chromium/src/ppapi.git
4 months agoPepper: Make pending_sync_msgs_ local to IO thread. master
teravest@chromium.org [Thu, 21 Aug 2014 23:19:07 +0000 (23:19 +0000)]
Pepper: Make pending_sync_msgs_ local to IO thread.

Adding support for PpapiHostMessage_OpenResource to SFI mode is tricky.
Validating the file tokens requires that a message to be sent to the browser
(!) to validate that information in the case of a compromised renderer.

This change is one step to allowing processing for the reply of the
PpapiHostMessage_OpenResource message to send a message to the browser for
constructing a NaClDesc properly.

This makes the encapsulation of NaClMessageScanner a bit worse, but also
reduces the amount of code for which the critical section under lock_ must be
held.

BUG=394130

Review URL: https://codereview.chromium.org/472073003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@291234 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoNaCl: Send fatal log messages via shared memory.
teravest@chromium.org [Wed, 20 Aug 2014 20:43:09 +0000 (20:43 +0000)]
NaCl: Send fatal log messages via shared memory.

nacl_helper reports the last few log messages to the renderer when a LOG_FATAL
message is logged in the plugin. This currently uses the "bootstrap" channel,
which we would like to get rid of.

This uses a shared memory segment instead of IPC, since something bad may have
happened in the plugin process, so we'd like to keep the behavior for handling
that case as simple as possible. For that reason, the shared memory segment is
mapped before the nexe is loaded.

I've tested this locally by calling RemoteLog(LOG_FATAL, ...) after StartModule
in ServiceRuntime, though I'd like to find some better way to test this as
there's a decent amount of logic here.

BUG=391039
R=jschuh@chromium.org, mseaborn@chromium.org

Review URL: https://codereview.chromium.org/469423002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@290910 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoAdd a X-Requested-With header to URL requests for PPAPI Flash (only).
tsepez@chromium.org [Wed, 20 Aug 2014 19:23:20 +0000 (19:23 +0000)]
Add a X-Requested-With header to URL requests for PPAPI Flash (only).

This is a revival of Trung's patch in https://codereview.chromium.org/10886047/.
The difference is that we provide the header only for the Shockwave Flash
plugin (plus a test plugin).

BUG=145090

Review URL: https://codereview.chromium.org/451923002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@290887 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: add Image Capture interfaces.
wuchengli@chromium.org [Sun, 17 Aug 2014 17:20:28 +0000 (17:20 +0000)]
Pepper: add Image Capture interfaces.

API proposal:
https://docs.google.com/a/chromium.org/document/d/1deeCrsvT9xA77UP4yWMhqbHgS3FvYCTF_8meIMQRrDo/edit#

BUG=387547
TEST=Build chrome, run PPAPITest, and run an app that uses ppapi.

Review URL: https://codereview.chromium.org/391323002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@290186 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months ago[PPAPI] Add target param for CompositorLayer::SetTexture().
penghuang@chromium.org [Fri, 15 Aug 2014 21:30:16 +0000 (21:30 +0000)]
[PPAPI] Add target param for CompositorLayer::SetTexture().

BUG=403504

Review URL: https://codereview.chromium.org/475123003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@290016 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoReturn from |Configure()| asynchronously, and allow |InitBuffers()| to return errors.
amistry@chromium.org [Fri, 15 Aug 2014 14:56:21 +0000 (14:56 +0000)]
Return from |Configure()| asynchronously, and allow |InitBuffers()| to return errors.

BUG=330851

Review URL: https://codereview.chromium.org/430943004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@289856 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoUsing PROFILE_ANY for vp8 and vp9
amogh.bihani@samsung.com [Fri, 15 Aug 2014 03:29:51 +0000 (03:29 +0000)]
Using PROFILE_ANY for vp8 and vp9

VP8 and VP9 do not take profile into account. Using PROFILE_MAIN is confusing.
This patch uses PROFILE_ANY for these codecs.

TBR=noelallen@chromium.org

BUG=361676

Review URL: https://codereview.chromium.org/418193003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@289765 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoppapi: Add way to ignore presubmit warning about including in NaCl SDK.
binji@chromium.org [Fri, 15 Aug 2014 00:37:03 +0000 (00:37 +0000)]
ppapi: Add way to ignore presubmit warning about including in NaCl SDK.

The ppapi presubmit checks that a modified interface exists in the NaCl SDK.
But not all interfaces should be included. It is currently a warning, but that
will prevent the CQ from submitting the CL.

Now we can add that file to an "IGNORE_LIST" and it will no longer emit a
presubmit warning.

BUG=403713
R=bbudge@chromium.org
TEST=modified pp_video_dev.h, ran presubmits, got warning. Added to ignore
list, ran again. No warning.

Review URL: https://codereview.chromium.org/475763003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@289734 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoAdd test support for transition from NPObject to gin bindings for Pepper
raymes@chromium.org [Mon, 11 Aug 2014 04:12:46 +0000 (04:12 +0000)]
Add test support for transition from NPObject to gin bindings for Pepper

This adds a function to PPB_Testing to run the garbage collector. This is
necessary when using gin-backed v8 objects so that we can test that
destruction works correctly when the gin-backed objects are no longer
referenced. It also modifies the tests to run garbage collection and also
to be careful with scoping such that any handles on the stack to v8 objects
are destroyed before running the GC.

BUG=351636

Review URL: https://codereview.chromium.org/445813002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@288647 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoAdd a ScopedPPVarArray class for managing arrays of PP_Vars.
raymes@chromium.org [Mon, 11 Aug 2014 03:08:24 +0000 (03:08 +0000)]
Add a ScopedPPVarArray class for managing arrays of PP_Vars.

This adds a ScopedPPVarArray class which automatically manages the memory
for an array of PP_Vars. There are two ways to construct a ScopedPPVarArray:
1) Pass in a size, which allocates an array of PP_Vars of that size.
2) Pass in an existing pointer to a PP_Var array and a size. This array
must have been allocated with PPB_Memory_Dev.

When the ScopedPPVarArray is destroyed, it will decrement a reference to each
of the vars in the array and then release the memory associated with the array.

Setting a value in the array will decrease the refcount of the existing var and
increase the refcount of the var that is set.

Calling Release() will pass ownership of the array memory and the PP_Var
references to the caller.

Copy and assignment of ScopedPPVarArray aren't currently supported.

This construct turns out to be useful in a bunch of places, but namely for the
synchronous scripting APIs.

BUG=351636

Review URL: https://codereview.chromium.org/441193002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@288642 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoAdd PepperTryCatch and V8ObjectVar classes
raymes@chromium.org [Mon, 11 Aug 2014 02:59:12 +0000 (02:59 +0000)]
Add PepperTryCatch and V8ObjectVar classes

This adds PepperTryCatch and V8ObjectVar classes which will be used in the
refactoring which replaces the NPAPI bindings with gin bindings for pepper.

V8ObjectVar is a fairly straightforward replacement for NPObjectVar.

PepperTryCatch (and subclasses) are used to conveniently catch and report
errors eminating from either pepper or V8. It also includes facilities for
conveniently converting between v8 objects and vars. It is essentially the
equivalent of some of the utilities in npapi_glue.h.

This also adds conversion for object vars to the V8VarConverter and some
unimplemented functions to the HostVarTracker to allow these new files to
compile.

BUG=351636

Review URL: https://codereview.chromium.org/421963008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@288637 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPPAPI: Revive ppp_messaging_proxy_perftest
dmichael@chromium.org [Fri, 8 Aug 2014 21:13:00 +0000 (21:13 +0000)]
PPAPI: Revive ppp_messaging_proxy_perftest

This test can measure PPP_Messaging performance and indirectly IPC
performance. It regressed due to a couple of minor changes; this
gets it running again.

