aboutsummaryrefslogtreecommitdiffstats
path: root/src/vnet/tcp
diff options
context:
space:
mode:
authorSrikanth A <srakula@cisco.com>2019-10-02 17:48:58 -0700
committerJohn Lo <loj@cisco.com>2019-10-10 16:37:40 +0000
commit02833ff3294f4abbd8e3d52b38446e0f8f533ffc (patch)
treee2e9de14d3939938e502347723006f7bb374a02d /src/vnet/tcp
parent8a047ed741072bdb8d93b0841473eae06ae3c9d0 (diff)
tcp: custom checksum calculations for Ipv4/Ipv6
Type: feature Based on the configuration, we can disable checksum offload capability and calculate checksum while pushing the TCP & IP header. This saves some cycles when VPP stack is used in legacy hardware devices. Signed-off-by: Srikanth A <srakula@cisco.com> Change-Id: Ic1b3fcf3040917e47ee65263694ebf7437ac5668 (cherry picked from commit 3642782a2748503f5b5ccf89d1575c1d489948ef)
Diffstat (limited to 'src/vnet/tcp')
-rw-r--r--src/vnet/tcp/tcp.h3
-rw-r--r--src/vnet/tcp/tcp_output.c116
2 files changed, 97 insertions, 22 deletions
diff --git a/src/vnet/tcp/tcp.h b/src/vnet/tcp/tcp.h
index a1b7d4cbd0d..1bf32818171 100644
--- a/src/vnet/tcp/tcp.h
+++ b/src/vnet/tcp/tcp.h
@@ -120,6 +120,7 @@ extern timer_expiration_handler tcp_timer_retransmit_syn_handler;
_(RATE_SAMPLE, "Conn does rate sampling") \
_(TRACK_BURST, "Track burst") \
_(ZERO_RWND_SENT, "Zero RWND sent") \
+ _(NO_CSUM_OFFLOAD, "No Checksum Offload") \
typedef enum _tcp_connection_flag_bits
{
@@ -1233,6 +1234,8 @@ vlib_buffer_push_tcp_net_order (vlib_buffer_t * b, u16 sp, u16 dp, u32 seq,
th->window = wnd;
th->checksum = 0;
th->urgent_pointer = 0;
+ vnet_buffer (b)->l4_hdr_offset = (u8 *) th - b->data;
+ b->flags |= VNET_BUFFER_F_L4_HDR_OFFSET_VALID;
return th;
}
diff --git a/src/vnet/tcp/tcp_output.c b/src/vnet/tcp/tcp_output.c
index 7ebea37fc43..a7b0e398d36 100644
--- a/src/vnet/tcp/tcp_output.c
+++ b/src/vnet/tcp/tcp_output.c
@@ -444,6 +444,78 @@ tcp_init_buffer (vlib_main_t * vm, vlib_buffer_t * b)
return vlib_buffer_make_headroom (b, TRANSPORT_MAX_HDRS_LEN);
}
+
+/* Compute TCP checksum in software when offloading is disabled for a connection */
+u16
+ip6_tcp_compute_checksum_custom (vlib_main_t * vm, vlib_buffer_t * p0,
+ ip46_address_t * src, ip46_address_t * dst)
+{
+ ip_csum_t sum0;
+ u16 payload_length_host_byte_order;
+ u32 i;
+
+ /* Initialize checksum with ip header. */
+ sum0 = clib_host_to_net_u16 (vlib_buffer_length_in_chain (vm, p0)) +
+ clib_host_to_net_u16 (IP_PROTOCOL_TCP);
+ payload_length_host_byte_order = vlib_buffer_length_in_chain (vm, p0);
