Add ragel validator to SCONS

Add ragel decoder/validator to SCONS

BUG=http://code.google.com/p/nativeclient/issues/detail?id=2597
TEST=scons decoder-test validator-test ; ls -al scons-out/*/staging

See also:
https://chromiumcodereview.appspot.com/9968039/
https://chromiumcodereview.appspot.com/9967015/
Committed: https://src.chromium.org/viewvc/native_client?view=rev&revision=8225

[khim, 2012-04-11 09:22:50]

This time all trybots passed on first try except two clearly misconfigured ones.
nacl-lucid_64-newlib-x86_32-pnacl-spec and nacl-lucid64_validator_opt have few
bad slaves thus only passed after few tries.

[pasko-google - do not use, 2012-04-11 09:37:59]

The commit message must be as the original one. More details can be appended,
such as, that it is try No. 3 and explanation for intentionally not including
the generated files. Also a link to previous reviews would be helpful. All the
usual stuff should be there, i.e. BUG= and TEST=.

[khim, 2012-04-11 09:44:08]

PTAL

[pasko-google - do not use, 2012-04-11 09:46:13]

https://chromiumcodereview.appspot.com/10031039/diff/12001/src/trusted/valida...
File src/trusted/validator_ragel/build.scons (right):

https://chromiumcodereview.appspot.com/10031039/diff/12001/src/trusted/valida...
src/trusted/validator_ragel/build.scons:22: if env.Bit('mac') or
env.Bit('asan'):
What is happening here? Why the ... are they using -Os and why does it fail?
Please, provide a comment.

[khim, 2012-04-11 09:47:34]

Comment in included in referenced bug.

[pasko-google - do not use, 2012-04-11 09:48:51]

On 2012/04/11 09:47:34, khim wrote:
> Comment in included in referenced bug.

Then reference the bug in the comment.

[khim, 2012-04-11 09:53:34]

The reference is there on Line 19.

[pasko-google - do not use, 2012-04-11 09:57:08]

On 2012/04/11 09:53:34, khim wrote:
> The reference is there on Line 19.

On that issue one has to scroll to a 3rd comment, and then the -Os is not
mentioned. Very informative.

[khim, 2012-04-11 10:00:50]

So what do you suggest?

[pasko-google - do not use, 2012-04-11 10:15:20]

On 2012/04/11 10:00:50, khim wrote:
> So what do you suggest?

Add to the comment on line 19:
XCode 4.2 takes too long to compile with -Os, LLVM for MacOS times out.

... and your comments do not cover the case with ASAN on non-MacOS systems, why
should we filter out there?

[khim, 2012-04-11 10:18:50]

Not sure why XCode comment should go here while MSVC comment is in the bug, but
whatever.

ASAN was broken in previous commit similarly to XCode 4.2, we don't have an asan
trybot and since ASAN uses LLVM as well I think the problem is likely similar
and should be fixed with the same change.

[pasko-google - do not use, 2012-04-11 10:34:21]

lgtm