BUG=

Review URL: https://codereview.chromium.org/453003002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@288433 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoFix PNaCl on Demand load faliures.
teravest@chromium.org [Fri, 8 Aug 2014 20:56:06 +0000 (20:56 +0000)]
Fix PNaCl on Demand load faliures.

r284961 broke PNaCl on-demand by loading resources retrieved by
component_uploader before sending a request to download a user pexe. This
change fixes the breakage by only loading resources after we've determined that
that there is no cached translated nexe.

BUG=401755
R=dmichael@chromium.org, dschuff@chromium.org

Review URL: https://codereview.chromium.org/453073002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@288426 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoMove StringToLowerASCII to base namespace
brettw@chromium.org [Thu, 7 Aug 2014 16:55:42 +0000 (16:55 +0000)]
Move StringToLowerASCII to base namespace

TBR=sky

Review URL: https://codereview.chromium.org/448853002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@288085 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: Fix caching for translated nexes.
teravest@chromium.org [Thu, 7 Aug 2014 00:15:34 +0000 (00:15 +0000)]
Pepper: Fix caching for translated nexes.

A refactor of FileDownloader (r284961) broke the caching behavior for nexes
that were translated from pexes; the cache would always end up with entries of
length 0, breaking future loads of the pexe. This is the translated nexe was
written to the wrong file handle, preventing caching of the result.

This change causes the translated nexe to be written to the correct file
handle, fixing the caching behavior.

I manually tested this change on the "game of life" demo at
https://gonativeclient.appspot.com/demo/life, verifying that a cached
translated nexe for the pexe was successfully loaded.

BUG=401121

Review URL: https://codereview.chromium.org/447013002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@287912 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: Remove trusted plugin loadable module.
teravest@chromium.org [Tue, 5 Aug 2014 19:53:20 +0000 (19:53 +0000)]
Pepper: Remove trusted plugin loadable module.

This change links the trusted plugin into the renderer, making it easier for
the trusted plugin to use libraries in Chromium. This removes the
ppGoogleNaClPlugin loadable module file on various platforms.

This is part of a larger effort to remove the "trusted plugin" used to
bootstrap NaCl plugins.

It introduces an "internal_module" interface for setting the value returned by
pp::Module::Get(). This is so that both the trusted plugin and the remoting
plugin can be linked into the renderer. However, I believe this is safe because
the trusted plugin runs in-process and the remoting plugin runs out-of-process.

BUG=394497
R=dmichael@chromium.org, mseaborn@chromium.org, phajdan.jr@chromium.org, thestig@chromium.org, wez@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=287071

Review URL: https://codereview.chromium.org/397243004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@287580 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoRevert "Pepper: Remove trusted plugin loadable module."
teravest@chromium.org [Fri, 1 Aug 2014 21:20:32 +0000 (21:20 +0000)]
Revert "Pepper: Remove trusted plugin loadable module."

The reverted change broke the ChromeOS builder, which depends on the loadable
module file existing. Argh.

This reverts commit b0b91c06bc2ed6afc0611f00d68f72b19dd06d26.

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@287090 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: Remove trusted plugin loadable module.
teravest@chromium.org [Fri, 1 Aug 2014 19:37:52 +0000 (19:37 +0000)]
Pepper: Remove trusted plugin loadable module.

This change links the trusted plugin into the renderer, making it easier for
the trusted plugin to use libraries in Chromium. This removes the
ppGoogleNaClPlugin loadable module file on various platforms.

This is part of a larger effort to remove the "trusted plugin" used to
bootstrap NaCl plugins.

It introduces an "internal_module" interface for setting the value returned by
pp::Module::Get(). This is so that both the trusted plugin and the remoting
plugin can be linked into the renderer. However, I believe this is safe because
the trusted plugin runs in-process and the remoting plugin runs out-of-process.

BUG=394497
R=dmichael@chromium.org, mseaborn@chromium.org, phajdan.jr@chromium.org, thestig@chromium.org, wez@chromium.org

Review URL: https://codereview.chromium.org/397243004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@287071 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: Move some histogram code out of plugin.
teravest@chromium.org [Thu, 31 Jul 2014 21:45:46 +0000 (21:45 +0000)]
Pepper: Move some histogram code out of plugin.

This is a tiny change that moves PNaCl cache hit histogram logic out of the
trusted plugin.

BUG=none

Review URL: https://codereview.chromium.org/423363003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@286897 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoExtract PepperFileSystemTypeToFileSystemType() from ppapi.
tfarina@chromium.org [Thu, 31 Jul 2014 12:56:29 +0000 (12:56 +0000)]
Extract PepperFileSystemTypeToFileSystemType() from ppapi.

This function is called only from within content to convert from a ppapi
enum type to webkit fileapi type (this one will be come a content enum
type in later change).

So anticipating that change from webkit/ to content/ we moved this
function from ppapi to content to make this future change a little bit
easier.

And it is also nice to do now because it removes the last include of
webkit/ from ppapi making it much more clearer in term of deps.

BUG=338338
TEST=blink_tests
R=dmichael@chromium.org,jam@chromium.org

Review URL: https://codereview.chromium.org/424213006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@286776 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoSupport configuring the audio buffer duration in the Pepper MediaStream API.
amistry@chromium.org [Thu, 31 Jul 2014 12:54:12 +0000 (12:54 +0000)]
Support configuring the audio buffer duration in the Pepper MediaStream API.

Increasing the buffer duration allows the Pepper plugin the sleep for
longer between buffers and reduces the per-sample overhead. This all
translates to lower power consumption, at the expense of latency.

BUG=330851

Review URL: https://codereview.chromium.org/414643003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@286774 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: Remove chatty printfs in trusted plugin.
teravest@chromium.org [Wed, 30 Jul 2014 19:17:16 +0000 (19:17 +0000)]
Pepper: Remove chatty printfs in trusted plugin.

There's a lot of unnecessary debug printing happening in the trusted plugin.
This cleans up some of it and simplifies some logic a little bit.

BUG=
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/426113002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@286577 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoCleanup PPAPI TestWebSocket::TestValidClose() for readability
tyoshino@chromium.org [Tue, 29 Jul 2014 09:35:59 +0000 (09:35 +0000)]
Cleanup PPAPI TestWebSocket::TestValidClose() for readability

BUG=396399

Review URL: https://codereview.chromium.org/416003002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@286147 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoMake IPC macros work with -Wunused-local-typedefs.
thakis@chromium.org [Tue, 29 Jul 2014 07:55:22 +0000 (07:55 +0000)]
Make IPC macros work with -Wunused-local-typedefs.

The typedef can be used by the IPC_MESSAGE_HANDLER macro, but that macro isn't
always used when IPC_BEGIN_MESSAGE_MAP is used. The alternatives are to either
pass ClassName::HanderFunction to IPC_MESSAGE_HANDLER (like what one has to do
to use IPC_MESSAGE_FORWARD), or to mark the typedef in IPC_BEGIN_MESSAGE_MAP
as ALLOW_UNUSED so that it isn't warned on.

BUG=321833
TBR=brettw

Review URL: https://codereview.chromium.org/422083002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@286115 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPrep change to use 'translate_pexe_with_build': 1 flag where needed.
jvoung@chromium.org [Tue, 29 Jul 2014 03:03:12 +0000 (03:03 +0000)]
Prep change to use 'translate_pexe_with_build': 1 flag where needed.

Paired with: https://codereview.chromium.org/415653002/
Once that rolls into Chrome, it will require this flag
set in order to translate a pexe -> nexe.

