Update paths for the new directory structure.

Update paths for the new directory structure.

Source paths in the gyp files and include paths in all files are
updated to the new directory structure.

Committed: 3472

[rtoy, 2013-02-04 18:17:38]

Please review this at your convenience.

One question I have:  The include paths are updated to something like
"api/armCOMM.h".  But should they be more like "openmax/dl/api/armCOMM.h"?  I
wasn't sure what the correct paths would be.

[ajm, 2013-02-04 22:56:27]

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/d...
File third_party/openmax/dl/dl.gyp (right):

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/d...
third_party/openmax/dl/dl.gyp:19: '.',
There are two other reasonable alternatives I can see here: using '..' or
'../..' and prefixing includes with "dl/" or "openmax/dl" respectively.

I think I prefer the former, as it makes less assumption on where the code will
be synced to, but still has _some_ kind of unique identifier. What do you think?

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/s...
File third_party/openmax/dl/sp/src/test/compare.c (right):

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/s...
third_party/openmax/dl/sp/src/test/compare.c:14: #include "compare.h"
I suppose you should use the qualified path here as well, i.e.
"dl/sp/src/test/compare.h"

and for the other test files.

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/s...
File third_party/openmax/dl/sp/src/test/test_fft.gyp (right):

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/s...
third_party/openmax/dl/sp/src/test/test_fft.gyp:20: '../../../',
Just realized, you could add this once to a target_defaults block at the top
instead of to each target.

[rtoy, 2013-02-04 23:01:19]

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/d...
File third_party/openmax/dl/dl.gyp (right):

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/d...
third_party/openmax/dl/dl.gyp:19: '.',
On 2013/02/04 22:56:27, ajm wrote:
> There are two other reasonable alternatives I can see here: using '..' or
> '../..' and prefixing includes with "dl/" or "openmax/dl" respectively.
> 
> I think I prefer the former, as it makes less assumption on where the code
will
> be synced to, but still has _some_ kind of unique identifier. What do you
think?

dl/ is fine with me.  I will make the appropriate changes.

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/s...
File third_party/openmax/dl/sp/src/test/compare.c (right):

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/s...
third_party/openmax/dl/sp/src/test/compare.c:14: #include "compare.h"
On 2013/02/04 22:56:27, ajm wrote:
> I suppose you should use the qualified path here as well, i.e.
> "dl/sp/src/test/compare.h"
> 
> and for the other test files.

Yes, I will update these accordingly.

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/s...
File third_party/openmax/dl/sp/src/test/test_fft.gyp (right):

https://webrtc-codereview.appspot.com/1095004/diff/1/third_party/openmax/dl/s...
third_party/openmax/dl/sp/src/test/test_fft.gyp:20: '../../../',
On 2013/02/04 22:56:27, ajm wrote:
> Just realized, you could add this once to a target_defaults block at the top
> instead of to each target.

That's much better.  Thanks.

[ajm, 2013-02-05 02:27:11]

lgtm with one possible change.

https://webrtc-codereview.appspot.com/1095004/diff/7001/third_party/openmax/d...
File third_party/openmax/dl/sp/src/test/test_fft.gyp (right):

https://webrtc-codereview.appspot.com/1095004/diff/7001/third_party/openmax/d...
third_party/openmax/dl/sp/src/test/test_fft.gyp:69: 'conditions': [
If you want, you could move this to target_defaults as well; I suppose it
doesn't hurt to be defined for the fixed-point tests.

[rtoy, 2013-02-05 17:10:19]

On 2013/02/05 02:27:11, ajm wrote:
> lgtm with one possible change.
> 
>
https://webrtc-codereview.appspot.com/1095004/diff/7001/third_party/openmax/d...
> File third_party/openmax/dl/sp/src/test/test_fft.gyp (right):
> 
>
https://webrtc-codereview.appspot.com/1095004/diff/7001/third_party/openmax/d...
> third_party/openmax/dl/sp/src/test/test_fft.gyp:69: 'conditions': [
> If you want, you could move this to target_defaults as well; I suppose it
> doesn't hurt to be defined for the fixed-point tests.

Done.  The fixed-point tests are fine since they don't use the value of
TWIDDLE_TABLE_ORDER to determine how many tests to run.

[Kevin, 2013-02-05 17:17:28]

lgtm

On 2013/02/05 17:10:19, rtoy wrote:
> On 2013/02/05 02:27:11, ajm wrote:
> > lgtm with one possible change.
> > 
> >
>
https://webrtc-codereview.appspot.com/1095004/diff/7001/third_party/openmax/d...
> > File third_party/openmax/dl/sp/src/test/test_fft.gyp (right):
> > 
> >
>
https://webrtc-codereview.appspot.com/1095004/diff/7001/third_party/openmax/d...
> > third_party/openmax/dl/sp/src/test/test_fft.gyp:69: 'conditions': [
> > If you want, you could move this to target_defaults as well; I suppose it
> > doesn't hurt to be defined for the fixed-point tests.
> 
> Done.  The fixed-point tests are fine since they don't use the value of
> TWIDDLE_TABLE_ORDER to determine how many tests to run.

[rtoy, 2013-02-05 17:23:33]

Thanks for reviewing this.