July 2024 - dev

[PATCH v2 0/4] Add vhost-user support to passt. (part 3)
by Laurent Vivier 15 Aug '24

15 Aug '24

This series of patches adds vhost-user support to passt and then allows passt to connect to QEMU network backend using virtqueue rather than a socket. With QEMU, rather than using to connect: -netdev stream,id=s,server=off,addr.type=unix,addr.path=/tmp/passt_1.socket we will use: -chardev socket,id=chr0,path=/tmp/passt_1.socket -netdev vhost-user,id=netdev0,chardev=chr0 -device virtio-net,netdev=netdev0 -object memory-backend-memfd,id=memfd0,share=on,size=$RAMSIZE -numa node,memdev=memfd0 The memory backend is needed to share data between passt and QEMU. Performance comparison between "-netdev stream" and "-netdev vhost-user": $ iperf3 -c localhost -p 10001 -t 60 -6 -u -b 50G socket: [ 5] 0.00-60.05 sec 95.6 GBytes 13.7 Gbits/sec 0.017 ms 6998988/10132413 (69%) receiver vhost-user: [ 5] 0.00-60.04 sec 237 GBytes 33.9 Gbits/sec 0.006 ms 53673/7813770 (0.69%) receiver $ iperf3 -c localhost -p 10001 -t 60 -4 -u -b 50G socket: [ 5] 0.00-60.05 sec 98.9 GBytes 14.1 Gbits/sec 0.018 ms 6260735/9501832 (66%) receiver vhost-user: [ 5] 0.00-60.05 sec 235 GBytes 33.7 Gbits/sec 0.008 ms 37581/7752699 (0.48%) receiver $ iperf3 -c localhost -p 10001 -t 60 -6 socket: [ 5] 0.00-60.00 sec 17.3 GBytes 2.48 Gbits/sec 0 sender [ 5] 0.00-60.06 sec 17.3 GBytes 2.48 Gbits/sec receiver vhost-user: [ 5] 0.00-60.00 sec 191 GBytes 27.4 Gbits/sec 0 sender [ 5] 0.00-60.05 sec 191 GBytes 27.3 Gbits/sec receiver $ iperf3 -c localhost -p 10001 -t 60 -4 socket: [ 5] 0.00-60.00 sec 15.6 GBytes 2.24 Gbits/sec 0 sender [ 5] 0.00-60.06 sec 15.6 GBytes 2.24 Gbits/sec receiver vhost-user: [ 5] 0.00-60.00 sec 189 GBytes 27.1 Gbits/sec 0 sender [ 5] 0.00-60.04 sec 189 GBytes 27.0 Gbits/sec receiver v2: - remove PATCH 4 - rewrite PATCH 2 and 3 to follow passt coding style - move some code from PATCH 3 to PATCH 4 (previously PATCH 5) - partially addressed David's comment on PATCH 5 Laurent Vivier (4): packet: replace struct desc by struct iovec vhost-user: introduce virtio API vhost-user: introduce vhost-user API vhost-user: add vhost-user Makefile | 4 +- checksum.c | 1 - conf.c | 24 +- iov.c | 1 - isolation.c | 15 +- packet.c | 97 ++-- packet.h | 16 +- passt.c | 16 +- passt.h | 10 + pcap.c | 1 - tap.c | 114 ++++- tap.h | 5 +- tcp.c | 17 +- tcp_vu.c | 560 +++++++++++++++++++++ tcp_vu.h | 12 + udp.c | 54 +- udp_internal.h | 39 ++ udp_vu.c | 240 +++++++++ udp_vu.h | 11 + util.h | 11 + vhost_user.c | 1273 ++++++++++++++++++++++++++++++++++++++++++++++++ vhost_user.h | 197 ++++++++ virtio.c | 605 +++++++++++++++++++++++ virtio.h | 190 ++++++++ 24 files changed, 3392 insertions(+), 121 deletions(-) create mode 100644 tcp_vu.c create mode 100644 tcp_vu.h create mode 100644 udp_internal.h create mode 100644 udp_vu.c create mode 100644 udp_vu.h create mode 100644 vhost_user.c create mode 100644 vhost_user.h create mode 100644 virtio.c create mode 100644 virtio.h -- 2.45.2