This currently include ppapi_nacl test nexes, and a few
nonsfi x86-32 nexes.

Also adds a couple more shim deps for the nonsfi tests.

BUG=none

Review URL: https://codereview.chromium.org/413763002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@286069 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: Remove LOAD_MODULE SRPC call in SFI mode.
teravest@chromium.org [Thu, 24 Jul 2014 23:58:56 +0000 (23:58 +0000)]
Pepper: Remove LOAD_MODULE SRPC call in SFI mode.

This change puts nexe information in the parameters for starting a NaCl
plugin instead of sending that information over SRPC. This may remove the
need for that IPC round trip entirely; perhaps we could report load failure
as part of the RPCs for starting sel_ldr or performing StartModule().

nacl_defines.gypi as added as a dependency in components/components_tests.gyp
because nacl_process_host.h now includes
"native_client/src/public/nacl_file_info.h", which requires nacl_defines.

BUG=333950
R=bbudge@chromium.org, brettw@chromium.org, hidehiko@chromium.org, jschuh@chromium.org, mseaborn@chromium.org, sky@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=285028

Review URL: https://codereview.chromium.org/332463003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@285418 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoMiscellaneous GN updates.
brettw@chromium.org [Thu, 24 Jul 2014 23:10:53 +0000 (23:10 +0000)]
Miscellaneous GN updates.

Removes some unnecessary prints accidentally checked in to the Android build.

Removes a GYP TODO now that angle has rolled.

Updates commented-out labels of targets that now exist.

R=dpranke@chromium.org

Review URL: https://codereview.chromium.org/419053002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@285400 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoRevert 285028 "Pepper: Remove LOAD_MODULE SRPC call in SFI mode."
bruening@google.com [Thu, 24 Jul 2014 02:35:07 +0000 (02:35 +0000)]
Revert 285028 "Pepper: Remove LOAD_MODULE SRPC call in SFI mode."

Reason for revert: r285028 seems to have broken the ThreadSanitizer build.

> Pepper: Remove LOAD_MODULE SRPC call in SFI mode.
>
> This change puts nexe information in the parameters for starting a NaCl
> plugin instead of sending that information over SRPC. This may remove the
> need for that IPC round trip entirely; perhaps we could report load failure
> as part of the RPCs for starting sel_ldr or performing StartModule().
>
> nacl_defines.gypi as added as a dependency in components/components_tests.gyp
> because nacl_process_host.h now includes
> "native_client/src/public/nacl_file_info.h", which requires nacl_defines.
>
> BUG=333950
> R=bbudge@chromium.org, brettw@chromium.org, hidehiko@chromium.org, jschuh@chromium.org, mseaborn@chromium.org, sky@chromium.org
>
> Review URL: https://codereview.chromium.org/332463003

TBR=teravest@chromium.org

Review URL: https://codereview.chromium.org/413063003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@285120 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: Remove LOAD_MODULE SRPC call in SFI mode.
teravest@chromium.org [Wed, 23 Jul 2014 20:22:49 +0000 (20:22 +0000)]
Pepper: Remove LOAD_MODULE SRPC call in SFI mode.

This change puts nexe information in the parameters for starting a NaCl
plugin instead of sending that information over SRPC. This may remove the
need for that IPC round trip entirely; perhaps we could report load failure
as part of the RPCs for starting sel_ldr or performing StartModule().

nacl_defines.gypi as added as a dependency in components/components_tests.gyp
because nacl_process_host.h now includes
"native_client/src/public/nacl_file_info.h", which requires nacl_defines.

BUG=333950
R=bbudge@chromium.org, brettw@chromium.org, hidehiko@chromium.org, jschuh@chromium.org, mseaborn@chromium.org, sky@chromium.org

Review URL: https://codereview.chromium.org/332463003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@285028 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoImprove content test GN build
brettw@chromium.org [Wed, 23 Jul 2014 20:21:50 +0000 (20:21 +0000)]
Improve content test GN build

Fix libxml clang warnings.

Add ppapi_unittest_shared target.

Rename ipc:test_support_ipc to ipc:test_support and use this target when needed (the places that wanted this incorrectly specified only "ipc".

R=hclam@chromium.org

Review URL: https://codereview.chromium.org/418453002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@285027 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoSome linking fixes for GN build.
brettw@chromium.org [Wed, 23 Jul 2014 17:15:38 +0000 (17:15 +0000)]
Some linking fixes for GN build.

Add some content_shell targets. content_shell itself doesn't link yet so it's commented out, but some assocaited libs now compile.

This fixes a variety of out-of-date sources lists to get some targets closer to linking in the GN build.

Add xdisplaycheckk target

Update GYP files for GN names

Rename events_ipc

R=hclam@chromium.org

Review URL: https://codereview.chromium.org/407243002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@284974 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 months agoPepper: Delete FileDownloader in trusted plugin.
teravest@chromium.org [Wed, 23 Jul 2014 16:58:15 +0000 (16:58 +0000)]
Pepper: Delete FileDownloader in trusted plugin.

This simplifies PnaclCoordinator considerably and reduces the total
amount of code in the trusted plugin.

BUG=239656

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=284684

R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/393693004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@284961 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoRevert 284684 "Pepper: Delete FileDownloader in trusted plugin."
sky@chromium.org [Tue, 22 Jul 2014 23:09:52 +0000 (23:09 +0000)]
Revert 284684 "Pepper: Delete FileDownloader in trusted plugin."

Reverting as broke chrome and chromeos dbg builders:
http://build.chromium.org/p/chromium.linux/builders/Linux%20Tests%20%28dbg%29%281%29/builds/32615/steps/browser_tests/logs/SuccessfulLoadUMA
[ RUN      ] NaClBrowserTestPnacl.SuccessfulLoadUMA
Xlib:  extension "RANDR" missing on display ":9".
Xlib:  extension "RANDR" missing on display ":9".
[18889:18889:0722/130648:WARNING:password_store_factory.cc(215)] Using
basic (unencrypted) store for password storage. See
http://code.google.com/p/chromium/wiki/LinuxPasswordStorage for more
information about password storage options.
HTTP server started on http://127.0.0.1:56115...
sending server_data: {"host": "127.0.0.1", "port": 56115} (36 bytes)
[9:10:0722/130650:ERROR:pnacl_translation_resource_host.cc(135)] Got
invalid platformfilefortransit
/usr/lib/gcc/x86_64-linux-gnu/4.6/../../../../include/c++/4.6/debug/vector:461:
    error: attempt to insert into container with an iterator from a
    different container.

Objects involved in the operation:
sequence "this" @ 0x0x392ba6c93260 {
}
iterator "__position" @ 0x0x7fff9ff24e20 {
  state = past-the-end;
  references sequence @ 0x0x7fff9ff24e20
}
[18889:18938:0722/130651:WARNING:raw_channel_posix.cc(214)] recvmsg:
Connection reset by peer
BrowserTestBase signal handler received SIGTERM. Backtrace:
#0 0x7fe4bf491cee base::debug::StackTrace::StackTrace()
#1 0x000004af93fa content::(anonymous
namespace)::DumpStackTraceSignalHandler()
#2 0x7fe4b68074a0 \u003Cunknown>
#3 0x7fe4b68b9a43 __poll
#4 0x7fe4b76faff6 \u003Cunknown>
#5 0x7fe4b76fb124 g_main_context_iteration
#6 0x7fe4bf44ff75 base::MessagePumpGlib::Run()
#7 0x7fe4bf54b1f0 base::MessageLoop::RunHandler()
#8 0x7fe4bf5b18e2 base::RunLoop::Run()
#9 0x000004b63969 content::RunThisRunLoop()
#10 0x000004b638fa content::RunMessageLoop()
#11 0x000004b0ea15 content::JavascriptTestObserver::Run()
#12 0x000001312927 NaClBrowserTestBase::RunJavascriptTest()
#13 0x000001312a48 NaClBrowserTestBase::RunLoadTest()
#14 0x00000130f514 (anonymous
namespace)::NaClBrowserTestPnacl_SuccessfulLoadUMA_Test::RunTestOnMai
...

