Revert the 'gn_migration.gypi' patches.

Revert the 'gn_migration.gypi' patches.

It looks like these patches might be confusing the
analyze step on the ios bots. I don't understand
the errors I'm seeing on them, but they seem to
have started when the first of these patches landed,
so I'm going to try reverting them and see what happens.

This reverts:

  "Rework how the gn migration targets are included from all.gyp."
  commit 91cb1747a84d80acc9fc5045c971986f1799e02b (#318599).

  "Fix a build failing in the x86 linux builds ..."
  commit 425ba073036f96f04aea09c9836826ccb2e124e4. (#318626)

  "Fix nacl-only targets in gn_migration.gypi"
  commit ad333257fe8e7d9b632f728de5fe6db9feb9a64d. (#318631)

TBR=brettw@chromium.org
BUG=461019
NOTRY=true

Committed: https://crrev.com/378e940e74d07896695b6bd1582de760bab7c3b1
Cr-Commit-Position: refs/heads/master@{#318636}

[Dirk Pranke, 2015-03-01 23:32:16]

I don't understand why this set of patches seems to still be causing problems on
the ios_dbg_simulator_ninja bot, but I'll see what happens if I revert them.

http://build.chromium.org/p/tryserver.chromium.mac/builders/ios_dbg_simulator...

The problems seem to start w/ the first patch landing, in 

http://build.chromium.org/p/tryserver.chromium.mac/builders/ios_dbg_simulator...

and continue off and on through the time when even the other patches landed,
i.e., I was still seeing issues as of 

http://build.chromium.org/p/tryserver.chromium.mac/builders/ios_dbg_simulator...

The problem seem to happen when analyze would determine when there is no work to
do, but instead of exiting cleanly, the steps fail; I didn't check every build,
but I saw no builds on the trybot pass that would've done no work (the only
green builds I saw were doing actual compiles).

Supposedly the analyze step is enabled on the bots but ignored, as of:

https://codereview.chromium.org/957493002

I did not examine all of the builds that occurred between that time and my
change to see if any of them should've done no work, but it sure seems likely
that some of them should've returned 'no work to do'.

[Dirk Pranke, 2015-03-01 23:32:26]

The CQ bit was checked by dpranke@chromium.org

[commit-bot: I haz the power, 2015-03-01 23:32:54]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/971633002/1

[Dirk Pranke, 2015-03-01 23:38:04]

Here's a link to the waterfall w/ builds 844 and 843 failing:

http://build.chromium.org/p/tryserver.chromium.mac/waterfall?last_time=142510...

Here's a link to the waterfall from around the time analyze was enabled:

http://build.chromium.org/p/tryserver.chromium.mac/waterfall?last_time=142506...

The first build w/ analyze enabled is build 641:

http://build.chromium.org/p/tryserver.chromium.mac/builders/ios_dbg_simulator...

[Dirk Pranke, 2015-03-02 01:38:55]

The CQ bit was checked by dpranke@chromium.org

[commit-bot: I haz the power, 2015-03-02 01:39:23]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/971633002/20001

[Dirk Pranke, 2015-03-02 01:39:30]

Setting NOTRY=true because I *think* the mac failures are goma-related, and not
due to my patch, but I am admittedly a bit nervous about this.

[commit-bot: I haz the power, 2015-03-02 01:39:45]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2015-03-02 01:40:23]

Patchset 2 (id:??) landed as
https://crrev.com/378e940e74d07896695b6bd1582de760bab7c3b1
Cr-Commit-Position: refs/heads/master@{#318636}