3 12

[net] tcp: add SO_PEEK_OFF socket option tor TCPv6
by jmaloy＠redhat.com 14 Aug '24

14 Aug '24

From: Jon Maloy <jmaloy(a)redhat.com> When we added the SO_PEEK_OFF socket option to TCP we forgot to add it even for TCP on IPv6. We do that here. Fixes: 05ea491641d3 ("tcp: add support for SO_PEEK_OFF socket option") Signed-off-by: Jon Maloy <jmaloy(a)redhat.com> --- net/ipv6/af_inet6.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index 90d2c7e3f5e9..ba69b86f1c7d 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -708,6 +708,7 @@ const struct proto_ops inet6_stream_ops = { .splice_eof = inet_splice_eof, .sendmsg_locked = tcp_sendmsg_locked, .splice_read = tcp_splice_read, + .set_peek_off = sk_set_peek_off, .read_sock = tcp_read_sock, .read_skb = tcp_read_skb, .peek_len = tcp_peek_len, -- 2.45.2

5 4

[PATCH 0/2] Avoid duplicated clock_gettime() in logging
by David Gibson 05 Aug '24

05 Aug '24

Remove a small nit in the logging code. David Gibson (2): log: Make logfile_write() private log: Avoid duplicate calls to clock_gettime() log.c | 344 +++++++++++++++++++++++++++++----------------------------- log.h | 1 - 2 files changed, 172 insertions(+), 173 deletions(-) -- 2.45.2

2 5

[PATCH v2] pasta.c: modify hostname when detaching new namespace
by Danish Prakash 30 Jul '24

30 Jul '24

When invoking pasta without any arguments, it's difficult to tell whether we are in the new namespace or not leaving users a bit confused. This change modifies the host namespace to add a prefix "pasta-" to make it a bit more obvious. Signed-off-by: Danish Prakash <contact(a)danishpraka.sh> --- pasta.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/pasta.c b/pasta.c index 31e1e000ee1b..7ce502d3f299 100644 --- a/pasta.c +++ b/pasta.c @@ -50,6 +50,8 @@ #include "netlink.h" #include "log.h" +#define HOSTNAME_PREFIX "pasta-" + /* PID of child, in case we created a namespace */ int pasta_child_pid; @@ -178,6 +180,7 @@ struct pasta_spawn_cmd_arg { */ static int pasta_spawn_cmd(void *arg) { + char hostname[HOST_NAME_MAX + 1] = HOSTNAME_PREFIX; const struct pasta_spawn_cmd_arg *a; sigset_t set; @@ -188,6 +191,13 @@ static int pasta_spawn_cmd(void *arg) if (write_file("/proc/sys/net/ipv4/ping_group_range", "0 0")) warn("Cannot set ping_group_range, ICMP requests might fail"); + if (!gethostname(hostname + sizeof(HOSTNAME_PREFIX) - 1, HOST_NAME_MAX + 1 - + sizeof(HOSTNAME_PREFIX)) || errno == ENAMETOOLONG) { + hostname[HOST_NAME_MAX] = '\0'; + if (sethostname(hostname, strlen(hostname))) + warn("Unable to set pasta-prefixed hostname"); + } + /* Wait for the parent to be ready: see main() */ sigemptyset(&set); sigaddset(&set, SIGUSR1); -- 2.45.2

2 1

[PATCH v2] Fix typo in README file
by AbdAlRahman Gad 29 Jul '24

29 Jul '24

- remove duplicated 'the' in the 'Services' section Signed-off-by: AbdAlRahman Gad <abdobngad(a)gmail.com> --- v2: Add Signed-off-by: tag README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index aa2afac..bb114ea 100644 --- a/README.md +++ b/README.md @@ -398,7 +398,7 @@ services: and nameserver using SLAAC * [DHCPv6 server](/passt/tree/dhcpv6.c): a simple implementation handing out one single IPv6 address to the guest or namespace, - namely, the the same address as the first one configured for the upstream host + namely, the same address as the first one configured for the upstream host interface, and passing the nameservers configured on the host ## Addresses -- 2.43.0

