chromium/src/ppapi.git
2 hours agoPepper: Cleanup PNaCl support in trusted plugin. master
teravest@chromium.org [Fri, 25 Apr 2014 03:08:27 +0000 (03:08 +0000)]
Pepper: Cleanup PNaCl support in trusted plugin.

This change cleans up some code in pnacl_coordinator and pnacl_resources to
remove code that can't be reached and simplify error reporting paths. This is
in preparation for moving all of this code out of the trusted plugin.

BUG=239656

Review URL: https://codereview.chromium.org/257463005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@266106 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 hours agoMove IPC::MessageFilter and router to a separate file
dmichael@chromium.org [Fri, 25 Apr 2014 00:07:30 +0000 (00:07 +0000)]
Move IPC::MessageFilter and router to a separate file

There are no changes in implementation in this CL. This is in preparation
for making IPC::Channel support filters on the Channel's thread.

BUG=364241
TBR=cpu@chromium.org,nduca@chromium.org

cpu: OWNERS for win8
nduca: OWNERS for components/tracing

Review URL: https://codereview.chromium.org/245443005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@266057 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 hours agoPepper: Move IME workaround from the trusted plugin.
teravest@chromium.org [Thu, 24 Apr 2014 23:43:26 +0000 (23:43 +0000)]
Pepper: Move IME workaround from the trusted plugin.

This moves a MACOSX-specific IME workaround outside the trusted plugin as part
of the effort to refactor all that code to components/nacl.

BUG=239656

Review URL: https://codereview.chromium.org/250063003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@266050 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

11 hours agoPepper: Move attributes out of the trusted plugin.
teravest@chromium.org [Thu, 24 Apr 2014 18:29:27 +0000 (18:29 +0000)]
Pepper: Move attributes out of the trusted plugin.

This moves attribute parsing and storage out of the trusted plugin.

There's some additional cleanup in here as well:
  * Unused fields in Plugin are removed.
  * Stale/unhelpful comments are dropped in the move.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/249083006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@265964 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

37 hours agoNon-SFI NaCl: Fix the crash for the AudioShared.
hidehiko@chromium.org [Wed, 23 Apr 2014 16:11:39 +0000 (16:11 +0000)]
Non-SFI NaCl: Fix the crash for the AudioShared.

It turned out that even for the non-SFI mode, we still need
to use PPB_AudioShared::SetThreadFunctions, although the
underlying function is pthread_create for both.
Otherwise the TLS in IRT for the created thread is not
initialized correctly.

BUG=359710
TEST=Run trybots.

Review URL: https://codereview.chromium.org/200283002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@265663 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

41 hours agoImplement ManifestService::StartupInitializationComplete().
hidehiko@chromium.org [Wed, 23 Apr 2014 11:53:13 +0000 (11:53 +0000)]
Implement ManifestService::StartupInitializationComplete().

This CL introduces StartupInitializationComplete() message between a plugin
and the renderer, focusing on non-SFI mode.
Currently PluginReverseInterface::StartupInitializationComplete() is called
directly in ServiceRuntime::InitReverseService in non-SFI mode. However, it
is too early. Since it blocks the renderer's main thread by sync PPAPI IPC
message, when we support the open_resource() for non-SFI mode, it would cause
a deadlock issue.
This CL delays the invocation of StartupInitializationComplete until when
PPAPI IPC channel is ready in plugin side to follow NaCl's manner in SFI-mode.

TBR=jln@chromium.org
TEST=Ran browser_tests --gtest_filter=*NonSfi* locally, and ran trybots.
BUG=358431

Review URL: https://codereview.chromium.org/238353016

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@265611 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

46 hours agoRevert of SDK building script now uses package_version to extract. (https://coderevie...
hidehiko@chromium.org [Wed, 23 Apr 2014 07:05:22 +0000 (07:05 +0000)]
Revert of SDK building script now uses package_version to extract. (https://codereview.chromium.org/235603004/)

Reason for revert:
This update breaks PPAPI*NonSfi tests on linux-32bits bot. See also crbug.com/365817

Original issue's description:
> NaCl: Update revision in DEPS, r13018 -> r13062.
>
> Chromium NaCl scripts now uses package_version to extract.
>
> Now that we have split up the toolchain into multiple tars, we should
> use the package versioning script to manage which tars to extract.
> Unfortunately, currently the SDK uses different directory names compared
> to what the NaCl toolchain directories normally use, so we also must map
> and move the toolchain directories to the SDK ones.
>
> TEST= trybots
>
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=265303

TBR=noelallen@chromium.org,bradnelson@chromium.org,jvoung@chromium.org,mseaborn@chromium.org,dyen@chromium.org
NOTREECHECKS=true
NOTRY=true

Review URL: https://codereview.chromium.org/247143007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@265559 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 days agoGPU/Pepper: Always return -1 for ID when we return a NULL buffer.
epenner@chromium.org [Wed, 23 Apr 2014 02:07:54 +0000 (02:07 +0000)]
GPU/Pepper: Always return -1 for ID when we return a NULL buffer.

In this particular case we could return NULL but also return
a valid ID. This can result in a crash if the caller looks at
the ID returned and assumes the buffer is also valid.

BUG=365373

Review URL: https://codereview.chromium.org/248193004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@265499 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 days agoPepper: Move some manifest processing for refactor.
teravest@chromium.org [Tue, 22 Apr 2014 22:42:08 +0000 (22:42 +0000)]
Pepper: Move some manifest processing for refactor.

This moves some of the manifest processing logic outside the trusted plugin.
This reduces the number of things we need to expose in PPB_NaCl_Private.

BUG=239656

Review URL: https://codereview.chromium.org/243353002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@265394 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 days agoNaCl: Update revision in DEPS, r13018 -> r13062.
dyen@chromium.org [Tue, 22 Apr 2014 18:28:04 +0000 (18:28 +0000)]
NaCl: Update revision in DEPS, r13018 -> r13062.

Chromium NaCl scripts now uses package_version to extract.

Now that we have split up the toolchain into multiple tars, we should
use the package versioning script to manage which tars to extract.
Unfortunately, currently the SDK uses different directory names compared
to what the NaCl toolchain directories normally use, so we also must map
and move the toolchain directories to the SDK ones.

TEST= trybots

Review URL: https://codereview.chromium.org/235603004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@265303 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 days agoPepper: Drop open time tracking from FileDownloader.
teravest@chromium.org [Mon, 21 Apr 2014 22:43:32 +0000 (22:43 +0000)]
Pepper: Drop open time tracking from FileDownloader.

This change reduces the size of the FileDownloader interface as part of moving
FileDownloader logic out of the trusted plugin.

BUG=239656

Review URL: https://codereview.chromium.org/245483002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@265102 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 days agoPepper: Remove FileDownloader DOWNLOAD_TO_BUFFER.
teravest@chromium.org [Mon, 21 Apr 2014 15:34:40 +0000 (15:34 +0000)]
Pepper: Remove FileDownloader DOWNLOAD_TO_BUFFER.

The semantics for downloading a NaCl manifest specified for a data URL are
pretty simple. This change adds an interface to parse a data URL to
PPB_NaCl_Private, and uses that instead of FileDownloader, simplifying that
class.

R=dmichael@chromium.org
TBR=eroman
BUG=239656

Review URL: https://codereview.chromium.org/242803004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@264999 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 days agoAdd test to make sure if PPB_Audio_Shared::StartThread() works.
hidehiko@chromium.org [Mon, 21 Apr 2014 10:36:21 +0000 (10:36 +0000)]
Add test to make sure if PPB_Audio_Shared::StartThread() works.

