September 2024 - dev

[PATCH v2 00/14] Clean up checksum and header generation for inbound packets
by David Gibson 10 Oct '25

10 Oct '25

The main packet "fast paths" for UDP and TCP mostly just forward packets rather than generating them from scratch. However the control paths for ICMP and DHCP sometimes generate packets more or less from scratch. Because these are relatively rare, it's not performance critical. The paths for sending these packets have some duplication of the header generation. There's also some layering violation in tap_ip_send() which both generates IP headers and updates the L4 (UDP or UCMP) checksum. Finally that checksum generation is a little awkward: it temporarily generates the IP pseudo header (or something close enough to serve) in the place of the actual header, generates the checksum, then replaces it with the real IP header. This approach seems to be causing miscompiles with some LTO optimization, because the stores to the pseudo header are being moved or elided across the code calculating the checksum. This series addresses all of these. We consolidate and clarify the packet sending helpers, and use them in some places there was previously duplicated code. In the process we use new checksum generation helpers which take a different approach which should avoid the LTO problems (this aspect I haven't tested yet though). Changes since v1: * Numerous minor style changes * Rename header generation helpers to make their behaviour clearer * Added several missing function doc comments * Corrected some erroneous statements and terms in comments David Gibson (14): Add csum_icmp6() helper for calculating ICMPv6 checksums Add csum_icmp4() helper for calculating ICMP checksums Add csum_udp6() helper for calculating UDP over IPv6 checksums Add csum_udp4() helper for calculating UDP over IPv4 checksums Add csum_ip4_header() helper to calculate IPv4 header checksums Add helpers for normal inbound packet destination addresses Remove support for TCP packets from tap_ip_send() tap: Remove unhelpeful vnet_pre optimization from tap_send() Split tap_ip_send() into IPv4 and IPv6 specific functions tap: Split tap_ip6_send() into UDP and ICMP variants ndp: Remove unneeded eh_source parameter ndp: Use tap_icmp6_send() helper tap: Split tap_ip4_send() into UDP and ICMP variants dhcp: Use tap_udp4_send() helper in dhcp() arp.c | 2 +- checksum.c | 120 ++++++++++++++++----- checksum.h | 15 ++- dhcp.c | 19 +--- dhcpv6.c | 21 +--- icmp.c | 12 +-- ndp.c | 28 +---- ndp.h | 3 +- tap.c | 312 ++++++++++++++++++++++++++++++++++------------------- tap.h | 19 +++- 10 files changed, 345 insertions(+), 206 deletions(-) -- 2.37.3

2 17

[PATCH v5 0/5] tcp: use csum_iov() in tcp_update_check_tcp[4|6]()
by Laurent Vivier 03 Oct '24

03 Oct '24

For vhost-user, we will need to spread TCP payload over several buffers. To re-use tcp_update_check_tcp[4|6](), provide an iovec rather than a pointer to a buffer. This series updates also csum_iov() and pcap_iov() to add an offset of bytes to skip in the iovec array. It's based on top of "tcp: Use tcp_payload_t rather than tcphdr" that is added in the series for convenience. v3 adds "udp: Update UDP checksum using an iovec array" that modifies csum_udp4() and csum_udp6() in the same way. Laurent Vivier (5): tcp: Use tcp_payload_t rather than tcphdr pcap: Add an offset argument in pcap_iov() checksum: Add an offset argument in csum_iov() tcp: Update TCP checksum using an iovec array udp: Update UDP checksum using an iovec array checksum.c | 46 +++++++++++----- checksum.h | 7 +-- iov.c | 1 - pcap.c | 5 +- pcap.h | 2 +- tap.c | 14 +++-- tap.h | 2 +- tcp.c | 140 +++++++++++++++++++++++++++++++++++++++---------- tcp_buf.c | 29 ---------- tcp_internal.h | 29 ++++++++++ udp.c | 17 ++++-- 11 files changed, 206 insertions(+), 86 deletions(-) -- 2.46.0

3 15

[PATCH] test: Add tests for passt in vhost-user mode
by Stefano Brivio 30 Sep '24

