Fix OutOfMemoryDeathTest.ViaSharedLibraries.

Fix OutOfMemoryDeathTest.ViaSharedLibraries.

This test was invalid on modern glibc:
asprintf errors EOVERFLOW without attempting an allocation
when the size is at least INT_MAX - 32. INT_MAX - 33 is the
largest size that can attempt to trigger allocation failure,
and will not necessarily accomplish that on some systems.

Create a library that is specified to wrap malloc trivially,
to accomplish the goal of the test without external dependencies
or inconsistently implemented behavior.

BUG=428211
Committed: https://crrev.com/9a4f22dbb7266733c40a632ece2ddc7eaf56e6ac
Cr-Commit-Position: refs/heads/master@{#324092}

[Ian Coolidge, 2015-04-07 06:42:23]

icoolidge@chromium.org changed reviewers:
+ danakj@chromium.org, mfomitchev@chromium.org, shess@chromium.org

[Ian Coolidge, 2015-04-07 06:43:55]

PTAL. mfomitchev perhaps you can give me some help verifying that I don't break
the iOS device build

[Scott Hess - ex-Googler, 2015-04-07 17:25:47]

LGTM.  I don't see how this could POSSIBLY break iOS!

I see that the previous version succeeded against ios_rel_device_ninja and
ios_dbg_simulator_ninja.  malloc_wrapper built fine, so it's just the signing
which was a problem.  Maybe it was in the wrong section of the gyp file.  You
could make a really boring movie about trying to pick the right section of a gyp
file.

[mfomitchev, 2015-04-07 17:34:52]

On 2015/04/07 06:43:55, Ian Coolidge wrote:
> PTAL. mfomitchev perhaps you can give me some help verifying that I don't
break
> the iOS device build

Unfortunately I can't be of much help with this. Perhaps just submit it and
watch the iOS_Device bot.

[Scott Hess - ex-Googler, 2015-04-07 17:37:58]

On 2015/04/07 17:34:52, mfomitchev wrote:
> On 2015/04/07 06:43:55, Ian Coolidge wrote:
> > PTAL. mfomitchev perhaps you can give me some help verifying that I don't
> break
> > the iOS device build
> 
> Unfortunately I can't be of much help with this. Perhaps just submit it and
> watch the iOS_Device bot.

I think for a case like this it's legit to ping the sheriffs and then use the
waterfall as your trybots.  It's annoying, but I'm not sure what else you can
really do.  I've been working with OSX for a long time but for signing issues
I'd have to just grovel around in the docs for hours, so I'm not sure it'll be
easy to find someone who knows for certain what's up.

[danakj, 2015-04-07 18:02:48]

can you summarize the diff vs the last patch set? (uploading as a new patch set
on the old CL makes this easier to see instead of opening a new CL)

[Ian Coolidge, 2015-04-07 18:18:44]

Uploaded older patch set as 4, and re-uploaded new patch set as 5, so
differences can be viewed side-by-side.

[danakj, 2015-04-07 18:21:37]

LGTM

[Ian Coolidge, 2015-04-07 18:22:13]

The CQ bit was checked by icoolidge@chromium.org

[Ian Coolidge, 2015-04-07 18:22:13]

The patchset sent to the CQ was uploaded after l-g-t-m from shess@chromium.org
Link to the patchset: https://codereview.chromium.org/1064793002/#ps80001
(title: "Fixed CL to avoid target enumeration on unsupported OS")

[commit-bot: I haz the power, 2015-04-07 18:23:14]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1064793002/80001

[commit-bot: I haz the power, 2015-04-07 19:28:22]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2015-04-07 19:30:07]

Patchset 5 (id:??) landed as
https://crrev.com/9a4f22dbb7266733c40a632ece2ddc7eaf56e6ac
Cr-Commit-Position: refs/heads/master@{#324092}