Add the MB config of the official win Clang bots.

Add the MB config of the official win Clang bots.

These bots don't yet exist, I'm creating them in a separate CL.

BUG=620849
Committed: https://crrev.com/7ed420a42d6ea759e585510605888570caeb0677
Cr-Commit-Position: refs/heads/master@{#411130}

[Sébastien Marchand, 2016-08-10 15:23:24]

Description was changed from

==========
Add the MB config of the official win Clang bots.

BUG=620849
==========

to

==========
Add the MB config of the official win Clang bots.

These bots don't yet exist, I'm creating them in a separate CL.

BUG=620849
==========

[Sébastien Marchand, 2016-08-10 15:23:24]

sebmarchand@chromium.org changed reviewers:
+ dpranke@chromium.org

[Sébastien Marchand, 2016-08-10 15:24:12]

sebmarchand@chromium.org changed reviewers:
+ hans@chromium.org

[Sébastien Marchand, 2016-08-10 15:24:13]

PTAL.

[hans, 2016-08-10 16:04:13]

lgtm

[Dirk Pranke, 2016-08-10 16:50:33]

lgtm, but we should be adding official.desktop.continuous builders, too, I
think?

[Sébastien Marchand, 2016-08-10 17:16:57]

There's already an continuous Official Chrome Clang builder on chromium.fyi, so
the one on official.desktop.continuous would probably be a copy of this one, not
sure if it's worth it (but we could probably move the existing one out of fyi).

We spoke about this with kerz@, the plan was to setup the official.desktop bot
and then see if we also want one on official.desktop.continuous (but the general
feeling was that we don't need it).

[Sébastien Marchand, 2016-08-10 17:17:02]

The CQ bit was checked by sebmarchand@chromium.org

[commit-bot: I haz the power, 2016-08-10 17:18:07]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[Dirk Pranke, 2016-08-10 18:28:06]

On 2016/08/10 17:16:57, Sébastien Marchand wrote:
> There's already an continuous Official Chrome Clang builder on chromium.fyi,
so
> the one on official.desktop.continuous would probably be a copy of this one,
not
> sure if it's worth it (but we could probably move the existing one out of
fyi).

We don't want the one on chromium.fyi to have a checkout of src-internal, so if
you
want to do a real build, it should move. In addition, I think any configuration
that
is on the official builders should have matches on some non-fyi public
waterfall,
so that they are monitored more actively.

In further addition -- :) -- I think that *any* bot that lives on
official.desktop should 
have a continuous counterpart. The problem with not doing that is that it's too
easy 
to miss breakages when they happen, and only find out about them when an 
official build is triggered, which is not when you want to find out :).

A corollary of this may be that things like win syzyasan -- which isn't shipped
anyway
-- shouldn't really be on official.desktop, and we should move it.

> We spoke about this with kerz@, the plan was to setup the official.desktop bot
> and then see if we also want one on official.desktop.continuous (but the
general
> feeling was that we don't need it).

I disagree :). I'm happy to discuss this w/ kerz@ if need be.

[Sébastien Marchand, 2016-08-10 18:36:37]

On 2016/08/10 18:28:06, Dirk Pranke wrote:
> On 2016/08/10 17:16:57, Sébastien Marchand wrote:
> > There's already an continuous Official Chrome Clang builder on chromium.fyi,
> so
> > the one on official.desktop.continuous would probably be a copy of this one,
> not
> > sure if it's worth it (but we could probably move the existing one out of
> fyi).
> 
> We don't want the one on chromium.fyi to have a checkout of src-internal, so
if
> you
> want to do a real build, it should move. In addition, I think any
configuration
> that
> is on the official builders should have matches on some non-fyi public
> waterfall,
> so that they are monitored more actively.
> 
> In further addition -- :) -- I think that *any* bot that lives on
> official.desktop should 
> have a continuous counterpart. The problem with not doing that is that it's
too
> easy 
> to miss breakages when they happen, and only find out about them when an 
> official build is triggered, which is not when you want to find out :).
> 
> A corollary of this may be that things like win syzyasan -- which isn't
shipped
> anyway
> -- shouldn't really be on official.desktop, and we should move it.