> Pepper: Delete FileDownloader in trusted plugin.
>
> This simplifies PnaclCoordinator considerably and reduces the total
> amount of code in the trusted plugin.
>
> BUG=239656
>
> Review URL: https://codereview.chromium.org/393693004

TBR=teravest@chromium.org

Review URL: https://codereview.chromium.org/406323003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@284791 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoRevert 284766 "Pepper: Simplify TempFile in trusted plugin."
sky@chromium.org [Tue, 22 Jul 2014 23:08:18 +0000 (23:08 +0000)]
Revert 284766 "Pepper: Simplify TempFile in trusted plugin."

Sorry I had to revert this. It's not because it caused problems, but
rather I needed to revert r284684 as it caused problems and your patch
touches the same files.

> Pepper: Simplify TempFile in trusted plugin.
>
> This change makes the behavior of internal_handle_ in TempFile easier to reason
> about. This change is possible after a large refactoring of PnaclCoordinator
> that happened as a result of removing FileDownloader.
>
> BUG=239656
> R=bbudge@chromium.org
>
> Review URL: https://codereview.chromium.org/413493002

TBR=teravest@chromium.org

Review URL: https://codereview.chromium.org/406383002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@284790 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Simplify TempFile in trusted plugin.
teravest@chromium.org [Tue, 22 Jul 2014 21:06:11 +0000 (21:06 +0000)]
Pepper: Simplify TempFile in trusted plugin.

This change makes the behavior of internal_handle_ in TempFile easier to reason
about. This change is possible after a large refactoring of PnaclCoordinator
that happened as a result of removing FileDownloader.

BUG=239656
R=bbudge@chromium.org

Review URL: https://codereview.chromium.org/413493002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@284766 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Delete FileDownloader in trusted plugin.
teravest@chromium.org [Tue, 22 Jul 2014 15:54:45 +0000 (15:54 +0000)]
Pepper: Delete FileDownloader in trusted plugin.

This simplifies PnaclCoordinator considerably and reduces the total
amount of code in the trusted plugin.

BUG=239656

Review URL: https://codereview.chromium.org/393693004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@284684 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoRemove an out of date comment in PPAPI code.
thestig@chromium.org [Mon, 21 Jul 2014 22:46:00 +0000 (22:46 +0000)]
Remove an out of date comment in PPAPI code.

NOTRY=true

Review URL: https://codereview.chromium.org/391413004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@284524 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoAdd to GN ppapi build.
brettw@chromium.org [Mon, 21 Jul 2014 17:25:31 +0000 (17:25 +0000)]
Add to GN ppapi build.

This should be everything except tests and examples.

R=bbudge@chromium.org

Review URL: https://codereview.chromium.org/403023002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@284451 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Remove unused ppapi.gypi file.
teravest@chromium.org [Thu, 17 Jul 2014 20:26:03 +0000 (20:26 +0000)]
Pepper: Remove unused ppapi.gypi file.

I came across this while trying to figure out how to make entry points work for
the trusted plugin. It's confusing to have these gypi variables around
when they're not used anywhere.

BUG=
R=bbudge@chromium.org

Review URL: https://codereview.chromium.org/405433006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@283855 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoAdd PPAPI interface for OpenGL ES 2.0 Vertex Array Objects.
zork@chromium.org [Thu, 17 Jul 2014 11:14:08 +0000 (11:14 +0000)]
Add PPAPI interface for OpenGL ES 2.0 Vertex Array Objects.

BUG=72612,370496

Review URL: https://codereview.chromium.org/397183002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@283750 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoUpdate and enable PPApiTest.URLLoaderTest3
yhirano@chromium.org [Thu, 17 Jul 2014 08:23:54 +0000 (08:23 +0000)]
Update and enable PPApiTest.URLLoaderTest3

The test was disabled[1] when XMLHttpRequest forbidden header names changed[2].
This CL updates and enables it.

1: https://src.chromium.org/viewvc/chrome?view=rev&revision=282507
2: https://src.chromium.org/viewvc/blink?view=rev&revision=178023

BUG=373120

Review URL: https://codereview.chromium.org/391213003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@283709 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Remove some gyp defines for trusted plugin.
teravest@chromium.org [Thu, 17 Jul 2014 06:46:34 +0000 (06:46 +0000)]
Pepper: Remove some gyp defines for trusted plugin.

As far as I can tell, these defines shouldn't have any effect. I'd like to make
the trusted plugin build more similar to what we have for the renderer so that
the transition will be easier when we get rid of the trusted plugin shared
library.

This mirrors a change made to native_client in 2012:
  http://git.chromium.org/gitweb/?p=native_client/src/native_client.git;a=commitdiff;h=c0a9077e3a604521f67f35c9bec3be031dd21465

BUG=394497

Review URL: https://codereview.chromium.org/391343006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@283668 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Change PPB_VideoDecoder::Reset behavior so plugin always manages textures.
bbudge@chromium.org [Wed, 16 Jul 2014 20:24:01 +0000 (20:24 +0000)]
Pepper: Change PPB_VideoDecoder::Reset behavior so plugin always manages textures.
Reset now will allow plugin to call RecyclePicture. The plugin must recycle any
texture it receives now. This makes plugin code simpler, since it can treat all
textures it receives in the same way (no special Reset handling).

Also fix a bug where Reset didn't return textures to the hardware decoder, which
already has this behavior.

BUG=281689

Review URL: https://codereview.chromium.org/390213002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@283483 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoShorten gyp target names for PNaCl IRT shim
dschuff@chromium.org [Wed, 16 Jul 2014 15:51:46 +0000 (15:51 +0000)]
Shorten gyp target names for PNaCl IRT shim

These are basically the longest generated names, and the trybots appear
to have shorter overall names than the buildbots, because try jobs pass
and then the builders break.

R=jvoung@chromium.org
TBR=binji@chromium.org, teravest@chromium.org
BUG= https://code.google.com/p/nativeclient/issues/detail?id=3846

Review URL: https://codereview.chromium.org/399523002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@283402 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoRemove return value for PpapiHostMsg_PPBVar_AddRefObject
jbauman@chromium.org [Mon, 14 Jul 2014 23:55:37 +0000 (23:55 +0000)]
Remove return value for PpapiHostMsg_PPBVar_AddRefObject

It's unused and not necessary.

BUG=391025

Review URL: https://codereview.chromium.org/380423002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@283090 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Clarify end-of-bitcode-stream behavior.
teravest@chromium.org [Fri, 11 Jul 2014 23:36:35 +0000 (23:36 +0000)]
Pepper: Clarify end-of-bitcode-stream behavior.

The PutBytes() method in PnaclTranslateThread is used for two purposes: giving
data to the translator processes and signaling the end of the stream.  The way
it's currently done is a bit confusing; this breaks the method into two
for clarity. This is to make code review easier to read for an upcoming change
which removes FileDownloader from the trusted plugin.