To NaCl in either SFI or non-SFI mode, PPB_Audio_Shraed::StartThread requires
ppapi_register_thread_creator invoked in advance.
Currently no test about it. This CL adds the test.

Tests are guarded by #ifdef __native_client__, because it
won't work and is meaningless for trusted plugins.

TEST=Ran browser_tests --gtest_filter=*AudioThreadCreation locally, and ran trybots.
BUG=359710

Review URL: https://codereview.chromium.org/240523002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@264974 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

6 days agoPepper: Move URL resolves outside trusted plugin.
teravest@chromium.org [Fri, 18 Apr 2014 16:53:04 +0000 (16:53 +0000)]
Pepper: Move URL resolves outside trusted plugin.

This moves plugin_base_url and manifest_base_url outside the trusted plugin,
pulling more state over to NexeLoadManager.

I had hoped to clean up SetIsInstalled() as part of this change, but we don't
have precise semantics for what is_installed means (yet), so I can't change how
that behaves.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/241173006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@264807 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

7 days agoClean up Pepper file system resource unit test.
bbudge@chromium.org [Fri, 18 Apr 2014 02:09:52 +0000 (02:09 +0000)]
Clean up Pepper file system resource unit test.

BUG=none

Review URL: https://codereview.chromium.org/238983010

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@264707 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

7 days agoAdd IPC Channel for new ManifestService.
hidehiko@chromium.org [Thu, 17 Apr 2014 12:17:15 +0000 (12:17 +0000)]
Add IPC Channel for new ManifestService.

This CL adds a new IPC Channel between NaCl plugin and the renderer process
with introducing ManifestService (in the plugin) and ManifestServiceChannel
(in the renderer) as its end points.
Currently, ManifestService is just an empty service. Its functions will be
added in following CLs. The service will be used only for non-SFI mode
as a first step. On other platforms, IPC Channel will not be created.

TEST=Ran trybots.
BUG=358431

Review URL: https://codereview.chromium.org/231793003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@264477 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

8 days agoPepper: Move PnaclOptions outside trusted plugin.
teravest@chromium.org [Thu, 17 Apr 2014 00:25:47 +0000 (00:25 +0000)]
Pepper: Move PnaclOptions outside trusted plugin.

I ran into this while moving the Manifest code out of the trusted plugin and
figured this was big enough to stand on its own.

We have to create a PP_PNaClOptions struct instead of PnaclOptions so that the
objects of that type can be passed to functions in PPB_NaCl_Private.

BUG=239656

Review URL: https://codereview.chromium.org/235983020

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@264365 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

8 days agoNaCl: Remove tests for NACL_NAME_SERVICE_LIST.
teravest@chromium.org [Wed, 16 Apr 2014 17:20:29 +0000 (17:20 +0000)]
NaCl: Remove tests for NACL_NAME_SERVICE_LIST.

Support for the NACL_NAME_SERVICE_LIST is being removed from NaCl, as it was
only ever used by tests. This change removes some tests that use this SRPC message
signature.

BUG=none
R=mseaborn@chromium.org

Review URL: https://codereview.chromium.org/239453008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@264239 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

9 days agoAllow the plugin to set an empty tickmarks array to clear tickmarks
raymes@chromium.org [Wed, 16 Apr 2014 03:13:05 +0000 (03:13 +0000)]
Allow the plugin to set an empty tickmarks array to clear tickmarks

Fix a bug which prevented empty tickmarks from being set to an empty
array (which clears them).

BUG=303491

Review URL: https://codereview.chromium.org/236733002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@264094 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

9 days agoPepper: Move time logic out of the trusted plugin.
teravest@chromium.org [Tue, 15 Apr 2014 18:57:18 +0000 (18:57 +0000)]
Pepper: Move time logic out of the trusted plugin.

This moves more of the time values stored for UMA histograms out of the trusted
plugin code at ppapi/native_client to components/nacl/renderer.

BUG=239656

Review URL: https://codereview.chromium.org/236363009

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263954 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

9 days agoPepper: Fix MODULE_PRINTF behavior.
teravest@chromium.org [Tue, 15 Apr 2014 18:39:51 +0000 (18:39 +0000)]
Pepper: Fix MODULE_PRINTF behavior.

An earlier change to use VLOG() as the destination for pluging logging messages
caused problems with MODULE_PRINTF() statements when NACL_PLUGIN_DEBUG is set.

This change alters MODULE_PRINTF() and PLUGIN_PRINTF() to log to stderr when
the NaCl interface isn't available, and tightens up where the PPB_NaCl_Private
interface is set.

BUG=

Review URL: https://codereview.chromium.org/236943004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263942 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

9 days agoRemoved dependency of native_client/toolchain from chromium pnacl support.
dyen@chromium.org [Tue, 15 Apr 2014 18:29:29 +0000 (18:29 +0000)]
Removed dependency of native_client/toolchain from chromium pnacl support.

R=dschuff@chromium.org, jvoung@chromium.org
BUG= https://code.google.com/p/nativeclient/issues/detail?id=3816

Review URL: https://codereview.chromium.org/235923011

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263935 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

10 days agoMore removal of GTK code.
erg@chromium.org [Tue, 15 Apr 2014 00:45:28 +0000 (00:45 +0000)]
More removal of GTK code.

BUG=297026

Review URL: https://codereview.chromium.org/232773008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263757 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

10 days agoRefactor some ResourceMessageReply usages.
bbudge@chromium.org [Mon, 14 Apr 2014 21:51:57 +0000 (21:51 +0000)]
Refactor some ResourceMessageReply usages.
There are some usages where these are allocated on the heap.
Mechanically refactor the ones where the scoped_ptr is like a nullable reference by adding a bool pending field.
Hand rewrite some of the previous factoring in PepperExtensionsCommonMessageFilter.

BUG=none

Review URL: https://codereview.chromium.org/231883002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263730 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

13 days agoEncrypted Media: Support VP9 playback.
xhwang@chromium.org [Fri, 11 Apr 2014 23:33:52 +0000 (23:33 +0000)]
Encrypted Media: Support VP9 playback.

Also adds a browser test with encrypted VP9 video. The test video file is checked in a separate CL.

IsTypeSupported() doesn't work with VP9 yet, which will be fixed in a followup CL.

BUG=361318

Review URL: https://codereview.chromium.org/234813008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263407 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

13 days agoUsed condvar timed wait instead of indefinite wait for WaitForSelLdrStart.
jvoung@google.com [Fri, 11 Apr 2014 15:10:25 +0000 (15:10 +0000)]
Used condvar timed wait instead of indefinite wait for WaitForSelLdrStart.

There are various reasons that the condvar signal will never happen:
- We currently do a CallOnMainThread in StartSelLdrContinuation
  to invoke the callback that does the signaling. However, that
  CallOnMainThread might not get scheduled as we are tearing down
  the embed.
- The IPC reply from the browser -> renderer might get completely
  dropped as the embed is being destroyed, so we may never even
  get to StartSelLdrContinuation.

Why care about the condvar being blocked at all?

The ~Plugin is blocked on ~PnaclCoordinator, which is
blocked on trying to join the PnaclTranslateThread's thread. However
if the surfaway is "early enough" then PnaclTranslateThread's thread
may be waiting for SelLdrStart (the condvar).