3 2

[PATCH] pasta.c: modify hostname when detaching new namespace
by Danish Prakash 29 Jul '24

29 Jul '24

When invoking pasta without any arguments, it's difficult to tell whether we are in the new namespace or not leaving users a bit confused. This change modifies the host namespace to add a prefix "pasta-" to make it a bit more obvious. Signed-off-by: Danish Prakash <contact(a)danishpraka.sh> --- pasta.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/pasta.c b/pasta.c index 31e1e00..840a2b1 100644 --- a/pasta.c +++ b/pasta.c @@ -180,6 +180,8 @@ static int pasta_spawn_cmd(void *arg) { const struct pasta_spawn_cmd_arg *a; sigset_t set; + char hostname[HOST_NAME_MAX+1], pasta_hostname[HOST_NAME_MAX+1]; + char *hostname_prefix = "pasta-"; /* We run in a detached PID and mount namespace: mount /proc over */ if (mount("", "/proc", "proc", 0, NULL)) @@ -188,6 +190,17 @@ static int pasta_spawn_cmd(void *arg) if (write_file("/proc/sys/net/ipv4/ping_group_range", "0 0")) warn("Cannot set ping_group_range, ICMP requests might fail"); + if (gethostname(hostname, HOST_NAME_MAX+1) == 0) { + if ((strlen(hostname) + strlen(hostname_prefix)) > HOST_NAME_MAX) { + hostname[strlen(hostname)-strlen(hostname_prefix)] = '\0'; + } + sprintf(pasta_hostname, "%s%s", hostname_prefix, hostname); + + if (sethostname(pasta_hostname, strlen(pasta_hostname)) != 0) { + warn("Unable to set pasta-prefixed hostname"); + } + } + /* Wait for the parent to be ready: see main() */ sigemptyset(&set); sigaddset(&set, SIGUSR1); -- 2.45.1

2 10

[PATCH 0/5] Fix assorted errors in the Qemu socket tap receive handler
by David Gibson 29 Jul '24

29 Jul '24

A long time ago Matej Hrica pointed out a possible buffer overrun when receiving data from the qemu socket. Stefano recently proposed a fix for this, but I don't think it's quite right. This series is a different approach to fixing that problem and a few adjacent ones. David Gibson (5): tap: Better report errors receiving from QEMU socket tap: Don't attempt to carry on if we get a bad frame length from qemu tap: Don't use EPOLLET on Qemu sockets tap: Correctly handle frames of odd length tap: Improve handling of partially received frames on qemu socket passt.h | 1 - tap.c | 72 ++++++++++++++++++++++++++++++--------------------------- util.h | 16 +++++++++++++ 3 files changed, 54 insertions(+), 35 deletions(-) -- 2.45.2

2 17

[PATCH 0/3] Fixes for log timestamps
by Stefano Brivio 29 Jul '24

29 Jul '24

Taken out of the "Minor assorted fixes, mostly logging and tests" series I recently posted. Stefano Brivio (3): log, util: Fix sub-second part in relative log time calculation log: Initialise timestamp for relative log time also if we use a log file log: Fetch log times with CLOCK_MONOTONIC, not CLOCK_REALTIME log.c | 46 +++++++++++++++++++++++----------------------- log.h | 1 + passt.c | 2 ++ util.c | 25 ++++++++++++++++++------- util.h | 1 + 5 files changed, 45 insertions(+), 30 deletions(-) -- 2.43.0

