libvpx.gyp: Remove redundant sources.

libvpx.gyp: Remove redundant sources.

A typo that went unnoticed at the usage sites of process_outputs_as_sources
(a missing 's' in outputs) resulted in the workaround of adding the output from
the libvpx asm offsets step as an explicit source file.

R=thakis@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=289715

[Tom Finegan, 2014-08-14 22:17:49]

Hi Nico,

This is a followup to https://codereview.chromium.org/465283003 (libvpx gyp
change) and https://codereview.chromium.org/472593002 (libvpx roll). 

The CL came about because Ami noticed that we had an error in our usage of
process_outputs_as_sources. This CL removes what (I think) are redundant sources
in libvpx.gyp.

Anyway, I've picked you as a reviewer since Ami suggested you as a good choice
for a gyp change review. PTAL when you can.

Thanks!

[Nico, 2014-08-14 22:29:22]

The "(very away)" is supposed to suggest similar ideas as Ami's handle…

https://codereview.chromium.org/469363004/diff/1/libvpx.gyp
File libvpx.gyp (left):

https://codereview.chromium.org/469363004/diff/1/libvpx.gyp#oldcode475
libvpx.gyp:475: '-s',
'<(ninja_obj_dir)/encoder/libvpx_asm_offsets_vp8.vp8_asm_enc_offsets.obj',
Huh, what's this needed for? Why are there 3 paths in this action?
(<(INTERMEDIATE_DIR)/..., <(PRODUCT_DIR)/obj/..., and this one)

https://codereview.chromium.org/469363004/diff/1/libvpx.gyp#oldcode574
libvpx.gyp:574: ],
This branch doesn't have an action with process_outputs_as_sources, do you want
to remove this one too?

[Tom Finegan, 2014-08-14 22:47:35]

> The "(very away)" is supposed to suggest similar ideas as Ami's handle…

"GONE FROM CHROMIUM" is a lot more clear when people have no idea about these
things-- anyway, sorry to drag you into this if you're trying to avoid
chrome/ium stuff. I really just need someone that's more comfortable with GYP
than I am, and who knows more about it. 

Can you suggest someone else I should ask to have a look, and then I'll leave
you alone? :)

Anyway, thanks for the review!

https://codereview.chromium.org/469363004/diff/1/libvpx.gyp
File libvpx.gyp (left):

https://codereview.chromium.org/469363004/diff/1/libvpx.gyp#oldcode475
libvpx.gyp:475: '-s',
'<(ninja_obj_dir)/encoder/libvpx_asm_offsets_vp8.vp8_asm_enc_offsets.obj',
On 2014/08/14 22:29:22, Nico (very away) wrote:
> Huh, what's this needed for? Why are there 3 paths in this action?
> (<(INTERMEDIATE_DIR)/..., <(PRODUCT_DIR)/obj/..., and this one)

There have been problems with search paths on various chrome (specifically,
android webview and webrtc) builders due to differences in environment setup.
The extra paths are a workaround.

Here's a CL with the addition of a path that has some explanation about what's
going on (not exactly the same thing, but somewhat similar at least).

https://codereview.chromium.org/450993002/

Basically, it's more search paths because of different output locations on
various platforms.

https://codereview.chromium.org/469363004/diff/1/libvpx.gyp#oldcode574
libvpx.gyp:574: ],
On 2014/08/14 22:29:22, Nico (very away) wrote:
> This branch doesn't have an action with process_outputs_as_sources, do you
want
> to remove this one too?

There's a process_outputs_as_sources from the inclusion of
unpack_lib_posix.gypi.

https://code.google.com/p/chromium/codesearch#chromium/src/third_party/libvpx...

[Nico, 2014-08-14 22:48:54]

https://codereview.chromium.org/469363004/diff/1/libvpx.gyp
File libvpx.gyp (left):

https://codereview.chromium.org/469363004/diff/1/libvpx.gyp#oldcode574
libvpx.gyp:574: ],
Ah! lgtm

[Tom Finegan, 2014-08-14 22:53:12]

Committed patchset #1 manually as r289715 (presubmit successful).

[Tom Finegan, 2014-08-15 06:04:21]

On 2014/08/14 22:53:12, Tom Finegan wrote:
> Committed patchset #1 manually as r289715 (presubmit successful).

Revert here:
https://codereview.chromium.org/474153002