_ether_addr(eth_mask->h_dest))
 			return -EINVAL;
 
 		if (!all_zeros_or_all_ones(eth_mask->h_proto))

-- 
1.7.11.3

[PATCH V1 net-next 1/3] net: ethtool: Add destination MAC address to flow steering API

From: Amir Vadai <hidden>
Date: 2012-12-12 12:13:37

From: Yan Burman <redacted>

Add ability to specify destination MAC address for L3/L4 flow spec
in order to be able to specify action for different VM's under vSwitch
configuration. This change is transparent to older userspace.

Signed-off-by: Yan Burman <redacted>
Signed-off-by: Amir Vadai <redacted>
---
 include/uapi/linux/ethtool.h | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/include/uapi/linux/ethtool.h b/include/uapi/linux/ethtool.h
index d3eaaaf..be8c41e 100644
--- a/include/uapi/linux/ethtool.h
+++ b/include/uapi/linux/ethtool.h

@@ -500,13 +500,15 @@ union ethtool_flow_union {
 	struct ethtool_ah_espip4_spec		esp_ip4_spec;
 	struct ethtool_usrip4_spec		usr_ip4_spec;
 	struct ethhdr				ether_spec;
-	__u8					hdata[60];
+	__u8					hdata[52];
 };
 
 struct ethtool_flow_ext {
-	__be16	vlan_etype;
-	__be16	vlan_tci;
-	__be32	data[2];
+	__u8		padding[2];
+	unsigned char	h_dest[ETH_ALEN];	/* destination eth addr	*/
+	__be16		vlan_etype;
+	__be16		vlan_tci;
+	__be32		data[2];
 };
 
 /**

@@ -1027,6 +1029,7 @@ enum ethtool_sfeatures_retval_bits {
 #define	ETHER_FLOW	0x12	/* spec only (ether_spec) */
 /* Flag to enable additional fields in struct ethtool_rx_flow_spec */
 #define	FLOW_EXT	0x80000000
+#define	FLOW_MAC_EXT	0x40000000
 
 /* L3-L4 network traffic flow hash options */
 #define	RXH_L2DA	(1 << 1)

-- 
1.7.11.3

[PATCH ETHTOOL] Added dst-mac parameter for L3/L4 flow spec rules. This is usefull in vSwitch configurations.

From: Amir Vadai <hidden>
Date: 2012-12-12 12:13:37

From: Yan Burman <redacted>

Signed-off-by: Yan Burman <redacted>
Signed-off-by: Amir Vadai <redacted>
---
 ethtool-copy.h | 11 +++++++----
 ethtool.8.in   |  6 ++++++
 ethtool.c      |  5 +++++
 rxclass.c      | 62 ++++++++++++++++++++++++++++++++++++++++------------------
 4 files changed, 61 insertions(+), 23 deletions(-)

diff --git a/ethtool-copy.h b/ethtool-copy.h
index 4801eef..d352f20 100644
--- a/ethtool-copy.h
+++ b/ethtool-copy.h

@@ -500,13 +500,15 @@ union ethtool_flow_union {
 	struct ethtool_ah_espip4_spec		esp_ip4_spec;
 	struct ethtool_usrip4_spec		usr_ip4_spec;
 	struct ethhdr				ether_spec;
-	__u8					hdata[60];
+	__u8					hdata[52];
 };
 
 struct ethtool_flow_ext {
-	__be16	vlan_etype;
-	__be16	vlan_tci;
-	__be32	data[2];
+	__u8		padding[2];
+	unsigned char	h_dest[ETH_ALEN];	/* destination eth addr	*/
+	__be16		vlan_etype;
+	__be16		vlan_tci;
+	__be32		data[2];
 };
 