Another reason that ~PnaclCoordinator on the main thread
could be blocked is that it attempts to acquire the
subprocess_mu_. However, the subprocess_mu_ is being held
by the translate thread while it is waiting on the condvar.
I tried shifting the lock acquisition until after the
process is started and to the point that the code is
actually trying to write to the shared field. However, that
doesn't help because there is still the thread join... and
we need the thread join because it protects against
the thread using fields of PnaclCoordinator while it's
being free'd.

Use timeout in background thread NaCl process startup condvar.
Don't wait too long though, because the renderer is hung
in the meantime.

There are other places we block indefinitely on condvars
(reverse service thread), which show up in crash reports
(see the BUG), but that's not addressed here.

BUG=350426
REPRO (but not TEST)= www/~jvoung/test_embed_kills/
not yet sure of how to make a test that isn't flaky, since you need to fuzz the surfaway time a bit to trigger (~150 ms) and the cond var timed-wait will make tests that do repro slow.

R=dschuff@chromium.org, teravest@chromium.org

Review URL: https://codereview.chromium.org/233443004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263247 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

13 days agoPepper: Remove NaClHttpResponseHeaders.
teravest@chromium.org [Fri, 11 Apr 2014 07:19:29 +0000 (07:19 +0000)]
Pepper: Remove NaClHttpResponseHeaders.

This change moves HTTP response header parsing to Chromium, removing the need
for having this logic in ppapi/native_client.

BUG=239656

Review URL: https://codereview.chromium.org/227913008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263168 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

13 days agoPepper: Make it possible to pass external file systems to NaCl modules.
thestig@chromium.org [Fri, 11 Apr 2014 07:18:32 +0000 (07:18 +0000)]
Pepper: Make it possible to pass external file systems to NaCl modules.

BUG=327171

Review URL: https://codereview.chromium.org/210593003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263166 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoPPAPI: Update pp_var.h documentation for Resource vars.
dmichael@chromium.org [Thu, 10 Apr 2014 23:58:41 +0000 (23:58 +0000)]
PPAPI: Update pp_var.h documentation for Resource vars.

BUG=361215
R=raymes@chromium.org

Review URL: https://codereview.chromium.org/229033003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263135 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoPPAPI: Move some of NexeFileDidOpen to NexeLoadManager
dmichael@chromium.org [Thu, 10 Apr 2014 20:59:32 +0000 (20:59 +0000)]
PPAPI: Move some of NexeFileDidOpen to NexeLoadManager

This also tickled a bug in the generator because I named a variable
with "http". Which made the generator tests run on presubmit. Which
tickled a bug in the tests :-). So those are fixed here as an aside.

BUG=239656

Review URL: https://codereview.chromium.org/231243002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263073 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoPepper: Remove Manifest::ResolveURL
teravest@chromium.org [Thu, 10 Apr 2014 19:18:19 +0000 (19:18 +0000)]
Pepper: Remove Manifest::ResolveURL

This need not be exposed as part of the Manifest interface.

Review URL: https://codereview.chromium.org/231853003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263045 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoPepper: Fix comments for PPB and PPP InputEvent.
teravest@chromium.org [Thu, 10 Apr 2014 18:40:07 +0000 (18:40 +0000)]
Pepper: Fix comments for PPB and PPP InputEvent.

The comments provided in PPB_InputEvent and PPP_InputEvent incorrectly describe
how handled events are delivered to handlers in web pages.

When a plugin handles an event, the event is not "bubbled" to any default
handlers. However, any listeners in a web page run before the event reaches the
plugin, so the plugin can't prevent any listeners from getting the event.

An attempt to make the implementation match the specification broke existing
plugins, so the only reasonable approach is to make the comments match the
behavior.

BUG=324352
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/228653006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@263034 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoApplied changes to chromium tree for new NaCl toolchain directories.
dyen@chromium.org [Wed, 9 Apr 2014 21:40:42 +0000 (21:40 +0000)]
Applied changes to chromium tree for new NaCl toolchain directories.

BUG= https://code.google.com/p/nativeclient/issues/detail?id=3816
TEST= trybots

Review URL: https://codereview.chromium.org/229283003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@262825 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoRefactor around start up of the plugin.
hidehiko@chromium.org [Wed, 9 Apr 2014 12:13:05 +0000 (12:13 +0000)]
Refactor around start up of the plugin.

This CL has some refactoring to prepare implementation of EmbedderService.
1) Moves irt_ppapi.c, plugin_main_irt.cc and plugin_main_irt.h to
   ppapi/nacl_irt.
2) Introduces irt_start.cc to handle startup of the plugin in SFI-mode.
   The file is extracted from irt_ppapi.cc. The main purpose of this separation
   is to avoid conflicting unistd.h. irt_ppapi.cc needs to include
   native_client/src/trusted/service_runtime/include/sys/unistd.h.
   However, start up code needs to include base/at_exit.h which indirectly
   include unistd.h. As a result they conflict, and
   NACL_ABI__SC_NACL_PNACL_MODE sometimes can be used, but sometimes not.
3) Moves AtExitManager, IOThread creation and global FD setting to
   irt_start.cc and plugin_startup.cc.

TEST=Ran trybots
BUG=358431

Review URL: https://codereview.chromium.org/225093006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@262649 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoPepper: Move some reporting into StartPpapiProxy.
teravest@chromium.org [Tue, 8 Apr 2014 16:16:52 +0000 (16:16 +0000)]
Pepper: Move some reporting into StartPpapiProxy.

Now that we're doing more error reporting and histogramming in NexeLoadManager,
we can move some of that type of logic into PPB_NaCl_Private.StartPpapiProxy.

BUG=239656

Review URL: https://codereview.chromium.org/226593002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@262433 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoPepper: Remove unnecessary file descriptor dup.
teravest@chromium.org [Mon, 7 Apr 2014 18:24:59 +0000 (18:24 +0000)]
Pepper: Remove unnecessary file descriptor dup.

There's an unnecessary file descriptor duplication in
Plugin::NaClManifestFileDidOpen with a confusing comment. This change removes
the call to DUP() which will make the change easier to read when moving this
logic to components/nacl.

BUG=239656

Review URL: https://codereview.chromium.org/227233008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@262163 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks ago[MIPS] Add missing support for NaCl browser tests
petarj@mips.com [Mon, 7 Apr 2014 17:51:04 +0000 (17:51 +0000)]
[MIPS] Add missing support for NaCl browser tests

- Define mips32.nexe for MIPS.
- Create NMFs for MIPS tests.

Review URL: https://codereview.chromium.org/227113008

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@262152 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoUse VLOG instead of vfprintf in PLUGIN_PRINTF.
teravest@chromium.org [Mon, 7 Apr 2014 17:47:03 +0000 (17:47 +0000)]
Use VLOG instead of vfprintf in PLUGIN_PRINTF.

The trusted plugin refactor will move code from ppapi/native_client to
components/nacl/renderer.

For consistency with the rest of Chrome, the refactored code will use VLOG
intead of PLUGIN_PRINTF. However, moving log statements over piecemeal will
make debugging difficult as multiple logfiles would need to be consulted
when debugging.

Instead, this changes the internals of PLUGIN_PRINTF to call VLOG instead of
vfprintf for logging. Any users who wish to see these debug statements must
then pass one of the following additional flags to Chrome:
  "--v=1" or
  "--vmodule=components/nacl/renderer*=1"
in addition to setting NACL_PLUGIN_DEBUG.

This change removes any effects of the environment variable "NACL_PLUGIN_LOG".
Once this change is submitted, I'll update any NaCl debugging documentation as
appropriate.

