Set V8_HAS_DECLSPEC_SELECTANY for clang-cl

Set V8_HAS_DECLSPEC_SELECTANY for clang-cl

This macro is used for defining static data members with
STATIC_CONST_MEMBER_DEFINITION. Clang-cl mimics MSVC's
behaviour here, so it also needs __declspec(selectany).

This change was prompted by Clang r237787 which changed
a bug where Clang would previously not emit symbols for
some static data members.

BUG=82385
LOG=N

Committed: https://crrev.com/e514015264f1e0193b57dee76487c85d24f7e4f1
Cr-Commit-Position: refs/heads/master@{#28563}

[hans, 2015-05-20 18:11:49]

hans@chromium.org changed reviewers:
+ bmeurer@chromium.org, rnk@chromium.org

[hans, 2015-05-20 18:11:49]

This should fix link errors such as
http://build.chromium.org/p/chromium.fyi/builders/CrWinClang%28dbg%29/builds/...

[Reid Kleckner, 2015-05-20 18:18:39]

lgtm

[Nico, 2015-05-20 18:40:17]

On 2015/05/20 18:18:39, Reid Kleckner wrote:
> lgtm

Can you check where that explicit clang check for added? Iirc I removed that a
while ago, but maybe I'm misremembering what exactly I did here.

[hans, 2015-05-20 18:49:57]

On 2015/05/20 18:40:17, Nico (afk until Thu May 21) wrote:
> On 2015/05/20 18:18:39, Reid Kleckner wrote:
> > lgtm
> 
> Can you check where that explicit clang check for added? Iirc I removed that a
> while ago, but maybe I'm misremembering what exactly I did here.

Looks like it's been there since https://codereview.chromium.org/23248006, which
if I'm reading it correctly is consolidating a bunch of different config .h
files. The __clang__ line came from src/globals.h, and seems to have been there
for a while.

[Nico, 2015-05-20 19:28:22]

thakis@chromium.org changed reviewers:
+ thakis@chromium.org

[Nico, 2015-05-20 19:28:22]

checked myself now that i'm on a laptop – I didn't change this as much as I
wanted.

In general, I think __clang__ checks should be kept to a minimum and the gcc and
msvc macro checks should instead mostly do the right thing in clang mode too. In
this case:

https://codereview.chromium.org/1145213004/diff/1/include/v8config.h
File include/v8config.h (right):

https://codereview.chromium.org/1145213004/diff/1/include/v8config.h#newcode273
include/v8config.h:273: #elif defined(_MSC_VER)
don't we support all the stuff in here too? Should this be an #if instead of an
#elif?

[hans, 2015-05-20 20:21:15]

New patch uploaded.

https://codereview.chromium.org/1145213004/diff/1/include/v8config.h
File include/v8config.h (right):

https://codereview.chromium.org/1145213004/diff/1/include/v8config.h#newcode273
include/v8config.h:273: #elif defined(_MSC_VER)
On 2015/05/20 19:28:22, Nico (afk until Thu May 21) wrote:
> don't we support all the stuff in here too? Should this be an #if instead of
an
> #elif?

Yeah, that would work.

[Nico, 2015-05-20 23:56:00]

On 2015/05/20 20:21:15, hans wrote:
> New patch uploaded.
> 
> https://codereview.chromium.org/1145213004/diff/1/include/v8config.h
> File include/v8config.h (right):
> 
>
https://codereview.chromium.org/1145213004/diff/1/include/v8config.h#newcode273
> include/v8config.h:273: #elif defined(_MSC_VER)
> On 2015/05/20 19:28:22, Nico (afk until Thu May 21) wrote:
> > don't we support all the stuff in here too? Should this be an #if instead of
> an
> > #elif?
> 
> Yeah, that would work.

Lgtm if it builds, thanks!

[Benedikt Meurer, 2015-05-21 04:43:49]

LGTM with nit.

https://codereview.chromium.org/1145213004/diff/20001/include/v8config.h
File include/v8config.h (right):

https://codereview.chromium.org/1145213004/diff/20001/include/v8config.h#newc...
include/v8config.h:281: #endif
Nit: keep the empty line before the #endif.

[hans, 2015-05-21 15:25:45]

https://codereview.chromium.org/1145213004/diff/20001/include/v8config.h
File include/v8config.h (right):

https://codereview.chromium.org/1145213004/diff/20001/include/v8config.h#newc...
include/v8config.h:281: #endif
On 2015/05/21 04:43:49, Benedikt Meurer wrote:
> Nit: keep the empty line before the #endif.

Done.

[hans, 2015-05-21 15:25:49]

The CQ bit was checked by hans@chromium.org

[hans, 2015-05-21 15:25:51]

The patchset sent to the CQ was uploaded after l-g-t-m from rnk@chromium.org,
bmeurer@chromium.org, thakis@chromium.org
Link to the patchset: https://codereview.chromium.org/1145213004/#ps40001
(title: "Fix newline")

[commit-bot: I haz the power, 2015-05-21 15:26:06]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1145213004/40001

[commit-bot: I haz the power, 2015-05-21 15:30:21]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2015-05-21 15:30:22]

Try jobs failed on following builders:
  v8_presubmit on tryserver.v8 (JOB_FAILED,
http://build.chromium.org/p/tryserver.v8/builders/v8_presubmit/builds/2874)

[hans, 2015-05-21 15:32:44]

The CQ bit was checked by hans@chromium.org

[commit-bot: I haz the power, 2015-05-21 15:33:10]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1145213004/40001

[commit-bot: I haz the power, 2015-05-21 16:21:34]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-05-21 16:21:48]

Patchset 3 (id:??) landed as
https://crrev.com/e514015264f1e0193b57dee76487c85d24f7e4f1
Cr-Commit-Position: refs/heads/master@{#28563}