From: Alexander Duyck <aduyck@mirantis.com>
To: intel-wired-lan@osuosl.org
Subject: [Intel-wired-lan] [next PATCH v3 08/15] i40e/i40evf: Do not write to descriptor unless we complete
Date: Sun, 24 Jan 2016 21:17:01 -0800 [thread overview]
Message-ID: <20160125051701.12004.23605.stgit@localhost.localdomain> (raw)
In-Reply-To: <20160125050602.12004.38884.stgit@localhost.localdomain>
This patch defers writing to the Tx descriptor bits until we know we have
successuflly completed a given operation. So for example we defer updating
the tunnelling portion of the context descriptor until we have fully
identified the type.
The advantage to this approach is that we can assemble values as we go
instead of having to try and cludge everything together all at once. As a
result we can significantly clean up the tunneling configuration for
instance as we can just do a pointer walk and do the math for the distance
between each set of points.
Signed-off-by: Alexander Duyck <aduyck@mirantis.com>
---
drivers/net/ethernet/intel/i40e/i40e_txrx.c | 81 +++++++++++++------------
drivers/net/ethernet/intel/i40evf/i40e_txrx.c | 81 +++++++++++++------------
2 files changed, 84 insertions(+), 78 deletions(-)
diff --git a/drivers/net/ethernet/intel/i40e/i40e_txrx.c b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
index 2e9ae018d639..713b39f4f6f8 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_txrx.c
@@ -2403,25 +2403,26 @@ static void i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
unsigned char *hdr;
} l4;
unsigned char *exthdr;
- u32 l4_tunnel = 0;
+ u32 offset, cmd = 0, tunnel = 0;
__be16 frag_off;
u8 l4_proto = 0;
ip.hdr = skb_network_header(skb);
l4.hdr = skb_transport_header(skb);
+ /* compute outer L2 header size */
+ offset = ((ip.hdr - skb->data) / 2) << I40E_TX_DESC_LENGTH_MACLEN_SHIFT;
+
if (skb->encapsulation) {
/* define outer network header type */
if (*tx_flags & I40E_TX_FLAGS_IPV4) {
- if (*tx_flags & I40E_TX_FLAGS_TSO) {
- *cd_tunneling |= I40E_TX_CTX_EXT_IP_IPV4;
- } else {
- *cd_tunneling |=
- I40E_TX_CTX_EXT_IP_IPV4_NO_CSUM;
- }
+ tunnel |= (*tx_flags & I40E_TX_FLAGS_TSO) ?
+ I40E_TX_CTX_EXT_IP_IPV4 :
+ I40E_TX_CTX_EXT_IP_IPV4_NO_CSUM;
+
l4_proto = ip.v4->protocol;
} else if (*tx_flags & I40E_TX_FLAGS_IPV6) {
- *cd_tunneling |= I40E_TX_CTX_EXT_IP_IPV6;
+ tunnel |= I40E_TX_CTX_EXT_IP_IPV6;
exthdr = ip.hdr + sizeof(*ip.v6);
l4_proto = ip.v6->nexthdr;
@@ -2430,33 +2431,38 @@ static void i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
&l4_proto, &frag_off);
}
+ /* compute outer L3 header size */
+ tunnel |= ((l4.hdr - ip.hdr) / 4) <<
+ I40E_TXD_CTX_QW0_EXT_IPLEN_SHIFT;
+
+ /* switch IP header pointer from outer to inner header */
+ ip.hdr = skb_inner_network_header(skb);
+
/* define outer transport */
switch (l4_proto) {
case IPPROTO_UDP:
- l4_tunnel = I40E_TXD_CTX_UDP_TUNNELING;
+ tunnel |= I40E_TXD_CTX_UDP_TUNNELING;
*tx_flags |= I40E_TX_FLAGS_TUNNEL;
break;
case IPPROTO_GRE:
- l4_tunnel = I40E_TXD_CTX_GRE_TUNNELING;