+
+ for (i = 0; i < ARRAY_LEN (src->ip6.as_uword); i++)
+ {
+ sum0 = ip_csum_with_carry
+ (sum0, clib_mem_unaligned (&src->ip6.as_uword[i], uword));
+ sum0 = ip_csum_with_carry
+ (sum0, clib_mem_unaligned (&dst->ip6.as_uword[i], uword));
+ }
+
+ return ip_calculate_l4_checksum (vm, p0, sum0,
+ payload_length_host_byte_order, NULL, 0,
+ NULL);
+}
+
+u16
+ip4_tcp_compute_checksum_custom (vlib_main_t * vm, vlib_buffer_t * p0,
+ ip46_address_t * src, ip46_address_t * dst)
+{
+ ip_csum_t sum0;
+ u32 payload_length_host_byte_order;
+
+ payload_length_host_byte_order = vlib_buffer_length_in_chain (vm, p0);
+ sum0 =
+ clib_host_to_net_u32 (payload_length_host_byte_order +
+ (IP_PROTOCOL_TCP << 16));
+
+ sum0 = ip_csum_with_carry (sum0, clib_mem_unaligned (&src->ip4, u32));
+ sum0 = ip_csum_with_carry (sum0, clib_mem_unaligned (&dst->ip4, u32));
+
+ return ip_calculate_l4_checksum (vm, p0, sum0,
+ payload_length_host_byte_order, NULL, 0,
+ NULL);
+}
+
+static inline u16
+tcp_compute_checksum (tcp_connection_t * tc, vlib_buffer_t * b)
+{
+ u16 checksum = 0;
+ if (PREDICT_FALSE (tc->flags & TCP_CONN_NO_CSUM_OFFLOAD))
+ {
+ tcp_worker_ctx_t *wrk = tcp_get_worker (tc->c_thread_index);
+ vlib_main_t *vm = wrk->vm;
+
+ if (tc->c_is_ip4)
+ checksum = ip4_tcp_compute_checksum_custom
+ (vm, b, &tc->c_lcl_ip, &tc->c_rmt_ip);
+ else
+ checksum = ip6_tcp_compute_checksum_custom
+ (vm, b, &tc->c_lcl_ip, &tc->c_rmt_ip);
+ }
+ else
+ {
+ b->flags |= VNET_BUFFER_F_OFFLOAD_TCP_CKSUM;
+ }
+ return checksum;
+}
+
+
/**
* Prepare ACK
*/
@@ -466,6 +538,9 @@ tcp_make_ack_i (tcp_connection_t * tc, vlib_buffer_t * b, tcp_state_t state,
tc->rcv_nxt, tcp_hdr_opts_len, flags, wnd);
tcp_options_write ((u8 *) (th + 1), snd_opts);
+
+ th->checksum = tcp_compute_checksum (tc, b);
+
vnet_buffer (b)->tcp.connection_index = tc->c_c_index;
if (wnd == 0)
@@ -517,6 +592,7 @@ tcp_make_syn (tcp_connection_t * tc, vlib_buffer_t * b)
initial_wnd);
vnet_buffer (b)->tcp.connection_index = tc->c_c_index;
tcp_options_write ((u8 *) (th + 1), &snd_opts);
+ th->checksum = tcp_compute_checksum (tc, b);
}
/**
@@ -541,6 +617,7 @@ tcp_make_synack (tcp_connection_t * tc, vlib_buffer_t * b)
tcp_options_write ((u8 *) (th + 1), snd_opts);
vnet_buffer (b)->tcp.connection_index = tc->c_c_index;
+ th->checksum = tcp_compute_checksum (tc, b);
}
always_inline void
@@ -786,7 +863,8 @@ tcp_send_reset_w_pkt (tcp_connection_t * tc, vlib_buffer_t * pkt,
{
ASSERT ((pkt_ih4->ip_version_and_header_length & 0xF0) == 0x40);
