GitLab

To fix tree-closer in Linux ASan LSan Tests (1) since
http://build.chromium.org/p/chromium.memory/builders/Linux%20ASan%20LSan%20Tests%20%281%29/builds/4338

errors like

Direct leak of 72 byte(s) in 1 object(s) allocated from:
    #0 0x44145b in operator new(unsigned long) /usr/local/google/work/chromium/src/third_party/llvm/projects/compiler-rt/lib/asan/asan_new_delete.cc:55
    #1 0x5e8afd in device::HidService::Create(scoped_refptr\u003Cbase::MessageLoopProxy>) device/hid/hid_service.cc:26
    #2 0x509324 in device::HidServiceTest_Create_Test::TestBody() device/hid/hid_service_unittest.cc:15

BUG=367145
NOTRY=true
NOTREECHECKS=true
TBR=rockot@chromium.org

Review URL: https://codereview.chromium.org/424623002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285848 0039d316-1c4b-4281-b951-d872f2087c98

BUG=397888
TBR=jchaffraix@chromium.org

Review URL: https://codereview.chromium.org/421843002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285847 0039d316-1c4b-4281-b951-d872f2087c98

R=maruel@chromium.org

Review URL: https://codereview.chromium.org/419973006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285842 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285841 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285840 0039d316-1c4b-4281-b951-d872f2087c98

http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog_blink.html?url=/trunk&range=178989:178990&mode=html
TBR=
BUG=

Review URL: https://codereview.chromium.org/421803002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285839 0039d316-1c4b-4281-b951-d872f2087c98

__THROW is not some magical thing, don't pretend it is. Don't add it to
extern "C" functions. Replace it with "throw ()" on the functions where it's
correct since it was defined to nothing on windows (the only place where this
file is used).

Fix an inconsequential enum constant mix-up in media (both enum constants
evaluated to the same value, and the enums had the same size, so it shouldn't
make a difference in practice, but the old code was comparing enum variables
with enum constants from a different enum).

BUG=82385
TBR=dalecurtis

Review URL: https://codereview.chromium.org/419323002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285838 0039d316-1c4b-4281-b951-d872f2087c98

. roots_ now always contains the root nodes. Previously I would add a
  fake id if a connection had roots_ and they were all removed. I'ved
  nuked that too.
. Fixed bug in ProcessNodeHierarchyChanged. Specifically if a
  connection added a new node that it didn't previously know about the
  set of known nodes needed to be updated. This is only matters for
  the window manager and may only come up in tests.
. GetUnknownNodes needed to check access policy.
. All children are now removed as part of an embed. Previously this
  wasn't done for the wm, but I can't see a good reason to keep
  that.
. Node::Destroy() would never complete if any of the children were
  from another connection.
. Client lib wasn't properly mapping in nodes from another
  connection.
. Updated a couple of tests appropriately.

BUG=397660
TEST=covered by tests
R=ben@chromium.org

Review URL: https://codereview.chromium.org/424533002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285837 0039d316-1c4b-4281-b951-d872f2087c98

This includes verification of:
* MojoCreateMessagePipeOptions
* MojoCreateDataPipeOptions
* MojoCreateSharedBufferOptions
* MojoDuplicateBufferHandleOptions

It also revamps options_validation.h completely (and updates tests).

R=darin@chromium.org

Review URL: https://codereview.chromium.org/414393002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285836 0039d316-1c4b-4281-b951-d872f2087c98

Remove or move/change #includes to appropriate places/ones.

BUG=

Review URL: https://codereview.chromium.org/416273005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285835 0039d316-1c4b-4281-b951-d872f2087c98

Hopefully the move will make DesktopWindowTreeHostX11Test.Deactivate no longer
flaky. Only a single window can be active at a time. I suspect that this test
sometimes fails when run in parallel with other tests

BUG=396491
TEST=None

Review URL: https://codereview.chromium.org/414123005

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285834 0039d316-1c4b-4281-b951-d872f2087c98

This CL fixes a DCHECK happening due to r284913. The CL changed the order
of window addition which caused window properties to update in a different
order. This is fixed by only running the DCHECKed code after the window
is parented beneath a root window.