Note that this change may introduce allocations on the logging path. I think it
should be unlikely that PLUGIN_PRINTF will receive inputs larger than 512
characters, but truncating the output is worse than heap allocations.

BUG=239656

Review URL: https://codereview.chromium.org/224933006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@262150 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoFix Win64 build breakage from r261748
jschuh@chromium.org [Sat, 5 Apr 2014 12:16:10 +0000 (12:16 +0000)]
Fix Win64 build breakage from r261748

R=penghuang@chromium.org, yzshen@chromium.org
TBR=penghuang@chromium.org, yzshen@chromium.org
NOTRY=true
NOTREECHECKS=true

Review URL: https://codereview.chromium.org/227063002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@262001 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoPepper: Move NexeDidCrash() to NexeLoadManager.
teravest@chromium.org [Sat, 5 Apr 2014 07:34:59 +0000 (07:34 +0000)]
Pepper: Move NexeDidCrash() to NexeLoadManager.

This change pulls more code out of ppapi/native_client to
compnents/nacl/renderer.

GetReadyTime() is removed from PPB_NaCl_Private here to simplify time
management. I don't want to mix time math from Chrome and NaCl, so plugin.cc
uses its own ready_time_ for some operations again, and SetReadyTime() stashes
the current value of base::Time::Now in NexeLoadManager.

BUG=239656

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=261820

Review URL: https://codereview.chromium.org/216103003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261978 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoSupport scoped types in PPAPI IDL.
mtomasz@chromium.org [Fri, 4 Apr 2014 23:02:35 +0000 (23:02 +0000)]
Support scoped types in PPAPI IDL.

IDL parser didn't support namespaces, which however are supported by JSON.
This patch adds grammar to allow adding a namespace as a prefix to any type.

TEST=json_schema_test.py + tested manually that it works.
BUG=158654

Review URL: https://codereview.chromium.org/197873009

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261911 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoRevert 261820 "Pepper: Move NexeDidCrash() to NexeLoadManager."
scottmg@chromium.org [Fri, 4 Apr 2014 20:40:46 +0000 (20:40 +0000)]
Revert 261820 "Pepper: Move NexeDidCrash() to NexeLoadManager."

Potentially causing failures (?)

http://build.chromium.org/p/chromium.memory/builders/Linux%20ASan%20Tests%20%28sandboxed%29/builds/509/steps/browser_tests/logs/BadNative

[18,1263494912:20:15:11.094751] NaClAppRuntimeHostSetup: too late
[18,1263494912:20:15:11.094843] NaClAppDescQuotaSetup: too late
[18,1267164928:20:15:11.095246] NaClAppStartModule: error loading module
[18,1267164928:20:15:11.095918] reap logs
POST-ABORT: LOG_FATAL abort exit

> Pepper: Move NexeDidCrash() to NexeLoadManager.
>
> This change pulls more code out of ppapi/native_client to
> compnents/nacl/renderer.
>
> GetReadyTime() is removed from PPB_NaCl_Private here to simplify time
> management. I don't want to mix time math from Chrome and NaCl, so plugin.cc
> uses its own ready_time_ for some operations again, and SetReadyTime() stashes
> the current value of base::Time::Now in NexeLoadManager.
>
> BUG=239656
>
> Review URL: https://codereview.chromium.org/216103003

TBR=teravest@chromium.org, tommycli@chromium.org

Review URL: https://codereview.chromium.org/226343004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261860 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoPepper: Move NexeDidCrash() to NexeLoadManager.
teravest@chromium.org [Fri, 4 Apr 2014 19:04:00 +0000 (19:04 +0000)]
Pepper: Move NexeDidCrash() to NexeLoadManager.

This change pulls more code out of ppapi/native_client to
compnents/nacl/renderer.

GetReadyTime() is removed from PPB_NaCl_Private here to simplify time
management. I don't want to mix time math from Chrome and NaCl, so plugin.cc
uses its own ready_time_ for some operations again, and SetReadyTime() stashes
the current value of base::Time::Now in NexeLoadManager.

BUG=239656

Review URL: https://codereview.chromium.org/216103003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261820 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoSync NaCl MediaStream Video example with ppapi/examples/media_stream_video
penghuang@chromium.org [Fri, 4 Apr 2014 14:05:09 +0000 (14:05 +0000)]
Sync NaCl MediaStream Video example with ppapi/examples/media_stream_video

It also fixes a bug in release build.

BUG=330851

Review URL: https://codereview.chromium.org/212533007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261748 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

2 weeks agoRemove unused function {anonymous}::GetNaClInterface() in pnacl_resources.cc
jvoung@chromium.org [Fri, 4 Apr 2014 06:16:36 +0000 (06:16 +0000)]
Remove unused function {anonymous}::GetNaClInterface() in pnacl_resources.cc

BUG=359779

Review URL: https://codereview.chromium.org/224603002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261687 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoMoves ppapi/nacl_irt/{README,irt_ppapi.h} to ppapi/nacl_irt/public.
hidehiko@chromium.org [Fri, 4 Apr 2014 03:52:24 +0000 (03:52 +0000)]
Moves ppapi/nacl_irt/{README,irt_ppapi.h} to ppapi/nacl_irt/public.

This is preparation to add open_resource() irt for non-SFI mode.
Some codes will be under ppapi/nacl_irt/src, so this CL moves the
files directly under ppapi/nacl_irt to its public directory.

TEST=Build the binary. Run trybot.
BUG=358431

Review URL: https://codereview.chromium.org/214883007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261656 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoPPAPI/NaCl: Move architecture histogram to NexeLoadManager
dmichael@chromium.org [Thu, 3 Apr 2014 23:03:52 +0000 (23:03 +0000)]
PPAPI/NaCl: Move architecture histogram to NexeLoadManager

BUG=239656

Review URL: https://codereview.chromium.org/224143003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261565 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoRemove redundant I420-to-YV12 conversion in MediaStreamVideoTrack.
hshi@chromium.org [Thu, 3 Apr 2014 19:10:43 +0000 (19:10 +0000)]
Remove redundant I420-to-YV12 conversion in MediaStreamVideoTrack.

The I420 and YV12 formats both are almost identical except for the order
of U and V planes in memory. They are effectively the same for renderers
that access U and V plane data via MediaFrame::data(size_t plane).

BUG=341452
BUG=355763
TEST=trybot
R=danakj@chromium.org, dmichael@chromium.org, fischman@chromium.org, perkj@chromium.org, piman@chromium.org

Review URL: https://codereview.chromium.org/222563002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261497 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoPepper: Move exit_status to NexeLoadManager.
teravest@chromium.org [Thu, 3 Apr 2014 16:44:56 +0000 (16:44 +0000)]
Pepper: Move exit_status to NexeLoadManager.

This allows us to take SetReadOnlyProperty out of PPB_NaCl_Private. Hooray.

BUG=236959
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/222203003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261447 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks ago[MIPS] Add missing mips.nexe entries in nmf
petarj@mips.com [Thu, 3 Apr 2014 16:22:47 +0000 (16:22 +0000)]
[MIPS] Add missing mips.nexe entries in nmf

Browser expects to know paths to mips native client executables, so these
need to be provided in manifest files. Lack of these causes the related
browser tests to fail on MIPS arch.

Review URL: https://codereview.chromium.org/219193017

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261439 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoPepper: Change uma presubmit to never block CQ.
teravest@chromium.org [Thu, 3 Apr 2014 15:34:41 +0000 (15:34 +0000)]
Pepper: Change uma presubmit to never block CQ.

