net/http: document that Server.Serve expects a hashable net.Listener (as a map key)

[jdef]

What version of Go are you using (go version)?

1.10

Does this issue reproduce with the latest release?

yes

What operating system and processor architecture are you using (go env)?

GOARCH="amd64"
GOBIN=""
GOCACHE="/home/vagrant/.cache/go-build"
GOEXE=""
GOHOSTARCH="amd64"
GOHOSTOS="linux"
GOOS="linux"
GOPATH="/home/vagrant/workspace-dss"
GORACE=""
GOROOT="/usr/local/go-1.10"
GOTMPDIR=""
GOTOOLDIR="/usr/local/go-1.10/pkg/tool/linux_amd64"
GCCGO="gccgo"
CC="gcc"
CXX="g++"
CGO_ENABLED="1"
CGO_CFLAGS="-g -O2"
CGO_CPPFLAGS=""
CGO_CXXFLAGS="-g -O2"
CGO_FFLAGS="-g -O2"
CGO_LDFLAGS="-g -O2"
PKG_CONFIG="pkg-config"
GOGCCFLAGS="-fPIC -m64 -pthread -fmessage-length=0 -fdebug-prefix-map=/tmp/go-build230598238=/tmp/go-build -gno-record-gcc-switches"

What did you do?

I implemented net.Listener with a struct that defined a func() field. I passed this struct to http.Server.Serve. The code compiled just fine and panicked at run-time because my listener implementation couldn't be used as a hash key (see trackListener in the http package).

What did you expect to see?

A working http.Server. Or else some documentation that clearly states the constraints for the kinds of listeners usable with http.Server.

What did you see instead?

Panic at run-time.

[meirf]

@jdef Thanks for your report. This is probably enough information to reproduce your issue, but it would be helpful if you had a snippet in the form of a https://play.golang.org link to save us time in reproducing and in guaranteeing we fix the exact issue.

[artyom]

Duplicate of #24812 ?

[jdef]

Yes, it's a duplicate. Thanks

…

On Tue, May 22, 2018, 6:29 AM Artyom Pervukhin ***@***.***> wrote: Duplicate of #24812 <#24812> ? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#25487 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ACPVLHUd9DtcI-g9BiIYzqQ25J6hO-qTks5t1BKjgaJpZM4UH8IO> .

[AlekSi]

@jdef Can you please close it then?

[meirf]

If this has already happened maybe this should stay open as a documentation issue so we don’t get more issues like this.

…

On Tue, May 22, 2018 at 11:47 AM Alexey Palazhchenko < ***@***.***> wrote: @jdef <https://github.com/jdef> Can you please close it then? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#25487 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AB8tqQa_F2sSJ9WUnKQjRCYVEZyeLjcaks5t1DL8gaJpZM4UH8IO> .

[odeke-em]

Great point @meirf! Might you or anyone else here be interested in sending a CL?

[artyom]

@meirf, @odeke-em looks like the linked issue already has this fixed, so we don’t need to change documentation.

[meirf]

@artyom Yes, looks like a fix for this was merged as shown in the linked issue. Sorry for misleading.

@jdef A fix for this has been merged, though not yet in an official release.

@odeke-em please close when you have a chance.