GitLab

https://boringssl.googlesource.com/boringssl/+log/c26692cfdd9b0c86f2573ad03461416c9702bd39..772a5bed7d5065b301c3979c325248689d36b92c

BUG=none

Review-Url: https://codereview.chromium.org/2672943002
Cr-Commit-Position: refs/heads/master@{#447912}

When error occurs during initializing the remote renderer,
CourierRenderer should not report the failure to media pipeline.
Instead, it should always report success to the media pipeline, and
then become a no-op renderer after that point.  The remoting session
will soon end, and a normal local renderer will be created instead.

BUG=687750

Review-Url: https://codereview.chromium.org/2669033004
Cr-Commit-Position: refs/heads/master@{#447911}

We're not working on this config at the moment, so having a red buildbot
testing it is just waste of cycles.

BUG=688123, 598772

Review-Url: https://codereview.chromium.org/2677593002
Cr-Commit-Position: refs/heads/master@{#447910}

This will measure how often the Form-Not-Secure warnings ("Login not secure" or
"Payment not secure" in the autofill dropdown) are clicked on to
show more information about the warning.

BUG=687823
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2665313002
Cr-Commit-Position: refs/heads/master@{#447909}

Currently there are no crash handlers for uncaught java exceptions
in child processes. Add a chrome debug url to crash the GPU process
with an uncaught java exception to aid in testing this code path.

The plumbing is same as chrome://gpucrash, except everything is
behind OS_ANDROID guards. The actual implementation of throwing
the exception is in base.

BUG=680775
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2664373002
Cr-Commit-Position: refs/heads/master@{#447908}

This is very similar to DomTreeExtractor (which we should remove
later) except it uses DOM.getFlattenedDocument.  This is useful for
very large documents which might otherwise blow up the JSON parse
which currently has a nesting depth limit of 200

BUG=546953

Review-Url: https://codereview.chromium.org/2673773002
Cr-Commit-Position: refs/heads/master@{#447907}

Currently, ResourceProvider members which are set once at init and never
modified are not clearly identified. This makes it hard to see what
members can be used in a threadsafe manner.

This change moves these members to a Settings struct and makes the
settings "const", making it clearer how these may be used.

R=vmpstr@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2662303004
Cr-Commit-Position: refs/heads/master@{#447906}

This fixes site-details related tests in Vulcanized mode as well as the site
details view when the --enable-site-settings runtime flag is supplied.

When in Vulcanized mode, the tests can no longer include site_details.html
directly, since all URLs redirect to the top-level doc. <site-details> was not
imported anywhere from settings.html's transitive deps, making it impossible for
the tests to pass in Vulcanized mode.

BUG=673825
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2669033008
Cr-Commit-Position: refs/heads/master@{#447905}

BUG=678847

Review-Url: https://codereview.chromium.org/2667283005
Cr-Commit-Position: refs/heads/master@{#447904}

CompositingRecorder::endCompositing has an optimization [1] to turn
patterns like [..., begin compositing, drawing, end compositing]
into [..., composited drawing] which takes advantage of an SkPicture
recording optimization. An SkPictureBuilder is not needed in this
case, as we can modify the underlying display list while creating the
new composited drawing. To do this, a DCHECK in DrawingRecorder that
the underlying list doesn't change has been suppressed with
DisableListModificationCheck.

The last user of DisableNullPaintPropertyChecks has been removed which
lets us delete code from PaintChunker.

[1] https://codereview.chromium.org/2186643002

BUG=628831
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_layout_tests_slimming_paint_v2

Review-Url: https://codereview.chromium.org/2675773003
Cr-Commit-Position: refs/heads/master@{#447903}

Modifies ServiceSyncObserver to recieve SyncService* to the notifying
service in all of it's methods.

Also adds OnShutdown() notification.

This will support adding a long-lived observer in
https://codereview.chromium.org/2653693004/

TBR=sdefresne,treib

BUG=

Review-Url: https://codereview.chromium.org/2657673004
Cr-Commit-Position: refs/heads/master@{#447902}

BUG=none

Review-Url: https://codereview.chromium.org/2655393003
Cr-Commit-Position: refs/heads/master@{#447901}

This CL lands support for QR/barcode detection in Mac by
adding the correspondent files in services/shape_detection.