 /**

@@ -1027,6 +1029,7 @@ enum ethtool_sfeatures_retval_bits {
 #define	ETHER_FLOW	0x12	/* spec only (ether_spec) */
 /* Flag to enable additional fields in struct ethtool_rx_flow_spec */
 #define	FLOW_EXT	0x80000000
+#define	FLOW_MAC_EXT	0x40000000
 
 /* L3-L4 network traffic flow hash options */
 #define	RXH_L2DA	(1 << 1)

diff --git a/ethtool.8.in b/ethtool.8.in
index e701919..a52e484 100644
--- a/ethtool.8.in
+++ b/ethtool.8.in

@@ -268,6 +268,7 @@ ethtool \- query or control network driver and hardware settings
 .BM vlan\-etype
 .BM vlan
 .BM user\-def
+.RB [ dst-mac \ \*(MA\ [ m \ \*(MA]]
 .BN action
 .BN loc
 .RB |

@@ -739,6 +740,11 @@ Includes the VLAN tag and an optional mask.
 .BI user\-def \ N \\fR\ [\\fPm \ N \\fR]\\fP
 Includes 64-bits of user-specific data and an optional mask.
 .TP
+.BR dst-mac \ \*(MA\ [ m \ \*(MA]
+Includes the destination MAC address, specified as 6 bytes in hexadecimal
+separated by colons, along with an optional mask.
+Valid for all IPv4 based flow-types.
+.TP
 .BI action \ N
 Specifies the Rx queue to send packets to, or some other action.
 .TS

diff --git a/ethtool.c b/ethtool.c
index 345c21c..55bc082 100644
--- a/ethtool.c
+++ b/ethtool.c

@@ -3231,6 +3231,10 @@ static int flow_spec_to_ntuple(struct ethtool_rx_flow_spec *fsp,
 	if (fsp->location != RX_CLS_LOC_ANY)
 		return -1;
 
+	/* destination MAC address in L3/L4 rules is not supported by ntuple */
+	if (fsp->flow_type & FLOW_MAC_EXT)
+		return -1;
+
 	/* verify ring cookie can transfer to action */
 	if (fsp->ring_cookie > INT_MAX && fsp->ring_cookie < (u64)(-2))
 		return -1;

@@ -3814,6 +3818,7 @@ static const struct option {
 	  "			[ vlan-etype %x [m %x] ]\n"
 	  "			[ vlan %x [m %x] ]\n"
 	  "			[ user-def %x [m %x] ]\n"
+	  "			[ dst-mac %x:%x:%x:%x:%x:%x [m %x:%x:%x:%x:%x:%x] ]\n"
 	  "			[ action %d ]\n"
 	  "			[ loc %d]] |\n"
 	  "		delete %d\n" },

diff --git a/rxclass.c b/rxclass.c
index e1633a8..1564b62 100644
--- a/rxclass.c
+++ b/rxclass.c

@@ -41,26 +41,38 @@ static void rxclass_print_ipv4_rule(__be32 sip, __be32 sipm, __be32 dip,
 
 static void rxclass_print_nfc_spec_ext(struct ethtool_rx_flow_spec *fsp)
 {
-	u64 data, datam;
-	__u16 etype, etypem, tci, tcim;
+	if (fsp->flow_type & FLOW_EXT) {
+		u64 data, datam;
+		__u16 etype, etypem, tci, tcim;
+		etype = ntohs(fsp->h_ext.vlan_etype);
+		etypem = ntohs(~fsp->m_ext.vlan_etype);
+		tci = ntohs(fsp->h_ext.vlan_tci);
+		tcim = ntohs(~fsp->m_ext.vlan_tci);
+		data = (u64)ntohl(fsp->h_ext.data[0]) << 32;
+		data = (u64)ntohl(fsp->h_ext.data[1]);
+		datam = (u64)ntohl(~fsp->m_ext.data[0]) << 32;
+		datam |= (u64)ntohl(~fsp->m_ext.data[1]);
 
