[PATCH net-next 0/4] sfc: EF100 TSO enhancements

[PATCH net-next 0/4] sfc: EF100 TSO enhancements

[Edward Cree]

Support TSO over encapsulation (with GSO_PARTIAL), and over VLANs
 (which the code already handled but we didn't advertise).  Also
 correct our handling of IPID mangling.

I couldn't find documentation of exactly what shaped SKBs we can
 get given, so patch #2 is slightly guesswork, but when I tested
 TSO over both underlay and (VxLAN) overlay, the checksums came
 out correctly, so at least in those cases the edits we're making
 must be the right ones.
Similarly, I'm not 100% sure I've correctly understood how FIXEDID
 and MANGLEID are supposed to work in patch #3.

Edward Cree (4):
  sfc: extend bitfield macros to 17 fields
  sfc: implement encap TSO on EF100
  sfc: only use fixed-id if the skb asks for it
  sfc: advertise our vlan features

 drivers/net/ethernet/sfc/bitfield.h  | 42 +++++++++++++++++---
 drivers/net/ethernet/sfc/ef100_nic.c | 17 ++++++--
 drivers/net/ethernet/sfc/ef100_tx.c  | 58 ++++++++++++++++------------
 3 files changed, 84 insertions(+), 33 deletions(-)

[PATCH net-next 1/4] sfc: extend bitfield macros to 17 fields

[Edward Cree]

We need EFX_POPULATE_OWORD_17 for an encap TSO descriptor on EF100.

Signed-off-by: Edward Cree <ecree@solarflare.com>
---
 drivers/net/ethernet/sfc/bitfield.h | 42 +++++++++++++++++++++++++----
 1 file changed, 37 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/sfc/bitfield.h b/drivers/net/ethernet/sfc/bitfield.h
index 2590cab53e54..64731eb5dd56 100644
--- a/drivers/net/ethernet/sfc/bitfield.h
+++ b/drivers/net/ethernet/sfc/bitfield.h
@@ -285,7 +285,11 @@ typedef union efx_oword {
 				 field10, value10,			\
 				 field11, value11,			\
 				 field12, value12,			\
-				 field13, value13)			\
+				 field13, value13,			\
+				 field14, value14,			\
+				 field15, value15,			\
+				 field16, value16,			\
+				 field17, value17)			\
 	(EFX_INSERT_FIELD_NATIVE((min), (max), field1, (value1)) |	\
 	 EFX_INSERT_FIELD_NATIVE((min), (max), field2, (value2)) |	\
 	 EFX_INSERT_FIELD_NATIVE((min), (max), field3, (value3)) |	\
@@ -298,7 +302,11 @@ typedef union efx_oword {
 	 EFX_INSERT_FIELD_NATIVE((min), (max), field10, (value10)) |	\
 	 EFX_INSERT_FIELD_NATIVE((min), (max), field11, (value11)) |	\
 	 EFX_INSERT_FIELD_NATIVE((min), (max), field12, (value12)) |	\
-	 EFX_INSERT_FIELD_NATIVE((min), (max), field13, (value13)))
+	 EFX_INSERT_FIELD_NATIVE((min), (max), field13, (value13)) |	\
+	 EFX_INSERT_FIELD_NATIVE((min), (max), field14, (value14)) |	\
+	 EFX_INSERT_FIELD_NATIVE((min), (max), field15, (value15)) |	\
+	 EFX_INSERT_FIELD_NATIVE((min), (max), field16, (value16)) |	\
+	 EFX_INSERT_FIELD_NATIVE((min), (max), field17, (value17)))
 
 #define EFX_INSERT_FIELDS64(...)				\
 	cpu_to_le64(EFX_INSERT_FIELDS_NATIVE(__VA_ARGS__))
@@ -340,7 +348,15 @@ typedef union efx_oword {
 #endif
 
 /* Populate an octword field with various numbers of arguments */
-#define EFX_POPULATE_OWORD_13 EFX_POPULATE_OWORD
+#define EFX_POPULATE_OWORD_17 EFX_POPULATE_OWORD
+#define EFX_POPULATE_OWORD_16(oword, ...) \
+	EFX_POPULATE_OWORD_17(oword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_OWORD_15(oword, ...) \
+	EFX_POPULATE_OWORD_16(oword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_OWORD_14(oword, ...) \
+	EFX_POPULATE_OWORD_15(oword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_OWORD_13(oword, ...) \
+	EFX_POPULATE_OWORD_14(oword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
 #define EFX_POPULATE_OWORD_12(oword, ...) \
 	EFX_POPULATE_OWORD_13(oword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
 #define EFX_POPULATE_OWORD_11(oword, ...) \
@@ -375,7 +391,15 @@ typedef union efx_oword {
 			     EFX_DWORD_3, 0xffffffff)
 
