syscall: Netlink message captured by syscall.ParseNetlinkMessage always has 0 pid

[lubronzhan]

Go version

go version go1.22.0 darwin/amd64

Output of go env in your module/workspace:

GO111MODULE='on'
GOARCH='amd64'
GOBIN=''
GOCACHE='/Users/lzhan/Library/Caches/go-build'
GOENV='/Users/lzhan/Library/Application Support/go/env'
GOEXE=''
GOEXPERIMENT=''
GOFLAGS=''
GOHOSTARCH='amd64'
GOHOSTOS='darwin'
GOINSECURE=''
GOMODCACHE='/Users/lzhan/work/pkg/mod'
GONOPROXY=''
GONOSUMDB=''
GOOS='linux'
GOPATH='/Users/lzhan/work'
GOPRIVATE=''
GOPROXY='https://proxy.golang.org,direct'
GOROOT='/usr/local/opt/go/libexec'
GOSUMDB='sum.golang.org'
GOTMPDIR=''
GOTOOLCHAIN='auto'
GOTOOLDIR='/usr/local/opt/go/libexec/pkg/tool/darwin_amd64'
GOVCS=''
GOVERSION='go1.22.0'
GCCGO='gccgo'
GOAMD64='v1'
AR='ar'
CC='cc'
CXX='c++'
CGO_ENABLED='1'
GOMOD='/dev/null'
GOWORK=''
CGO_CFLAGS='-O2 -g'
CGO_CPPFLAGS=''
CGO_CXXFLAGS='-O2 -g'
CGO_FFLAGS='-O2 -g'
CGO_LDFLAGS='-O2 -g'
PKG_CONFIG='pkg-config'
GOGCCFLAGS='-fPIC -arch x86_64 -m64 -pthread -fno-caret-diagnostics -Qunused-arguments -fmessage-length=0 -ffile-prefix-map=/var/folders/yz/2cfkf6v56cg_p1srf2dsk6pm0000gp/T/go-build2040811365=/tmp/go-build -gno-record-gcc-switches -fno-common'
 ~/work/src/gitlab.eng.vmware.com/TKG/bolt/yaml   topic/lubron

What did you do?

I followed this example and wrote a simple program to subscribe netlink message using syscall.ParseNetlinkMessage But the pid of the message is always 0

package main

import (
    "fmt"
    "syscall"
)

func main() {
    l, _ := ListenNetlink()

    for {
        msgs, err := l.ReadMsgs()
        if err != nil {
            fmt.Println("Could not read netlink: %s", err)
        }

        for _, m := range msgs {
            fmt.Println(m.Header.Pid)
        }
    }
}

type NetlinkListener struct {
    fd int
    sa *syscall.SockaddrNetlink
}

func ListenNetlink() (*NetlinkListener, error) {
    groups := (1 << (syscall.RTNLGRP_LINK - 1)) |
        (1 << (syscall.RTNLGRP_IPV4_IFADDR - 1)) |
        (1 << (syscall.RTNLGRP_IPV6_IFADDR - 1))

    s, err := syscall.Socket(syscall.AF_NETLINK, syscall.SOCK_DGRAM,
        syscall.NETLINK_ROUTE)
    if err != nil {
        return nil, fmt.Errorf("socket: %s", err)
    }

    saddr := &syscall.SockaddrNetlink{
        Family: syscall.AF_NETLINK,
        Pid:    uint32(0),
        Groups: uint32(groups),
    }

    err = syscall.Bind(s, saddr)
    if err != nil {
        return nil, fmt.Errorf("bind: %s", err)
    }

    return &NetlinkListener{fd: s, sa: saddr}, nil
}

func (l *NetlinkListener) ReadMsgs() ([]syscall.NetlinkMessage, error) {
    defer func() {
        recover()
    }()

    pkt := make([]byte, 2048)

    n, err := syscall.Read(l.fd, pkt)
    if err != nil {
        return nil, fmt.Errorf("read: %s", err)
    }

    msgs, err := syscall.ParseNetlinkMessage(pkt[:n])
    if err != nil {
        return nil, fmt.Errorf("parse: %s", err)
    }

    return msgs, nil
}
}

Then I tried create kubernetes pod on the node, CNI is calico

What did you see happen?

sender PID is always 0

What did you expect to see?

The pid should be the id of process who sent out the request. For example, in my case it should be calico.

If I use rtmon and extract the message using iproute2, I could see correct pid

[ianlancetaylor]

Just checking: you are reporting that you are using darwin/amd64, but SockaddrNetlink is only defined on Linux.

[lubronzhan]

Sorry, I compiled using GOOS=linux and GOARCH=amd64. Let me update the description