- dev - passt

[PATCH v2 0/1] RFC: IOV tail helpers
by David Gibson 08 Nov '24

08 Nov '24

As mentioned here's my draft patch adding a helper to pull headers off an IOV, checking for contiguity, alignment and so forth. It's based around the idea of an "iov tail", that is an iov from which you've already consumed (in some sense) some data from the beginning. It uses an explicit offset to track this, so we don't have to copy an entire iovec array in order to trim a bit off it. But, it's built so that if you pull enough off the front to remove some entire buffers, it won't repeatedly have to step through them working out later offsets. It may be somewhat overengineered, let me know what you think of the interface. v2: * Substantial changes to comments and names, to make the purpose clearer David Gibson (1): iov: iov tail helpers iov.c | 93 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ iov.h | 76 ++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 169 insertions(+) -- 2.47.0

2 2

[PATCH 0/1] RFC: IOV tail helpers
by David Gibson 08 Nov '24

08 Nov '24

As mentioned here's my draft patch adding a helper to pull headers off an IOV, checking for contiguity, alignment and so forth. It's based around the idea of an "iov tail", that is an iov from which you've already consumed (in some sense) some data from the beginning. It uses an explicit offset to track this, so we don't have to copy an entire iovec array in order to trim a bit off it. But, it's built so that if you pull enough off the front to remove some entire buffers, it won't repeatedly have to step through them working out later offsets. It may be somewhat overengineered, let me know what you think of the interface. David Gibson (1): iov: iov tail helpers iov.c | 83 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ iov.h | 24 +++++++++++++++++ 2 files changed, 107 insertions(+) -- 2.47.0

2 5

[PATCH 0/2] Small fixes for nstool
by David Gibson 08 Nov '24

08 Nov '24

While working on the exeter tests, I noticed a couple of things missing in nstool. These make sense standalone, even if we don't have an urgent need for them without the exeter tests. David Gibson (2): test: Rename propagating signal handler test: Make nstool hold robust against interruptions to control clients test/nstool.c | 40 ++++++++++++++++++++++++++++------------ 1 file changed, 28 insertions(+), 12 deletions(-) -- 2.47.0

2 6

[PATCH 0/6] Fix some more static checker warnings
by Stefano Brivio 08 Nov '24

08 Nov '24

Stefano Brivio (6): dhcpv6: Use for loop instead of goto to avoid false positive cppcheck warning dhcpv6: Turn some option headers pointers to const tap: Cast TAP_BUF_BYTES - ETH_MAX_MTU to ssize_t, not TAP_BUF_BYTES util: Define small and big thresholds for socket buffers as unsigned long long passt: Use NOLINT clang-tidy block instead of NOLINTNEXTLINE tap, tcp, util: Add some missing SOCK_CLOEXEC flags dhcpv6.c | 51 +++++++++++++++++++++++---------------------------- passt.c | 3 ++- tap.c | 7 ++++--- tcp.c | 2 +- util.c | 3 ++- util.h | 9 ++++++--- 6 files changed, 38 insertions(+), 37 deletions(-) -- 2.43.0

2 12

[PATCH 0/8] Avoid running cppcheck on system headers
by David Gibson 08 Nov '24

08 Nov '24

It turns out cppcheck has inbuilt knowledge of the C library, and isn't typically given the system headers. Avoiding doing so reduces the runtime to less than half of what it currently is. For non-obvious reasons, this change also exposes some new warnings. Some are real, one is a cppcheck bug. Fix and/or workaround these then make the change to the cppcheck options. This is based on my earlier series with clangd configuration and fixes. David Gibson (8): linux_dep: Generalise tcp_info.h to handling Linux extension compatibility log: Only check for FALLOC_FL_COLLAPSE_RANGE availability at runtime linux_dep: Move close_range() conditional handling to linux_dep.h linux_dep: Fix CLOSE_RANGE_UNSHARE availability handling ndp: Use const pointer for ndp_ns packet udp: Don't dereference uflow before NULL check in udp_reply_sock_handler() util: Work around cppcheck bug 6936 cppcheck: Don't check the system headers Makefile | 17 ++--------------- tcp_info.h => linux_dep.h | 32 ++++++++++++++++++++++++++++---- log.c | 9 +-------- ndp.c | 4 ++-- tcp.c | 2 +- udp.c | 5 +++-- util.h | 29 ++++++++++------------------- 7 files changed, 47 insertions(+), 51 deletions(-) rename tcp_info.h => linux_dep.h (85%) -- 2.47.0