BUG=370556

Review URL: https://codereview.chromium.org/383213005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282734 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Make sure 'out' params are initialized for all code paths.
bbudge@chromium.org [Fri, 11 Jul 2014 21:00:15 +0000 (21:00 +0000)]
Pepper: Make sure 'out' params are initialized for all code paths.
Change PPB_ImageData_Proxy and PPB_Instance_Proxy to always initialize
'out' params, to avoid sending uninitialized memory from renderer to
plugin.

BUG=391023

Review URL: https://codereview.chromium.org/383143003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282692 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Add GL_TEXTURE_EXTERNAL_OES texture target to PPB_VideoDecoder docs and example.
bbudge@chromium.org [Fri, 11 Jul 2014 20:01:47 +0000 (20:01 +0000)]
Pepper: Add GL_TEXTURE_EXTERNAL_OES texture target to PPB_VideoDecoder docs and example.

GL_TEXTURE_2D is used by x86 non-OSX platforms and by software fallback.
GL_TEXTURE_RECTANGLE_ARB will be used on OSX hardware.
GL_TEXTURE_EXTERNAL_OES is used on ARM ChromeOS.

BUG=281689

Review URL: https://codereview.chromium.org/354763003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282674 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months ago[NaCl SDK] docs: Remove references to obsolete debugging env vars.
jvoung@chromium.org [Fri, 11 Jul 2014 18:08:44 +0000 (18:08 +0000)]
[NaCl SDK] docs: Remove references to obsolete debugging env vars.

Make also complained about some broken links and stopped w/
warnings as errors.

I assume a bunch these went away w/ the really old proxy switch =)

BUG=266395
NOTRY=true

Review URL: https://codereview.chromium.org/383523004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282653 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months ago[PPAPI] Make the NaCl SDK presubmit check a warning instead of an error.
binji@chromium.org [Fri, 11 Jul 2014 16:43:57 +0000 (16:43 +0000)]
[PPAPI] Make the NaCl SDK presubmit check a warning instead of an error.

There are some PPAPI dev interfaces that should not be added to the SDK, so it
is better to warn instead of making it a presubmit error.

BUG=none
R=teravest@chromium.org
TEST=made change to ppapi/cpp/buffer_dev.h, ran git cl presubmit

Review URL: https://codereview.chromium.org/379923010

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282637 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoEnsures that input Pepper Flash supports the newly added AudioBus interface.
henrika@chromium.org [Thu, 10 Jul 2014 16:44:01 +0000 (16:44 +0000)]
Ensures that input Pepper Flash supports the newly added AudioBus interface.

This CL (https://codereview.chromium.org/344583002) has modified the data structure in shared memory and the original CL did not take the input side of Pepper Flash into account. Pepper Flash recording was therefore broken. This CL ensures that Pepper knows about the new audio structure and accounts for it using by mapping shared memory to an audio bus and then interleaves the data into a plain byte vector before sending it to the Pepper callback.

Adding vrk@ as TBR for the added dependency in media.

R=piman@chromium.org, xians@chromium.org
TBR=vrk
BUG=387303
TEST=www.youtube.com/my_webcam

Review URL: https://codereview.chromium.org/371273004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282357 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Initialize PP_ImageDataDesc when creating ImageData.
bbudge@chromium.org [Thu, 10 Jul 2014 02:54:11 +0000 (02:54 +0000)]
Pepper: Initialize PP_ImageDataDesc when creating ImageData.
Clears memory so we don't send unitialized memory to the
renderer.

BUG=391023

Review URL: https://codereview.chromium.org/383543004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282222 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Small PnaclResources cleanup
teravest@chromium.org [Wed, 9 Jul 2014 23:39:22 +0000 (23:39 +0000)]
Pepper: Small PnaclResources cleanup

This fixes a place where the wrong URL is used for the ld process when
performing PNaCl translation. I don't believe the effect of this is severe,
though it will cause the wrong url to be displayed when it's running.

This also removes an unnecessary string constant usage in the trusted plugin.

BUG=

Review URL: https://codereview.chromium.org/373003005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282176 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoFixes for re-enabling more MSVC level 4 warnings: ppapi/ edition
pkasting@chromium.org [Wed, 9 Jul 2014 21:41:02 +0000 (21:41 +0000)]
Fixes for re-enabling more MSVC level 4 warnings: ppapi/ edition

This contains fixes for the following sorts of issues:
* Assignment inside conditional
* Possibly-uninitialized local variable
* Signedness mismatch

This also contains a small number of other cleanups/simplifications to nearby
code.

BUG=81439
TEST=none
R=teravest@chromium.org

Review URL: https://codereview.chromium.org/375133003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@282133 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoMove webpreferences.* from webkit/ to content/
tfarina@chromium.org [Wed, 9 Jul 2014 06:25:05 +0000 (06:25 +0000)]
Move webpreferences.* from webkit/ to content/

BUG=338338
TEST=None
R=jam@chromium.org
TBR=boliu@chromium.org # for android_webview
TBR=bauerb@chromium.org # for components/plugins
TBR=dmichael@chromium.org # for ppapi

Review URL: https://codereview.chromium.org/357203003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@281978 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Narrow dev channel whitelisting behavior.
teravest@chromium.org [Mon, 7 Jul 2014 21:30:21 +0000 (21:30 +0000)]
Pepper: Narrow dev channel whitelisting behavior.

An earlier change whitelisted all "dev channel" Pepper APIs for two select
application origins. This change narrows the whitelisting behavior to only
grant access to the Compositor and VideoDecode APIs. This makes this instance
of whitelisting easier to remove in the future.

BUG=386743

Review URL: https://codereview.chromium.org/346263006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@281588 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Remove program_url in trusted plugin.
teravest@chromium.org [Thu, 3 Jul 2014 00:22:03 +0000 (00:22 +0000)]
Pepper: Remove program_url in trusted plugin.

This reduces the amount of state in the trusted plugin (slightly).

BUG=239656

Review URL: https://codereview.chromium.org/367153004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@281127 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoAdd support for strong future sync points.
jbauman@chromium.org [Tue, 1 Jul 2014 00:27:22 +0000 (00:27 +0000)]
Add support for strong future sync points.

If InsertFutureSyncPoint() is used (this is only allowed from the browser process), the sync point will not be retired automatically but can be retired later manually with RetireSyncPoint. Any command buffer that waits on it before it's retired can wait indefinitely.

BUG=365454

Review URL: https://codereview.chromium.org/284233008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280708 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: SelLdrLauncherChrome cleanup.
teravest@chromium.org [Mon, 30 Jun 2014 22:56:28 +0000 (22:56 +0000)]
Pepper: SelLdrLauncherChrome cleanup.

The indirection provided by SelLdrLauncherChrome::Start() isn't much help.
Instead, we just expose the channel so that ServiceRuntime can pass it to
LaunchSelLdr directly.

BUG=239656

Review URL: https://codereview.chromium.org/360743002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280686 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoEnable mmap and identity-based validation caching on pnacl-{llc,ld}.nexe
jvoung@chromium.org [Mon, 30 Jun 2014 16:18:53 +0000 (16:18 +0000)]
Enable mmap and identity-based validation caching on pnacl-{llc,ld}.nexe

Register the executables w/ the validation cache on open and hook up the
file info just like the other NaCl module loads.

The non-executable files are still opened without the windows-specific
FLAG_EXEC.

BUG= https://code.google.com/p/nativeclient/issues/detail?id=3614
TEST= ValidationCacheOfTranslatorNexes

Review URL: https://codereview.chromium.org/356923004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280605 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Remove PPP_ManifestService.
teravest@chromium.org [Fri, 27 Jun 2014 02:07:29 +0000 (02:07 +0000)]
Pepper: Remove PPP_ManifestService.

