Re-enable as many clang warning as possible

Re-enable as many clang warning as possible

ICU has historically had many clang warnings disabled, but some of the
underlying causes have been fixed. One of the VC++ specific warning
disables was recently removed (crrev.com/2494793002) and this change
removes a bunch more, for clang and VC++. For some of them there are
specific icu bugs that have been fixed that allow this, and for others
I just have the evidence that on Linux and Windows icu compiles with
clang and VC++ with these changes.

BUG=21515
R=thakis@chromium.org

Committed: https://chromium.googlesource.com/chromium/deps/icu/+/3c89207bdd2c44b78ec69b7d52f9768475835bd9

[brucedawson, 2016-11-14 19:05:21]

Description was changed from

==========
Re-enable as many clang warning as possible

ICU has historically had many clang warnings disabled, but some of the
underlying causes have been fixed.

BUG=21515
==========

to

==========
Re-enable as many clang warning as possible

ICU has historically had many clang warnings disabled, but some of the
underlying causes have been fixed. One of the VC++ specific warning
disables was recently removed (crrev.com/2494793002) and this change
removes a bunch more, for clang and VC++. For some of them there are
specific icu bugs that have been fixed that allow this, and for others
I just have the evidence that on Linux and Windows icu compiles with
clang and VC++ with these changes.

BUG=21515
==========

[brucedawson, 2016-11-14 19:05:21]

brucedawson@chromium.org changed reviewers:
+ jshin@chromium.org

[brucedawson, 2016-11-14 19:06:12]

It looks like we can remove quite a few warning-disable options for icu. PTAL
jungshik.

[Nico, 2016-11-14 19:10:48]

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

[Nico, 2016-11-14 19:10:48]

Lgtm if it builds :-)

[brucedawson, 2016-11-14 19:16:13]

On 2016/11/14 19:10:48, Nico wrote:
> Lgtm if it builds :-)

I've tried debug/release x component/non-component on Linux/Windows using clang
- all of them work.

[brucedawson, 2016-11-14 19:33:17]

Description was changed from

==========
Re-enable as many clang warning as possible

ICU has historically had many clang warnings disabled, but some of the
underlying causes have been fixed. One of the VC++ specific warning
disables was recently removed (crrev.com/2494793002) and this change
removes a bunch more, for clang and VC++. For some of them there are
specific icu bugs that have been fixed that allow this, and for others
I just have the evidence that on Linux and Windows icu compiles with
clang and VC++ with these changes.

BUG=21515
==========

to

==========
Re-enable as many clang warning as possible

ICU has historically had many clang warnings disabled, but some of the
underlying causes have been fixed. One of the VC++ specific warning
disables was recently removed (crrev.com/2494793002) and this change
removes a bunch more, for clang and VC++. For some of them there are
specific icu bugs that have been fixed that allow this, and for others
I just have the evidence that on Linux and Windows icu compiles with
clang and VC++ with these changes.

BUG=21515
R=thakis@chromium.org

Committed:
https://chromium.googlesource.com/chromium/deps/icu/+/3c89207bdd2c44b78ec69b7...
==========

[brucedawson, 2016-11-14 19:33:19]

Committed patchset #3 (id:40001) manually as
3c89207bdd2c44b78ec69b7d52f9768475835bd9 (presubmit successful).

[jungshik at Google, 2016-11-15 18:06:36]

Thank you for removing warnings I forgot to remove updating ICU. 

Below is one post-landing comment.

https://codereview.chromium.org/2492553006/diff/40001/BUILD.gn
File BUILD.gn (left):

https://codereview.chromium.org/2492553006/diff/40001/BUILD.gn#oldcode77
BUILD.gn:77: # http://bugs.icu-project.org/trac/ticket/12821
Did you mean to remove this comment? It's for wd4333 we still have to keep.

[brucedawson, 2016-11-15 19:53:48]

https://codereview.chromium.org/2492553006/diff/40001/BUILD.gn
File BUILD.gn (left):

https://codereview.chromium.org/2492553006/diff/40001/BUILD.gn#oldcode77
BUILD.gn:77: # http://bugs.icu-project.org/trac/ticket/12821
On 2016/11/15 18:06:36, jungshik at google wrote:
> Did you mean to remove this comment? It's for wd4333 we still have to keep. 

Good catch. I accidentally associated the comment with the previous line instead
of with the subsequent line. I'll put it back.