This changes the presubmit rule for adding pepper interface hashes to UMA to a
notification instead of a warning. This is because the presubmit is a bit too
broad, and was blocking some changes from being submitted through the CQ.

Using a notification message should be fine, since the impact of missing a
histogram is relatively low (we'll just see a number for a while), and the hash
can always be added later.

BUG=351389
R=xhwang@chromium.org

Review URL: https://codereview.chromium.org/221973003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261411 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoAdds first PPAPI test for non-SFI mode.
hidehiko@chromium.org [Thu, 3 Apr 2014 12:01:49 +0000 (12:01 +0000)]
Adds first PPAPI test for non-SFI mode.

This CL adds simple non-SFI PPAPI testing framework
with enabling one test "Core".
The testee nexe file for non-SFI mode is build by
translating a pexe file.

BUG=358425
TEST=Run trybots, run test locally with target_arch=ia32, and run test with ASAN.

Review URL: https://codereview.chromium.org/212633003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261357 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoRemove Dev permission check from TextInputController.
nona@chromium.org [Thu, 3 Apr 2014 07:07:46 +0000 (07:07 +0000)]
Remove Dev permission check from TextInputController.

The API is already moved to stable. No need to check dev permission any more.

BUG=343625
TEST=trybots

Review URL: https://codereview.chromium.org/220583002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261293 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoAdd the missing entry for PPB_Find_Private to interfaces_ppb_private.h
raymes@google.com [Thu, 3 Apr 2014 01:29:18 +0000 (01:29 +0000)]
Add the missing entry for PPB_Find_Private to interfaces_ppb_private.h

BUG=358820
TBR=teravest@chromium.org

Review URL: https://codereview.chromium.org/222923006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261251 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoRemove unused PPB_Graphics2D_Dev interface
raymes@google.com [Thu, 3 Apr 2014 00:39:47 +0000 (00:39 +0000)]
Remove unused PPB_Graphics2D_Dev interface

This may be revived later to implement fast PDF scrolling but that's too far out so get rid of it for the time being.

BUG=303491
R=jschuh@chromium.org, piman@chromium.org, teravest@chromium.org

Review URL: https://codereview.chromium.org/216213002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261242 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoRefactor gpu::Buffer to allow different types of backing
piman@chromium.org [Wed, 2 Apr 2014 23:54:09 +0000 (23:54 +0000)]
Refactor gpu::Buffer to allow different types of backing

This is so that we can use either base::SharedMemory or mojo::SharedBuffer to
back gpu::Buffer.

BUG=333157
R=darin@chromium.org, viettrungluu@chromium.org

Review URL: https://codereview.chromium.org/213353005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261228 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoPepper: LoadAbort and DeadNexe out of plugin.cc.
teravest@chromium.org [Wed, 2 Apr 2014 16:22:18 +0000 (16:22 +0000)]
Pepper: LoadAbort and DeadNexe out of plugin.cc.

This moves more error reporting function logic into NexeLoadManager from
plugin.cc in the trusted plugin.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/219133006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261135 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agogpu: Lose context when BeginQueryEXT fails to allocate.
danakj@chromium.org [Wed, 2 Apr 2014 14:43:13 +0000 (14:43 +0000)]
gpu: Lose context when BeginQueryEXT fails to allocate.

Instead of crashes, raise a GL_OUT_OF_MEMORY error. Since compositor
does not want to deal with these errors and it would leave it in a
bad state, add the ability to lose the context when GL_OUT_OF_MEMORY
occurs.

R=dmichael@chromium.org, jamesr@chromium.org, piman@chromium.org, piman
BUG=351587

Review URL: https://codereview.chromium.org/199443004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261120 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoRevert of gpu: Raise GL_OUT_OF_MEMORY when BeginQueryEXT fails to allocate. (https...
pfeldman@chromium.org [Wed, 2 Apr 2014 04:56:18 +0000 (04:56 +0000)]
Revert of gpu: Raise GL_OUT_OF_MEMORY when BeginQueryEXT fails to allocate. (https://codereview.chromium.org/199443004/)

Reason for revert:
Build failure in content/common/gpu/client/gl_helper_benchmark.cc.

http://build.chromium.org/p/chromium/builders/Linux%20x64/builds/63241/steps/compile/logs/stdio

Original issue's description:
> gpu: Lose context when BeginQueryEXT fails to allocate.
>
> Instead of crashes, raise a GL_OUT_OF_MEMORY error. Since compositor
> does not want to deal with these errors and it would leave it in a
> bad state, add the ability to lose the context when GL_OUT_OF_MEMORY
> occurs.
>
> R=piman
> BUG=351587
>
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=261064

TBR=piman@chromium.org,dmichael@chromium.org,jamesr@chromium.org,danakj@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=351587

Review URL: https://codereview.chromium.org/221783002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261075 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agogpu: Lose context when BeginQueryEXT fails to allocate.
danakj@chromium.org [Wed, 2 Apr 2014 04:16:29 +0000 (04:16 +0000)]
gpu: Lose context when BeginQueryEXT fails to allocate.

Instead of crashes, raise a GL_OUT_OF_MEMORY error. Since compositor
does not want to deal with these errors and it would leave it in a
bad state, add the ability to lose the context when GL_OUT_OF_MEMORY
occurs.

R=piman
BUG=351587

Review URL: https://codereview.chromium.org/199443004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@261064 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoPepper: Simplify desc mgmt in NexeFileDidOpen.
teravest@chromium.org [Tue, 1 Apr 2014 20:22:04 +0000 (20:22 +0000)]
Pepper: Simplify desc mgmt in NexeFileDidOpen.

This removes an unnecessary file descriptor duplication that takes place. Duplicating real file descriptors is more complicated than doing so for native_client's version of them.

I'm sending this as its own review to make sure its sane before moving much of
the NexeFileDidOpen logic to components/nacl/renderer.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/219053006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@260936 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoRemove CommandBuffer::GetTransferBuffer.
piman@chromium.org [Tue, 1 Apr 2014 09:06:43 +0000 (09:06 +0000)]
Remove CommandBuffer::GetTransferBuffer.

The only time we call GetTransferBuffer is on Pepper, right after
CreateTransferBuffer, so combine the two. This lets us dramatically remove all
the maintenance burden (maps) and extra IPCs.

Also separate service-only calls out of CommandBuffer, into CommandBufferServiceBase

BUG=None

Review URL: https://codereview.chromium.org/215803002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@260806 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoPepper: Move ReportLoadSuccess to PPB_NaCl_Private.
teravest@chromium.org [Mon, 31 Mar 2014 16:25:56 +0000 (16:25 +0000)]
Pepper: Move ReportLoadSuccess to PPB_NaCl_Private.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/216673005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@260557 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoMojo: Move C gles2 header files to mojo/public/c/gles2.
viettrungluu@chromium.org [Sun, 30 Mar 2014 20:40:17 +0000 (20:40 +0000)]
Mojo: Move C gles2 header files to mojo/public/c/gles2.

(C++ files still to do.)

Also, as a drive-by, update the generated copyright notice in
build_gles2_cmd_buffer.py to the correct, current one (no "(c)").

R=darin@chromium.org
TBR=piman@chromium.org

Review URL: https://codereview.chromium.org/218193002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@260459 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoAdd a PPB_Find_Private function to set the tickmarks on the page.
raymes@chromium.org [Sun, 30 Mar 2014 17:21:08 +0000 (17:21 +0000)]
Add a PPB_Find_Private function to set the tickmarks on the page.

This adds a PPB_Find_Private function to set the tickmarks on the page.