2 7

[PATCH] Fix typo in README file
by AbdAlRahman Gad 27 Jul '24

27 Jul '24

- remove duplicated 'the' in the 'Services' section --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index aa2afac..bb114ea 100644 --- a/README.md +++ b/README.md @@ -398,7 +398,7 @@ services: and nameserver using SLAAC * [DHCPv6 server](/passt/tree/dhcpv6.c): a simple implementation handing out one single IPv6 address to the guest or namespace, - namely, the the same address as the first one configured for the upstream host + namely, the same address as the first one configured for the upstream host interface, and passing the nameservers configured on the host ## Addresses -- 2.43.0

1 0

passt: new version 2024_07_26.57a21d2 available
by Stefano Brivio 26 Jul '24

26 Jul '24

The new version with tag 2024_07_26.57a21d2 includes the following changes: 57a21d2 tap: Improve handling of partially received frames on qemu socket 37e3b24 tap: Correctly handle frames of odd length 4684f60 tap: Don't use EPOLLET on Qemu sockets 9e3f235 tap: Don't attempt to carry on if we get a bad frame length from qemu a06db27 tap: Better report errors receiving from QEMU socket 77c092e log: Fetch log times with CLOCK_MONOTONIC, not CLOCK_REALTIME e5c37ba log: Initialise timestamp for relative log time also if we use a log file 327d9d4 log, util: Fix sub-second part in relative log time calculation 2ce1d37 test/lib/perf_report: Fix highlight e9a5423 test: Fix spurious test failure with systemd-resolved becf81a fwd: Broaden what we consider for DNS specific forwarding rules 0ada84e fwd: Refactor tests in fwd_nat_from_tap() for clarity 4a333c8 conf: Accept addresses enclosed by square brackets in port forwarding specifiers 6ff702f tap: Exit if we fail to bind a UNIX domain socket with explicit path f72d35a test: iperf3 3.16 introduces multiple threads, drop our own implementation of that 606e0c7 test: Update names of symbols and slabinfo entries f16f8f5 test: Fix memory/passt tests, --netns-only is not a valid option for passt 1cd7730 log: Drop newlines in the middle of the perror()-like messages 1329558 tcp: Change SO_PEEK_OFF support message to debug() d19b396 tap: Don't quit if pasta gets EIO on writev() to tap, interface might be down a09aeb4 tcp: Correctly update SO_PEEK_OFF when tcp_send_frames() drops frames 9cb6b50 tcp: probe for SO_PEEK_OFF both in tcpv4 and tcp6 882599e udp: Rename UDP listening sockets d29fa08 udp: Remove rdelta port forwarding maps d89b3aa udp: Remove obsolete socket tracking 898f797 udp: Direct datagrams from host to guest via flow table b7ad193 udp: Find or create flows for datagrams from tap interface 8126f7a udp: Remove obsolete splice tracking e0647ad udp: Handle "spliced" datagrams with per-flow sockets a45a7e9 udp: Create flows for datagrams from originating sockets 8abd06e fwd: Update flow forwarding logic for UDP c000f2a flow, icmp: Use general flow forwarding rules for ICMP 060f24e flow, tcp: Flow based NAT and port forwarding for TCP 4cd753e icmp: Manage outbound socket address via flow table 781164e flow: Helper to create sockets based on flowside 2faf6fc icmp: Eliminate icmp_id_map 2f40a01 icmp: Look up ping flows using flow hash 6d76278 icmp: Obtain destination addresses from the flowsides 5cffb1b icmp: Remove redundant id field from flow table entry 508adde tcp: Re-use flow hash for initial sequence number generation acca423 flow, tcp: Generalise TCP hash table to general flow hash table 163a339 tcp, flow: Replace TCP specific hash function with general flow hash f19a8f7 tcp_splice: Eliminate SPLICE_V6 flag 528a651 tcp: Simplify endpoint