 /* Populate a quadword field with various numbers of arguments */
-#define EFX_POPULATE_QWORD_13 EFX_POPULATE_QWORD
+#define EFX_POPULATE_QWORD_17 EFX_POPULATE_QWORD
+#define EFX_POPULATE_QWORD_16(qword, ...) \
+	EFX_POPULATE_QWORD_17(qword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_QWORD_15(qword, ...) \
+	EFX_POPULATE_QWORD_16(qword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_QWORD_14(qword, ...) \
+	EFX_POPULATE_QWORD_15(qword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_QWORD_13(qword, ...) \
+	EFX_POPULATE_QWORD_14(qword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
 #define EFX_POPULATE_QWORD_12(qword, ...) \
 	EFX_POPULATE_QWORD_13(qword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
 #define EFX_POPULATE_QWORD_11(qword, ...) \
@@ -408,7 +432,15 @@ typedef union efx_oword {
 			     EFX_DWORD_1, 0xffffffff)
 
 /* Populate a dword field with various numbers of arguments */
-#define EFX_POPULATE_DWORD_13 EFX_POPULATE_DWORD
+#define EFX_POPULATE_DWORD_17 EFX_POPULATE_DWORD
+#define EFX_POPULATE_DWORD_16(dword, ...) \
+	EFX_POPULATE_DWORD_17(dword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_DWORD_15(dword, ...) \
+	EFX_POPULATE_DWORD_16(dword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_DWORD_14(dword, ...) \
+	EFX_POPULATE_DWORD_15(dword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
+#define EFX_POPULATE_DWORD_13(dword, ...) \
+	EFX_POPULATE_DWORD_14(dword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
 #define EFX_POPULATE_DWORD_12(dword, ...) \
 	EFX_POPULATE_DWORD_13(dword, EFX_DUMMY_FIELD, 0, __VA_ARGS__)
 #define EFX_POPULATE_DWORD_11(dword, ...) \

[PATCH net-next 2/4] sfc: implement encap TSO on EF100

[Edward Cree]

The NIC only needs to know where the headers it has to edit (TCP and
 inner and outer IPv4) are, which fits GSO_PARTIAL nicely.
It also supports non-PARTIAL offload of UDP tunnels, again just
 needing to be told the outer transport offset so that it can edit
 the UDP length field.
(It's not clear to me whether the stack will ever use the non-PARTIAL
 version with the netdev feature flags we're setting here.)

Signed-off-by: Edward Cree <ecree@solarflare.com>
---
 drivers/net/ethernet/sfc/ef100_nic.c | 13 ++++++--
 drivers/net/ethernet/sfc/ef100_tx.c  | 45 ++++++++++++++++------------
 2 files changed, 37 insertions(+), 21 deletions(-)

diff --git a/drivers/net/ethernet/sfc/ef100_nic.c b/drivers/net/ethernet/sfc/ef100_nic.c
index 3148fe770356..bf92cdc60cda 100644
--- a/drivers/net/ethernet/sfc/ef100_nic.c
+++ b/drivers/net/ethernet/sfc/ef100_nic.c
@@ -182,8 +182,16 @@ static int efx_ef100_init_datapath_caps(struct efx_nic *efx)
 	if (rc)
 		return rc;
 
-	if (efx_ef100_has_cap(nic_data->datapath_caps2, TX_TSO_V3))
-		efx->net_dev->features |= NETIF_F_TSO | NETIF_F_TSO6;
+	if (efx_ef100_has_cap(nic_data->datapath_caps2, TX_TSO_V3)) {
+		struct net_device *net_dev = efx->net_dev;
+		netdev_features_t tso = NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_GSO_PARTIAL |
+					NETIF_F_GSO_UDP_TUNNEL | NETIF_F_GSO_UDP_TUNNEL_CSUM;
+
+		net_dev->features |= tso;
+		net_dev->hw_features |= tso;
+		net_dev->hw_enc_features |= tso;
+		net_dev->gso_partial_features |= NETIF_F_GSO_UDP_TUNNEL | NETIF_F_GSO_UDP_TUNNEL_CSUM;
+	}
 	efx->num_mac_stats = MCDI_WORD(outbuf,
 				       GET_CAPABILITIES_V4_OUT_MAC_STATS_NUM_STATS);
 	netif_dbg(efx, probe, efx->net_dev,
@@ -1101,6 +1109,7 @@ static int ef100_probe_main(struct efx_nic *efx)
 	nic_data->efx = efx;
 	net_dev->features |= efx->type->offload_features;
 	net_dev->hw_features |= efx->type->offload_features;
+	net_dev->hw_enc_features |= efx->type->offload_features;
 