BUG=396911

Review URL: https://codereview.chromium.org/417923003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285833 0039d316-1c4b-4281-b951-d872f2087c98

Touch exploration was too buggy in M37, we don't want to launch it.
The plan is to land this change and merge to M37, then revert it to aim for
a launch of touch exploration in M38.

BUG=396193

Review URL: https://codereview.chromium.org/414493004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285832 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285826 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285824 0039d316-1c4b-4281-b951-d872f2087c98

http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog_blink.html?url=/trunk&range=178971:178988&mode=html
TBR=
BUG=

Review URL: https://codereview.chromium.org/426463002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285823 0039d316-1c4b-4281-b951-d872f2087c98

BUG=396340

Review URL: https://codereview.chromium.org/414633002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285819 0039d316-1c4b-4281-b951-d872f2087c98

Instead of giving BrowserAccessibilityManagerWin a pointer
to a LegacyRenderWidgetHostHWND, which has been a source of
problems due to complicated lifecycles, just have BAMW
request the parent HWND and parent IAccessible from its
delegate whenever needed, with the understanding that these
may be NULL sometimes.

Note: this change fixes a DrMemory UAF detected in the
patch that migrated accessibility from RVH to RFH
(codereview.chromium.org/273423004), and also fixes
a potential concern in another fix to
LegacyRenderWidgetHostHWND (codereview.chromium.org/387353004).

BUG=368298,393665

Review URL: https://codereview.chromium.org/415633002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285818 0039d316-1c4b-4281-b951-d872f2087c98

BUG=392666

Review URL: https://codereview.chromium.org/422583002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285817 0039d316-1c4b-4281-b951-d872f2087c98

Enable bluetooth on chromeos, win, and mac, but disable on linux unless in the whitelist. The whitelist includes the HotRod apps.

BUG=396117

Review URL: https://codereview.chromium.org/406293002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285816 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285815 0039d316-1c4b-4281-b951-d872f2087c98

Revert of NaCl: Update revision in DEPS, r13506 -> r13528 (https://codereview.chromium.org/417233002/)

Reason for revert:
This failed on the "Win x64 Builder (dbg)" buildbot due to long filename issues.

I rolled a too-old revision of NaCl that didn't include Jan's revert of r13511 (which switches the PNaCl toolchain from Cygwin to MinGW).

Original issue's description:
> NaCl: Update revision in DEPS, r13506 -> r13528
> 
> This pulls in the following Native Client changes:
> 
> r13507: (noelallen) Reapply patch which was reverted due to conflict.
> r13508: (dschuff) PNaCl Driver: do not generate temp file names beyond Windows' MAX_PATH limit
> r13509: (jvoung) Add a flag to pnacl-finalize that allows users to keep function names.
> r13510: (dschuff) Disable flaky sandboxed translator tests on mac
> r13511: (dschuff) Update revision for PNaCl r13466->r13509
> r13512: (ncbray) Fix untrusted.gypi and build_nexe.py
> r13513: (noelallen) Cleanup bionic build script.
> r13514: (dyen) Added initial pass of nacl_irt_code_data_alloc.
> r13515: (sbc) Use custom exception class for errors in package_version.
> r13516: (hidehiko) Non-SFI mode: SCONS_NONSFI_TC clean up.
> r13517: (hidehiko) Non-SFI Mode: Refactoring. Provide non-SFI version of nacl_irt_query_core.
> r13518: (binji) Revert "Non-SFI mode: SCONS_NONSFI_TC clean up."
> r13519: (bradnelson) Make the trybots fail early if 'git cl try' is mistakenly used.
> r13520: (bradnelson) Make build_nexe.py rebuild when script itself has changed.
> r13521: (noelallen) Updates GN dep, and cleans up some toolchain prep definitions to make it easier to debug what's going on.
> r13523: (teravest) Add embedder interface for fatal log messages.
> r13524: (bradnelson) Passing --config-name to build_nexe.py for all arches.
> r13525: (teravest) Create a sel_ldr_launcher_base target.
> r13526: (teravest) Revert "Add embedder interface for fatal log messages."
> r13527: (hidehiko) Non-SFI mode: Get rid of socketcall on ARM.
> r13528: (teravest) Add embedder interface for fatal log messages.
> 
> BUG=none
> TEST=browser_tests and nacl_integration
> CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_nacl_sdk_build
> TBR=bradnelson@google.com
> 
> Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=285812