30 Sep '24

Run functional and performance tests for vhost-user mode as well. For functional tests, we add passt_vu and passt_vu_in_ns as symbolic links to their non-vhost-user counterparts, as no differences are intended but we want to distinguish them in test logs. For performance tests, instead, we add separate perf/passt_vu_tcp and perf/passt_vu_udp files, as we need longer test duration, as well as higher UDP sending bandwidths and larger TCP windows, to actually get the highest throughput vhost-user mode offers. For valgrind tests, vhost-user mode needs two extra system calls: statx and readlink. Add them as EXTRA_SYSCALLS for the valgrind target. Signed-off-by: Stefano Brivio <sbrivio(a)redhat.com> --- This is based on: [PATCH v5 0/4] Add vhost-user support to passt. (part 3) [PATCH] test: Kernel binary can now be passed via the KERNEL environmental variable and currently shows an issue reported by valgrind, as passt starts and gets a connection from QEMU: accepted connection from PID 4806 ==4804== Conditional jump or move depends on uninitialised value(s) ==4804== at 0x13AE26: vu_set_mem_table_exec (vhost_user.c:459) ==4804== by 0x13C756: vu_control_handler (vhost_user.c:1232) ==4804== by 0x10CE46: main (passt.c:380) ==4804== which I didn't really investigate. This is the relevant code: static bool vu_set_mem_table_exec(struct vu_dev *vdev, struct vhost_user_msg *msg) { struct vhost_user_memory m = msg->payload.memory, *memory = &m; unsigned int i; for (i = 0; i < vdev->nregions; i++) { <-- on top of that, once the tests terminate QEMU, we get: qemu-system-x86_64: Failed to set msg fds. qemu-system-x86_64: vhost VQ 0 ring restore failed: -22: Invalid argument (22) qemu-system-x86_64: Failed to set msg fds. qemu-system-x86_64: vhost VQ 1 ring restore failed: -22: Invalid argument (22) qemu-system-x86_64: terminating on signal 15 from pid 4220 (<unknown process>) About performance tests: I didn't try hard to get throughput figures as good as the ones Laurent reported by running iperf3 without arguments on his setup. We should probably play with that a bit more. Laurent, please feel free to carry this as part of your vhost-user series. Makefile | 3 +- test/lib/perf_report | 15 +++++++++ test/lib/setup | 77 ++++++++++++++++++++++++++++++++++++-------- test/run | 21 ++++++++++++ 4 files changed, 101 insertions(+), 15 deletions(-) diff --git a/Makefile b/Makefile index 1e8910d..ce8aa43 100644 --- a/Makefile +++ b/Makefile @@ -138,7 +138,8 @@ qrap: $(QRAP_SRCS) passt.h valgrind: EXTRA_SYSCALLS += rt_sigprocmask rt_sigtimedwait rt_sigaction \ rt_sigreturn getpid gettid kill clock_gettime mmap \ - mmap2 munmap open unlink gettimeofday futex + mmap2 munmap open unlink gettimeofday futex statx \ + readlink valgrind: FLAGS += -g -DVALGRIND valgrind: all diff --git a/test/lib/perf_report b/test/lib/perf_report index d1ef50b..c4ec817 100755 --- a/test/lib/perf_report +++ b/test/lib/perf_report @@ -49,6 +49,21 @@ td:empty { visibility: hidden; } __passt_tcp_LINE__ __passt_udp_LINE__ </table> +</li><li><p>passt with vhost-user support</p> +<table class="passt" width="70%"> + <tr> + <th/> + <th id="perf_passt_vu_tcp" colspan="__passt_vu_tcp_cols__">TCP, __passt_vu_tcp_threads__ at __passt_vu_tcp_freq__ GHz</th> + <th id="perf_passt_vu_udp" colspan="__passt_vu_udp_cols__">UDP, __passt_vu_udp_threads__ at __passt_vu_udp_freq__ GHz</th> + </tr> + <tr> + <td align="right">MTU:</td> + __passt_vu_tcp_header__ + __passt_vu_udp_header__ + </tr> + __passt_vu_tcp_LINE__ __passt_vu_udp_LINE__ +</table> + <style type="text/CSS"> table.pasta_local td { border: 0px solid; padding: 6px; line-height: 1; } table.pasta_local td { text-align: right; } diff --git a/test/lib/setup b/test/lib/setup index 5338393..3409bd2 100755 --- a/test/lib/setup +++ b/test/lib/setup @@ -15,8 +15,7 @@ INITRAMFS="${BASEPATH}/mbuto.img" VCPUS="$( [ $(nproc) -ge 8 ] && echo 6 || echo $(( $(nproc) / 2 + 1 )) )" -__mem_kib="$(sed -n 's/MemTotal:[ ]*$[0-9]*$ kB/\1/p' /proc/meminfo)" -VMEM="$((${__mem_kib} / 1024 / 4))" +MEM_KIB="$(sed -n 's/MemTotal:[ ]*$[0-9]*$ kB/\1/p' /proc/meminfo)" QEMU_ARCH="$(uname -m)" [ "${QEMU_ARCH}" = "i686" ] && QEMU_ARCH=i386 @@ -46,6 +45,7 @@ setup_passt() { [ ${PCAP} -eq 1 ] && __opts="${__opts} -p ${LOGDIR}/passt.pcap" [ ${DEBUG} -eq 1 ] && __opts="${__opts} -d" [ ${TRACE} -eq 1 ] && __opts="${__opts} --trace" + [ ${VHOST_USER} -eq 1 ] && __opts="${__opts} --vhost-user" context_run passt "make clean" context_run passt "make valgrind" @@ -54,16 +54,29 @@ setup_passt() { # pidfile isn't created until passt is listening wait_for [ -f "${STATESETUP}/passt.pid" ] + __vmem="$((${MEM_KIB} / 1024 / 4))" + if [ ${VHOST_USER} -eq 1 ]; then + __vmem="$(((${__vmem} + 500) / 1000))G" + __qemu_netdev=" \ + -chardev socket,id=c,path=${STATESETUP}/passt.socket \ + -netdev vhost-user,id=v,chardev=c \ + -device virtio-net,netdev=v \ + -object memory-backend-memfd,id=m,share=on,size=${__vmem} \ + -numa node,memdev=m" + else + __qemu_netdev="-device virtio-net-pci,netdev=s \ + -netdev stream,id=s,server=off,addr.type=unix,addr.path=${STATESETUP}/passt.socket" + fi + GUEST_CID=94557 context_run_bg qemu 'qemu-system-'"${QEMU_ARCH}" \ ' -machine accel=kvm' \ - ' -m '${VMEM}' -cpu host -smp '${VCPUS} \ + ' -m '${__vmem}' -cpu host -smp '${VCPUS} \ ' -kernel '"${KERNEL}" \ ' -initrd '${INITRAMFS}' -nographic -serial stdio' \ ' -nodefaults' \ ' -append "console=ttyS0 mitigations=off apparmor=0" ' \ - ' -device virtio-net-pci,netdev=s0 ' \ - " -netdev stream,id=s0,server=off,addr.type=unix,addr.path=${STATESETUP}/passt.socket " \ + " ${__qemu_netdev}" \ " -pidfile ${STATESETUP}/qemu.pid" \ " -device vhost-vsock-pci,guest-cid=$GUEST_CID" @@ -142,6 +155,7 @@ setup_passt_in_ns() { [ ${PCAP} -eq 1 ] && __opts="${__opts} -p ${LOGDIR}/passt_in_pasta.pcap" [ ${DEBUG} -eq 1 ] && __opts="${__opts} -d" [ ${TRACE} -eq 1 ] && __opts="${__opts} --trace" + [ ${VHOST_USER} -eq 1 ] && __opts="${__opts} --vhost-user" if [ ${VALGRIND} -eq 1 ]; then context_run passt "make clean" @@ -154,17 +168,30 @@ setup_passt_in_ns() { fi wait_for [ -f "${STATESETUP}/passt.pid" ] + __vmem="$((${MEM_KIB} / 1024 / 4))" + if [ ${VHOST_USER} -eq 1 ]; then + __vmem="$(((${__vmem} + 500) / 1000))G" + __qemu_netdev=" \ + -chardev socket,id=c,path=${STATESETUP}/passt.socket \ + -netdev vhost-user,id=v,chardev=c \ + -device virtio-net,netdev=v \ + -object memory-backend-memfd,id=m,share=on,size=${__vmem} \ + -numa node,memdev=m" + else + __qemu_netdev="-device virtio-net-pci,netdev=s \ + -netdev stream,id=s,server=off,addr.type=unix,addr.path=${STATESETUP}/passt.socket" + fi + GUEST_CID=94557 context_run_bg qemu 'qemu-system-'"${QEMU_ARCH}" \ ' -machine accel=kvm' \ ' -M accel=kvm:tcg' \ - ' -m '${VMEM}' -cpu host -smp '${VCPUS} \ + ' -m '${__vmem}' -cpu host -smp '${VCPUS} \ ' -kernel '"${KERNEL}" \ ' -initrd '${INITRAMFS}' -nographic -serial stdio' \ ' -nodefaults' \ ' -append "console=ttyS0 mitigations=off apparmor=0" ' \ - ' -device virtio-net-pci,netdev=s0 ' \ - " -netdev stream,id=s0,server=off,addr.type=unix,addr.path=${STATESETUP}/passt.socket " \ + " ${__qemu_netdev}" \ " -pidfile ${STATESETUP}/qemu.pid" \ " -device vhost-vsock-pci,guest-cid=$GUEST_CID" @@ -214,6 +241,7 @@ setup_two_guests() { [ ${PCAP} -eq 1 ] && __opts="${__opts} -p ${LOGDIR}/passt_1.pcap" [ ${DEBUG} -eq 1 ] && __opts="${__opts} -d" [ ${TRACE} -eq 1 ] && __opts="${__opts} --trace" + [ ${VHOST_USER} -eq 1 ] && __opts="${__opts} --vhost-user" context_run_bg passt_1 "./passt -s ${STATESETUP}/passt_1.socket -P ${STATESETUP}/passt_1.pid -f ${__opts} -t 10001 -u 10001" wait_for [ -f "${STATESETUP}/passt_1.pid" ] @@ -222,33 +250,54 @@ setup_two_guests() { [ ${PCAP} -eq 1 ] && __opts="${__opts} -p ${LOGDIR}/passt_2.pcap" [ ${DEBUG} -eq 1 ] && __opts="${__opts} -d" [ ${TRACE} -eq 1 ] && __opts="${__opts} --trace" + [ ${VHOST_USER} -eq 1 ] && __opts="${__opts} --vhost-user" context_run_bg passt_2 "./passt -s ${STATESETUP}/passt_2.socket -P ${STATESETUP}/passt_2.pid -f ${__opts} -t 10004 -u 10004" wait_for [ -f "${STATESETUP}/passt_2.pid" ] + __vmem="$((${MEM_KIB} / 1024 / 4))" + if [ ${VHOST_USER} -eq 1 ]; then + __vmem="$(((${__vmem} + 500) / 1000))G" + __qemu_netdev1=" \ + -chardev socket,id=c,path=${STATESETUP}/passt_1.socket \ + -netdev vhost-user,id=v,chardev=c \ + -device virtio-net,netdev=v \ + -object memory-backend-memfd,id=m,share=on,size=${__vmem} \ + -numa node,memdev=m" + __qemu_netdev1=" \ + -chardev socket,id=c,path=${STATESETUP}/passt_2.socket \ + -netdev vhost-user,id=v,chardev=c \ + -device virtio-net,netdev=v \ + -object memory-backend-memfd,id=m,share=on,size=${__vmem} \ + -numa node,memdev=m" + else + __qemu_netdev1="-device virtio-net-pci,netdev=s \ + -netdev stream,id=s,server=off,addr.type=unix,addr.path=${STATESETUP}/passt_1.socket" + __qemu_netdev2="-device virtio-net-pci,netdev=s \ + -netdev stream,id=s,server=off,addr.type=unix,addr.path=${STATESETUP}/passt_2.socket" + fi + GUEST_1_CID=94557 context_run_bg qemu_1 'qemu-system-'"${QEMU_ARCH}" \ ' -M accel=kvm:tcg' \ - ' -m '${VMEM}' -cpu host -smp '${VCPUS} \ + ' -m '${__vmem}' -cpu host -smp '${VCPUS} \ ' -kernel '"${KERNEL}" \ ' -initrd '${INITRAMFS}' -nographic -serial stdio' \ ' -nodefaults' \ ' -append "console=ttyS0 mitigations=off apparmor=0" ' \ - ' -device virtio-net-pci,netdev=s0 ' \ - " -netdev stream,id=s0,server=off,addr.type=unix,addr.path=${STATESETUP}/passt_1.socket " \ + " ${__qemu_netdev1}" \ " -pidfile ${STATESETUP}/qemu_1.pid" \ " -device vhost-vsock-pci,guest-cid=$GUEST_1_CID" GUEST_2_CID=94558 context_run_bg qemu_2 'qemu-system-'"${QEMU_ARCH}" \ ' -M accel=kvm:tcg' \ - ' -m '${VMEM}' -cpu host -smp '${VCPUS} \ + ' -m '${__vmem}' -cpu host -smp '${VCPUS} \ ' -kernel '"${KERNEL}" \ ' -initrd '${INITRAMFS}' -nographic -serial stdio' \ ' -nodefaults' \ ' -append "console=ttyS0 mitigations=off apparmor=0" ' \ - ' -device virtio-net-pci,netdev=s0 ' \ - " -netdev stream,id=s0,server=off,addr.type=unix,addr.path=${STATESETUP}/passt_2.socket " \ + " ${__qemu_netdev2}" \ " -pidfile ${STATESETUP}/qemu_2.pid" \ " -device vhost-vsock-pci,guest-cid=$GUEST_2_CID" diff --git a/test/run b/test/run index 547a729..662a29c 100755 --- a/test/run +++ b/test/run @@ -93,6 +93,7 @@ run() { test memory/passt teardown memory + VHOST_USER=0 setup passt test passt/ndp test passt/dhcp @@ -115,7 +116,18 @@ run() { test two_guests/basic teardown two_guests + VHOST_USER=1 + setup passt_in_ns + test passt_vu/ndp + test passt_vu_in_ns/dhcp + test passt_vu_in_ns/icmp + test passt_vu_in_ns/tcp + test passt_vu_in_ns/udp + test passt_vu_in_ns/shutdown + teardown passt_in_ns + VALGRIND=0 + VHOST_USER=0 setup passt_in_ns test passt/ndp test passt_in_ns/dhcp @@ -126,6 +138,15 @@ run() { test passt_in_ns/shutdown teardown passt_in_ns + VHOST_USER=1 + setup passt_in_ns + test passt_vu/ndp + test passt_vu_in_ns/dhcp + test perf/passt_vu_tcp + test perf/passt_vu_udp + test passt_vu_in_ns/shutdown + teardown passt_in_ns + # TODO: Make those faster by at least pre-installing gcc and make on # non-x86 images, then re-enable. skip_distro() { -- 2.43.0