 	/* Populate design-parameter defaults */
 	nic_data->tso_max_hdr_len = ESE_EF100_DP_GZ_TSO_MAX_HDR_LEN_DEFAULT;
diff --git a/drivers/net/ethernet/sfc/ef100_tx.c b/drivers/net/ethernet/sfc/ef100_tx.c
index a90e5a9d2a37..d267b12bdaa0 100644
--- a/drivers/net/ethernet/sfc/ef100_tx.c
+++ b/drivers/net/ethernet/sfc/ef100_tx.c
@@ -54,8 +54,6 @@ static bool ef100_tx_can_tso(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
 	struct efx_nic *efx = tx_queue->efx;
 	struct ef100_nic_data *nic_data;
 	struct efx_tx_buffer *buffer;
-	struct tcphdr *tcphdr;
-	struct iphdr *iphdr;
 	size_t header_len;
 	u32 mss;
 
@@ -98,20 +96,6 @@ static bool ef100_tx_can_tso(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
 	buffer->unmap_len = 0;
 	buffer->skb = skb;
 	++tx_queue->insert_count;
-
-	/* Adjust the TCP checksum to exclude the total length, since we set
-	 * ED_INNER_IP_LEN in the descriptor.
-	 */
-	tcphdr = tcp_hdr(skb);
-	if (skb_is_gso_v6(skb)) {
-		tcphdr->check = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
-						 &ipv6_hdr(skb)->daddr,
-						 0, IPPROTO_TCP, 0);
-	} else {
-		iphdr = ip_hdr(skb);
-		tcphdr->check = ~csum_tcpudp_magic(iphdr->saddr, iphdr->daddr,
-						   0, IPPROTO_TCP, 0);
-	}
 	return true;
 }
 
@@ -209,17 +193,35 @@ static void ef100_make_tso_desc(struct efx_nic *efx,
 		ESE_GZ_TX_DESC_IP4_ID_INC_MOD16;
 	u16 vlan_enable =  efx->net_dev->features & NETIF_F_HW_VLAN_CTAG_TX ?
 		skb_vlan_tag_present(skb) : 0;
+	bool gso_partial = skb_shinfo(skb)->gso_type & SKB_GSO_PARTIAL;
 	unsigned int len, ip_offset, tcp_offset, payload_segs;
+	unsigned int outer_ip_offset, outer_l4_offset;
 	u16 vlan_tci = skb_vlan_tag_get(skb);
 	u32 mss = skb_shinfo(skb)->gso_size;
+	bool encap = skb->encapsulation;
+	struct tcphdr *tcp;
+	u32 paylen;
 
 	len = skb->len - buffer->len;
 	/* We use 1 for the TSO descriptor and 1 for the header */
 	payload_segs = segment_count - 2;
-	ip_offset =  skb_network_offset(skb);
-	tcp_offset = skb_transport_offset(skb);
+	if (encap) {
+		outer_ip_offset = skb_network_offset(skb);
+		outer_l4_offset = skb_transport_offset(skb);
+		ip_offset = skb_inner_network_offset(skb);
+		tcp_offset = skb_inner_transport_offset(skb);
+	} else {
+		ip_offset =  skb_network_offset(skb);
+		tcp_offset = skb_transport_offset(skb);
+		outer_ip_offset = outer_l4_offset = 0;
+	}
+
+	/* subtract TCP payload length from inner checksum */
+	tcp = (void *)skb->data + tcp_offset;
+	paylen = skb->len - tcp_offset;
+	csum_replace_by_diff(&tcp->check, (__force __wsum)htonl(paylen));
 
-	EFX_POPULATE_OWORD_13(*txd,
+	EFX_POPULATE_OWORD_17(*txd,
 			      ESF_GZ_TX_DESC_TYPE, ESE_GZ_TX_DESC_TYPE_TSO,
 			      ESF_GZ_TX_TSO_MSS, mss,
 			      ESF_GZ_TX_TSO_HDR_NUM_SEGS, 1,
@@ -231,6 +233,11 @@ static void ef100_make_tso_desc(struct efx_nic *efx,
 			      ESF_GZ_TX_TSO_INNER_L4_OFF_W, tcp_offset >> 1,
 			      ESF_GZ_TX_TSO_ED_INNER_IP4_ID, mangleid,
 			      ESF_GZ_TX_TSO_ED_INNER_IP_LEN, 1,
+			      ESF_GZ_TX_TSO_OUTER_L3_OFF_W, outer_ip_offset >> 1,
+			      ESF_GZ_TX_TSO_OUTER_L4_OFF_W, outer_l4_offset >> 1,
+			      ESF_GZ_TX_TSO_ED_OUTER_UDP_LEN, encap && !gso_partial,
+			      ESF_GZ_TX_TSO_ED_OUTER_IP4_ID, encap ? mangleid :
+								     ESE_GZ_TX_DESC_IP4_ID_NO_OP,
 			      ESF_GZ_TX_TSO_VLAN_INSERT_EN, vlan_enable,
 			      ESF_GZ_TX_TSO_VLAN_INSERT_TCI, vlan_tci
 		);