TBR=bradnelson@google.com
NOTREECHECKS=true
NOTRY=true
BUG=none

Review URL: https://codereview.chromium.org/419373002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285814 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285813 0039d316-1c4b-4281-b951-d872f2087c98

This pulls in the following Native Client changes:

r13507: (noelallen) Reapply patch which was reverted due to conflict.
r13508: (dschuff) PNaCl Driver: do not generate temp file names beyond Windows' MAX_PATH limit
r13509: (jvoung) Add a flag to pnacl-finalize that allows users to keep function names.
r13510: (dschuff) Disable flaky sandboxed translator tests on mac
r13511: (dschuff) Update revision for PNaCl r13466->r13509
r13512: (ncbray) Fix untrusted.gypi and build_nexe.py
r13513: (noelallen) Cleanup bionic build script.
r13514: (dyen) Added initial pass of nacl_irt_code_data_alloc.
r13515: (sbc) Use custom exception class for errors in package_version.
r13516: (hidehiko) Non-SFI mode: SCONS_NONSFI_TC clean up.
r13517: (hidehiko) Non-SFI Mode: Refactoring. Provide non-SFI version of nacl_irt_query_core.
r13518: (binji) Revert "Non-SFI mode: SCONS_NONSFI_TC clean up."
r13519: (bradnelson) Make the trybots fail early if 'git cl try' is mistakenly used.
r13520: (bradnelson) Make build_nexe.py rebuild when script itself has changed.
r13521: (noelallen) Updates GN dep, and cleans up some toolchain prep definitions to make it easier to debug what's going on.
r13523: (teravest) Add embedder interface for fatal log messages.
r13524: (bradnelson) Passing --config-name to build_nexe.py for all arches.
r13525: (teravest) Create a sel_ldr_launcher_base target.
r13526: (teravest) Revert "Add embedder interface for fatal log messages."
r13527: (hidehiko) Non-SFI mode: Get rid of socketcall on ARM.
r13528: (teravest) Add embedder interface for fatal log messages.

BUG=none
TEST=browser_tests and nacl_integration
CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_nacl_sdk_build
TBR=bradnelson@google.com

Review URL: https://codereview.chromium.org/417233002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285812 0039d316-1c4b-4281-b951-d872f2087c98

Plus some minor style fixes, mainly clang-format the header.

BUG=331299
TEST=content_unittests, content_shell. No functional changes
TBR=dpranke@chromium.org

Review URL: https://codereview.chromium.org/421703003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285811 0039d316-1c4b-4281-b951-d872f2087c98

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285810 0039d316-1c4b-4281-b951-d872f2087c98

http://build.chromium.org/f/chromium/perf/dashboard/ui/changelog_blink.html?url=/trunk&range=178930:178970&mode=html
TBR=haraken@chromium.org,jchaffraix@chromium.org

Review URL: https://codereview.chromium.org/419353002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285809 0039d316-1c4b-4281-b951-d872f2087c98

BUG=396447

Review URL: https://codereview.chromium.org/417203002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285808 0039d316-1c4b-4281-b951-d872f2087c98

The fix was picked up in the r282874 libvpx roll.
BUG=350982

Review URL: https://codereview.chromium.org/423643002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285807 0039d316-1c4b-4281-b951-d872f2087c98

Before, wifi_config_view expected to read only a plain PKCS#11 ID from the shill::kEapCertIdProperty. However the format changed to '<slot_id>:<key_id>' since the system token is enabled in ChromeOS. This lead to the view not correctly pre-selecting the previously configured certificate.
With this change, the property is correctly parsed and the pre-selection works again for certificates from the user's slot.

