Update make for hello examples

Update make for hello examples

Simplify the makefile for the hello_world_XXX examples.
Update create_nmf to support direct PNaCl NMF files.
Update common.mk to support PNaCl

NOTRY=true
R=binji@chromium.org
BUG=169044
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=175921

[noelallen1, 2012-12-30 23:30:16]

PTaL

[binji, 2013-01-02 21:02:26]

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
File native_client_sdk/src/examples/hello_world/Makefile (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
native_client_sdk/src/examples/hello_world/Makefile:48: LIBS=$(DEPS) ppapi
pthread
where is DEPS?

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
File native_client_sdk/src/examples/hello_world_gles/Makefile (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
native_client_sdk/src/examples/hello_world_gles/Makefile:48: LIBS=ppapi_gles2
ppapi pthread
no DEPS here?

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
File native_client_sdk/src/examples/hello_world_stdio/Makefile (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
native_client_sdk/src/examples/hello_world_stdio/Makefile:48: DEPS=ppapi_main
why do we need to rebuild ppapi_main?

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
File native_client_sdk/src/tools/common.mk (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
native_client_sdk/src/tools/common.mk:278: $(warning No valid Chrome found at
CHROME_PATH=$(CHROME_PATH))
tabs...? Actually, most of this seems to be untabbed. Is that just a bug in
Rietveld? Is it legal to use 8 spaces instead of a tab character?

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
native_client_sdk/src/tools/common.mk:297: CHROME_ARGS+=--enable-nacl
--enable-pnacl --incognito --ppapi-out-of-process
I'm starting to wonder about --incognito, actually. It can reduce confusion due
to Chrome caching, but it can introduce confusion when trying to use the
filesystem. Should we encourage it?

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
File native_client_sdk/src/tools/create_nmf.py (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
native_client_sdk/src/tools/create_nmf.py:542: #  if options.debug_mode:
uncomment

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
native_client_sdk/src/tools/create_nmf.py:581: if options.stage_dependencies and
not nmf.pnacl:
is this necessary? Shouldn't the flow be the same as newlib?

[noelallen1, 2013-01-07 20:01:30]

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
File native_client_sdk/src/examples/hello_world/Makefile (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
native_client_sdk/src/examples/hello_world/Makefile:48: LIBS=$(DEPS) ppapi
pthread
On 2013/01/02 21:02:26, binji wrote:
> where is DEPS?

Done.

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
File native_client_sdk/src/examples/hello_world_gles/Makefile (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
native_client_sdk/src/examples/hello_world_gles/Makefile:48: LIBS=ppapi_gles2
ppapi pthread
On 2013/01/02 21:02:26, binji wrote:
> no DEPS here?

Done.

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
File native_client_sdk/src/examples/hello_world_stdio/Makefile (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
native_client_sdk/src/examples/hello_world_stdio/Makefile:48: DEPS=ppapi_main
We provide the DEPS in case someone wants to modify ppapi_main and have it
rebuild (this was really useful for printf debugging).  In the future we may
want to do this for ppapi_cpp as well.

On 2013/01/02 21:02:26, binji wrote:
> why do we need to rebuild ppapi_main?

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
File native_client_sdk/src/tools/common.mk (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
native_client_sdk/src/tools/common.mk:278: $(warning No valid Chrome found at
CHROME_PATH=$(CHROME_PATH))
Strange, should be tabs.  It's a tab in my source.  I will double check.

On 2013/01/02 21:02:26, binji wrote:
> tabs...? Actually, most of this seems to be untabbed. Is that just a bug in
> Rietveld? Is it legal to use 8 spaces instead of a tab character?

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
native_client_sdk/src/tools/common.mk:297: CHROME_ARGS+=--enable-nacl
--enable-pnacl --incognito --ppapi-out-of-process
Good question.  A "temp" user-data-dir might work better.

On 2013/01/02 21:02:26, binji wrote:
> I'm starting to wonder about --incognito, actually. It can reduce confusion
due
> to Chrome caching, but it can introduce confusion when trying to use the
> filesystem. Should we encourage it?

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
File native_client_sdk/src/tools/create_nmf.py (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
native_client_sdk/src/tools/create_nmf.py:542: #  if options.debug_mode:
On 2013/01/02 21:02:26, binji wrote:
> uncomment

Done.

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/tools/c...
native_client_sdk/src/tools/create_nmf.py:581: if options.stage_dependencies and
not nmf.pnacl:
Unfortunately no.  obj-dump fails on bit code files as they are not standard ELF
files.

On 2013/01/02 21:02:26, binji wrote:
> is this necessary? Shouldn't the flow be the same as newlib?

[binji, 2013-01-08 23:39:32]

lgtm

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
File native_client_sdk/src/examples/hello_world_stdio/Makefile (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
native_client_sdk/src/examples/hello_world_stdio/Makefile:48: DEPS=ppapi_main
OK, seems like we'd want to include any buildable library here then
(nacl_mounts?)

On 2013/01/07 20:01:30, noelallen1 wrote:
> We provide the DEPS in case someone wants to modify ppapi_main and have it
> rebuild (this was really useful for printf debugging).  In the future we may
> want to do this for ppapi_cpp as well.
> 
> On 2013/01/02 21:02:26, binji wrote:
> > why do we need to rebuild ppapi_main?
>

[noelallen_use_chromium, 2013-01-09 17:48:45]

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
File native_client_sdk/src/examples/hello_world_stdio/Makefile (right):

https://codereview.chromium.org/11725003/diff/1/native_client_sdk/src/example...
native_client_sdk/src/examples/hello_world_stdio/Makefile:48: DEPS=ppapi_main
Yes.  nacl_mounts should be in there.  It was removed because it wasn't
supporting all versions (host, arm, newlib, glibc)  But I believe that now
works.

On 2013/01/08 23:39:32, binji wrote:
> OK, seems like we'd want to include any buildable library here then
> (nacl_mounts?)
> 
> On 2013/01/07 20:01:30, noelallen1 wrote:
> > We provide the DEPS in case someone wants to modify ppapi_main and have it
> > rebuild (this was really useful for printf debugging).  In the future we may
> > want to do this for ppapi_cpp as well.
> > 
> > On 2013/01/02 21:02:26, binji wrote:
> > > why do we need to rebuild ppapi_main?
> > 
>

[commit-bot: I haz the power, 2013-01-09 20:51:10]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/noelallen@chromium.org/11725003/16004

[commit-bot: I haz the power, 2013-01-09 20:51:26]

Presubmit check for 11725003-16004 failed and returned exit status 1.


Running presubmit commit checks ...
Running pylint on 48 files.

** Presubmit Warnings **
Found lines longer than 80 characters (first 5 shown).
  native_client_sdk/src/tools/common.mk, line 229, 81 chars

** Presubmit ERRORS **
Run the command: svn pset svn:eol-style LF \
  native_client_sdk/src/tools/common.mk

Presubmit checks took 10.5s to calculate.

[commit-bot: I haz the power, 2013-01-09 22:27:01]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/noelallen@chromium.org/11725003/16004

[commit-bot: I haz the power, 2013-01-09 22:27:13]

Presubmit check for 11725003-16004 failed and returned exit status 1.


Running presubmit commit checks ...
Running pylint on 48 files.

** Presubmit Warnings **
Found lines longer than 80 characters (first 5 shown).
  native_client_sdk/src/tools/common.mk, line 229, 81 chars

** Presubmit ERRORS **
Run the command: svn pset svn:eol-style LF \
  native_client_sdk/src/tools/common.mk

Presubmit checks took 7.2s to calculate.