1 0

[PATCH] test: Kernel binary can now be passed via the KERNEL environmental variable
by Stefano Brivio 30 Sep '24

30 Sep '24

This is quite useful at least for myself as I'm usually running tests using a guest kernel that's not the same as the one on the host. Signed-off-by: Stefano Brivio <sbrivio(a)redhat.com> --- test/lib/setup | 8 ++++---- test/lib/term | 2 +- test/run | 3 +++ 3 files changed, 8 insertions(+), 5 deletions(-) diff --git a/test/lib/setup b/test/lib/setup index d764138..5338393 100755 --- a/test/lib/setup +++ b/test/lib/setup @@ -58,7 +58,7 @@ setup_passt() { context_run_bg qemu 'qemu-system-'"${QEMU_ARCH}" \ ' -machine accel=kvm' \ ' -m '${VMEM}' -cpu host -smp '${VCPUS} \ - ' -kernel ' "/boot/vmlinuz-$(uname -r)" \ + ' -kernel '"${KERNEL}" \ ' -initrd '${INITRAMFS}' -nographic -serial stdio' \ ' -nodefaults' \ ' -append "console=ttyS0 mitigations=off apparmor=0" ' \ @@ -159,7 +159,7 @@ setup_passt_in_ns() { ' -machine accel=kvm' \ ' -M accel=kvm:tcg' \ ' -m '${VMEM}' -cpu host -smp '${VCPUS} \ - ' -kernel ' "/boot/vmlinuz-$(uname -r)" \ + ' -kernel '"${KERNEL}" \ ' -initrd '${INITRAMFS}' -nographic -serial stdio' \ ' -nodefaults' \ ' -append "console=ttyS0 mitigations=off apparmor=0" ' \ @@ -230,7 +230,7 @@ setup_two_guests() { context_run_bg qemu_1 'qemu-system-'"${QEMU_ARCH}" \ ' -M accel=kvm:tcg' \ ' -m '${VMEM}' -cpu host -smp '${VCPUS} \ - ' -kernel ' "/boot/vmlinuz-$(uname -r)" \ + ' -kernel '"${KERNEL}" \ ' -initrd '${INITRAMFS}' -nographic -serial stdio' \ ' -nodefaults' \ ' -append "console=ttyS0 mitigations=off apparmor=0" ' \ @@ -243,7 +243,7 @@ setup_two_guests() { context_run_bg qemu_2 'qemu-system-'"${QEMU_ARCH}" \ ' -M accel=kvm:tcg' \ ' -m '${VMEM}' -cpu host -smp '${VCPUS} \ - ' -kernel ' "/boot/vmlinuz-$(uname -r)" \ + ' -kernel '"${KERNEL}" \ ' -initrd '${INITRAMFS}' -nographic -serial stdio' \ ' -nodefaults' \ ' -append "console=ttyS0 mitigations=off apparmor=0" ' \ diff --git a/test/lib/term b/test/lib/term index 3834092..0fa0936 100755 --- a/test/lib/term +++ b/test/lib/term @@ -664,7 +664,7 @@ pause_continue() { # run_term() - Start tmux session, running entry point, with recording if needed run_term() { - TMUX="tmux new-session -s passt_test -eSTATEBASE=$STATEBASE -ePCAP=$PCAP -eDEBUG=$DEBUG" + TMUX="tmux new-session -s passt_test -eSTATEBASE=$STATEBASE -ePCAP=$PCAP -eDEBUG=$DEBUG -eKERNEL=$KERNEL" if [ ${CI} -eq 1 ]; then printf '\e[8;50;240t' diff --git a/test/run b/test/run index cd6d707..547a729 100755 --- a/test/run +++ b/test/run @@ -38,6 +38,9 @@ TRACE=${TRACE:-0} # If set, tell passt and pasta to take packet captures PCAP=${PCAP:-0} +# Custom kernel to boot guests with, if given +KERNEL=${KERNEL:-"/boot/vmlinuz-$(uname -r)"} + COMMIT="$(git log --oneline --no-decorate -1)" . lib/util -- 2.43.0

