Gyp changes for ninja iOS build.

Gyp changes for ninja iOS build.

Various tweaks to make ninja happy with our gyp files.
ninja iOS builds don't CODE_SIGN yet.
ninja only supports i386/simulator for now.
pass parent_generator to mac_build.gypi for iOS sub-ninja builds

BUG=236517
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=198084

[justincohen, 2013-05-01 13:24:39]

All of these changes fall within ios blocks in gyp, except for landmines.py
which just acknowledges ninja-ios as a GENERATOR

Thanks!

[stuartmorgan, 2013-05-01 14:03:42]

I think I'm just missing background; why is ninja-ios a separate thing?

https://codereview.chromium.org/14577007/diff/5001/breakpad/breakpad.gyp
File breakpad/breakpad.gyp (right):

https://codereview.chromium.org/14577007/diff/5001/breakpad/breakpad.gyp#newc...
breakpad/breakpad.gyp:700: 'outputs': ['compile_breakpad_utilities'],
No outputs should also force it every time. What problem are you solving here?

https://codereview.chromium.org/14577007/diff/5001/build/ios/mac_build.gypi
File build/ios/mac_build.gypi (right):

https://codereview.chromium.org/14577007/diff/5001/build/ios/mac_build.gypi#n...
build/ios/mac_build.gypi:43: 'parent_generator%': 0,
0?

[justincohen, 2013-05-01 14:08:57]

ninja-ios was the recommendation from thakis@.  -ios defines the flavor, so the
ninja knows how to do things like build the app bundle for ios instead of from
mac.  This was cleaner than looking at the GYP_DEFINES for an OS value.

https://codereview.chromium.org/14577007/diff/5001/breakpad/breakpad.gyp
File breakpad/breakpad.gyp (right):

https://codereview.chromium.org/14577007/diff/5001/breakpad/breakpad.gyp#newc...
breakpad/breakpad.gyp:700: 'outputs': ['compile_breakpad_utilities'],
ninja doesn't like no outputs.

On 2013/05/01 14:03:42, stuartmorgan wrote:
> No outputs should also force it every time. What problem are you solving here?

https://codereview.chromium.org/14577007/diff/5001/build/ios/mac_build.gypi
File build/ios/mac_build.gypi (right):

https://codereview.chromium.org/14577007/diff/5001/build/ios/mac_build.gypi#n...
build/ios/mac_build.gypi:43: 'parent_generator%': 0,
It's just the default value.  It shouldn't matter what it is, since if it's the
default value, the parent gyp isn't using it.  Otherwise the compile is going to
fail regardless.

On 2013/05/01 14:03:42, stuartmorgan wrote:
> 0?

[Mark Mentovai, 2013-05-01 18:12:10]

Perhaps ninja should be made to understand what an empty 'outputs' list means
instead of some of these changes.

https://codereview.chromium.org/14577007/diff/5001/breakpad/breakpad.gyp
File breakpad/breakpad.gyp (right):

https://codereview.chromium.org/14577007/diff/5001/breakpad/breakpad.gyp#newc...
breakpad/breakpad.gyp:699: # A nonexistant name forces the ninja cmd every time.
spelling: nonexistent

spelling: command, not cmd (don’t abbreviate, not everyone’s a native English
speaker that will understand this shorthand).

meaning: I see through the spelling mistake and am a native English speaker and
I still don’t know what “forces the ninja command” is supposed to mean. Is this
even ninja-specific?

https://codereview.chromium.org/14577007/diff/5001/build/common.gypi
File build/common.gypi (right):

https://codereview.chromium.org/14577007/diff/5001/build/common.gypi#newcode4352
build/common.gypi:4352: 'SDKROOT': 'iphonesimulator<(ios_sdk)',  # -isysroot
Where do you normally set SDKROOT to the simulator SDK for simulator builds?

https://codereview.chromium.org/14577007/diff/5001/testing/iossim/iossim.gyp
File testing/iossim/iossim.gyp (right):

https://codereview.chromium.org/14577007/diff/5001/testing/iossim/iossim.gyp#...
testing/iossim/iossim.gyp:74: # A nonexistant name forces the ninja cmd every
time.
As before.

https://codereview.chromium.org/14577007/diff/5001/third_party/protobuf/proto...
File third_party/protobuf/protobuf.gyp (right):

https://codereview.chromium.org/14577007/diff/5001/third_party/protobuf/proto...
third_party/protobuf/protobuf.gyp:60: # A nonexistant name forces the ninja cmd
every time.
Here too.

https://codereview.chromium.org/14577007/diff/5001/third_party/protobuf/proto...
third_party/protobuf/protobuf.gyp:62: '<(ninja_product_dir)/protoc',
So what…a real output and a fake output? Why?

[justincohen, 2013-05-02 02:59:56]

I removed the empty output changes and will try to make those changes within
ninja.py instead.

https://codereview.chromium.org/14577007/diff/5001/build/common.gypi
File build/common.gypi (right):

https://codereview.chromium.org/14577007/diff/5001/build/common.gypi#newcode4352
build/common.gypi:4352: 'SDKROOT': 'iphonesimulator<(ios_sdk)',  # -isysroot
It's done on the command line passed to xcodebuild, -sdk iphonesimulator6.1

On 2013/05/01 18:12:10, Mark Mentovai wrote:
> Where do you normally set SDKROOT to the simulator SDK for simulator builds?

[Mark Mentovai, 2013-05-02 03:18:48]

LGTM

[justincohen, 2013-05-02 12:11:18]

stuartmorgan@ changed mac_build to pass include dir instead.

[Mark Mentovai, 2013-05-02 13:19:22]

LG otherwise

https://codereview.chromium.org/14577007/diff/21004/build/ios/mac_build.gypi
File build/ios/mac_build.gypi (right):

https://codereview.chromium.org/14577007/diff/21004/build/ios/mac_build.gypi#...
build/ios/mac_build.gypi:52: },
This line doesn’t belong.

[justincohen, 2013-05-02 13:21:41]

https://codereview.chromium.org/14577007/diff/21004/build/ios/mac_build.gypi
File build/ios/mac_build.gypi (right):

https://codereview.chromium.org/14577007/diff/21004/build/ios/mac_build.gypi#...
build/ios/mac_build.gypi:52: },
On 2013/05/02 13:19:22, Mark Mentovai wrote:
> This line doesn’t belong.

Done.

[justincohen, 2013-05-02 15:04:38]

I misunderstood stuartmorgan's request for the downstream CL associated with
this.  Reverting back to what I had in patch set 5 which already had mento's
LGTM.

PTAL stuartmorgan!

[stuartmorgan, 2013-05-02 15:27:09]

LGTM

[commit-bot: I haz the power, 2013-05-02 15:48:11]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/justincohen@google.com/14577007/27004

[commit-bot: I haz the power, 2013-05-02 16:59:46]

Retried try job too often on linux_chromeos for step(s) browser_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_chro...

[commit-bot: I haz the power, 2013-05-02 17:36:47]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/justincohen@google.com/14577007/27004

[commit-bot: I haz the power, 2013-05-02 18:55:32]

Retried try job too often on linux_chromeos for step(s) ash_unittests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_chro...

[commit-bot: I haz the power, 2013-05-02 19:34:35]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/justincohen@google.com/14577007/27004

[commit-bot: I haz the power, 2013-05-02 22:25:14]

Retried try job too often on linux_chromeos for step(s) browser_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_chro...

[commit-bot: I haz the power, 2013-05-03 04:05:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/justincohen@google.com/14577007/27004

[commit-bot: I haz the power, 2013-05-03 11:06:43]

Change committed as 198084