+ tunnel |= I40E_TXD_CTX_GRE_TUNNELING;
*tx_flags |= I40E_TX_FLAGS_TUNNEL;
break;
default:
return;
}
+ /* compute tunnel header size */
+ tunnel |= ((ip.hdr - l4.hdr) / 2) <<
+ I40E_TXD_CTX_QW0_NATLEN_SHIFT;
+
+ /* record tunnel offload values */
+ *cd_tunneling |= tunnel;
+
/* switch L4 header pointer from outer to inner */
- ip.hdr = skb_inner_network_header(skb);
l4.hdr = skb_inner_transport_header(skb);
l4_proto = 0;
- /* Now set the ctx descriptor fields */
- *cd_tunneling |= (skb_network_header_len(skb) >> 2) <<
- I40E_TXD_CTX_QW0_EXT_IPLEN_SHIFT |
- l4_tunnel |
- ((skb_inner_network_offset(skb) -
- skb_transport_offset(skb)) >> 1) <<
- I40E_TXD_CTX_QW0_NATLEN_SHIFT;
-
/* reset type as we transition from outer to inner headers */
*tx_flags &= ~(I40E_TX_FLAGS_IPV4 | I40E_TX_FLAGS_IPV6);
if (ip.v4->version == 4)
@@ -2471,13 +2477,11 @@ static void i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
/* the stack computes the IP header already, the only time we
* need the hardware to recompute it is in the case of TSO.
*/
- if (*tx_flags & I40E_TX_FLAGS_TSO) {
- *td_cmd |= I40E_TX_DESC_CMD_IIPT_IPV4_CSUM;
- } else {
- *td_cmd |= I40E_TX_DESC_CMD_IIPT_IPV4;
- }
+ cmd |= (*tx_flags & I40E_TX_FLAGS_TSO) ?
+ I40E_TX_DESC_CMD_IIPT_IPV4_CSUM :
+ I40E_TX_DESC_CMD_IIPT_IPV4;
} else if (*tx_flags & I40E_TX_FLAGS_IPV6) {
- *td_cmd |= I40E_TX_DESC_CMD_IIPT_IPV6;
+ cmd |= I40E_TX_DESC_CMD_IIPT_IPV6;
exthdr = ip.hdr + sizeof(*ip.v6);
l4_proto = ip.v6->nexthdr;
@@ -2486,35 +2490,34 @@ static void i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
&l4_proto, &frag_off);
}
- /* Now set the td_offset for IP header length */
- *td_offset = ((l4.hdr - ip.hdr) / 4) << I40E_TX_DESC_LENGTH_IPLEN_SHIFT;
- /* words in MACLEN + dwords in IPLEN + dwords in L4Len */
- *td_offset |= (skb_network_offset(skb) >> 1) <<
- I40E_TX_DESC_LENGTH_MACLEN_SHIFT;
+ /* compute inner L3 header size */
+ offset |= ((l4.hdr - ip.hdr) / 4) << I40E_TX_DESC_LENGTH_IPLEN_SHIFT;
/* Enable L4 checksum offloads */
switch (l4_proto) {
case IPPROTO_TCP:
/* enable checksum offloads */
- *td_cmd |= I40E_TX_DESC_CMD_L4T_EOFT_TCP;
- *td_offset |= l4.tcp->doff <<
- I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
+ cmd |= I40E_TX_DESC_CMD_L4T_EOFT_TCP;
+ offset |= l4.tcp->doff << I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
break;
case IPPROTO_SCTP:
/* enable SCTP checksum offload */
- *td_cmd |= I40E_TX_DESC_CMD_L4T_EOFT_SCTP;
- *td_offset |= (sizeof(struct sctphdr) >> 2) <<
- I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
+ cmd |= I40E_TX_DESC_CMD_L4T_EOFT_SCTP;
+ offset |= (sizeof(struct sctphdr) >> 2) <<
+ I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
break;
case IPPROTO_UDP:
/* enable UDP checksum offload */
- *td_cmd |= I40E_TX_DESC_CMD_L4T_EOFT_UDP;
- *td_offset |= (sizeof(struct udphdr) >> 2) <<
- I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
+ cmd |= I40E_TX_DESC_CMD_L4T_EOFT_UDP;
+ offset |= (sizeof(struct udphdr) >> 2) <<