Now that enough logic has been moved out of the trusted plugin, we can get rid
of PPP_ManifestService and make the code in the Non-SFI case much clearer.

CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_rel_precise32
BUG=239656

Review URL: https://codereview.chromium.org/356053004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280204 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoNaCl: begin reducing ServiceRuntime's dependance on Plugin in the trusted plugin.
ncbray@chromium.org [Fri, 27 Jun 2014 01:41:42 +0000 (01:41 +0000)]
NaCl: begin reducing ServiceRuntime's dependance on Plugin in the trusted plugin.

BUG=

Review URL: https://codereview.chromium.org/342553004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280197 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoAdd preliminary support for UBsan's vptr in Chromium (-fsanitize=vptr).
byoungyoung@google.com [Thu, 26 Jun 2014 23:31:40 +0000 (23:31 +0000)]
Add preliminary support for UBsan's vptr in Chromium (-fsanitize=vptr).

Can be enabled using ubsan_vptr=1, but not working right now due to
1) Clang's missing library and 2) Clang's link bug.

BUG=174801

TBR=kbr@chromium.org, mseaborn@chromium.org
R=glider@chromium.org, inferno@chromium.org

Review URL: https://codereview.chromium.org/357753002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280162 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoMove PPB_TrueTypeFont_Dev host from renderer to browser.
bbudge@chromium.org [Thu, 26 Jun 2014 21:28:18 +0000 (21:28 +0000)]
Move PPB_TrueTypeFont_Dev host from renderer to browser.

On Linux, we can now use FontConfig in the browser process since it
is thread-safe, so we don't need to use SandboxIPC from the renderer.
In the browser, the needed functionality is split off from
SandboxIPCHandler into a static function,
MatchFontFaceWithFallback in a new pair of files, font_utils_linux.*.

This change also moves any potentially blocking font creation and
reading to the browser's blocking thread pool.

I reworked the PepperTrueTypeFont base class to make Create simpler.
- Added an Initialize method which returns the font descriptor.
- Removed the Describe method.

I reworked the resource to delay calls to the host until we receive the desc
and initialization is complete. Describe will now wait until Initialize completes,
and the host uses a SequencedTaskRunner to serialize tasks, so Initialize
completes before GetTableTags and GetTable calls complete.

The Describe method can be implemented without IPC since we have the desc
on the plugin side.

BUG=382729

Review URL: https://codereview.chromium.org/337203003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280107 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Move Pnacl init time out of trusted plugin.
teravest@chromium.org [Thu, 26 Jun 2014 20:08:31 +0000 (20:08 +0000)]
Pepper: Move Pnacl init time out of trusted plugin.

This change moves the pnacl init time out of the trusted plugin. It's part of a
series of changes to get rid of the trusted plugin entirely.

BUG=239656
R=bbudge@chromium.org

Review URL: https://codereview.chromium.org/356923002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280082 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Remove method in PluginReverseInterface.
teravest@chromium.org [Thu, 26 Jun 2014 20:03:04 +0000 (20:03 +0000)]
Pepper: Remove method in PluginReverseInterface.

CloseManifestEntry has been removed from nacl::ReverseInterface, so we don't
need this empty method anymore.

BUG=
R=bbudge@chromium.org

Review URL: https://codereview.chromium.org/357863002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280079 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoExport ResourceMessageParams::SerializedHandles
hans@chromium.org [Thu, 26 Jun 2014 17:36:39 +0000 (17:36 +0000)]
Export ResourceMessageParams::SerializedHandles

The inner class needs to be exported in case any uses in inline methods
or vtable of the outer class causes it to potentially be referenced across
a shared library boundary.

On non-Windows, the visibility attribute is already inherited by nested
classes, but on Windows, they must be marked explicitly.

BUG=82385
TEST=use Clang to do component release build of chrome on Windows

Review URL: https://codereview.chromium.org/359503003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@280038 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoTidy up stray semicolons and missing include guard in message files.
tsepez@chromium.org [Thu, 26 Jun 2014 06:08:40 +0000 (06:08 +0000)]
Tidy up stray semicolons and missing include guard in message files.

These manifested when I was making an ipc utility for my own use. They
are wrong even if they don't break the builders.

R=jam@chromium.org

Review URL: https://codereview.chromium.org/354493007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@279919 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Refactor OpenManifestEntry.
teravest@chromium.org [Wed, 25 Jun 2014 22:01:20 +0000 (22:01 +0000)]
Pepper: Refactor OpenManifestEntry.

This change pulls out more OpenManifestEntry code from service_runtime, making
PPB_NaCl_Private a smallter interface, and preparing for pulling all of that
logic out of the trusted plugin.

An earlier version of this change was reverted after breaking Non-SFI mode, but
the Non-SFI case no longer uses this codepath.

CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_rel_precise32
BUG=239656

Review URL: https://codereview.chromium.org/357773002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@279832 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoNaCl: clean up nexe loading logic in trusted plugin.
ncbray@chromium.org [Wed, 25 Jun 2014 18:35:15 +0000 (18:35 +0000)]
NaCl: clean up nexe loading logic in trusted plugin.

CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_rel_precise32
BUG=none

Review URL: https://codereview.chromium.org/338353008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@279762 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPepper: Remove callback for resource loading.
teravest@chromium.org [Tue, 24 Jun 2014 20:47:07 +0000 (20:47 +0000)]
Pepper: Remove callback for resource loading.

Using CompletionCallbacks for PNaClResources::StartLoad and
PNaClResources::ReadResourceInfo doesn't accomplish much and makes
the code more complicated. This is in preparation for a larger
change to get rid of FileDownloader for streaming pexe downloads.

BUG=239656

Review URL: https://codereview.chromium.org/358443002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@279485 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoRemove bogus include paths from NaCl gyp files.
sergeyu@chromium.org [Tue, 24 Jun 2014 00:12:43 +0000 (00:12 +0000)]
Remove bogus include paths from NaCl gyp files.

BUG=384752

Review URL: https://codereview.chromium.org/336023003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@279241 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoPpapi: Avoid entering resource three times in BindGraphics().
penghuang@chromium.org [Mon, 23 Jun 2014 23:34:04 +0000 (23:34 +0000)]
Ppapi: Avoid entering resource three times in BindGraphics().

BUG=None

Review URL: https://codereview.chromium.org/355493002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@279227 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 months agoRemove LoadModule SRPC for non-SFI mode.
hidehiko@chromium.org [Mon, 23 Jun 2014 09:44:08 +0000 (09:44 +0000)]
Remove LoadModule SRPC for non-SFI mode.

This CL removes LoadModule SRPC for non-SFI mode. Instead, non-SFI mode
starts to pass nexe file to LaunchSelLdr, which is eventually passed to
NaClListener::OnStart().

TEST=Ran browser_tests --gtest_filter=*NonSfi*:*NonSFI* locally and trybots
BUG=333950
CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_rel_precise32

Review URL: https://codereview.chromium.org/337463002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@279069 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPepper: Simplify OpenResource() for Non-SFI.
teravest@chromium.org [Fri, 20 Jun 2014 23:31:34 +0000 (23:31 +0000)]
Pepper: Simplify OpenResource() for Non-SFI.

The Non-SFI implementation of OpenResource is pretty complicated. This is
because the logic to support that operation was entirely in the trusted
plugin. Now, we can perform the necessary logic entirely in Chromium, so
it can be made much simpler.

CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_rel_precise32
BUG=239656

Review URL: https://codereview.chromium.org/339213003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278859 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months ago[PPAPI] Add browser tests for compositor API
penghuang@chromium.org [Fri, 20 Jun 2014 18:42:39 +0000 (18:42 +0000)]
[PPAPI] Add browser tests for compositor API

 And fix a bug found with the tests.
 BindGraphics() does not work for a device which is in the same type with the current bound device.