BUG=358366, 396181

Review URL: https://codereview.chromium.org/410113002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285806 0039d316-1c4b-4281-b951-d872f2087c98

Can be reenabled after this test was moved to browser_tests.

BUG=396497
TBR=joaodasilva@chromium.org

Review URL: https://codereview.chromium.org/420963002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285805 0039d316-1c4b-4281-b951-d872f2087c98

It adds a new TLSSocket type in the extensions:: namespace, that acts similarly
to TCP/UDPSocket.  Failures to enable TLS will auto-close the socket.

API Proposal: https://docs.google.com/a/google.com/document/d/1XSjBtLjyGXQmgB0XE4kxuvM0p2yKI_8T5aOiKDJ9lFg/edit

BUG=132896
TESTED=some, mostly with a custom chrome app, and with the included test.

Review URL: https://codereview.chromium.org/76403004

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285804 0039d316-1c4b-4281-b951-d872f2087c98

Notification promos currently have start, end times, and a number of maximum
views. However, the length of time a promo can be visible was not capped.
This change adds a flag to cap the length of time a promo is visible by
max_seconds_.

BUG=None
Test=None

Review URL: https://codereview.chromium.org/419033007

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285803 0039d316-1c4b-4281-b951-d872f2087c98

BUG=381820

Review URL: https://codereview.chromium.org/420603014

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285802 0039d316-1c4b-4281-b951-d872f2087c98

This is the refactoring to prepare newlib switching.
By this CL, we can easily use elf_loader defined in native_client repository, even for making a working demo.

TEST=Ran trybots.
BUG=358465
CQ_EXTRA_TRYBOTS=tryserver.chromium:linux_rel_precise32

Review URL: https://codereview.chromium.org/418133006

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285801 0039d316-1c4b-4281-b951-d872f2087c98

BUG=None
TEST=gn gen out/Debug_gn && ninja -C out/Debug_gn extensions/browser
R=brettw@chromium.org
TBR=yoz@chromium.org

Review URL: https://codereview.chromium.org/418763002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285800 0039d316-1c4b-4281-b951-d872f2087c98

BUG=367364

Review URL: https://codereview.chromium.org/423543002

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285799 0039d316-1c4b-4281-b951-d872f2087c98

RSA properties such as n, e, d, p, q, dp, dq, qi are big integers encoded as base64 url-safe, big-endian octet strings.

 * Reject big integers that contain leading zeros, since by the JWA rules they must be minimal.
 * Reject big integers that are the empty octet string (since 0 is the minimal representation of the big-endian number 0, not empty string).

This also changes the exception message and type when one of the optional parameters p, q, dp, dq, qi are missing. Before it would give an OperationError, because NSS was unable to infer the missing parameters. Now it gives a DataError explaining that the parameter is required.

BUG=374927,383998

Review URL: https://codereview.chromium.org/416993009

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285798 0039d316-1c4b-4281-b951-d872f2087c98

This is an optimization to speed up in-page navigations. We want those to be
blazingly fast, and currently the pixel readback done while taking a
screenshot takes 40ms on a GPU on Pixel. This may be reverted once
screenshotting is optimizaed.

Also changed the background color shown bu GestureNav when there is no
screenshot available from gray to white to make it look less broken.

BUG=379983

Review URL: https://codereview.chromium.org/417343003

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285797 0039d316-1c4b-4281-b951-d872f2087c98

The toolbar's superview chromeContentView was incorrectly sized and was too
small. This happened because chromeContentView's layout was not being updated
in -[BrowserWindowController layoutSubviews].

I updated the unit tests in browser_window_controller_unittest.mm with 2 changes:
- All layout checking should use window coordinates, to ensure consistency.
- Layout checking should ensure that the frame of each view is within the
bounds of the superview.

BUG=396740

Review URL: https://codereview.chromium.org/419683008

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@285796 0039d316-1c4b-4281-b951-d872f2087c98