[PATCH net-next 3/4] sfc: only use fixed-id if the skb asks for it

[Edward Cree]

AIUI, the NETIF_F_TSO_MANGLEID flag is a signal to the stack that a
 driver may _need_ to mangle IDs in order to do TSO, and conversely
 a signal from the stack that the driver is permitted to do so.
Since we support both fixed and incrementing IPIDs, we should rely
 on the SKB_GSO_FIXEDID flag on a per-skb basis, rather than using
 the MANGLEID feature to make all TSOs fixed-id.
Includes other minor cleanups of ef100_make_tso_desc() coding style.

Signed-off-by: Edward Cree <ecree@solarflare.com>
---
 drivers/net/ethernet/sfc/ef100_nic.c |  2 +-
 drivers/net/ethernet/sfc/ef100_tx.c  | 13 +++++++------
 2 files changed, 8 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/sfc/ef100_nic.c b/drivers/net/ethernet/sfc/ef100_nic.c
index bf92cdc60cda..8a187a16ac89 100644
--- a/drivers/net/ethernet/sfc/ef100_nic.c
+++ b/drivers/net/ethernet/sfc/ef100_nic.c
@@ -694,7 +694,7 @@ static unsigned int ef100_check_caps(const struct efx_nic *efx,
 #define EF100_OFFLOAD_FEATURES	(NETIF_F_HW_CSUM | NETIF_F_RXCSUM |	\
 	NETIF_F_HIGHDMA | NETIF_F_SG | NETIF_F_FRAGLIST | NETIF_F_NTUPLE | \
 	NETIF_F_RXHASH | NETIF_F_RXFCS | NETIF_F_TSO_ECN | NETIF_F_RXALL | \
-	NETIF_F_TSO_MANGLEID | NETIF_F_HW_VLAN_CTAG_TX)
+	NETIF_F_HW_VLAN_CTAG_TX)
 
 const struct efx_nic_type ef100_pf_nic_type = {
 	.revision = EFX_REV_EF100,
diff --git a/drivers/net/ethernet/sfc/ef100_tx.c b/drivers/net/ethernet/sfc/ef100_tx.c
index d267b12bdaa0..ad0ad9bad423 100644
--- a/drivers/net/ethernet/sfc/ef100_tx.c
+++ b/drivers/net/ethernet/sfc/ef100_tx.c
@@ -187,21 +187,22 @@ static void ef100_make_tso_desc(struct efx_nic *efx,
 				struct efx_tx_buffer *buffer, efx_oword_t *txd,
 				unsigned int segment_count)
 {
-	u32 mangleid = (efx->net_dev->features & NETIF_F_TSO_MANGLEID) ||
-		skb_shinfo(skb)->gso_type & SKB_GSO_TCP_FIXEDID ?
-		ESE_GZ_TX_DESC_IP4_ID_NO_OP :
-		ESE_GZ_TX_DESC_IP4_ID_INC_MOD16;
-	u16 vlan_enable =  efx->net_dev->features & NETIF_F_HW_VLAN_CTAG_TX ?
-		skb_vlan_tag_present(skb) : 0;
 	bool gso_partial = skb_shinfo(skb)->gso_type & SKB_GSO_PARTIAL;
 	unsigned int len, ip_offset, tcp_offset, payload_segs;
+	u32 mangleid = ESE_GZ_TX_DESC_IP4_ID_INC_MOD16;
 	unsigned int outer_ip_offset, outer_l4_offset;
 	u16 vlan_tci = skb_vlan_tag_get(skb);
 	u32 mss = skb_shinfo(skb)->gso_size;
 	bool encap = skb->encapsulation;
+	u16 vlan_enable = 0;
 	struct tcphdr *tcp;
 	u32 paylen;
 
+	if (skb_shinfo(skb)->gso_type & SKB_GSO_TCP_FIXEDID)
+		mangleid = ESE_GZ_TX_DESC_IP4_ID_NO_OP;
+	if (efx->net_dev->features & NETIF_F_HW_VLAN_CTAG_TX)
+		vlan_enable = skb_vlan_tag_present(skb);
+
 	len = skb->len - buffer->len;
 	/* We use 1 for the TSO descriptor and 1 for the header */
 	payload_segs = segment_count - 2;

[PATCH net-next 4/4] sfc: advertise our vlan features

[Edward Cree]