+ I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
break;
default:
break;
}
+
+ *td_cmd |= cmd;
+ *td_offset |= offset;
}
/**
diff --git a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
index d37cba28ecfa..35aa07f8b9da 100644
--- a/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
+++ b/drivers/net/ethernet/intel/i40evf/i40e_txrx.c
@@ -1620,25 +1620,26 @@ static void i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
unsigned char *hdr;
} l4;
unsigned char *exthdr;
- u32 l4_tunnel = 0;
+ u32 offset, cmd = 0, tunnel = 0;
__be16 frag_off;
u8 l4_proto = 0;
ip.hdr = skb_network_header(skb);
l4.hdr = skb_transport_header(skb);
+ /* compute outer L2 header size */
+ offset = ((ip.hdr - skb->data) / 2) << I40E_TX_DESC_LENGTH_MACLEN_SHIFT;
+
if (skb->encapsulation) {
/* define outer network header type */
if (*tx_flags & I40E_TX_FLAGS_IPV4) {
- if (*tx_flags & I40E_TX_FLAGS_TSO) {
- *cd_tunneling |= I40E_TX_CTX_EXT_IP_IPV4;
- } else {
- *cd_tunneling |=
- I40E_TX_CTX_EXT_IP_IPV4_NO_CSUM;
- }
+ tunnel |= (*tx_flags & I40E_TX_FLAGS_TSO) ?
+ I40E_TX_CTX_EXT_IP_IPV4 :
+ I40E_TX_CTX_EXT_IP_IPV4_NO_CSUM;
+
l4_proto = ip.v4->protocol;
} else if (*tx_flags & I40E_TX_FLAGS_IPV6) {
- *cd_tunneling |= I40E_TX_CTX_EXT_IP_IPV6;
+ tunnel |= I40E_TX_CTX_EXT_IP_IPV6;
exthdr = ip.hdr + sizeof(*ip.v6);
l4_proto = ip.v6->nexthdr;
@@ -1647,33 +1648,38 @@ static void i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
&l4_proto, &frag_off);
}
+ /* compute outer L3 header size */
+ tunnel |= ((l4.hdr - ip.hdr) / 4) <<
+ I40E_TXD_CTX_QW0_EXT_IPLEN_SHIFT;
+
+ /* switch IP header pointer from outer to inner header */
+ ip.hdr = skb_inner_network_header(skb);
+
/* define outer transport */
switch (l4_proto) {
case IPPROTO_UDP:
- l4_tunnel = I40E_TXD_CTX_UDP_TUNNELING;
+ tunnel |= I40E_TXD_CTX_UDP_TUNNELING;
*tx_flags |= I40E_TX_FLAGS_TUNNEL;
break;
case IPPROTO_GRE:
- l4_tunnel = I40E_TXD_CTX_GRE_TUNNELING;
+ tunnel |= I40E_TXD_CTX_GRE_TUNNELING;
*tx_flags |= I40E_TX_FLAGS_TUNNEL;
break;
default:
return;
}
+ /* compute tunnel header size */
+ tunnel |= ((ip.hdr - l4.hdr) / 2) <<
+ I40E_TXD_CTX_QW0_NATLEN_SHIFT;
+
+ /* record tunnel offload values */
+ *cd_tunneling |= tunnel;
+
/* switch L4 header pointer from outer to inner */
- ip.hdr = skb_inner_network_header(skb);
l4.hdr = skb_inner_transport_header(skb);
l4_proto = 0;
- /* Now set the ctx descriptor fields */
- *cd_tunneling |= (skb_network_header_len(skb) >> 2) <<
- I40E_TXD_CTX_QW0_EXT_IPLEN_SHIFT |
- l4_tunnel |
- ((skb_inner_network_offset(skb) -
- skb_transport_offset(skb)) >> 1) <<
- I40E_TXD_CTX_QW0_NATLEN_SHIFT;
-
/* reset type as we transition from outer to inner headers */
*tx_flags &= ~(I40E_TX_FLAGS_IPV4 | I40E_TX_FLAGS_IPV6);
if (ip.v4->version == 4)
@@ -1688,13 +1694,11 @@ static void i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
/* the stack computes the IP header already, the only time we
* need the hardware to recompute it is in the case of TSO.