We ship the win-asan build to 5% of the Canary population every day :).

> 
> > We spoke about this with kerz@, the plan was to setup the official.desktop
bot
> > and then see if we also want one on official.desktop.continuous (but the
> general
> > feeling was that we don't need it).
> 
> I disagree :). I'm happy to discuss this w/ kerz@ if need be.

Yeah, I also agree that if we start shipping this then we'll want to add its
continuous
counterpart, for now the plan is to not automatically ship these builds (but
people
would be able to opt-in to receive them), we'll also tell the TE team to ignore
these bots,
the Lexan team is responsible of checking them.

Ideally they should live on a desktop.official.fyi waterfall, but this doesn't
exist and
last time I've checked there was no real desire to create it.

To be clear, I really don't mind adding the continuous counterpart of these bots
if we
have enough machines for this, but I'm not sure that we do ?

[hans, 2016-08-10 18:43:26]

On 2016/08/10 18:28:06, Dirk Pranke wrote:
> On 2016/08/10 17:16:57, Sébastien Marchand wrote:
> > We spoke about this with kerz@, the plan was to setup the official.desktop
bot
> > and then see if we also want one on official.desktop.continuous (but the
> general
> > feeling was that we don't need it).
> 
> I disagree :). I'm happy to discuss this w/ kerz@ if need be.

Our argument was that we already have continuous official Windows clang build-
and test bots, and adding exactly the same configs to
official.desktop.continuous seems like a lot of duplicated bots.

[Dirk Pranke, 2016-08-10 18:52:24]

On 2016/08/10 18:36:37, Sébastien Marchand wrote:
> We ship the win-asan build to 5% of the Canary population every day :).

I did not know that! Never mind, then ...

> > I disagree :). I'm happy to discuss this w/ kerz@ if need be.
> 
> Yeah, I also agree that if we start shipping this then we'll want to add its
> continuous
> counterpart, for now the plan is to not automatically ship these builds (but
> people
> would be able to opt-in to receive them), we'll also tell the TE team to
ignore
> these bots,
> the Lexan team is responsible of checking them.
> 
> Ideally they should live on a desktop.official.fyi waterfall, but this doesn't
> exist and
> last time I've checked there was no real desire to create it.

Hrm. We should discuss further, but we don't need to block this CL on this.

> To be clear, I really don't mind adding the continuous counterpart of these
bots
> if we
> have enough machines for this, but I'm not sure that we do ?

I don't believe that's a problem.

[Dirk Pranke, 2016-08-10 18:53:18]

On 2016/08/10 18:43:26, hans wrote:
> Our argument was that we already have continuous official Windows clang build-
> and test bots, and adding exactly the same configs to
> official.desktop.continuous seems like a lot of duplicated bots.

See the parts above about src-internal. Either we change the fyi configs (so
that
they would no longer be duplicating things), or they need to move.

[commit-bot: I haz the power, 2016-08-10 20:16:40]

Description was changed from

==========
Add the MB config of the official win Clang bots.

These bots don't yet exist, I'm creating them in a separate CL.

BUG=620849
==========

to

==========
Add the MB config of the official win Clang bots.

These bots don't yet exist, I'm creating them in a separate CL.

BUG=620849
==========

[commit-bot: I haz the power, 2016-08-10 20:16:42]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-08-10 20:18:48]

Description was changed from

==========
Add the MB config of the official win Clang bots.

These bots don't yet exist, I'm creating them in a separate CL.

BUG=620849
==========

to

==========
Add the MB config of the official win Clang bots.

These bots don't yet exist, I'm creating them in a separate CL.

BUG=620849

Committed: https://crrev.com/7ed420a42d6ea759e585510605888570caeb0677
Cr-Commit-Position: refs/heads/master@{#411130}
==========

[commit-bot: I haz the power, 2016-08-10 20:18:50]

Patchset 1 (id:??) landed as
https://crrev.com/7ed420a42d6ea759e585510605888570caeb0677
Cr-Commit-Position: refs/heads/master@{#411130}