code review 7812052: net: enable IPv6 tests on Windows

net: enable IPv6 tests on Windows

Also removes redundant tests that run Go 1.0 non-IPv6 support
Windows code on IPv6 enabled Windows kernels.

[mikio, 2013-03-27 11:39:25]

Hello alex.brainman@gmail.com, golang-dev@googlegroups.com (cc:
golang-dev@googlegroups.com),

I'd like you to review this change to
https://code.google.com/p/go

[bradfitz, 2013-03-27 16:22:20]

https://codereview.appspot.com/7812052/diff/5001/src/pkg/net/unicast_posix_te...
File src/pkg/net/unicast_posix_test.go (left):

https://codereview.appspot.com/7812052/diff/5001/src/pkg/net/unicast_posix_te...
src/pkg/net/unicast_posix_test.go:104: func TestSimpleTCPListener(t *testing.T)
{
why are these being deleted? does not match CL description.

[mikio, 2013-03-28 02:01:45]

PTAL

https://codereview.appspot.com/7812052/diff/5001/src/pkg/net/unicast_posix_te...
File src/pkg/net/unicast_posix_test.go (left):

https://codereview.appspot.com/7812052/diff/5001/src/pkg/net/unicast_posix_te...
src/pkg/net/unicast_posix_test.go:104: func TestSimpleTCPListener(t *testing.T)
{
On 2013/03/27 16:22:20, bradfitz wrote:
> why are these being deleted? does not match CL description.

updated the CL description.

those two guys are just for running Go 1.0 Windows net code on the IPv6 enabled
kernels. i don't remember why I did make separate test funcs for that purpose.
anyway, we don't need it anymore.

[brainman, 2013-03-28 03:01:40]

LGTM

https://codereview.appspot.com/7812052/diff/10001/src/pkg/net/multicast_test.go
File src/pkg/net/multicast_test.go (right):

https://codereview.appspot.com/7812052/diff/10001/src/pkg/net/multicast_test....
src/pkg/net/multicast_test.go:96: case "plan9", "solaris":
It skips this test anyway:

--- SKIP: TestIPv6MulticastListener (0.00 seconds)
	multicast_test.go:103: skipping test; must be root

[dave_cheney.net, 2013-03-28 03:10:55]

LGTM if you revert the removal of those tests (please see the comment) and
update the description.

I'm uneasy with twiddling with these tests so close to the 1.1 release. If there
is a failure on any platform once you commit this change, I would expect it to
be undone.

https://codereview.appspot.com/7812052/diff/10001/src/pkg/net/multicast_test.go
File src/pkg/net/multicast_test.go (right):

https://codereview.appspot.com/7812052/diff/10001/src/pkg/net/multicast_test....
src/pkg/net/multicast_test.go:96: case "plan9", "solaris":
On 2013/03/28 03:01:40, brainman wrote:
> It skips this test anyway:
> 
> --- SKIP: TestIPv6MulticastListener (0.00 seconds)
> 	multicast_test.go:103: skipping test; must be root

Does that mean if you try this as someone with Administrator local rights, it
will attempt the test ?

https://codereview.appspot.com/7812052/diff/10001/src/pkg/net/unicast_posix_t...
File src/pkg/net/unicast_posix_test.go (left):

https://codereview.appspot.com/7812052/diff/10001/src/pkg/net/unicast_posix_t...
src/pkg/net/unicast_posix_test.go:104: func TestSimpleTCPListener(t *testing.T)
{
If these tests pass, please leave them in place for this proposal. Add a
TODO(mikioh) to remove them later (maybe after go 1.1 ships)

[brainman, 2013-03-28 03:18:00]

On 2013/03/28 03:10:55, dfc wrote:
> On 2013/03/28 03:01:40, brainman wrote:
> > It skips this test anyway:
> > 
> > --- SKIP: TestIPv6MulticastListener (0.00 seconds)
> > 	multicast_test.go:103: skipping test; must be root
> 
> Does that mean if you try this as someone with Administrator local rights, it
> will attempt the test ?
> 

No, it will not. os.Getuid (used for the check) does nothing on Windows. It only
returns -1.

Alex

[dave_cheney.net, 2013-03-28 03:19:09]

> No, it will not. os.Getuid (used for the check) does nothing on Windows.
> It only returns -1.

I did not know that. Thank you.

[mikio, 2013-03-28 04:05:05]

On 2013/03/28 03:19:09, dfc wrote:
> > No, it will not. os.Getuid (used for the check) does nothing on Windows.
> > It only returns -1.
> 
> I did not know that. Thank you.

me too, thanks.

[mikio, 2013-03-28 04:27:57]

On 2013/03/28 03:10:55, dfc wrote:

> I'm uneasy with twiddling with these tests so close to the 1.1 release.
> If there is a failure on any platform once you commit this change,
> I would expect it to be undone.

i failed to grub what you were trying to say.
are you saying that you see any possibility of flaw on this CL?

[mikio, 2013-03-28 04:30:39]

slipped.

On 2013/03/28 04:27:57, mikio wrote:

> i failed to grub what you were trying to say.
> are you saying that you see any possibility of flaw on this CL?

i mean, if so i will postpone this CL for the side bet.

[dave_cheney.net, 2013-03-28 04:33:47]

>> i failed to grub what you were trying to say.
>> are you saying that you see any possibility of flaw on this CL?
>
>
> i mean, if so i will postpone this CL for the side bet.

Hello

To explain myself. I am find with submitting this change, but if it
breaks one of the builders, please roll it back, rather than trying to
correct the breakage.

[mikio, 2013-03-28 04:41:16]

yup, thanks.

On 2013/03/28 04:33:47, dfc wrote:
> >> i failed to grub what you were trying to say.
> >> are you saying that you see any possibility of flaw on this CL?
> >
> >
> > i mean, if so i will postpone this CL for the side bet.
> 
> Hello
> 
> To explain myself. I am find with submitting this change, but if it
> breaks one of the builders, please roll it back, rather than trying to
> correct the breakage.

[mikio, 2013-03-29 02:46:56]

*** Submitted as https://code.google.com/p/go/source/detail?r=47934a0f435e ***

net: enable IPv6 tests on Windows

Also removes redundant tests that run Go 1.0 non-IPv6 support
Windows code on IPv6 enabled Windows kernels.

R=alex.brainman, golang-dev, bradfitz, dave
CC=golang-dev
https://codereview.appspot.com/7812052