BUG=374383
R=piman@chromium.org, raymes@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=278728

Review URL: https://codereview.chromium.org/324983005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278779 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoRevert 278728 "[PPAPI] Add browser tests for compositor API"
yzshen@chromium.org [Fri, 20 Jun 2014 18:08:31 +0000 (18:08 +0000)]
Revert 278728 "[PPAPI] Add browser tests for compositor API"

> [PPAPI] Add browser tests for compositor API
>
>  And fix a bug found with the tests.
>  BindGraphics() does not work for a device which is in the same type with the current bound device.
>
> BUG=374383
> R=piman@chromium.org, raymes@chromium.org
>
> Review URL: https://codereview.chromium.org/324983005

TBR=penghuang@chromium.org

Review URL: https://codereview.chromium.org/342323006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278765 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months ago[PPAPI] Add browser tests for compositor API
penghuang@chromium.org [Fri, 20 Jun 2014 17:00:46 +0000 (17:00 +0000)]
[PPAPI] Add browser tests for compositor API

 And fix a bug found with the tests.
 BindGraphics() does not work for a device which is in the same type with the current bound device.

BUG=374383
R=piman@chromium.org, raymes@chromium.org

Review URL: https://codereview.chromium.org/324983005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278728 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPepper: Whitelist dev channel APIs for some apps.
hamaji@chromium.org [Fri, 20 Jun 2014 07:51:42 +0000 (07:51 +0000)]
Pepper: Whitelist dev channel APIs for some apps.

Some applications need access to dev channel APIs before they will
become stable. As written, this change only performs whitelisting for NaCl
apps; I don't think we have any need for enabling this in the non-NaCl case.

I need to find a good way to test this.

BUG=383937
R=dmichael@chromium.org, jln@chromium.org, piman@chromium.org

Review URL: https://codereview.chromium.org/338523007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278645 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPPAPI: Move PPB_TraceEvent to have dev-channel permissions
dmichael@chromium.org [Thu, 19 Jun 2014 21:53:39 +0000 (21:53 +0000)]
PPAPI: Move PPB_TraceEvent to have dev-channel permissions

PPB_TraceEvent should be available to developers who want to profile
their application on Dev channel or Canary.

BUG=386636

Review URL: https://codereview.chromium.org/349443002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278499 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoQuiet output when untarring the pnacl translator from gyp builds.
dyen@chromium.org [Thu, 19 Jun 2014 19:05:24 +0000 (19:05 +0000)]
Quiet output when untarring the pnacl translator from gyp builds.

BUG= https://code.google.com/p/chromium/issues/detail?id=370981
TEST= Run "ninja -C out/Debug untar_pnacl_translator"

Review URL: https://codereview.chromium.org/341893003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278451 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoRemoved ManagedMemoryStats.
vmpstr@chromium.org [Thu, 19 Jun 2014 07:01:41 +0000 (07:01 +0000)]
Removed ManagedMemoryStats.

As of r277056, the gpu memory manager is no longer using
managed memory stats produced. This patch removes the managed
memory stats calculation from the tile manager. As well, this
patch removes ununsed gpu managed memory stats plumbing.

BUG=377065

Review URL: https://codereview.chromium.org/342483007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278294 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPepper: Add latency calculation to ppapi/examples/video_decode.
bbudge@chromium.org [Thu, 19 Jun 2014 06:29:43 +0000 (06:29 +0000)]
Pepper: Add latency calculation to ppapi/examples/video_decode.

Also, flip picture vertically (it was upside down).

BUG=281689

Review URL: https://codereview.chromium.org/340163002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278281 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoAdd GetScrollOffset function to PPB_View
raymes@chromium.org [Thu, 19 Jun 2014 01:48:14 +0000 (01:48 +0000)]
Add GetScrollOffset function to PPB_View

This adds a function to PPB_View which allows plugins to know the scroll offset
of the page when they are in view. This is useful for OOP PDF which uses the
scroll offset of the window it is contained in to determine the document's
scroll location. A web page can send scroll location via postMessage but
this is slow. Sending the offset directly via view messages is much faster
and seems reasonable.

We don't send the scroll offset in the cases where the plugin is off screen
to avoid any more additional IPC traffic than what currently exists.

BUG=303491

Review URL: https://codereview.chromium.org/329033003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278236 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months ago[PPAPI] Using ProduceTextureDirectCHROMIUM() to avoid binding and restoring GL_TEXTUR...
penghuang@chromium.org [Wed, 18 Jun 2014 20:48:42 +0000 (20:48 +0000)]
[PPAPI] Using ProduceTextureDirectCHROMIUM() to avoid binding and restoring GL_TEXTURE_2D target.

BUG=374383
R=bbudge@chromium.org, danakj@chromium.org

Review URL: https://codereview.chromium.org/341893004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278165 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPepper: Small cleanup in PNaClCoordinator.
teravest@chromium.org [Wed, 18 Jun 2014 19:19:48 +0000 (19:19 +0000)]
Pepper: Small cleanup in PNaClCoordinator.

This simplifies the case where we fail to open all the necessary temporary
files as part of PNaCl translation.

RunTranslate() isn't called if num_object_files_opened_ != split_module_count_
anyway, so this should be equivalent to the existing behavior.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/335343006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278143 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPPAPI: Implement synchronous postMessage
dmichael@chromium.org [Wed, 18 Jun 2014 16:44:00 +0000 (16:44 +0000)]
PPAPI: Implement synchronous postMessage

BUG=367896

Review URL: https://codereview.chromium.org/264303002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278102 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPossible narrowing conversion build fix for Compositor API example.
pgal.u-szeged@partner.samsung.com [Wed, 18 Jun 2014 15:08:23 +0000 (15:08 +0000)]
Possible narrowing conversion build fix for Compositor API example.

In the compositor API example gcc 4.9 reports a possible
type narrowing when using fabs call. To fix this use fabsf instead.

BUG=385988
R=dmichael

Review URL: https://codereview.chromium.org/335383003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@278076 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoRemove redundant use_nonsfi_mode from SelLdrStartParams.
hidehiko@chromium.org [Tue, 17 Jun 2014 21:13:26 +0000 (21:13 +0000)]
Remove redundant use_nonsfi_mode from SelLdrStartParams.

The params is used for ServiceRuntime::StartSelLdr(). ServiceRuntime
already knows if it is for nonsfi or not. So the use_nonsfi_mode in
SelLdrStartParams is just redundant. Remove it.

BUG=333950
TEST=Ran browser_tests --gtest_filter=*NonSfi*:*NonSFI* locally, and trybots.
CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_rel_precise32

Review URL: https://codereview.chromium.org/332333003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277864 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPepper: Add VP9 support to PPB_VideoDecoder API.
bbudge@chromium.org [Tue, 17 Jun 2014 20:40:10 +0000 (20:40 +0000)]
Pepper: Add VP9 support to PPB_VideoDecoder API.
Improve documentation for Reset and Flush.
Use VpxVideoDecoder in implementation code.

BUG=281689

Review URL: https://codereview.chromium.org/336833003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277860 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoFix race condition on ManifestService initialization.
hidehiko@chromium.org [Tue, 17 Jun 2014 20:06:29 +0000 (20:06 +0000)]
Fix race condition on ManifestService initialization.