Most of the work is reshuffling code that was specific to
Face detection, to suit both Face and Barcode detection:

- Removed shape_detection_service_dispatcher.h and merged
 its two methods as statics in render_process_host_impl.cc.

- Moved common service/ code to detection_utils_mac.h/mm
 (1 helper function)

** Important note re. testing **
Chromium Mac bots seem to have troubles running CoreImage
methods, probably because they are virtual machines (no GPU).
This prevents testing the Barcode detection API properly, and also
renders void the current Face detection test. So, the plan is:
a) Remove the current Mac-Face detection test, because is not
testing anything, really (there was a bug in the error-return path
that is cleaned up here).
b) land (imminently) new unit tests for face and barcode, running
in Mac Gpu bots.

This CL does a); b) will be done in a subsequent CL.

BUG=665150
TEST=
./out/gn/Chromium.app/Contents/MacOS/Chromium --enable-blink-features=ShapeDetection,GeometryInterfaces  https://codepen.io/miguelao/full/wgrYjZ
NOTE: unittests for {face/barcode}_detection_impl_mac.mm coming
imminently after this).

Review-Url: https://codereview.chromium.org/2655303005
Cr-Commit-Position: refs/heads/master@{#447900}

BUG=none

Review-Url: https://codereview.chromium.org/2644233007
Cr-Commit-Position: refs/heads/master@{#447899}

Make a copy of the sink-id before removing it from the map. This is
because the id variable is owned by the GpuCompositorFrameSink in the
map. So when the sink is removed from the map, the id is also destroyed.
But the map can continue to iterate and try to use it. Making a copy of
it avoids this.

BUG=none

Review-Url: https://codereview.chromium.org/2673823002
Cr-Commit-Position: refs/heads/master@{#447898}

This patch initialized the underlying ThreadSpecific<WTFThreadData>*
in a separate code path during initialization, to avoid a branch in
wtfThreadData().

BUG=621786

Review-Url: https://codereview.chromium.org/2646003003
Cr-Commit-Position: refs/heads/master@{#447897}

Revert of Convert make_qualified_names and make_element_factory to use JSON5. (patchset #5 id:80001 of https://codereview.chromium.org/2645283006/ )

Reason for revert:
Unexpected compile failure --
https://build.chromium.org/p/chromium/builders/Linux%20x64/builds/32431

Original issue's description:
> Convert make_qualified_names and make_element_factory to use JSON5.
>
> Convert make_qualified_names.py, make_element_factory.py,
> make_element_type_helpers.py, and make_element_lookup_trie.py
> to use JSON5 config format.
>
> Converted these config files accordingly:
>   Source/core/html/HTMLAttributeNames.in
>   Source/core/html/HTMLTagNames.in
>   Source/core/html/parser/MathMLAttributeNames.in
>   Source/core/html/parser/MathMLTagNames.in
>   Source/core/svg/SVGAttributeNames.in
>   Source/core/svg/SVGTagNames.in
>   Source/core/svg/xlinkattrs.in
>   Source/core/xml/xmlattrs.in
>   Source/core/xml/xmlnsattrs.in
>
> Note that these files are updated together because:
> 1. make_element_factory.py depends on make_qualified_names.py
> 2. HTMLTagNames.in is common input to the scripts make_element_factory,
> make_element_type_helpers and make_element_lookup_trie
>
> Removed FIXME in make_element_type_helpers.py since that case
> no longer occurs.
>
> Also fixed bug in json5_generator where parameters should not be
> validated if default_parameters is not provided.
> (caught when I had typo on default_parameters in the script)
>
> BUG=677884
>
> Review-Url: https://codereview.chromium.org/2645283006
> Cr-Original-Commit-Position: refs/heads/master@{#447450}
> Committed: https://chromium.googlesource.com/chromium/src/+/f5ca5654596eeda38c2a796bbbad299d92b7f008
> Review-Url: https://codereview.chromium.org/2645283006
> Cr-Commit-Position: refs/heads/master@{#447891}
> Committed: https://chromium.googlesource.com/chromium/src/+/701ccf95a0ec04657b59ad240e28be4085fe033f

TBR=nainar@chromium.org,sashab@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true
BUG=677884

Review-Url: https://codereview.chromium.org/2677603002
Cr-Commit-Position: refs/heads/master@{#447896}

Revert of Allow using goma on Windows with symbol_level == 2 (patchset #2 id:20001 of https://codereview.chromium.org/2661023010/ )

Reason for revert:
It broke amd64-generic Trusty informational build at generate_build_files:
ERROR at //build/config/compiler/compiler.gni:115:3: Assertion failed.
  assert(is_win_fastlink || !use_goma,
  ^-----
Goma builds that use symbol_level 2 must use is_win_fastlink.
See //BUILD.gn:11:1: whence it was imported.
import("//build/config/compiler/compiler.gni")
^--------------------------------------------
GN gen failed: 1
step returned non-zero exit code: 1
@@@STEP_FAILURE@@@

See crbug.com/688203 for more info.