2 17

[PATCH 00/12] Minor fixups for or inspired by clangd and related tools
by David Gibson 07 Nov '24

07 Nov '24

I've been experimenting with Zed and clangd recently. Currently it generates an enormous number of largely spurious errors and warnings on the passt code base. Mostly that's due to its default configurations not suiting us. This series adds some configuration that addresses a number of those warnings, though there remain many more for now. Some of the warnings also look reasonable, so I have a grab bag of fixes or workarounds for some of those two. David Gibson (12): clang: Add .clang-format file Makefile: Simplify exclusion of qrap from static checks clang: Move clang-tidy configuration from Makefile to .clang-tidy arch: Avoid explicit access to 'environ' flow: Correct type of flowside_at_sidx() netlink: RTA_PAYLOAD() returns int, not size_t Makefile: Move NETNS_RUN_DIR definition to C code seccomp: Simplify handling of AUDIT_ARCH Makefile: Use -DARCH for qrap only Makefile: Don't attempt to auto-detect stack size clang: Add rudimentary clangd configuration util: Remove unused ffsl() function .clang-format | 126 ++++++++++++++++++++++++++++++++++++++++++++++ .clang-tidy | 93 ++++++++++++++++++++++++++++++++++ .clangd | 3 ++ Makefile | 136 +++----------------------------------------------- arch.c | 2 +- conf.c | 2 + flow_table.h | 2 +- netlink.c | 4 +- seccomp.sh | 14 +++++- util.h | 5 +- 10 files changed, 247 insertions(+), 140 deletions(-) create mode 100644 .clang-format create mode 100644 .clang-tidy create mode 100644 .clangd -- 2.47.0