Signed-off-by: Edward Cree <ecree@solarflare.com>
---
 drivers/net/ethernet/sfc/ef100_nic.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/net/ethernet/sfc/ef100_nic.c b/drivers/net/ethernet/sfc/ef100_nic.c
index 8a187a16ac89..cd93c5ffd45c 100644
--- a/drivers/net/ethernet/sfc/ef100_nic.c
+++ b/drivers/net/ethernet/sfc/ef100_nic.c
@@ -1110,6 +1110,8 @@ static int ef100_probe_main(struct efx_nic *efx)
 	net_dev->features |= efx->type->offload_features;
 	net_dev->hw_features |= efx->type->offload_features;
 	net_dev->hw_enc_features |= efx->type->offload_features;
+	net_dev->vlan_features |= NETIF_F_HW_CSUM | NETIF_F_SG |
+				  NETIF_F_HIGHDMA | NETIF_F_ALL_TSO;
 
 	/* Populate design-parameter defaults */
 	nic_data->tso_max_hdr_len = ESE_EF100_DP_GZ_TSO_MAX_HDR_LEN_DEFAULT;

Re: [PATCH net-next 2/4] sfc: implement encap TSO on EF100

[Willem de Bruijn]

On Wed, Oct 28, 2020 at 9:39 PM Edward Cree <ecree@solarflare.com> wrote:
>
> The NIC only needs to know where the headers it has to edit (TCP and
>  inner and outer IPv4) are, which fits GSO_PARTIAL nicely.
> It also supports non-PARTIAL offload of UDP tunnels, again just
>  needing to be told the outer transport offset so that it can edit
>  the UDP length field.
> (It's not clear to me whether the stack will ever use the non-PARTIAL
>  version with the netdev feature flags we're setting here.)
>
> Signed-off-by: Edward Cree <ecree@solarflare.com>
> ---
>  drivers/net/ethernet/sfc/ef100_nic.c | 13 ++++++--
>  drivers/net/ethernet/sfc/ef100_tx.c  | 45 ++++++++++++++++------------
>  2 files changed, 37 insertions(+), 21 deletions(-)
>
> diff --git a/drivers/net/ethernet/sfc/ef100_nic.c b/drivers/net/ethernet/sfc/ef100_nic.c
> index 3148fe770356..bf92cdc60cda 100644
> --- a/drivers/net/ethernet/sfc/ef100_nic.c
> +++ b/drivers/net/ethernet/sfc/ef100_nic.c
> @@ -182,8 +182,16 @@ static int efx_ef100_init_datapath_caps(struct efx_nic *efx)
>         if (rc)
>                 return rc;
>
> -       if (efx_ef100_has_cap(nic_data->datapath_caps2, TX_TSO_V3))
> -               efx->net_dev->features |= NETIF_F_TSO | NETIF_F_TSO6;
> +       if (efx_ef100_has_cap(nic_data->datapath_caps2, TX_TSO_V3)) {
> +               struct net_device *net_dev = efx->net_dev;
> +               netdev_features_t tso = NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_GSO_PARTIAL |
> +                                       NETIF_F_GSO_UDP_TUNNEL | NETIF_F_GSO_UDP_TUNNEL_CSUM;
> +
> +               net_dev->features |= tso;
> +               net_dev->hw_features |= tso;
> +               net_dev->hw_enc_features |= tso;
> +               net_dev->gso_partial_features |= NETIF_F_GSO_UDP_TUNNEL | NETIF_F_GSO_UDP_TUNNEL_CSUM;
> +       }
>         efx->num_mac_stats = MCDI_WORD(outbuf,
>                                        GET_CAPABILITIES_V4_OUT_MAC_STATS_NUM_STATS);
>         netif_dbg(efx, probe, efx->net_dev,
> @@ -1101,6 +1109,7 @@ static int ef100_probe_main(struct efx_nic *efx)
>         nic_data->efx = efx;
>         net_dev->features |= efx->type->offload_features;
>         net_dev->hw_features |= efx->type->offload_features;
> +       net_dev->hw_enc_features |= efx->type->offload_features;
>
>         /* Populate design-parameter defaults */
>         nic_data->tso_max_hdr_len = ESE_EF100_DP_GZ_TSO_MAX_HDR_LEN_DEFAULT;
> diff --git a/drivers/net/ethernet/sfc/ef100_tx.c b/drivers/net/ethernet/sfc/ef100_tx.c
> index a90e5a9d2a37..d267b12bdaa0 100644
> --- a/drivers/net/ethernet/sfc/ef100_tx.c
> +++ b/drivers/net/ethernet/sfc/ef100_tx.c
> @@ -54,8 +54,6 @@ static bool ef100_tx_can_tso(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