BUG=303491
TBR=viettrungluu@chromium.org

Review URL: https://codereview.chromium.org/195893044

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@260443 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoPPAPI: Bring PPB_MediaStream*Track to stable.
dmichael@chromium.org [Sat, 29 Mar 2014 00:07:30 +0000 (00:07 +0000)]
PPAPI: Bring PPB_MediaStream*Track to stable.

R=teravest@chromium.org
TBR=binji
BUG=330851,353717

Review URL: https://codereview.chromium.org/212023003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@260314 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

3 weeks agoReport some PNaCl UMA earlier, instead of in one chunk at the end.
jvoung@google.com [Fri, 28 Mar 2014 15:43:39 +0000 (15:43 +0000)]
Report some PNaCl UMA earlier, instead of in one chunk at the end.

This way, perhaps we can determine how many load-compiler-process
events happen vs load-linker-process events, and see if there is
any drop off in counts between the stages.

BUG=http://code.google.com/p/nativeclient/issues/detail?id=3824
R=teravest@chromium.org

Review URL: https://codereview.chromium.org/215553002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@260136 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoCheck NaCl debug mask (not just flag) before choosing PNaCl debug URL.
jvoung@google.com [Fri, 28 Mar 2014 02:47:41 +0000 (02:47 +0000)]
Check NaCl debug mask (not just flag) before choosing PNaCl debug URL.

That way, the debug URL is chosen only if the app will be debugged.

BUG=http://code.google.com/p/nativeclient/issues/detail?id=3765
R=bradnelson@google.com, tsepez@chromium.org
TBR=bradnelson@chromium.org, teravest@chromium.org

Review URL: https://codereview.chromium.org/212103002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@260067 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoPepper: Move is_installed state to NexeLoadManager.
teravest@chromium.org [Thu, 27 Mar 2014 21:10:59 +0000 (21:10 +0000)]
Pepper: Move is_installed state to NexeLoadManager.

This will make it easier to move more code out of ppapi/native_client,
especially UMA/Histogram support.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/214973002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259989 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoRedirect NaCl plugin logs to different files for different processes.
halyavin@google.com [Thu, 27 Mar 2014 20:22:04 +0000 (20:22 +0000)]
Redirect NaCl plugin logs to different files for different processes.

Currently these logs are not useful because different renderer processes
overwrite each other logs.

TEST= launch chrome with --no-sandbox flag and NACL_PLUGIN_LOG variable
set to some path and NACL_PLUGIN_DEBUG set to 1. Observe different
log files for different render processes.
BUG= None

Review URL: https://codereview.chromium.org/212633002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259964 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoPepper: More ReportLoadError logic in PPB_NaCl_Private.
teravest@chromium.org [Thu, 27 Mar 2014 15:53:26 +0000 (15:53 +0000)]
Pepper: More ReportLoadError logic in PPB_NaCl_Private.

This moves more state from the trusted plugin in ppapi/native_client to
components/nacl/renderer: nacl_ready_state_ and nexe_error_reported_.

The primary motivation is so that all of the logic for ReportLoadError()
happens in components/nacl.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/212813005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259888 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoFix resource leak code path around OpenManifestEntry.
hidehiko@chromium.org [Thu, 27 Mar 2014 07:39:10 +0000 (07:39 +0000)]
Fix resource leak code path around OpenManifestEntry.

Currently, there are some code paths which causes resource leak
in OpenManifestEntry (and its related methods).
There are mainly two cases:
1) If pnacl_options.translate() is true in original code,
open_cont is just leaked.
2) If stream_cc is created but not called, e.g. Plugin::StreamAsFile()
is failed somehow, or it is pNaCl related url, then the open_cont is
leaked.

It seems the root cause is PP_ComletionCallback/pp::CompletionCallback
usage. PP_CompletionCallback is designed for C, rather than C++,
so it does not manage the resources. pp::CompletionCallback is its
very simple wrapper. So, even when it is destructed without the invocation,
it just leakes the data.
Note that, WeakRef does not help such a case, because its resource management
fires only when the callback is invoked.
Probably, more generic wider solution would be minimize the usage
of these callbacks, such as the just interface to the PPAPI, and instead
base::Callback would be used internally in the code of chrome tree.

This CL just focuses on the OpenManifestEntry related code.

BUG=n/a
TEST=Ran trybots.

Review URL: https://codereview.chromium.org/211133003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259812 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoPepper: organize ppapi VideoDecoder messages.
bbudge@chromium.org [Wed, 26 Mar 2014 19:22:24 +0000 (19:22 +0000)]
Pepper: organize ppapi VideoDecoder messages.
1) Move messages for the existing Dev Video Decoder APIs to be together, so
things don't get too confusing when I add the new VideoDecoder API.
2) Consolidate Flash messages.
3) Remove extra OS_NACL guards.
4) Remove some messages from the NaCl untrusted build that aren't proxied.

BUG=281689

Review URL: https://codereview.chromium.org/211573007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259657 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoReplace --ppapi-keep-alive-throttle command line switch with IPC parameter.
scheib@chromium.org [Wed, 26 Mar 2014 18:28:30 +0000 (18:28 +0000)]
Replace --ppapi-keep-alive-throttle command line switch with IPC parameter.

Part of a broad effort to reduce the number of command line switches.

This is the third land of this change - handling DISABLE_NACL builds.

BUG=350510

Review URL: https://codereview.chromium.org/211513002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259637 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoPass string PP_Vars to JS as string primitives
raymes@chromium.org [Wed, 26 Mar 2014 05:04:16 +0000 (05:04 +0000)]
Pass string PP_Vars to JS as string primitives

Previously we sent PP_Var strings to JS (via PostMessage) as string objects
in order to allow references to strins in the graph to maintained (string
objects in JS can be referenced but string primitives cannot). However
maintaining reference information is hugely useful and it is more natural
for JS clients to receive a string primitive. This CL ensures that all
strings received via PostMessage are string primitives.

BUG=350031

Review URL: https://codereview.chromium.org/201553002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259498 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoRename PPB_Find_Dev to PPB_Find_Private
raymes@chromium.org [Wed, 26 Mar 2014 05:03:34 +0000 (05:03 +0000)]
Rename PPB_Find_Dev to PPB_Find_Private

This makes PPB_Find_Dev a private API and renames it as such. There is no intention to make this a public API.

BUG=303491

Review URL: https://codereview.chromium.org/197623005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259497 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoGPU: Reference count gpu::Buffer.
epenner@chromium.org [Wed, 26 Mar 2014 04:53:58 +0000 (04:53 +0000)]
GPU: Reference count gpu::Buffer.

We tend to duplicate shared memory within the same process
when we don't need to, causing extra page faults etc when
accessing the memory.

With reference counting we can more easily re-use blocks of
shared memory in the same process, and make security easier
(objects in shared memory can hold a reference to their
memory and prevent use-after-free).

BUG=177063, 353822

Review URL: https://codereview.chromium.org/209353002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259491 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoPPAPI/NaCl: Define NexeLoadManager class.
teravest@chromium.org [Tue, 25 Mar 2014 22:19:54 +0000 (22:19 +0000)]
PPAPI/NaCl: Define NexeLoadManager class.

This adds an object that tracks the lifetime of the "Plugin" class in the
trusted plugin. State is kept in this class as long as the plugin is alive, to
make accounting much simpler for PPB_NaCl_Private.

This is derived from https://codereview.chromium.org/180213006/ but adds a
bugfix that initializes NexeLoadManager earlier.