2 2

[PATCH v2 0/4] tcp: use csum_iov() in tcp_update_check_tcp[4|6]()
by Laurent Vivier 30 Sep '24

30 Sep '24

For vhost-user, we will need to spread TCP payload over several buffers. To re-use tcp_update_check_tcp[4|6](), provide an iovec rather than a pointer to a buffer. This series updates also csum_iov() and pcap_iov() to add an offset of bytes to skip in the iovec array. It's based on top of "tcp: Use tcp_payload_t rather than tcphdr" that is added in the series for convenience. Laurent Vivier (4): tcp: Use tcp_payload_t rather than tcphdr pcap: Add an offset argument in pcap_iov() checksum: Add an offset argument in csum_iov() tcp: Update TCP checksum using an iovec array checksum.c | 17 +++++-- checksum.h | 3 +- pcap.c | 5 +- pcap.h | 2 +- tcp.c | 134 ++++++++++++++++++++++++++++++++++++------------- tcp_buf.c | 29 ----------- tcp_internal.h | 29 +++++++++++ 7 files changed, 147 insertions(+), 72 deletions(-) -- 2.46.0

3 12

[PATCH v4 0/5] tcp: use csum_iov() in tcp_update_check_tcp[4|6]()
by Laurent Vivier 27 Sep '24

27 Sep '24

For vhost-user, we will need to spread TCP payload over several buffers. To re-use tcp_update_check_tcp[4|6](), provide an iovec rather than a pointer to a buffer. This series updates also csum_iov() and pcap_iov() to add an offset of bytes to skip in the iovec array. It's based on top of "tcp: Use tcp_payload_t rather than tcphdr" that is added in the series for convenience. v3 adds "udp: Update UDP checksum using an iovec array" that modifies csum_udp4() and csum_udp6() in the same way. Laurent Vivier (5): tcp: Use tcp_payload_t rather than tcphdr pcap: Add an offset argument in pcap_iov() checksum: Add an offset argument in csum_iov() tcp: Update TCP checksum using an iovec array udp: Update UDP checksum using an iovec array checksum.c | 46 ++++++++++----- checksum.h | 7 ++- pcap.c | 5 +- pcap.h | 2 +- tap.c | 14 ++++- tap.h | 2 +- tcp.c | 156 +++++++++++++++++++++++++++++++++++++------------ tcp_buf.c | 29 --------- tcp_internal.h | 29 +++++++++ udp.c | 17 ++++-- 10 files changed, 214 insertions(+), 93 deletions(-) -- 2.46.0

2 6

[PATCH v3 0/5] tcp: use csum_iov() in tcp_update_check_tcp[4|6]()
by Laurent Vivier 26 Sep '24