2 16

[PATCH v3] tcp: unify payload and flags l2 frames array
by Jon Maloy 07 Nov '24

07 Nov '24

In order to reduce static memory and code footprint, we merge the array for l2 flag frames into the one for payload frames. This change also ensures that no flag message will be sent out over the l2 media bypassing already queued payload messages. Performance measurements with iperf3, where we force all traffic via the tap queue, show no significant difference: Dual traffic both directions sinmultaneously, with patch: ======================================================== host->ns: -------- [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-100.00 sec 36.3 GBytes 3.12 Gbits/sec 4759 sender [ 5] 0.00-100.04 sec 36.3 GBytes 3.11 Gbits/sec receiver ns->host: --------- [ ID] Interval Transfer Bitrate [ 5] 0.00-100.00 sec 321 GBytes 27.6 Gbits/sec receiver Dual traffic both directions sinmultaneously, without patch: ============================================================ host->ns: -------- [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-100.00 sec 35.0 GBytes 3.01 Gbits/sec 6001 sender [ 5] 0.00-100.04 sec 34.8 GBytes 2.99 Gbits/sec receiver ns->host -------- [ ID] Interval Transfer Bitrate [ 5] 0.00-100.00 sec 345 GBytes 29.6 Gbits/sec receiver Single connection, with patch: ============================== host->ns: --------- [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-100.00 sec 138 GBytes 11.8 Gbits/sec 922 sender [ 5] 0.00-100.04 sec 138 GBytes 11.8 Gbits/sec receiver ns->host: ----------- [ ID] Interval Transfer Bitrate [ 5] 0.00-100.00 sec 430 GBytes 36.9 Gbits/sec receiver Single connection, without patch: ================================= host->ns: ------------ [ ID] Interval Transfer Bitrate Retr [ 5] 0.00-100.00 sec 139 GBytes 11.9 Gbits/sec 900 sender [ 5] 0.00-100.04 sec 139 GBytes 11.9 Gbits/sec receiver ns->host: --------- [ ID] Interval Transfer Bitrate [ 5] 0.00-100.00 sec 440 GBytes 37.8 Gbits/sec receiver Signed-off-by: Jon Maloy <jmaloy(a)redhat.com> ---- v2: - Adapted to and rebased on latest release. - Removed redundant tcp_flags_push() function. - Added measurement results. v3: - Rebased on latest release. - Fixes based on comments from D. Gibson --- tcp.c | 1 - tcp_buf.c | 70 ++++++++++++-------------------------------------- tcp_buf.h | 1 - tcp_internal.h | 15 ----------- 4 files changed, 17 insertions(+), 70 deletions(-) diff --git a/tcp.c b/tcp.c index 4e0a17e..b652900 100644 --- a/tcp.c +++ b/tcp.c @@ -937,7 +937,6 @@ bool tcp_flow_defer(const struct tcp_tap_conn *conn) /* cppcheck-suppress [constParameterPointer, unmatchedSuppression] */ void tcp_defer_handler(struct ctx *c) { - tcp_flags_flush(c); tcp_payload_flush(c); } diff --git a/tcp_buf.c b/tcp_buf.c index 274e313..d29c1a9 100644 --- a/tcp_buf.c +++ b/tcp_buf.c @@ -20,7 +20,7 @@ #include <netinet/ip.h> -#include <linux/tcp.h> +#include <netinet/tcp.h> #include "util.h" #include "ip.h" @@ -59,22 +59,10 @@ static_assert(MSS6 <= sizeof(tcp_payload[0].data), "MSS6 is greater than 65516") static struct tcp_tap_conn *tcp_frame_conns[TCP_FRAMES_MEM]; static unsigned int tcp_payload_used; -static struct tap_hdr tcp_flags_tap_hdr[TCP_FRAMES_MEM]; -/* IPv4 headers for TCP segment without payload */ -static struct iphdr tcp4_flags_ip[TCP_FRAMES_MEM]; -/* TCP segments without payload for IPv4 frames */ -static struct tcp_flags_t tcp_flags[TCP_FRAMES_MEM]; - -static unsigned int tcp_flags_used; - -/* IPv6 headers for TCP segment without payload */ -static struct ipv6hdr tcp6_flags_ip[TCP_FRAMES_MEM]; - /* recvmsg()/sendmsg() data for tap */ static struct iovec iov_sock [TCP_FRAMES_MEM + 1]; static struct iovec tcp_l2_iov[TCP_FRAMES_MEM][TCP_NUM_IOVS]; -static struct iovec tcp_l2_flags_iov[TCP_FRAMES_MEM][TCP_NUM_IOVS]; /** * tcp_update_l2_buf() - Update Ethernet header buffers with addresses @@ -103,15 +91,6 @@ void tcp_sock_iov_init(const struct ctx *c) for (i = 0; i < ARRAY_SIZE(tcp_payload); i++) { tcp6_payload_ip[i] = ip6; tcp4_payload_ip[i] = iph; - tcp_payload[i].th.doff = sizeof(struct tcphdr) / 4; - tcp_payload[i].th.ack = 1; - } - - for (i = 0; i < ARRAY_SIZE(tcp_flags); i++) { - tcp6_flags_ip[i] = ip6; - tcp4_flags_ip[i] = iph; - tcp_flags[i].th.doff = sizeof(struct tcphdr) / 4; - tcp_flags[i].th.ack = 1; } for (i = 0; i < TCP_FRAMES_MEM; i++) { @@ -121,25 +100,6 @@ void tcp_sock_iov_init(const struct ctx *c) iov[TCP_IOV_ETH].iov_len = sizeof(struct ethhdr); iov[TCP_IOV_PAYLOAD].iov_base = &tcp_payload[i]; } - - for (i = 0; i < TCP_FRAMES_MEM; i++) { - struct iovec *iov = tcp_l2_flags_iov[i]; - - iov[TCP_IOV_TAP] = tap_hdr_iov(c, &tcp_flags_tap_hdr[i]); - iov[TCP_IOV_ETH].iov_len = sizeof(struct ethhdr); - iov[TCP_IOV_PAYLOAD].iov_base = &tcp_flags[i]; - } -} - -/** - * tcp_flags_flush() - Send out buffers for segments with no data (flags) - * @c: Execution context - */ -void tcp_flags_flush(const struct ctx *c) -{ - tap_send_frames(c, &tcp_l2_flags_iov[0][0], TCP_NUM_IOVS, - tcp_flags_used); - tcp_flags_used = 0; } /** @@ -171,7 +131,7 @@ static void tcp_revert_seq(const struct ctx *c, struct tcp_tap_conn **conns, } /** - * tcp_payload_flush() - Send out buffers for segments with data + * tcp_payload_flush() - Send out buffers for segments with data or flags * @c: Execution context */ void tcp_payload_flush(const struct ctx *c) @@ -197,37 +157,35 @@ void tcp_payload_flush(const struct ctx *c) */ int tcp_buf_send_flag(const struct ctx *c, struct tcp_tap_conn *conn, int flags) { - struct tcp_flags_t *payload; + struct tcp_payload_t *payload; struct iovec *iov; size_t optlen; size_t l4len; uint32_t seq; int ret; - iov = tcp_l2_flags_iov[tcp_flags_used]; + iov = tcp_l2_iov[tcp_payload_used]; if (CONN_V4(conn)) { - iov[TCP_IOV_IP] = IOV_OF_LVALUE(tcp4_flags_ip[tcp_flags_used]); + iov[TCP_IOV_IP] = IOV_OF_LVALUE(tcp4_payload_ip[tcp_payload_used]); iov[TCP_IOV_ETH].iov_base = &tcp4_eth_src; } else { - iov[TCP_IOV_IP] = IOV_OF_LVALUE(tcp6_flags_ip[tcp_flags_used]); + iov[TCP_IOV_IP] = IOV_OF_LVALUE(tcp6_payload_ip[tcp_payload_used]); iov[TCP_IOV_ETH].iov_base = &tcp6_eth_src; } payload = iov[TCP_IOV_PAYLOAD].iov_base; seq = conn->seq_to_tap; ret = tcp_prepare_flags(c, conn, flags, &payload->th, - &payload->opts, &optlen); + (struct tcp_syn_opts *)&payload->data, &optlen); if (ret <= 0) return ret; - tcp_flags_used++; + tcp_payload_used++; l4len = tcp_l2_buf_fill_headers(conn, iov, optlen, NULL, seq, false); iov[TCP_IOV_PAYLOAD].iov_len = l4len; - if (flags & DUP_ACK) { - struct iovec *dup_iov; + struct iovec *dup_iov = tcp_l2_iov[tcp_payload_used++]; - dup_iov = tcp_l2_flags_iov[tcp_flags_used++]; memcpy(dup_iov[TCP_IOV_TAP].iov_base, iov[TCP_IOV_TAP].iov_base, iov[TCP_IOV_TAP].iov_len); dup_iov[TCP_IOV_ETH].iov_base = iov[TCP_IOV_ETH].iov_base; @@ -237,8 +195,8 @@ int tcp_buf_send_flag(const struct ctx *c, struct tcp_tap_conn *conn, int flags) dup_iov[TCP_IOV_PAYLOAD].iov_len = l4len; } - if (tcp_flags_used > TCP_FRAMES_MEM - 2) - tcp_flags_flush(c); + if (tcp_payload_used > TCP_FRAMES_MEM - 2) + tcp_payload_flush(c); return 0; } @@ -254,6 +212,7 @@ int tcp_buf_send_flag(const struct ctx *c, struct tcp_tap_conn *conn, int flags) static void tcp_data_to_tap(const struct ctx *c, struct tcp_tap_conn *conn, ssize_t dlen, int no_csum, uint32_t seq) { + struct tcp_payload_t *payload; const uint16_t *check = NULL; struct iovec *iov; size_t l4len; @@ -274,6 +233,11 @@ static void tcp_data_to_tap(const struct ctx *c, struct tcp_tap_conn *conn, iov[TCP_IOV_IP] = IOV_OF_LVALUE(tcp6_payload_ip[tcp_payload_used]); iov[TCP_IOV_ETH].iov_base = &tcp6_eth_src; } + payload = iov[TCP_IOV_PAYLOAD].iov_base; + payload->th.th_off = sizeof(struct tcphdr) / 4; + payload->th.th_x2 = 0; + payload->th.th_flags = 0; + payload->th.ack = 1; l4len = tcp_l2_buf_fill_headers(conn, iov, dlen, check, seq, false); iov[TCP_IOV_PAYLOAD].iov_len = l4len; if (++tcp_payload_used > TCP_FRAMES_MEM - 1) diff --git a/tcp_buf.h b/tcp_buf.h index 49c04d4..54f5e53 100644 --- a/tcp_buf.h +++ b/tcp_buf.h @@ -7,7 +7,6 @@ #define TCP_BUF_H void tcp_sock_iov_init(const struct ctx *c); -void tcp_flags_flush(const struct ctx *c); void tcp_payload_flush(const struct ctx *c); int tcp_buf_data_from_sock(const struct ctx *c, struct tcp_tap_conn *conn); int tcp_buf_send_flag(const struct ctx *c, struct tcp_tap_conn *conn, int flags); diff --git a/tcp_internal.h b/tcp_internal.h index a5a47df..c846f60 100644 --- a/tcp_internal.h +++ b/tcp_internal.h @@ -134,21 +134,6 @@ struct tcp_syn_opts { .ws = TCP_OPT_WS(ws_), \ }) -/** - * struct tcp_flags_t - TCP header and data to send zero-length - * segments (flags) - * @th: TCP header - * @opts TCP options - */ -struct tcp_flags_t { - struct tcphdr th; - struct tcp_syn_opts opts; -#ifdef __AVX2__ -} __attribute__ ((packed, aligned(32))); -#else -} __attribute__ ((packed, aligned(__alignof__(unsigned int)))); -#endif - extern char tcp_buf_discard [MAX_WINDOW]; void conn_flag_do(const struct ctx *c, struct tcp_tap_conn *conn, -- 2.45.2