-	if (!(fsp->flow_type & FLOW_EXT))
-		return;
+		fprintf(stdout,
+			"\tVLAN EtherType: 0x%x mask: 0x%x\n"
+			"\tVLAN: 0x%x mask: 0x%x\n"
+			"\tUser-defined: 0x%llx mask: 0x%llx\n",
+			etype, etypem, tci, tcim, data, datam);
+	}
 
-	etype = ntohs(fsp->h_ext.vlan_etype);
-	etypem = ntohs(~fsp->m_ext.vlan_etype);
-	tci = ntohs(fsp->h_ext.vlan_tci);
-	tcim = ntohs(~fsp->m_ext.vlan_tci);
-	data = (u64)ntohl(fsp->h_ext.data[0]) << 32;
-	data = (u64)ntohl(fsp->h_ext.data[1]);
-	datam = (u64)ntohl(~fsp->m_ext.data[0]) << 32;
-	datam |= (u64)ntohl(~fsp->m_ext.data[1]);
+	if (fsp->flow_type & FLOW_MAC_EXT) {
+		unsigned char *dmac, *dmacm;
 
-	fprintf(stdout,
-		"\tVLAN EtherType: 0x%x mask: 0x%x\n"
-		"\tVLAN: 0x%x mask: 0x%x\n"
-		"\tUser-defined: 0x%llx mask: 0x%llx\n",
-		etype, etypem, tci, tcim, data, datam);
+		dmac = fsp->h_ext.h_dest;
+		dmacm = fsp->m_ext.h_dest;
+
+		fprintf(stdout,
+			"\tDest MAC addr: %02X:%02X:%02X:%02X:%02X:%02X"
+			" mask: %02X:%02X:%02X:%02X:%02X:%02X\n",
+			dmac[0], dmac[1], dmac[2], dmac[3], dmac[4],
+			dmac[5], dmacm[0], dmacm[1], dmacm[2], dmacm[3],
+			dmacm[4], dmacm[5]);
+	}
 }
 
 static void rxclass_print_nfc_rule(struct ethtool_rx_flow_spec *fsp)

@@ -70,7 +82,7 @@ static void rxclass_print_nfc_rule(struct ethtool_rx_flow_spec *fsp)
 
 	fprintf(stdout,	"Filter: %d\n", fsp->location);
 
-	flow_type = fsp->flow_type & ~FLOW_EXT;
+	flow_type = fsp->flow_type & ~(FLOW_EXT | FLOW_MAC_EXT);
 
 	invert_flow_mask(fsp);

@@ -172,7 +184,7 @@ static void rxclass_print_nfc_rule(struct ethtool_rx_flow_spec *fsp)
 static void rxclass_print_rule(struct ethtool_rx_flow_spec *fsp)
 {
 	/* print the rule in this location */
-	switch (fsp->flow_type & ~FLOW_EXT) {
+	switch (fsp->flow_type & ~(FLOW_EXT | FLOW_MAC_EXT)) {
 	case TCP_V4_FLOW:
 	case UDP_V4_FLOW:
 	case SCTP_V4_FLOW:

@@ -533,6 +545,7 @@ typedef enum {
 #define NTUPLE_FLAG_VLAN	0x100
 #define NTUPLE_FLAG_UDEF	0x200
 #define NTUPLE_FLAG_VETH	0x400
+#define NFC_FLAG_MAC_ADDR	0x800
 
 struct rule_opts {
 	const char	*name;