SyncMessageFilter::Send() returns false immediately, if the IPC connection
is not yet established. As connecting is done asynchronously, there is
no guarantee that the connection is established on the first Send() invocation.
By this CL, Send() blocks the caller thread if the connection is not yet
established.
Note that currently the ratio should be probably low, because there are some
more initialization steps between the ManifestService creation and the first
Send() invocation. We're switching to changing the initialization procedure,
and then this race would be hit more easily.

TEST=Ran browser_tests --gtest_filter=*NonSfi*:*NonSFI* locally, and trybots. Also, locally modified the code to delay OnFilterAdded with and without this CL, and made sure this CL works well.
BUG=333950
CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_rel_precise32

Review URL: https://codereview.chromium.org/334593004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277840 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoRemove --disable-extensions-resource-whitelist.
pkasting@chromium.org [Tue, 17 Jun 2014 02:26:26 +0000 (02:26 +0000)]
Remove --disable-extensions-resource-whitelist.

BUG=357720
TEST=none
R=kalman@chromium.org

Review URL: https://codereview.chromium.org/324793002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277625 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPepper: Fix texture management in VideoDecoderShim on Reset.
bbudge@chromium.org [Mon, 16 Jun 2014 22:01:20 +0000 (22:01 +0000)]
Pepper: Fix texture management in VideoDecoderShim on Reset.

- Use a hash_set to track available textures. This prevents weird behavior
if the plugin recycles a texture twice.
- Fix Reset, so all textures are made available on completion.
- Fix the plugin, which had a bug that allowed pictures to jump the queue
and didn't behave correctly on Reset.

BUG=281689

Review URL: https://codereview.chromium.org/337743003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277548 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoRemove apps in Pepper support and the PPB_Alarms_Dev API.
yzshen@chromium.org [Mon, 16 Jun 2014 19:09:38 +0000 (19:09 +0000)]
Remove apps in Pepper support and the PPB_Alarms_Dev API.

BUG=366304
TEST=None

Review URL: https://codereview.chromium.org/314823002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277513 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPepper: Move more UMA stuff out of trusted plugin.
teravest@chromium.org [Mon, 16 Jun 2014 14:22:19 +0000 (14:22 +0000)]
Pepper: Move more UMA stuff out of trusted plugin.

This is yet another pretty simple change just lifting more lines out of the
trusted plugin. I added a method to PPB_NaCl_Private that's safe to call from
a background thread (there are other methods that are safe as well) to keep the
call sites simple.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/331973002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277424 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months ago[PPAPI] Compositor API implementation.
penghuang@chromium.org [Mon, 16 Jun 2014 14:20:26 +0000 (14:20 +0000)]
[PPAPI] Compositor API implementation.

Implement the compositor API which allows a plugin to combine different sources of visual data efficiently, such as PPB_ImageData and OpengGL texture.

API Proposal http://goo.gl/V7xcu3

BUG=374383

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=277208

Review URL: https://codereview.chromium.org/298023004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277422 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoRevert 277208 "[PPAPI] Compositor API implementation."
tzik@chromium.org [Mon, 16 Jun 2014 07:21:26 +0000 (07:21 +0000)]
Revert 277208 "[PPAPI] Compositor API implementation."

This CL seems to cause perf bot failure.
The log said nacl_helper-data size gained exceeds expectation.

http://build.chromium.org/p/chromium/builders/Linux/builds/50560
https://chromeperf.appspot.com/report?masters=Chromium&bots=chromium-rel-linux&tests=sizes%2Fnacl_helper-data&rev=277212&checked=core

> [PPAPI] Compositor API implementation.
>
> Implement the compositor API which allows a plugin to combine different sources of visual data efficiently, such as PPB_ImageData and OpengGL texture.
>
> API Proposal http://goo.gl/V7xcu3
>
> BUG=374383
>
> Review URL: https://codereview.chromium.org/298023004

TBR=penghuang@chromium.org

Review URL: https://codereview.chromium.org/331123003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277372 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months ago[PPAPI] Compositor API implementation.
penghuang@chromium.org [Sat, 14 Jun 2014 08:41:25 +0000 (08:41 +0000)]
[PPAPI] Compositor API implementation.

Implement the compositor API which allows a plugin to combine different sources of visual data efficiently, such as PPB_ImageData and OpengGL texture.

API Proposal http://goo.gl/V7xcu3

BUG=374383

Review URL: https://codereview.chromium.org/298023004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277208 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoPepper: Refactor some coordinator histograms.
teravest@chromium.org [Fri, 13 Jun 2014 19:40:43 +0000 (19:40 +0000)]
Pepper: Refactor some coordinator histograms.

We have a bunch of histograms that log data at the same time as
ReportTranslationFinished(), so we may as well move some of that logic out of
the trusted plugin. We use a different time range for translation for other
metrics, so I had to add a "HistogramTimeTranslate" method in addition to the
existing ones for "small", "medium", and "large" times.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/307403005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277073 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 months agoRevert of Revert of Implement software fallback for PPB_VideoDecoder. (https://codere...
schenney@chromium.org [Fri, 13 Jun 2014 15:00:14 +0000 (15:00 +0000)]
Revert of Revert of Implement software fallback for PPB_VideoDecoder. (https://codereview.chromium.org/337683002/)

Reason for revert:
Revert of revert, which apparently wasn't a full revert.

Original issue's description:
> Revert of Implement software fallback for PPB_VideoDecoder. (https://codereview.chromium.org/311853005/)
>
> Reason for revert:
> Broke blink Linux tests compile.http://build.chromium.org/p/chromium.webkit/builders/Linux%20Tests/builds/37259
>
> Original issue's description:
> > Implement software fallback for PPB_VideoDecoder.
> > This modifies the proxy to implement software fallback mode.
> > The main change is to the host, which now can work with
> > media::VideoDecoders.
> >
> > media::VideoDecoder works differently from media::VideoDecodeAccelerator
> > so an adapter object, content::VideoDecoderShim is defined. It lives on the main thread and drives the actual decoder on the media thread via a child DecoderImpl class, which sends back frames of video. VideoDecoderShim receives those and converts frames to GL textures.
> >
> > gpu::Mailboxes are used so the host can create textures that are aliased
> > to the plugin's textures.
> >
> > The test plugin has been changed to include bitstream data for VP8 in order to
> > test the software decoder. The data is in ppapi/examples/video_decode/testdata.h
> > alongside the H264 data. The file diff is too large for this site but is structured
> > something like this:
> >
> > const unsigned char kData[] = {
> > #if defined USE_VP8_TESTDATA_INSTEAD_OF_H264
> > ...  lots of VP8 data
> >
> > #else  // !USE_VP8_TESTDATA_INSTEAD_OF_H264
> > ...  lots of H264 data
> >
> > #endif  // USE_VP8_TESTDATA_INSTEAD_OF_H264
> > };
> >
> >
> > There is a TODO to convert the example to load a file. I'm not sure how to go
> > about that but am willing to do the work if someone can point the way.
> >
> > BUG=281689
> > R=dmichael@chromium.org, fischman@chromium.org, sievers@chromium.org, tsepez@chromium.org
> >
> > Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=277012
>
> TBR=dmichael@chromium.org,fischman@chromium.org,igorc@chromium.org,piman@chromium.org,sievers@chromium.org,tsepez@chromium.org,bbudge@chromium.org
> NOTREECHECKS=true
> NOTRY=true
> BUG=281689
>
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=277015

TBR=dmichael@chromium.org,fischman@chromium.org,igorc@chromium.org,piman@chromium.org,sievers@chromium.org,tsepez@chromium.org,bbudge@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=281689

Review URL: https://codereview.chromium.org/333903002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@277020 4ff67af0-8c30-449e-8e8b-ad334ec8d88c