26 Sep '24

For vhost-user, we will need to spread TCP payload over several buffers. To re-use tcp_update_check_tcp[4|6](), provide an iovec rather than a pointer to a buffer. This series updates also csum_iov() and pcap_iov() to add an offset of bytes to skip in the iovec array. It's based on top of "tcp: Use tcp_payload_t rather than tcphdr" that is added in the series for convenience. v3 adds "udp: Update UDP checksum using an iovec array" that modifies csum_udp4() and csum_udp6() in the same way. Laurent Vivier (5): tcp: Use tcp_payload_t rather than tcphdr pcap: Add an offset argument in pcap_iov() checksum: Add an offset argument in csum_iov() tcp: Update TCP checksum using an iovec array udp: Update UDP checksum using an iovec array checksum.c | 46 +++++++++++----- checksum.h | 7 ++- pcap.c | 5 +- pcap.h | 2 +- tap.c | 14 ++++- tap.h | 2 +- tcp.c | 145 +++++++++++++++++++++++++++++++++++++------------ tcp_buf.c | 29 ---------- tcp_internal.h | 29 ++++++++++ udp.c | 17 ++++-- 10 files changed, 203 insertions(+), 93 deletions(-) -- 2.46.0

1 6

Problematic throughput numbers
by Jon Maloy 26 Sep '24

26 Sep '24

I made many runs with iperf3 ns->host, and the results are puzzling me. Over and over again, I see throughput practically collapse, with a two orders of magnitude decrease. Just to make sure this wasn't something introduced by me, I went back to the master branch and disabled the SO_PEEK_OFF feature. The result was the same. The below log shows typical run, but it is highly variable. Sometimes almost the whole series is in the 50-70 Gb/s range, and sometimes almost all in the 100-300 Mb/s range. When I added the kernel fix it didn't seem to make any difference. To me this is really worrying, and should be investigated. ///jon pasta NS->host (master branch, SO_PEEK_OFF disabled ----------------------------------------------------------- Server listening on 5201 (test #2) ----------------------------------------------------------- Accepted connection from 127.0.0.1, port 48354 [ 5] local 127.0.0.1 port 5201 connected to 127.0.0.1 port 48360 [ ID] Interval Transfer Bitrate [ 5] 0.00-1.00 sec 7.61 GBytes 65.3 Gbits/sec [ 5] 1.00-2.00 sec 7.59 GBytes 65.2 Gbits/sec [ 5] 2.00-3.00 sec 7.55 GBytes 64.8 Gbits/sec [ 5] 3.00-4.00 sec 7.69 GBytes 66.1 Gbits/sec [ 5] 4.00-5.00 sec 7.55 GBytes 64.8 Gbits/sec [ 5] 5.00-6.00 sec 7.60 GBytes 65.3 Gbits/sec [ 5] 6.00-7.00 sec 7.52 GBytes 64.6 Gbits/sec [ 5] 7.00-8.00 sec 1.02 GBytes 8.73 Gbits/sec [ 5] 8.00-9.00 sec 3.50 MBytes 29.4 Mbits/sec [ 5] 9.00-10.00 sec 17.2 MBytes 145 Mbits/sec [ 5] 10.00-11.00 sec 64.0 MBytes 537 Mbits/sec [ 5] 11.00-12.00 sec 24.1 MBytes 202 Mbits/sec [ 5] 12.00-13.00 sec 25.4 MBytes 213 Mbits/sec [ 5] 13.00-14.00 sec 24.9 MBytes 209 Mbits/sec [ 5] 14.00-15.00 sec 21.4 MBytes 179 Mbits/sec [ 5] 15.00-16.00 sec 32.0 MBytes 268 Mbits/sec [ 5] 16.00-17.00 sec 10.9 MBytes 91.2 Mbits/sec [ 5] 17.00-18.00 sec 27.6 MBytes 232 Mbits/sec [ 5] 18.00-19.00 sec 75.6 MBytes 634 Mbits/sec [ 5] 19.00-20.00 sec 21.1 MBytes 177 Mbits/sec [ 5] 20.00-21.00 sec 109 MBytes 912 Mbits/sec [ 5] 21.00-22.00 sec 23.9 MBytes 200 Mbits/sec [ 5] 22.00-23.00 sec 62.6 MBytes 525 Mbits/sec [ 5] 23.00-24.00 sec 34.5 MBytes 289 Mbits/sec [ 5] 24.00-25.00 sec 54.4 MBytes 456 Mbits/sec [ 5] 25.00-26.00 sec 14.1 MBytes 118 Mbits/sec [ 5] 26.00-27.00 sec 38.8 MBytes 325 Mbits/sec [ 5] 27.00-28.00 sec 95.0 MBytes 797 Mbits/sec [ 5] 28.00-29.00 sec 23.2 MBytes 195 Mbits/sec [ 5] 29.00-30.00 sec 71.9 MBytes 603 Mbits/sec [ 5] 30.00-31.00 sec 28.8 MBytes 241 Mbits/sec [ 5] 31.00-32.00 sec 34.8 MBytes 292 Mbits/sec [ 5] 32.00-33.00 sec 19.4 MBytes 163 Mbits/sec [ 5] 33.00-34.00 sec 39.1 MBytes 328 Mbits/sec [ 5] 34.00-35.00 sec 31.4 MBytes 263 Mbits/sec [ 5] 35.00-36.00 sec 28.2 MBytes 237 Mbits/sec [ 5] 36.00-37.00 sec 48.5 MBytes 407 Mbits/sec [ 5] 37.00-38.00 sec 23.4 MBytes 196 Mbits/sec [ 5] 38.00-39.00 sec 71.4 MBytes 599 Mbits/sec [ 5] 39.00-40.00 sec 41.4 MBytes 347 Mbits/sec [ 5] 40.00-41.00 sec 15.2 MBytes 128 Mbits/sec [ 5] 41.00-42.00 sec 34.1 MBytes 286 Mbits/sec [ 5] 42.00-43.00 sec 31.1 MBytes 261 Mbits/sec [ 5] 43.00-44.00 sec 61.5 MBytes 516 Mbits/sec [ 5] 44.00-45.00 sec 38.2 MBytes 321 Mbits/sec [ 5] 45.00-46.00 sec 34.2 MBytes 287 Mbits/sec [ 5] 46.00-47.00 sec 33.8 MBytes 283 Mbits/sec [ 5] 47.00-48.00 sec 7.12 MBytes 59.8 Mbits/sec [ 5] 48.00-49.00 sec 97.4 MBytes 817 Mbits/sec [ 5] 49.00-50.00 sec 13.6 MBytes 114 Mbits/sec - - - - - - - - - - - - - - - - - - - - - - - - - [ ID] Interval Transfer Bitrate [ 5] 0.00-50.00 sec 55.7 GBytes 9.57 Gbits/sec receiver ----------------------------------------------------------- Server listening on 5201 (test #3) -----------------------------------------------------------