@@ -571,6 +584,9 @@ static const struct rule_opts rule_nfc_tcp_ip4[] = {
 	{ "user-def", OPT_BE64, NTUPLE_FLAG_UDEF,
 	  offsetof(struct ethtool_rx_flow_spec, h_ext.data),
 	  offsetof(struct ethtool_rx_flow_spec, m_ext.data) },
+	{ "dst-mac", OPT_MAC, NFC_FLAG_MAC_ADDR,
+	  offsetof(struct ethtool_rx_flow_spec, h_ext.h_dest),
+	  offsetof(struct ethtool_rx_flow_spec, m_ext.h_dest) },
 };
 
 static const struct rule_opts rule_nfc_esp_ip4[] = {

@@ -599,6 +615,9 @@ static const struct rule_opts rule_nfc_esp_ip4[] = {
 	{ "user-def", OPT_BE64, NTUPLE_FLAG_UDEF,
 	  offsetof(struct ethtool_rx_flow_spec, h_ext.data),
 	  offsetof(struct ethtool_rx_flow_spec, m_ext.data) },
+	{ "dst-mac", OPT_MAC, NFC_FLAG_MAC_ADDR,
+	  offsetof(struct ethtool_rx_flow_spec, h_ext.h_dest),
+	  offsetof(struct ethtool_rx_flow_spec, m_ext.h_dest) },
 };
 
 static const struct rule_opts rule_nfc_usr_ip4[] = {

@@ -639,6 +658,9 @@ static const struct rule_opts rule_nfc_usr_ip4[] = {
 	{ "user-def", OPT_BE64, NTUPLE_FLAG_UDEF,
 	  offsetof(struct ethtool_rx_flow_spec, h_ext.data),
 	  offsetof(struct ethtool_rx_flow_spec, m_ext.data) },
+	{ "dst-mac", OPT_MAC, NFC_FLAG_MAC_ADDR,
+	  offsetof(struct ethtool_rx_flow_spec, h_ext.h_dest),
+	  offsetof(struct ethtool_rx_flow_spec, m_ext.h_dest) },
 };
 
 static const struct rule_opts rule_nfc_ether[] = {

@@ -1063,6 +1085,8 @@ int rxclass_parse_ruleopts(struct cmd_context *ctx,
 		fsp->h_u.usr_ip4_spec.ip_ver = ETH_RX_NFC_IP4;
 	if (flags & (NTUPLE_FLAG_VLAN | NTUPLE_FLAG_UDEF | NTUPLE_FLAG_VETH))
 		fsp->flow_type |= FLOW_EXT;
+	if (flags & NFC_FLAG_MAC_ADDR)
+		fsp->flow_type |= FLOW_MAC_EXT;
 
 	return 0;

-- 
1.7.11.3

[PATCH V1 net-next 3/3] net/mlx4_en: Add support for destination MAC in steering rules

From: Amir Vadai <hidden>
Date: 2012-12-12 12:13:37

From: Yan Burman <redacted>

Implement destination MAC rule extension for L3/L4 rules in
flow steering. Usefull for vSwitch/macvlan configurations.

Signed-off-by: Yan Burman <redacted>
Signed-off-by: Amir Vadai <redacted>
---
 drivers/net/ethernet/mellanox/mlx4/en_ethtool.c | 21 +++++++++++++++------
 1 file changed, 15 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
index cc7bb25..03447da 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_ethtool.c

@@ -617,7 +617,13 @@ static int mlx4_en_validate_flow(struct net_device *dev,
 	if (cmd->fs.location >= MAX_NUM_OF_FS_RULES)
 		return -EINVAL;
 
-	switch (cmd->fs.flow_type & ~FLOW_EXT) {
+	if (cmd->fs.flow_type & FLOW_MAC_EXT) {
+		/* dest mac mask must be ff:ff:ff:ff:ff:ff */
+		if (!is_broadcast_ether_addr(cmd->fs.m_ext.h_dest))
+			return -EINVAL;
+	}
+
+	switch (cmd->fs.flow_type & ~(FLOW_EXT | FLOW_MAC_EXT)) {
 	case TCP_V4_FLOW:
 	case UDP_V4_FLOW:
 		if (cmd->fs.m_u.tcp_ip4_spec.tos)