BUG=239656

Review URL: https://codereview.chromium.org/208913002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259362 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoRe-enable disabled size asserts in ppapi_tests test_struct_sizes.cc.
bbudge@chromium.org [Tue, 25 Mar 2014 02:46:20 +0000 (02:46 +0000)]
Re-enable disabled size asserts in ppapi_tests test_struct_sizes.cc.
Khronos headers have been updated.

BUG=177779

Review URL: https://codereview.chromium.org/205263005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@259114 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoCurrently it automatically loads pepper_spinning_cube_demo and only supports those...
yzshen@chromium.org [Sat, 22 Mar 2014 04:11:12 +0000 (04:11 +0000)]
Currently it automatically loads pepper_spinning_cube_demo and only supports those APIs that needed by it.

BUG=None
TEST=None

Review URL: https://codereview.chromium.org/178953003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258773 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

4 weeks agoAdd IsCrashReportingEnabled call to Pepper UMA API
elijahtaylor@chromium.org [Fri, 21 Mar 2014 22:44:01 +0000 (22:44 +0000)]
Add IsCrashReportingEnabled call to Pepper UMA API

This mirrors the existing metricsPrivate JS API.

BUG=341130

Review URL: https://codereview.chromium.org/195793026

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258702 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoPepper: Move AddToConsole() to PPB_NaCl_Private.
teravest@chromium.org [Thu, 20 Mar 2014 21:43:34 +0000 (21:43 +0000)]
Pepper: Move AddToConsole() to PPB_NaCl_Private.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/203373005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258417 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoFix DCHECK errors on non-SFI mode.
hidehiko@chromium.org [Thu, 20 Mar 2014 19:11:07 +0000 (19:11 +0000)]
Fix DCHECK errors on non-SFI mode.

This CL is extracted from crrev.com/189553005.
The fixes are:
1) AtExitManager is conflicting. In non-SFI mode, AtExitManager is already
instantiated in nacl_helper. To avoid the conflicting, skip to create it in
plugin_main_irt.cc.
2) Allowed creating Singleton on the plugin's main thread.
3) SRPC may be successfully done, and then NOTREACHED catch it in
nacl_listener. It is false alarm, so removed it.

BUG=https://code.google.com/p/nativeclient/issues/detail?id=3734
TETS=Run trybots.

Review URL: https://codereview.chromium.org/206033006

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258367 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoPepper: Small cleanup in FileDownloader.
teravest@chromium.org [Thu, 20 Mar 2014 17:44:52 +0000 (17:44 +0000)]
Pepper: Small cleanup in FileDownloader.

This change removes some methods that encapsulate the values of mode_. This
reduces the size of the public interface of FileDownloader, and I think it
makes the code easier to read.

BUG=239656
R=bbudge@chromium.org

Review URL: https://codereview.chromium.org/202523005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258331 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoRevert of Add very simple loading test for non-SFI NaCl. (https://codereview.chromium...
pneubeck@chromium.org [Thu, 20 Mar 2014 09:08:35 +0000 (09:08 +0000)]
Revert of Add very simple loading test for non-SFI NaCl. (https://codereview.chromium.org/189553005/)

Reason for revert:
Broken linking on ChromiumOS
http://build.chromium.org/p/chromium.chromiumos/builders/ChromiumOS%20%28x86%29/builds/20529/steps/BuildPackages/logs/stdio

Original issue's description:
> Add very simple loading test for non-SFI NaCl.
>
> This is the first step to add test cases for non-SFI NaCl.
> The testee binary is build by a new target rule introduced by this CL, which is relocatable/libc-free binary. The code has a minimum set of handlers for irt_ppapi, to load the plugin successfully done with communicating hosts (browser/renderer processes).
>
> Along this change, some DCHECK errors are found. So they're fixed, too.
> 1) AtExitManager is conflicting. In non-SFI mode, AtExitManager is already instantiated in nacl_helper. To avoid the conflicting, skip to create it in plugin_main_irt.cc.
> 2) Allowed creating Singleton on the plugin's main thread.
> 3) SRPC may be successfully done, and then NOTREACHED catch it in nacl_listener. It is false alarm, so removed it.
>
> BUG=https://code.google.com/p/nativeclient/issues/detail?id=3734
> TEST=run ./browser_tests --gtest_filter=NaCl*NonSFI* locally. Run trybots.
>
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257118
>
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257743
>
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=258200

TBR=mseaborn@chromium.org,dmichael@chromium.org,hidehiko@chromium.org
NOTREECHECKS=true
NOTRY=true
BUG=https://code.google.com/p/nativeclient/issues/detail?id=3734

Review URL: https://codereview.chromium.org/206223002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258258 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoAdd very simple loading test for non-SFI NaCl.
hidehiko@chromium.org [Thu, 20 Mar 2014 05:40:10 +0000 (05:40 +0000)]
Add very simple loading test for non-SFI NaCl.

This is the first step to add test cases for non-SFI NaCl.
The testee binary is build by a new target rule introduced by this CL, which is relocatable/libc-free binary. The code has a minimum set of handlers for irt_ppapi, to load the plugin successfully done with communicating hosts (browser/renderer processes).

Along this change, some DCHECK errors are found. So they're fixed, too.
1) AtExitManager is conflicting. In non-SFI mode, AtExitManager is already instantiated in nacl_helper. To avoid the conflicting, skip to create it in plugin_main_irt.cc.
2) Allowed creating Singleton on the plugin's main thread.
3) SRPC may be successfully done, and then NOTREACHED catch it in nacl_listener. It is false alarm, so removed it.

BUG=https://code.google.com/p/nativeclient/issues/detail?id=3734
TEST=run ./browser_tests --gtest_filter=NaCl*NonSFI* locally. Run trybots.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257118

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257743

Review URL: https://codereview.chromium.org/189553005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258200 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoThis proxies PPB_Find_Dev so that it can be used out of process. It also adds a funct...
raymes@google.com [Thu, 20 Mar 2014 00:27:54 +0000 (00:27 +0000)]
This proxies PPB_Find_Dev so that it can be used out of process. It also adds a function named SetPluginToHandleFindRequests which can be called by embedded plugins which want to handle browser find. This is needed for the new out of process PDF plugin, which will be embedded in an extensions page.

BUG=303491
R=jam@chromium.org, jschuh@chromium.org, yzshen@chromium.org

Review URL: https://codereview.chromium.org/188323002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258171 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoPepper: Make PPB ReportLoadError set lastError.
teravest@chromium.org [Wed, 19 Mar 2014 18:58:00 +0000 (18:58 +0000)]
Pepper: Make PPB ReportLoadError set lastError.

This is part of a series of changes to move more error handling logic to
components/nacl/renderer from ppapi/native_client. I would have moved over some
of the console logging work as part of this change as well, but it started to
look pretty large.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/197873030

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258050 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoPepper: Fix VLOG line wrapping in generated thunks.
teravest@chromium.org [Wed, 19 Mar 2014 17:55:28 +0000 (17:55 +0000)]
Pepper: Fix VLOG line wrapping in generated thunks.

This fixes generated thunks to have nicer line wrapping for long function names.

Previously, thunk files kept getting generated for
ppapi/thunk/ppb_display_color_profile_private_thunk.cc,
and the generated file looked much worse than what was
checked in.

BUG=
R=bbudge@chromium.org

Review URL: https://codereview.chromium.org/204863002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258033 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoPepper: Move GetUrlScheme to PPB_NaCl_Private.
teravest@chromium.org [Wed, 19 Mar 2014 17:23:18 +0000 (17:23 +0000)]
Pepper: Move GetUrlScheme to PPB_NaCl_Private.