>         struct efx_nic *efx = tx_queue->efx;
>         struct ef100_nic_data *nic_data;
>         struct efx_tx_buffer *buffer;
> -       struct tcphdr *tcphdr;
> -       struct iphdr *iphdr;
>         size_t header_len;
>         u32 mss;
>
> @@ -98,20 +96,6 @@ static bool ef100_tx_can_tso(struct efx_tx_queue *tx_queue, struct sk_buff *skb)
>         buffer->unmap_len = 0;
>         buffer->skb = skb;
>         ++tx_queue->insert_count;
> -
> -       /* Adjust the TCP checksum to exclude the total length, since we set
> -        * ED_INNER_IP_LEN in the descriptor.
> -        */
> -       tcphdr = tcp_hdr(skb);
> -       if (skb_is_gso_v6(skb)) {
> -               tcphdr->check = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
> -                                                &ipv6_hdr(skb)->daddr,
> -                                                0, IPPROTO_TCP, 0);
> -       } else {
> -               iphdr = ip_hdr(skb);
> -               tcphdr->check = ~csum_tcpudp_magic(iphdr->saddr, iphdr->daddr,
> -                                                  0, IPPROTO_TCP, 0);
> -       }
>         return true;
>  }
>
> @@ -209,17 +193,35 @@ static void ef100_make_tso_desc(struct efx_nic *efx,
>                 ESE_GZ_TX_DESC_IP4_ID_INC_MOD16;
>         u16 vlan_enable =  efx->net_dev->features & NETIF_F_HW_VLAN_CTAG_TX ?
>                 skb_vlan_tag_present(skb) : 0;
> +       bool gso_partial = skb_shinfo(skb)->gso_type & SKB_GSO_PARTIAL;
>         unsigned int len, ip_offset, tcp_offset, payload_segs;
> +       unsigned int outer_ip_offset, outer_l4_offset;
>         u16 vlan_tci = skb_vlan_tag_get(skb);
>         u32 mss = skb_shinfo(skb)->gso_size;
> +       bool encap = skb->encapsulation;
> +       struct tcphdr *tcp;
> +       u32 paylen;
>
>         len = skb->len - buffer->len;
>         /* We use 1 for the TSO descriptor and 1 for the header */
>         payload_segs = segment_count - 2;
> -       ip_offset =  skb_network_offset(skb);
> -       tcp_offset = skb_transport_offset(skb);
> +       if (encap) {
> +               outer_ip_offset = skb_network_offset(skb);
> +               outer_l4_offset = skb_transport_offset(skb);
> +               ip_offset = skb_inner_network_offset(skb);
> +               tcp_offset = skb_inner_transport_offset(skb);
> +       } else {
> +               ip_offset =  skb_network_offset(skb);
> +               tcp_offset = skb_transport_offset(skb);
> +               outer_ip_offset = outer_l4_offset = 0;
> +       }
> +
> +       /* subtract TCP payload length from inner checksum */
> +       tcp = (void *)skb->data + tcp_offset;
> +       paylen = skb->len - tcp_offset;
> +       csum_replace_by_diff(&tcp->check, (__force __wsum)htonl(paylen));
>
> -       EFX_POPULATE_OWORD_13(*txd,
> +       EFX_POPULATE_OWORD_17(*txd,
>                               ESF_GZ_TX_DESC_TYPE, ESE_GZ_TX_DESC_TYPE_TSO,
>                               ESF_GZ_TX_TSO_MSS, mss,
>                               ESF_GZ_TX_TSO_HDR_NUM_SEGS, 1,
> @@ -231,6 +233,11 @@ static void ef100_make_tso_desc(struct efx_nic *efx,
>                               ESF_GZ_TX_TSO_INNER_L4_OFF_W, tcp_offset >> 1,
>                               ESF_GZ_TX_TSO_ED_INNER_IP4_ID, mangleid,
>                               ESF_GZ_TX_TSO_ED_INNER_IP_LEN, 1,
> +                             ESF_GZ_TX_TSO_OUTER_L3_OFF_W, outer_ip_offset >> 1,
> +                             ESF_GZ_TX_TSO_OUTER_L4_OFF_W, outer_l4_offset >> 1,
> +                             ESF_GZ_TX_TSO_ED_OUTER_UDP_LEN, encap && !gso_partial,