*/
- if (*tx_flags & I40E_TX_FLAGS_TSO) {
- *td_cmd |= I40E_TX_DESC_CMD_IIPT_IPV4_CSUM;
- } else {
- *td_cmd |= I40E_TX_DESC_CMD_IIPT_IPV4;
- }
+ cmd |= (*tx_flags & I40E_TX_FLAGS_TSO) ?
+ I40E_TX_DESC_CMD_IIPT_IPV4_CSUM :
+ I40E_TX_DESC_CMD_IIPT_IPV4;
} else if (*tx_flags & I40E_TX_FLAGS_IPV6) {
- *td_cmd |= I40E_TX_DESC_CMD_IIPT_IPV6;
+ cmd |= I40E_TX_DESC_CMD_IIPT_IPV6;
exthdr = ip.hdr + sizeof(*ip.v6);
l4_proto = ip.v6->nexthdr;
@@ -1703,35 +1707,34 @@ static void i40e_tx_enable_csum(struct sk_buff *skb, u32 *tx_flags,
&l4_proto, &frag_off);
}
- /* Now set the td_offset for IP header length */
- *td_offset = ((l4.hdr - ip.hdr) / 4) << I40E_TX_DESC_LENGTH_IPLEN_SHIFT;
- /* words in MACLEN + dwords in IPLEN + dwords in L4Len */
- *td_offset |= (skb_network_offset(skb) >> 1) <<
- I40E_TX_DESC_LENGTH_MACLEN_SHIFT;
+ /* compute inner L3 header size */
+ offset |= ((l4.hdr - ip.hdr) / 4) << I40E_TX_DESC_LENGTH_IPLEN_SHIFT;
/* Enable L4 checksum offloads */
switch (l4_proto) {
case IPPROTO_TCP:
/* enable checksum offloads */
- *td_cmd |= I40E_TX_DESC_CMD_L4T_EOFT_TCP;
- *td_offset |= l4.tcp->doff <<
- I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
+ cmd |= I40E_TX_DESC_CMD_L4T_EOFT_TCP;
+ offset |= l4.tcp->doff << I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
break;
case IPPROTO_SCTP:
/* enable SCTP checksum offload */
- *td_cmd |= I40E_TX_DESC_CMD_L4T_EOFT_SCTP;
- *td_offset |= (sizeof(struct sctphdr) >> 2) <<
- I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
+ cmd |= I40E_TX_DESC_CMD_L4T_EOFT_SCTP;
+ offset |= (sizeof(struct sctphdr) >> 2) <<
+ I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
break;
case IPPROTO_UDP:
/* enable UDP checksum offload */
- *td_cmd |= I40E_TX_DESC_CMD_L4T_EOFT_UDP;
- *td_offset |= (sizeof(struct udphdr) >> 2) <<
- I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
+ cmd |= I40E_TX_DESC_CMD_L4T_EOFT_UDP;
+ offset |= (sizeof(struct udphdr) >> 2) <<
+ I40E_TX_DESC_LENGTH_L4_FC_LEN_SHIFT;
break;
default:
break;
}
+
+ *td_cmd |= cmd;
+ *td_offset |= offset;
}
/**
next prev parent reply other threads:[~2016-01-25 5:17 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-01-25 5:16 [Intel-wired-lan] [next PATCH v3 00/15] TSO and checksum fixes for i40e Alexander Duyck
2016-01-25 5:16 ` [Intel-wired-lan] [next PATCH v3 01/15] i40e/i40evf: Drop outer checksum offload that was not requested Alexander Duyck
2016-01-27 16:00 ` Bowers, AndrewX
2016-01-25 5:16 ` [Intel-wired-lan] [next PATCH v3 02/15] i40e/i40evf: Use u64 values instead of casting them in TSO function Alexander Duyck
2016-01-27 16:03 ` Bowers, AndrewX