This is part of a series of changes to clean up FileDownloader so that it's
less dependent on Plugin.

As an added benefit, the code for GetUrlScheme is much shorter once it's on the
Chromium side.

BUG=239656
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/203523003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@258027 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoMove Blink dependency out of plugin proxy code.
bbudge@chromium.org [Tue, 18 Mar 2014 22:51:32 +0000 (22:51 +0000)]
Move Blink dependency out of plugin proxy code.
Moves dependency on WebSocket.h from plugin to host.

BUG=350097

Review URL: https://codereview.chromium.org/202143007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257796 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoRevert 257743 "Add very simple loading test for non-SFI NaCl."
yzshen@chromium.org [Tue, 18 Mar 2014 22:28:36 +0000 (22:28 +0000)]
Revert 257743 "Add very simple loading test for non-SFI NaCl."

Broke Linux ASan+LSan Tests (1).

> Add very simple loading test for non-SFI NaCl.
>
> This is the first step to add test cases for non-SFI NaCl.
> The testee binary is build by a new target rule introduced by this CL, which is relocatable/libc-free binary. The code has a minimum set of handlers for irt_ppapi, to load the plugin successfully done with communicating hosts (browser/renderer processes).
>
> Along this change, some DCHECK errors are found. So they're fixed, too.
> 1) AtExitManager is conflicting. In non-SFI mode, AtExitManager is already instantiated in nacl_helper. To avoid the conflicting, skip to create it in plugin_main_irt.cc.
> 2) Allowed creating Singleton on the plugin's main thread.
> 3) SRPC may be successfully done, and then NOTREACHED catch it in nacl_listener. It is false alarm, so removed it.
>
> BUG=https://code.google.com/p/nativeclient/issues/detail?id=3734
> TEST=run ./browser_tests --gtest_filter=NaCl*NonSFI* locally. Run trybots.
>
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257118
>
> Review URL: https://codereview.chromium.org/189553005

TBR=hidehiko@chromium.org

Review URL: https://codereview.chromium.org/203993003

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257787 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoReplace command buffer FlushSync with WaitForTokenInRange and WaitForGetOffsetInRange
jbauman@chromium.org [Tue, 18 Mar 2014 21:52:36 +0000 (21:52 +0000)]
Replace command buffer FlushSync with WaitForTokenInRange and WaitForGetOffsetInRange

This allows the command-buffer interface to be more clear about what's happening, so we'll later be able to create IPCs that wait specifically for one of these events to happen. Currently both function are still implemented by looping on the GetStateFast IPC.

BUG=349632

Review URL: https://codereview.chromium.org/189123004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257771 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoPepper: Clean up ErrorInfo use in ServiceRuntime.
teravest@chromium.org [Tue, 18 Mar 2014 21:51:38 +0000 (21:51 +0000)]
Pepper: Clean up ErrorInfo use in ServiceRuntime.

There are places where ErrorInfo is passed around needlessly. There are
places in ServiceRuntime that don't use ErrorInfo with ReportLoadError(). This
cleans up that use a bit to make the code easier to read and remove some cases
of duplicate logging. I came across this while moving Manifest out of the
trusted plugin.

BUG=239656

Review URL: https://codereview.chromium.org/198053007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257770 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoAdd very simple loading test for non-SFI NaCl.
hidehiko@chromium.org [Tue, 18 Mar 2014 20:49:53 +0000 (20:49 +0000)]
Add very simple loading test for non-SFI NaCl.

This is the first step to add test cases for non-SFI NaCl.
The testee binary is build by a new target rule introduced by this CL, which is relocatable/libc-free binary. The code has a minimum set of handlers for irt_ppapi, to load the plugin successfully done with communicating hosts (browser/renderer processes).

Along this change, some DCHECK errors are found. So they're fixed, too.
1) AtExitManager is conflicting. In non-SFI mode, AtExitManager is already instantiated in nacl_helper. To avoid the conflicting, skip to create it in plugin_main_irt.cc.
2) Allowed creating Singleton on the plugin's main thread.
3) SRPC may be successfully done, and then NOTREACHED catch it in nacl_listener. It is false alarm, so removed it.

BUG=https://code.google.com/p/nativeclient/issues/detail?id=3734
TEST=run ./browser_tests --gtest_filter=NaCl*NonSFI* locally. Run trybots.

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257118

Review URL: https://codereview.chromium.org/189553005

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257743 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoFix "unreachable code" warnings (MSVC warning 4702) in ppapi/.
pkasting@chromium.org [Tue, 18 Mar 2014 20:48:42 +0000 (20:48 +0000)]
Fix "unreachable code" warnings (MSVC warning 4702) in ppapi/.

BUG=346399
TEST=none
R=dmichael@chromium.org

Review URL: https://codereview.chromium.org/197203004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257741 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoFix PNaCl debug commandline flags, and use the pexe with debug info in test.
jvoung@google.com [Tue, 18 Mar 2014 20:35:11 +0000 (20:35 +0000)]
Fix PNaCl debug commandline flags, and use the pexe with debug info in test.

Now that the PNaCl translator which supports reading
llvm-formatted bitcode files is DEPS'ed in, we can change
the debug test to actually use a pexe with debug info. In
the process, fix the way the commandline flags were set
up (avoid <<'ing a string that starts with a null byte).

Also try to turn on the test for 64-bit windows instead,
since there should only be a problem with 32-bit windows.

BUG= http://code.google.com/p/nativeclient/issues/detail?id=3765
R=bradnelson@chromium.org, bradnelson@google.com

Review URL: https://codereview.chromium.org/203103002

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257729 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoRevert 257524 "Move IsStringASCII/UTF8 to base namespace."
brettw@chromium.org [Mon, 17 Mar 2014 23:55:43 +0000 (23:55 +0000)]
Revert 257524 "Move IsStringASCII/UTF8 to base namespace."

> Move IsStringASCII/UTF8 to base namespace.
>
> Use StringPiece for IsStringUTF8.
>
> TBR=sky
>
> Review URL: https://codereview.chromium.org/196793010

TBR=brettw@chromium.org

Review URL: https://codereview.chromium.org/198163004

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257533 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoMove IsStringASCII/UTF8 to base namespace.
brettw@chromium.org [Mon, 17 Mar 2014 23:13:58 +0000 (23:13 +0000)]
Move IsStringASCII/UTF8 to base namespace.

Use StringPiece for IsStringUTF8.

TBR=sky

Review URL: https://codereview.chromium.org/196793010

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257524 4ff67af0-8c30-449e-8e8b-ad334ec8d88c

5 weeks agoBump the min PNaCl version of M35 to what we send out to M34. Delete fallback.
jvoung@google.com [Mon, 17 Mar 2014 18:04:06 +0000 (18:04 +0000)]
Bump the min PNaCl version of M35 to what we send out to M34. Delete fallback.

By the time M35 rolls around, most users will have upgraded to the M34
PNaCl. However, for those that haven't we can use the on-demand update.
Given this minimum PNaCl version, we don't need to attempt a
fallback IPC (from < M34) when the M34 IPC isn't available.

Cleanup for:
BUG=https://code.google.com/p/nativeclient/issues/detail?id=3777
R=cpu@chromium.org, teravest@chromium.org

Review URL: https://codereview.chromium.org/197023007

git-svn-id: http://src.chromium.org/svn/trunk/src/ppapi@257465 4ff67af0-8c30-449e-8e8b-ad334ec8d88c