Specify pnacl_newlib_dir to SCons when running nacl_integration

Specify pnacl_newlib_dir to SCons when running nacl_integration

The PNaCl toolchain will soon be required by SCons to build the IRT. Even
though nacl_integration doesn't use the IRT built by SCons, it still wants to
set up the environment as part of the SCons global initialization, and that
fails if the compiler is not found.

Also remove the extra download, since the toolchain in the native_client directory (which is already synced by gclient) is now used.

R=ncbray@chromium.org
BUG= https://code.google.com/p/nativeclient/issues/detail?id=4088

Committed: https://crrev.com/a8469893a29933a288bccd858ebf0e080dbb9585
Cr-Commit-Position: refs/heads/master@{#321286}

[Nick Bray (chromium), 2015-03-18 20:23:35]

What bots don't have this toolchain already?  (Naively, they should all get it
via gclient sync.)  Is this an actual fix, or preventative?

Also, beware of --optional-pnacl in build/download_nacl_toolchains.py (for the
gyp build).  I haven't had time to track down if that's still used anywhere.

[Derek Schuff, 2015-03-18 23:29:19]

PTAL. i removed the download entirely and fixed the actual problem.

[Nick Bray (chromium), 2015-03-18 23:34:12]

https://codereview.chromium.org/1021563002/diff/20001/chrome/test/nacl_test_i...
File chrome/test/nacl_test_injection/buildbot_chrome_nacl_stage.py (right):

https://codereview.chromium.org/1021563002/diff/20001/chrome/test/nacl_test_i...
chrome/test/nacl_test_injection/buildbot_chrome_nacl_stage.py:117:
pnacl_newlib_dir = os.path.join(toolchain_dir, 'pnacl_newlib' % arch_name)
... I suspect there's a bug here that should cause an exception.  (not all
arguments converted during string formatting)

[Derek Schuff, 2015-03-18 23:40:18]

https://codereview.chromium.org/1021563002/diff/20001/chrome/test/nacl_test_i...
File chrome/test/nacl_test_injection/buildbot_chrome_nacl_stage.py (right):

https://codereview.chromium.org/1021563002/diff/20001/chrome/test/nacl_test_i...
chrome/test/nacl_test_injection/buildbot_chrome_nacl_stage.py:117:
pnacl_newlib_dir = os.path.join(toolchain_dir, 'pnacl_newlib' % arch_name)
On 2015/03/18 23:34:12, Nick Bray (chromium) wrote:
> ... I suspect there's a bug here that should cause an exception.  (not all
> arguments converted during string formatting)

oops, you're right.

[Nick Bray (chromium), 2015-03-19 00:19:02]

LGTM-ish, but the fact patch #2 passed the trybots makes me think the trybots
aren't actually testing this patch.  Bounce it off linux_rel or something?

[Derek Schuff, 2015-03-19 00:36:25]

On 2015/03/19 00:19:02, Nick Bray (chromium) wrote:
> LGTM-ish, but the fact patch #2 passed the trybots makes me think the trybots
> aren't actually testing this patch.  Bounce it off linux_rel or something?

yeah it's definitely not testing it. i did test it locally too.

[Derek Schuff, 2015-03-19 03:57:54]

The CQ bit was checked by dschuff@chromium.org

[commit-bot: I haz the power, 2015-03-19 03:58:12]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1021563002/40001

[commit-bot: I haz the power, 2015-03-19 04:01:32]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-03-19 04:02:19]

Patchset 3 (id:??) landed as
https://crrev.com/a8469893a29933a288bccd858ebf0e080dbb9585
Cr-Commit-Position: refs/heads/master@{#321286}