3 2

spliced performance
by Jon Maloy 06 Nov '24

06 Nov '24

I re-ran the splice ns->host test using splice, and it works flawlessly. It seems like Stefano's fix solved the problem. /jon [ 5] 90.00-91.00 sec 4.06 GBytes 34.9 Gbits/sec [ 5] 91.00-92.00 sec 4.05 GBytes 34.8 Gbits/sec [ 5] 92.00-93.00 sec 4.04 GBytes 34.7 Gbits/sec [ 5] 93.00-94.00 sec 3.93 GBytes 33.8 Gbits/sec [ 5] 94.00-95.00 sec 3.92 GBytes 33.6 Gbits/sec [ 5] 95.00-96.00 sec 3.95 GBytes 33.9 Gbits/sec [ 5] 96.00-97.00 sec 3.91 GBytes 33.6 Gbits/sec [ 5] 97.00-98.00 sec 3.92 GBytes 33.7 Gbits/sec [ 5] 98.00-99.00 sec 3.67 GBytes 31.5 Gbits/sec [ 5] 99.00-100.00 sec 3.64 GBytes 31.3 Gbits/sec [ 5] 100.00-100.00 sec 1.88 MBytes 35.0 Gbits/sec - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate [ 5] 0.00-100.00 sec 389 GBytes 33.4 Gbits/sec receiver ----------------------------------------------------------- Server listening on 5201 (test #2) -----------------------------------------------------------

