x/net/ipv4: multicast ICMP tests fail when net.inet.icmp.bmcastecho=0

[mikioh]

--- FAIL: TestReadWriteMulticastIPPayloadICMP-2 (0.00 seconds)
    mocktransponder_test.go:42: rcvd cmsg: ttl: 1, src: 127.0.0.1, dst: 224.0.0.254, ifindex: 7
    multicast_test.go:106: got type=echo, code=0; expected type=echo reply, code=0
--- FAIL: TestReadWriteMulticastIPDatagram-2 (0.00 seconds)
    mocktransponder_test.go:73: rcvd cmsg: ttl: 1, src: 127.0.0.1, dst: 224.0.0.254, ifindex: 7
    mocktransponder_test.go:74: rcvd hdr: ver: 4, hdrlen: 20, tos: 0x1, totallen: 43, id: 0x5b5a, fragoff: 0x0, ttl: 1, proto: 1, cksum: 0xfe77, src: 127.0.0.1, dst: 224.0.0.254
    multicast_test.go:165: got type=echo, code=0; expected type=echo reply, code=0
FAIL

[davecheney]

Comment 1:

Which host/os, etc ?

[mikioh]

Comment 2:

FreeBSD 9.x.
Just indicates loopback'd sent packets but got no replies from protocol stack inside the
kernel.
Hm, loopback marker's bug maybe.

[mikioh]

Comment 3:

Nope, I completely forgot that old RFC says "An ICMP Echo Request destined to an IP
broadcast or IP multicast address MAY be silently discarded" and almost all
implementations follow that direction. Will fix the test case.

[mikioh]

Comment 4:

Of course, net.inet.icmp.bmcastech=0 by default on BSD bros.

[gopherbot]

Comment 5 by atkaaz:

I have only one related fail but on Windows 7 64bit, possibly because I've disabled some
things in this system (but can't remember what and where but ie. maybe multicast is
disabled, but it's not because the firewall is silently blocking it(without logging)
because I've disabled the firewall and still happens)
ok      mime/multipart  0.863s
--- FAIL: TestIPv4MulticastListener (0.02 seconds)
        multicast_test.go:53: First ListenMulticastUDP on <nil> failed: listen udp 224.0.0.254: set
sockopt: An unknown, invalid, or unsupported option or level was specified in a
getsockopt or setso
ckopt call.
FAIL
FAIL    net     3.160s
ok      net/http        5.568s
this was the go version used in the test:
http://code.google.com/p/go/source/detail?r=749d021640436ab33eb8dfdb25bbec472e80af73
go version devel +749d02164043 Thu Aug 15 14:38:32 2013 -0400 windows/amd64

[mikioh]

Comment 6:

Sorry, but that's net package's fault. Please open new issue because we focus on the
issue of code.google.com/p/.go.net/ipv4 package here.

[gopherbot]

Comment 7 by atkaaz:

Ok sorry, I thought since it's the same multicast_test.go file(but different line), this
would be the place. I'll open a new issue, thanks.

[mikioh]

Comment 8:

Labels changed: added priority-soon, removed priority-triage.

Owner changed to @mikioh.

[rsc]

Comment 9:

Labels changed: added go1.3maybe.

[rsc]

Comment 10:

Labels changed: added release-none, removed go1.3maybe.

[rsc]

Comment 11:

Labels changed: added repo-net.

[mikioh]

Comment 12:

https://golang.org/cl/21360043/

[gopherbot]

Comment 13:

CL https://golang.org/cl/21360043 references this issue.

[mikioh]

Comment 14:

This issue was closed by revision golang/net@8ecd624.

Status changed to Fixed.