validation using flowside information e2ea10e tcp: Manage outbound address via flow table 52d45f1 tcp: Obtain guest address from flowside f9fe212 tcp, flow: Remove redundant information, repack connection structures 4e2d36e flow: Common address information for target side 8012f5f flow: Common address information for initiating side ba74b1f doc: Extend zero-recv test with methods using msghdr 01e5611 doc: Test behaviour of closing duplicate UDP sockets 66a02c9 tcp_splice: Use parameterised macros for per-side event/flag bits 5235c47 flow: Introduce flow_foreach_sidei() macro 71d7985 flow, tcp_splice: Prefer 'sidei' for variables referring to side index 9b125e7 flow, icmp, tcp: Clean up helpers for getting flow from index 2fa91ee udp: Handle errors on UDP sockets 6bd8283 util: Add AF_UNSPEC support to sockaddr_ntop() 4e1f850 udp, tcp: Tweak handling of no_udp and no_tcp flags 272d1d0 udp: Make udp_sock_recv static f79c423 conf: Don't configure port forwarding for a disabled protocol a740e16 tcp: handle shrunk window advertisements from guest e63d281 tcp: leverage support of SO_PEEK_OFF socket option when available 8bd57bf doc: Trivial fix for reuseaddr-priority ec2691a doc: Test behaviour of zero length datagram recv()s 299c407 doc: Add program to document and test assumptions about SO_REUSEADDR be0214c udp: Consolidate datagram batching 69e5393 udp: Move some more of sock_handler tasks into sub-functions c6c61a9 udp: Don't repeatedly initialise udp[46]_eth_hdr 55aff45 udp: Unify udp[46]_l2_iov 9f9b15f udp: Unify udp[46]_mh_splice fbd78b6 udp: Rename IOV and mmsghdr arrays f62c33d udp: Pass full epoll reference through more of sock handler path 8f8eb73 flow: Add flow_sidx_valid() helper 74c1c5e util: sock_l4() determine protocol from epoll type rather than the reverse b625ed5 conf: Use the right maximum buffer size for c->sock_path 403a7c1 tcp_splice: Check return value of setsockopt() for SO_RCVLOWAT 21ee1eb conf: Copy up to MAXDNSRCH - 1 bytes, not MAXDNSRCH https://passt.top/passt/log/?qt=range&q=2024_06_24.1ee2eca..2024_07_26.57a2… Packages: - Alpine Linux: https://pkgs.alpinelinux.org/packages?name=passt - Arch Linux: https://www.archlinux.org/packages/extra/x86_64/passt/ https://archlinuxarm.org/packages/aarch64/passt https://archlinuxarm.org/packages/armv7h/passt - Chimera: https://pkgs.chimera-linux.org/packages?name=passt - Clear Linux: https://github.com/clearlinux-pkgs/passt/ - Debian tracker: https://tracker.debian.org/pkg/passt - Copr (CentOS Stream, EPEL, Fedora, Mageia): https://copr.fedorainfracloud.org/coprs/sbrivio/passt/build/7793866/ permanent mirror: https://passt.top/builds/copr/0^20240726.g57a21d2/ - Fedora updates: https://bodhi.fedoraproject.org/updates/?packages=passt - Gentoo versions: https://packages.gentoo.org/packages/net-misc/passt - GNU Guix: https://packages.guix.gnu.org/packages/passt/ - NixOS: https://github.com/NixOS/nixpkgs/tree/nixos-unstable/pkgs/by-name/pa/passt - openSUSE: https://software.opensuse.org/package/passt - PLD Linux: https://git.pld-linux.org/cgi-bin/gitweb.cgi?p=packages/passt.git - Solus: https://github.com/getsolus/packages/tree/main/packages/p/passt - Ubuntu tracker: https://launchpad.net/ubuntu/+source/passt - Void Linux: https://voidlinux.org/packages/?q=passt - Static builds: - Package for other RPM-based distributions, x86_64 only: https://passt.top/builds/latest/x86_64/passt-g57a21d2-1.x86_64.rpm - x86_64 static binaries: https://passt.top/builds/latest/x86_64/ - Debian package, from x86_64 static build: https://passt.top/builds/latest/x86_64/passt_57a21d2-1_all.deb -- Stefano

1 0