2 1

[PATCH] fwd: Squash different-signedness comparison warning
by David Gibson 06 Nov '24

06 Nov '24

On certain architectures we get a warning about comparison between different signedness integers in fwd_probe_ephemeral(). This is because NUM_PORTS evaluates to an unsigned integer. It's a fixed value, though and we know it will fit in a signed long on anything reasonable, so add a cast to suppress the warning. Signed-off-by: David Gibson <david(a)gibson.dropbear.id.au> --- fwd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fwd.c b/fwd.c index c71f5e1..0b7f8b1 100644 --- a/fwd.c +++ b/fwd.c @@ -75,8 +75,8 @@ void fwd_probe_ephemeral(void) if (*end || errno) goto parse_err; - if (min < 0 || min >= NUM_PORTS || - max < 0 || max >= NUM_PORTS) + if (min < 0 || min >= (long)NUM_PORTS || + max < 0 || max >= (long)NUM_PORTS) goto parse_err; fwd_ephemeral_min = min; -- 2.47.0

1 0

[PATCH] test: Adjust misplaced sleeps in two_guests code
by David Gibson 06 Nov '24

06 Nov '24

Most of our transfer tests using socat use 'sleep' waaiting for the server side to be ready before starting the client. However in two_guests/basic the sleep is in the wrong place: rather than being between starting the server and starting the client, it's after waiting for the server to complete. This causes occasional hangs when the client runs before the server is ready - in that case the receiving guest sends an RST, which we don't (currently) propagate back to the sender. Signed-off-by: David Gibson <david(a)gibson.dropbear.id.au> --- test/two_guests/basic | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/test/two_guests/basic b/test/two_guests/basic index 9ba5efe..e2338ff 100644 --- a/test/two_guests/basic +++ b/test/two_guests/basic @@ -52,33 +52,33 @@ check [ "__ADDR2_6__" = "__HOST_ADDR6__" ] test TCP/IPv4: guest 1 > guest 2 g1out GW1 ip -j -4 route show|jq -rM '.[] | select(.dst == "default").gateway' guest2b socat -u TCP4-LISTEN:10004 OPEN:msg,create,trunc +sleep 1 guest1 echo "Hello_from_guest_1" | socat -u STDIN TCP4:__GW1__:10004 guest2w -sleep 1 g2out MSG2 cat msg check [ "__MSG2__" = "Hello_from_guest_1" ] test TCP/IPv6: guest 2 > guest 1 g2out GW2_6 ip -j -6 route show|jq -rM '.[] | select(.dst == "default").gateway' guest1b socat -u TCP6-LISTEN:10001 OPEN:msg,create,trunc +sleep 1 guest2 echo "Hello_from_guest_2" | socat -u STDIN TCP6:[__GW2_6__%__IFNAME2__]:10001 guest1w -sleep 1 g1out MSG1 cat msg check [ "__MSG1__" = "Hello_from_guest_2" ] test UDP/IPv4: guest 1 > guest 2 guest2b socat -u TCP4-LISTEN:10004 OPEN:msg,create,trunc +sleep 1 guest1 echo "Hello_from_guest_1" | socat -u STDIN TCP4:__GW1__:10004 guest2w -sleep 1 g2out MSG2 cat msg check [ "__MSG2__" = "Hello_from_guest_1" ] test UDP/IPv6: guest 2 > guest 1 guest1b socat -u TCP6-LISTEN:10001 OPEN:msg,create,trunc +sleep 1 guest2 echo "Hello_from_guest_2" | socat -u STDIN TCP6:[__GW2_6__%__IFNAME2__]:10001 guest1w -sleep 1 g1out MSG1 cat msg check [ "__MSG1__" = "Hello_from_guest_2" ] -- 2.47.0

2 1

2025

2024

2023

2022

2021

dev