2016-01-25 5:16 ` [Intel-wired-lan] [next PATCH v3 03/15] i40e/i40evf: Factor out L4 header and checksum from L3 bits in TSO path Alexander Duyck
2016-01-27 16:09 ` Bowers, AndrewX
2016-01-25 5:16 ` [Intel-wired-lan] [next PATCH v3 04/15] i40e/i40evf: Consolidate all header changes into TSO function Alexander Duyck
2016-01-27 16:14 ` Bowers, AndrewX
2016-01-25 5:16 ` [Intel-wired-lan] [next PATCH v3 05/15] i40e/i40evf: Replace header pointers with unions of pointers in Tx checksum path Alexander Duyck
2016-01-27 16:27 ` Bowers, AndrewX
2016-01-25 5:16 ` [Intel-wired-lan] [next PATCH v3 06/15] i40e/i40evf: Add support for IPv4 encapsulated in IPv6 Alexander Duyck
2016-01-27 18:04 ` Bowers, AndrewX
2016-01-25 5:16 ` [Intel-wired-lan] [next PATCH v3 07/15] i40e/i40evf: Handle IPv6 extension headers in checksum offload Alexander Duyck
2016-01-27 18:05 ` Bowers, AndrewX
2016-01-25 5:17 ` Alexander Duyck [this message]
2016-01-27 18:07 ` [Intel-wired-lan] [next PATCH v3 08/15] i40e/i40evf: Do not write to descriptor unless we complete Bowers, AndrewX
2016-01-25 5:17 ` [Intel-wired-lan] [next PATCH v3 09/15] i40e/i40evf: Add exception handling for Tx checksum Alexander Duyck
2016-01-27 18:08 ` Bowers, AndrewX
2016-01-25 5:17 ` [Intel-wired-lan] [next PATCH v3 10/15] i40e/i40evf: Clean-up Rx packet checksum handling Alexander Duyck
2016-01-27 18:09 ` Bowers, AndrewX
2016-01-25 5:17 ` [Intel-wired-lan] [next PATCH v3 11/15] i40e/i40evf: Enable support for SKB_GSO_UDP_TUNNEL_CSUM Alexander Duyck
2016-01-27 18:17 ` Bowers, AndrewX
2016-02-02 22:49 ` Jesse Brandeburg
2016-02-02 23:10 ` Jesse Brandeburg
2016-02-02 23:18 ` Jesse Brandeburg
2016-02-03 0:06 ` Alexander Duyck
2016-01-25 5:17 ` [Intel-wired-lan] [next PATCH v3 12/15] i40e: Fix ATR in relation to tunnels Alexander Duyck
2016-01-25 19:27 ` Patil, Kiran
2016-01-25 22:21 ` Alexander Duyck
2016-01-26 0:16 ` Patil, Kiran
2016-01-27 18:18 ` Bowers, AndrewX
2016-01-25 5:17 ` [Intel-wired-lan] [next PATCH v3 13/15] i40e: Do not drop support for IPv6 VXLAN or GENEVE tunnels Alexander Duyck
2016-01-27 18:26 ` Bowers, AndrewX
2016-01-25 5:17 ` [Intel-wired-lan] [next PATCH v3 14/15] i40e: Update feature flags to reflect newly enabled features Alexander Duyck
2016-01-27 18:44 ` Bowers, AndrewX
2016-01-25 5:17 ` [Intel-wired-lan] [next PATCH v3 15/15] i40evf: " Alexander Duyck
2016-01-25 19:43 ` Singhai, Anjali
2016-01-27 18:45 ` Bowers, AndrewX
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160125051701.12004.23605.stgit@localhost.localdomain \
--to=aduyck@mirantis.com \
--cc=intel-wired-lan@osuosl.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.