Extract common macros and start a base library

Extract common macros and start a base library

BUG=v8:3015
R=svenpanne@chromium.org
LOG=n

Committed: https://code.google.com/p/v8/source/detail?r=20905

[Sven Panne, 2014-04-23 11:38:34]

LGTM with a nit

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h
File src/base/macros.h (right):

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h#newcode31
src/base/macros.h:31: #include "../../include/v8stdint.h"
Just a note: I think we have to reconsider our include paths a bit when this is
a real library. Relative paths leaving the library won't work then...

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h#newcode50
src/base/macros.h:50: static_cast<size_t>(!(sizeof(a) % sizeof(*(a)))))
This has been here before, but anyway: The divisor seems to be some kind of
safeguard, but when exactly should it trigger? Either add an example or remove
it.

[jochen (gone - plz use gerrit), 2014-04-23 11:50:15]

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h
File src/base/macros.h (right):

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h#newcode31
src/base/macros.h:31: #include "../../include/v8stdint.h"
On 2014/04/23 11:38:35, Sven Panne wrote:
> Just a note: I think we have to reconsider our include paths a bit when this
is
> a real library. Relative paths leaving the library won't work then...

Yes

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h#newcode50
src/base/macros.h:50: static_cast<size_t>(!(sizeof(a) % sizeof(*(a)))))
On 2014/04/23 11:38:35, Sven Panne wrote:
> This has been here before, but anyway: The divisor seems to be some kind of
> safeguard, but when exactly should it trigger? Either add an example or remove
> it.

See
https://code.google.com/p/chromium/codesearch#chromium/src/base/macros.h&l=107

[jochen (gone - plz use gerrit), 2014-04-23 11:51:36]

Committed patchset #1 manually as r20905 (presubmit successful).

[Nico, 2015-06-16 21:02:46]

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

[Nico, 2015-06-16 21:02:47]

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h
File src/base/macros.h (right):

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h#newcode39
src/base/macros.h:39: // Here we simply use the non-zero value 4, which seems to
work.
Is this really still needed? Other places use offsetof() fine, and this comment
has been in globals.h (where you got it from) since "initial import"...

[Sven Panne, 2015-06-17 07:42:14]

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h
File src/base/macros.h (right):

https://codereview.chromium.org/249183003/diff/1/src/base/macros.h#newcode39
src/base/macros.h:39: // Here we simply use the non-zero value 4, which seems to
work.
On 2015/06/16 21:02:46, Nico wrote:
> Is this really still needed? [...]

Alas, yes. :-/ I made a mechanical CL
(https://codereview.chromium.org/1173343006/) which replaces OFFSET_OF with
offsetof where possible, but the rest needs more work. As it is, things are a
bit adventurous, exploring interesting corner cases of the C++ spec... :-P