ih4 = vlib_buffer_push_ip4 (vm, b, &pkt_ih4->dst_address,
- &pkt_ih4->src_address, IP_PROTOCOL_TCP, 1);
+ &pkt_ih4->src_address, IP_PROTOCOL_TCP,
+ (!(tc->flags & TCP_CONN_NO_CSUM_OFFLOAD)));
th->checksum = ip4_tcp_udp_compute_checksum (vm, b, ih4);
}
else
@@ -833,6 +911,7 @@ tcp_send_reset (tcp_connection_t * tc)
tc->rcv_nxt, tcp_hdr_opts_len, flags,
advertise_wnd);
opts_write_len = tcp_options_write ((u8 *) (th + 1), &tc->snd_opts);
+ th->checksum = tcp_compute_checksum (tc, b);
ASSERT (opts_write_len == tc->snd_opts_len);
vnet_buffer (b)->tcp.connection_index = tc->c_c_index;
tcp_enqueue_to_output (wrk, b, bi, tc->c_is_ip4);
@@ -851,7 +930,8 @@ tcp_push_ip_hdr (tcp_worker_ctx_t * wrk, tcp_connection_t * tc,
{
ip4_header_t *ih;
ih = vlib_buffer_push_ip4 (vm, b, &tc->c_lcl_ip4,
- &tc->c_rmt_ip4, IP_PROTOCOL_TCP, 1);
+ &tc->c_rmt_ip4, IP_PROTOCOL_TCP,
+ (!(tc->flags & TCP_CONN_NO_CSUM_OFFLOAD)));
th->checksum = ip4_tcp_udp_compute_checksum (vm, b, ih);
}
else
@@ -1082,6 +1162,9 @@ tcp_push_hdr_i (tcp_connection_t * tc, vlib_buffer_t * b, u32 snd_nxt,
tc->bytes_out += data_len;
tc->data_segs_out += 1;
+
+ th->checksum = tcp_compute_checksum (tc, b);
+
TCP_EVT (TCP_EVT_PKTIZE, tc);
}
@@ -2160,30 +2243,19 @@ always_inline void
tcp_output_push_ip (vlib_main_t * vm, vlib_buffer_t * b0,
tcp_connection_t * tc0, u8 is_ip4)
{
- tcp_header_t *th0 = 0;
+ u8 __clib_unused *ih0;
+ tcp_header_t __clib_unused *th0 = vlib_buffer_get_current (b0);
- th0 = vlib_buffer_get_current (b0);
TCP_EVT (TCP_EVT_OUTPUT, tc0, th0->flags, b0->current_length);
+
if (is_ip4)
- {
- vlib_buffer_push_ip4 (vm, b0, &tc0->c_lcl_ip4, &tc0->c_rmt_ip4,
- IP_PROTOCOL_TCP, 1);
- b0->flags |= VNET_BUFFER_F_OFFLOAD_TCP_CKSUM;
- vnet_buffer (b0)->l4_hdr_offset = (u8 *) th0 - b0->data;
- th0->checksum = 0;
- }
+ ih0 = vlib_buffer_push_ip4 (vm, b0, &tc0->c_lcl_ip4, &tc0->c_rmt_ip4,
+ IP_PROTOCOL_TCP,
+ (!(tc0->flags & TCP_CONN_NO_CSUM_OFFLOAD)));
else
- {
- ip6_header_t *ih0;
- ih0 = vlib_buffer_push_ip6 (vm, b0, &tc0->c_lcl_ip6,
- &tc0->c_rmt_ip6, IP_PROTOCOL_TCP);
- b0->flags |= VNET_BUFFER_F_OFFLOAD_TCP_CKSUM;
- vnet_buffer (b0)->l3_hdr_offset = (u8 *) ih0 - b0->data;
- vnet_buffer (b0)->l4_hdr_offset = (u8 *) th0 - b0->data;
- b0->flags |=
- VNET_BUFFER_F_L3_HDR_OFFSET_VALID | VNET_BUFFER_F_L4_HDR_OFFSET_VALID;
- th0->checksum = 0;
- }
+ ih0 = vlib_buffer_push_ip6 (vm, b0, &tc0->c_lcl_ip6, &tc0->c_rmt_ip6,
+ IP_PROTOCOL_TCP);
+
}
always_inline void