@@ -745,7 +751,6 @@ static int mlx4_en_ethtool_to_net_trans_rule(struct net_device *dev,
 					     struct list_head *rule_list_h)
 {
 	int err;
-	u64 mac;
 	__be64 be_mac;
 	struct ethhdr *eth_spec;
 	struct mlx4_en_priv *priv = netdev_priv(dev);

@@ -760,12 +765,16 @@ static int mlx4_en_ethtool_to_net_trans_rule(struct net_device *dev,
 	if (!spec_l2)
 		return -ENOMEM;
 
-	mac = priv->mac & MLX4_MAC_MASK;
-	be_mac = cpu_to_be64(mac << 16);
+	if (cmd->fs.flow_type & FLOW_MAC_EXT) {
+		memcpy(&be_mac, cmd->fs.h_ext.h_dest, ETH_ALEN);
+	} else {
+		u64 mac = priv->mac & MLX4_MAC_MASK;
+		be_mac = cpu_to_be64(mac << 16);
+	}
 
 	spec_l2->id = MLX4_NET_TRANS_RULE_ID_ETH;
 	memcpy(spec_l2->eth.dst_mac_msk, &mac_msk, ETH_ALEN);
-	if ((cmd->fs.flow_type & ~FLOW_EXT) != ETHER_FLOW)
+	if ((cmd->fs.flow_type & ~(FLOW_EXT | FLOW_MAC_EXT)) != ETHER_FLOW)
 		memcpy(spec_l2->eth.dst_mac, &be_mac, ETH_ALEN);
 
 	if ((cmd->fs.flow_type & FLOW_EXT) && cmd->fs.m_ext.vlan_tci) {

@@ -775,7 +784,7 @@ static int mlx4_en_ethtool_to_net_trans_rule(struct net_device *dev,
 
 	list_add_tail(&spec_l2->list, rule_list_h);
 
-	switch (cmd->fs.flow_type & ~FLOW_EXT) {
+	switch (cmd->fs.flow_type & ~(FLOW_EXT | FLOW_MAC_EXT)) {
 	case ETHER_FLOW:
 		eth_spec = &cmd->fs.h_u.ether_spec;
 		memcpy(&spec_l2->eth.dst_mac, eth_spec->h_dest, ETH_ALEN);

-- 
1.7.11.3

Re: [PATCH V1 net-next 0/4] Add destination MAC address to ethtool flow steering

From: David Miller <davem@davemloft.net>
Date: 2012-12-12 18:03:11

From: Amir Vadai <redacted>
Date: Wed, 12 Dec 2012 14:13:16 +0200

From: Yan Burman <redacted>

In vSwitch configuration it is often beneficial to create flow steering
rules for L3/L4 traffic based on VM port. This requires destination MAC
address of that port to be present. Note that today the mlx4_en driver 
adds the mac address of itself to the flow spec, where under the new
ethtool flag suggested here it doesn't.

It may also be useful in macvlan devices.

These patches add kernel support for the new field (does not break old
userspace compatibility, so new ethtool will work on old kernels and
old ethtool will work with new kernels).

Also present here is the ethtool userspace patch.

See more details here http ://marc.info/?t=134977576500003

Kernel side applied to net-next, thanks.

Re: [PATCH V1 net-next 1/3] net: ethtool: Add destination MAC address to flow steering API

From: Ben Hutchings <hidden>
Date: 2012-12-12 18:28:06

On Wed, 2012-12-12 at 14:13 +0200, Amir Vadai wrote:

quoted hunk

From: Yan Burman <redacted>

Add ability to specify destination MAC address for L3/L4 flow spec
in order to be able to specify action for different VM's under vSwitch
configuration. This change is transparent to older userspace.