This is a boolean field to signal whether the NIC needs to fix up the
udp length field ?

Which in the case of GSO_PARTIAL has already been resolved by the gso
layer (in __skb_udp_tunnel_segment).

Just curious, is this ever expected to be true? Not based on current
advertised features, right?

> +                             ESF_GZ_TX_TSO_ED_OUTER_IP4_ID, encap ? mangleid :
> +                                                                    ESE_GZ_TX_DESC_IP4_ID_NO_OP,
>                               ESF_GZ_TX_TSO_VLAN_INSERT_EN, vlan_enable,
>                               ESF_GZ_TX_TSO_VLAN_INSERT_TCI, vlan_tci
>                 );
>

Re: [PATCH net-next 2/4] sfc: implement encap TSO on EF100

[Edward Cree]

On 30/10/2020 15:49, Willem de Bruijn wrote:
> On Wed, Oct 28, 2020 at 9:39 PM Edward Cree <ecree@solarflare.com> wrote:
>> +                             ESF_GZ_TX_TSO_ED_OUTER_UDP_LEN, encap && !gso_partial,
> 
> This is a boolean field to signal whether the NIC needs to fix up the
> udp length field ?
Yes.

> Which in the case of GSO_PARTIAL has already been resolved by the gso
> layer (in __skb_udp_tunnel_segment).
Indeed.

> Just curious, is this ever expected to be true? Not based on current
> advertised features, right?
As I mentioned in the patch description and cover letter, I'm not
 entirely certain.  I don't _think_ the stack will ever give us an
 encap skb without GSO_PARTIAL with the features we've advertised,
 but since the hardware supports it I thought it better to handle
 that case anyway, just in case I'm mistaken.

-ed

Re: [PATCH net-next 2/4] sfc: implement encap TSO on EF100

[Willem de Bruijn]

On Fri, Oct 30, 2020 at 12:16 PM Edward Cree <ecree@solarflare.com> wrote:
>
> On 30/10/2020 15:49, Willem de Bruijn wrote:
> > On Wed, Oct 28, 2020 at 9:39 PM Edward Cree <ecree@solarflare.com> wrote:
> >> +                             ESF_GZ_TX_TSO_ED_OUTER_UDP_LEN, encap && !gso_partial,
> >
> > This is a boolean field to signal whether the NIC needs to fix up the
> > udp length field ?
> Yes.

Thanks

> > Which in the case of GSO_PARTIAL has already been resolved by the gso
> > layer (in __skb_udp_tunnel_segment).
> Indeed.
>
> > Just curious, is this ever expected to be true? Not based on current
> > advertised features, right?
> As I mentioned in the patch description and cover letter, I'm not
>  entirely certain.  I don't _think_ the stack will ever give us an
>  encap skb without GSO_PARTIAL with the features we've advertised,

That's my understanding too.

>  but since the hardware supports it I thought it better to handle
>  that case anyway, just in case I'm mistaken.

Then you could (as follow-up) advertise without GSO_PARTIAL and avoid
the whole transition through the gso layer?

Re: [PATCH net-next 2/4] sfc: implement encap TSO on EF100

[Edward Cree]

On 30/10/2020 16:26, Willem de Bruijn wrote:
> Then you could (as follow-up) advertise without GSO_PARTIAL and avoid
> the whole transition through the gso layer?

The thing is, non-PARTIAL offload only supports tunnels that the NIC
 understands (single-layer UDP tunnels), but AIUI GSO_PARTIAL can
 support all sorts of other things, such as gretaps (though we'd need
 some more advertised features, I haven't figured out quite which
 ones yet but when I do and can test it I'll send a follow-up) and
 nested tunnels (as long as we don't need to edit the 'middle' IP ID,
 e.g. if it's DF or IPv6).  So we definitely want to advertise
 GSO_PARTIAL support.
But possibly I don't need to have NETIF_F_GSO_UDP_TUNNEL[_CSUM] in
 net_dev->gso_partial_features?

-ed

Re: [PATCH net-next 2/4] sfc: implement encap TSO on EF100

[Willem de Bruijn]

On Fri, Oct 30, 2020 at 12:43 PM Edward Cree <ecree@solarflare.com> wrote:
>
> On 30/10/2020 16:26, Willem de Bruijn wrote:
> > Then you could (as follow-up) advertise without GSO_PARTIAL and avoid
> > the whole transition through the gso layer?
>
> The thing is, non-PARTIAL offload only supports tunnels that the NIC
>  understands (single-layer UDP tunnels), but AIUI GSO_PARTIAL can
>  support all sorts of other things, such as gretaps (though we'd need
>  some more advertised features, I haven't figured out quite which
>  ones yet but when I do and can test it I'll send a follow-up) and
>  nested tunnels (as long as we don't need to edit the 'middle' IP ID,
>  e.g. if it's DF or IPv6).  So we definitely want to advertise
>  GSO_PARTIAL support.
> But possibly I don't need to have NETIF_F_GSO_UDP_TUNNEL[_CSUM] in
>  net_dev->gso_partial_features?

If the device can handle fixing the odd last segment length, indeed.

If you remove them from gso_partial_flags, then gso_features_check
won't mask them out


        /* Support for GSO partial features requires software
         * intervention before we can actually process the packets
         * so we need to strip support for any partial features now
         * and we can pull them back in after we have partially
         * segmented the frame.
         */
        if (!(skb_shinfo(skb)->gso_type & SKB_GSO_PARTIAL))
                features &= ~dev->gso_partial_features;

as called in validate_xmit_skb prior to testing whether the packet can
be passed to the nic as is in netif_needs_gso.

Until adding other tunnel types like NETIF_F_GSO_GRE_CSUM, for this
device gso_partial_features would then be 0 and thus
NETIF_F_GSO_PARTIAL is not needed at all?