3 4

[PATCH 0/2] Don't expose container loopback services to the host
by David Gibson 26 Sep '24

26 Sep '24

podman issue #24045 pointed out that pasta's spliced forwarding logic can expose services within the namespace bound only to 127.0.0.1 or ::1 to the host. However, the namespace probably expects those to only be accessible to itself, so that's probably not what we want. This changes our forwarding logic so as not to do this. Note that the podman tests will currently fail with this series, I've submitted podman PR https://github.com/containers/podman/pull/24064 to fix that. Link: https://github.com/containers/podman/issues/24045 David Gibson (2): test: Clarify test for spliced inbound transfers fwd: Direct inbound spliced forwards to the guest's external address fwd.c | 24 +++++++++++++++--------- test/passt_in_ns/tcp | 8 ++++---- test/passt_in_ns/udp | 4 ++-- test/pasta/tcp | 16 ++++++++-------- test/pasta/udp | 8 ++++---- 5 files changed, 33 insertions(+), 27 deletions(-) -- 2.46.1

2 5

[PATCH 0/4] Use inany addresses through port forwarding configuration
by David Gibson 25 Sep '24

25 Sep '24

We use inany addresses, which can represent either an IPv4 or IPv6 address heavily through the flow table. However, for the configuration of port forwarding: both the command line parsing and initialization of listening sockets, we use address family+(void *) pairs instead. Using inany through these paths makes them slightly neater now, but more importantly will make it easier to add more robust tracking of forwarding configuration. David Gibson (4): udp: Don't attempt to get dual-stack sockets in nonsensical cases util, pif: Replace sock_l4() with pif_sock_l4() tcp, udp: Make {tcp,udp}_sock_init() take an inany address inany: Add inany_pton() helper conf.c | 23 ++++++++--------------- inany.c | 20 ++++++++++++++++++++ inany.h | 1 + pif.c | 42 ++++++++++++++++++++++++++++++++++++++++++ pif.h | 3 +++ tcp.c | 45 +++++++++++++++++++++++---------------------- tcp.h | 2 +- udp.c | 50 ++++++++++++++++++++++---------------------------- udp.h | 4 ++-- util.c | 52 ---------------------------------------------------- util.h | 3 --- 11 files changed, 122 insertions(+), 123 deletions(-) -- 2.46.1

2 5