Signed-off-by: Yan Burman <redacted>
Signed-off-by: Amir Vadai <redacted>
---
 include/uapi/linux/ethtool.h | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/include/uapi/linux/ethtool.h b/include/uapi/linux/ethtool.h
index d3eaaaf..be8c41e 100644
--- a/include/uapi/linux/ethtool.h
+++ b/include/uapi/linux/ethtool.h

@@ -500,13 +500,15 @@ union ethtool_flow_union {
 	struct ethtool_ah_espip4_spec		esp_ip4_spec;
 	struct ethtool_usrip4_spec		usr_ip4_spec;
 	struct ethhdr				ether_spec;
-	__u8					hdata[60];
+	__u8					hdata[52];
 };
 
 struct ethtool_flow_ext {
-	__be16	vlan_etype;
-	__be16	vlan_tci;
-	__be32	data[2];
+	__u8		padding[2];
+	unsigned char	h_dest[ETH_ALEN];	/* destination eth addr	*/
+	__be16		vlan_etype;
+	__be16		vlan_tci;
+	__be32		data[2];
 };
 
 /**

@@ -1027,6 +1029,7 @@ enum ethtool_sfeatures_retval_bits {
 #define	ETHER_FLOW	0x12	/* spec only (ether_spec) */
 /* Flag to enable additional fields in struct ethtool_rx_flow_spec */
 #define	FLOW_EXT	0x80000000
+#define	FLOW_MAC_EXT	0x40000000

Please can you send another patch that adds kernel-doc to struct
ethtool_flow_ext explaining which fields are dependent on which flags.

Ben.

 /* L3-L4 network traffic flow hash options */
 #define	RXH_L2DA	(1 << 1)

-- 
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.

Re: [PATCH V1 net-next 1/3] net: ethtool: Add destination MAC address to flow steering API

From: Amir Vadai <hidden>
Date: 2012-12-12 20:48:07

On 12/12/2012 20:28, Ben Hutchings wrote:

On Wed, 2012-12-12 at 14:13 +0200, Amir Vadai wrote:

quoted

From: Yan Burman <redacted>

Add ability to specify destination MAC address for L3/L4 flow spec
in order to be able to specify action for different VM's under vSwitch
configuration. This change is transparent to older userspace.

Signed-off-by: Yan Burman <redacted>
Signed-off-by: Amir Vadai <redacted>
---
  include/uapi/linux/ethtool.h | 11 +++++++----
  1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/include/uapi/linux/ethtool.h b/include/uapi/linux/ethtool.h
index d3eaaaf..be8c41e 100644
--- a/include/uapi/linux/ethtool.h
+++ b/include/uapi/linux/ethtool.h

@@ -500,13 +500,15 @@ union ethtool_flow_union {
  	struct ethtool_ah_espip4_spec		esp_ip4_spec;
  	struct ethtool_usrip4_spec		usr_ip4_spec;
  	struct ethhdr				ether_spec;
-	__u8					hdata[60];
+	__u8					hdata[52];
  };

  struct ethtool_flow_ext {
-	__be16	vlan_etype;
-	__be16	vlan_tci;
-	__be32	data[2];
+	__u8		padding[2];
+	unsigned char	h_dest[ETH_ALEN];	/* destination eth addr	*/
+	__be16		vlan_etype;
+	__be16		vlan_tci;
+	__be32		data[2];
  };

  /**

@@ -1027,6 +1029,7 @@ enum ethtool_sfeatures_retval_bits {
  #define	ETHER_FLOW	0x12	/* spec only (ether_spec) */
  /* Flag to enable additional fields in struct ethtool_rx_flow_spec */
  #define	FLOW_EXT	0x80000000
+#define	FLOW_MAC_EXT	0x40000000

Please can you send another patch that adds kernel-doc to struct
ethtool_flow_ext explaining which fields are dependent on which flags.

Ben.

quoted

  /* L3-L4 network traffic flow hash options */
  #define	RXH_L2DA	(1 << 1)

Sure. Will send it.

Amir

`h`	back out one level
`j`	next message in thread
`k`	previous message in thread
`l`	drill in
`Esc`	close help / fold thread tree
`?`	toggle this help