Re: [PATCH net-next 2/4] sfc: implement encap TSO on EF100

[Edward Cree]

On 30/10/2020 17:33, Willem de Bruijn wrote:
> On Fri, Oct 30, 2020 at 12:43 PM Edward Cree <ecree@solarflare.com> wrote:
>> But possibly I don't need to have NETIF_F_GSO_UDP_TUNNEL[_CSUM] in
>>  net_dev->gso_partial_features?
> If the device can handle fixing the odd last segment length, indeed.
It can, but...
I thought Linux didn't give drivers odd last segments any more.  At
 least I'm fairly sure that in the (non-PARTIAL) non-encap TSO tests
 I've done, the GSO skbs we've been given have all been a whole
 number of mss long.
Which means I haven't been able to test that the device gets it right.

> Until adding other tunnel types like NETIF_F_GSO_GRE_CSUM, for this
> device gso_partial_features would then be 0 and thus
> NETIF_F_GSO_PARTIAL is not needed at all?
Unless the kernel supports GSO_PARTIAL of nested tunnels.  The device
 will handle (say) VxLAN-in-VxLAN just fine, as long as you don't want
 it to update the middle IPID; and being able to cope with crazy
 things like that was (I thought) part of the point of GSO_PARTIAL.
Indeed, given that GSO_PARTIAL is supposed to be a non-protocol-
 ossified design, it seems a bit silly to me that we have to specify
 a list of other NETIF_F_GSO_foo, rather than just being able to say
 "I can handle anything of the form ETH-IP-gubbins-IP-TCP" and let
 the kernel put whatever it likes — VxLAN, GRE, fou-over-geneve with
 cherry and sprinkles — in the 'gubbins'.  Wasn't that what 'less is
 more' was supposed to be all about?

-ed

Re: [PATCH net-next 2/4] sfc: implement encap TSO on EF100

[Willem de Bruijn]

On Fri, Oct 30, 2020 at 2:14 PM Edward Cree <ecree@solarflare.com> wrote:
>
> On 30/10/2020 17:33, Willem de Bruijn wrote:
> > On Fri, Oct 30, 2020 at 12:43 PM Edward Cree <ecree@solarflare.com> wrote:
> >> But possibly I don't need to have NETIF_F_GSO_UDP_TUNNEL[_CSUM] in
> >>  net_dev->gso_partial_features?
> > If the device can handle fixing the odd last segment length, indeed.
> It can, but...
> I thought Linux didn't give drivers odd last segments any more.  At
>  least I'm fairly sure that in the (non-PARTIAL) non-encap TSO tests
>  I've done, the GSO skbs we've been given have all been a whole
>  number of mss long.
> Which means I haven't been able to test that the device gets it right.

Perhaps the local TCP stack tries to avoid it. Off the top of my head
not sure if this is assured in all edge cases.

The forwarding path is another wildcard.

> > Until adding other tunnel types like NETIF_F_GSO_GRE_CSUM, for this
> > device gso_partial_features would then be 0 and thus
> > NETIF_F_GSO_PARTIAL is not needed at all?
> Unless the kernel supports GSO_PARTIAL of nested tunnels.  The device
>  will handle (say) VxLAN-in-VxLAN just fine, as long as you don't want
>  it to update the middle IPID; and being able to cope with crazy
>  things like that was (I thought) part of the point of GSO_PARTIAL.

Oh right.

> Indeed, given that GSO_PARTIAL is supposed to be a non-protocol-
>  ossified design, it seems a bit silly to me that we have to specify
>  a list of other NETIF_F_GSO_foo, rather than just being able to say
>  "I can handle anything of the form ETH-IP-gubbins-IP-TCP" and let
>  the kernel put whatever it likes — VxLAN, GRE, fou-over-geneve with
>  cherry and sprinkles — in the 'gubbins'.  Wasn't that what 'less is
>  more' was supposed to be all about?
>
> -ed

Re: [PATCH net-next 0/4] sfc: EF100 TSO enhancements

[Jakub Kicinski]

On Wed, 28 Oct 2020 20:41:21 +0000 Edward Cree wrote:
> Support TSO over encapsulation (with GSO_PARTIAL), and over VLANs
>  (which the code already handled but we didn't advertise).  Also
>  correct our handling of IPID mangling.
> 
> I couldn't find documentation of exactly what shaped SKBs we can
>  get given, so patch #2 is slightly guesswork, but when I tested
>  TSO over both underlay and (VxLAN) overlay, the checksums came
>  out correctly, so at least in those cases the edits we're making
>  must be the right ones.
> Similarly, I'm not 100% sure I've correctly understood how FIXEDID
>  and MANGLEID are supposed to work in patch #3.

Applied, thanks everyone!