Original issue's description:
> Allow using goma on Windows with symbol_level == 2
>
> Traditionally goma for Chrome has been less useful on Windows than on
> other platforms because it was incompatible with full debug information.
> Building with goma requires using /Z7 instead of /Zi, and this causes
> the linker's memory usage and runtime to blow up as all of the debug
> information is merged.
>
> However, /debug:fastlink makes this work. Because it doesn't merge all
> of the debug information it makes goma and /Z7 practical. Full release
> component builds can be done in less than fifteen minutes, with
> incremental builds taking just a few seconds. Without goma a full
> release component build of Chrome can easily take 40+ minutes, even on a
> Z840.
>
> Goma's speedup comes from massively parallelizing the compile phase,
> however even with /debug:fastlink the linking phases are longer with the
> /Z7 switch that is required by goma. A /debug:fastlink of chrome.dll in
> a component build goes from ~32 seconds to ~110 seconds on a Z620 when
> /Z7 (goma) is selected. This penalty will be reduced by VC++ 2017 which
> claims 30% speed improvements on /debug:fastlink.
>
> So, if you frequently need to do full relinks then goma may still be a
> bad choice. However for most scenarios this change should make goma a
> good choice for component builds of Chrome, even with full debug
> information enabled. To make use of this ability you need to explicitly
> specify the switches below - symbol_level will otherwise default to 1
> when use_goma == true.
>
>     use_goma = true
>     is_win_fastlink = true
>     symbol_level = 2
>
> The fastlink PDBs work well for most scenarios but there are a few
> scenarios (rare for most people) where they do not work:
> - Copying PDBs to another machine (fails due to references to the .obj
> files)
> - Reporting on all symbols/globals with SymbolSort or similar fails
> - ETW tracing fails
> - VS heap profiling fails
> Ideally mspdbcmf.exe would let us create "normal" PDBs when needed but
> in practice this appears to be unusable with /Z7 created PDBs.
>
> R=scottmg@chromium.org
>
> Review-Url: https://codereview.chromium.org/2661023010
> Cr-Commit-Position: refs/heads/master@{#447892}
> Committed: https://chromium.googlesource.com/chromium/src/+/4a3cadb2d998fb8f477d0ebc7ff13619b123a805

TBR=scottmg@chromium.org,dpranke@chromium.org,brucedawson@chromium.org
# Skipping CQ checks because original CL landed less than 1 days ago.
NOPRESUBMIT=true
NOTREECHECKS=true
NOTRY=true

Review-Url: https://codereview.chromium.org/2669373002
Cr-Commit-Position: refs/heads/master@{#447895}

https://chromium.googlesource.com/external/github.com/catapult-project/catapult.git/+log/b61c782e8622..53604dda661b

$ git log b61c782e8..53604dda6 --date=short --no-merges --format='%ad %ae %s'
2017-02-02 benjhayden Merge Breakdown diagnostics.
2017-02-02 benjhayden Add chromium and v8 commit positions to RevisionInfo.
2017-02-02 benjhayden Fix related histogram links.

Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls

CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.android:android_optional_gpu_tests_rel
TBR=catapult-sheriff@chromium.org

Review-Url: https://codereview.chromium.org/2673723003
Cr-Commit-Position: refs/heads/master@{#447894}

Make the GetTransferMatrix and GetRangeAdjustMatrix functions part of
gfx::ColorSpace instead of gfx::ColorSpaceTransform, and call them from
GLRenderer to compute the YUV to RGB matrix and offset.

Do not actually use the new computation results yet, because they break
pixel tests. Rather, add a DCHECK that the new results are very close to
the previous values being used (within 1/255.f).

Fix the transfer matrix for SMPTE170M to have Kb=0.114 instead of 0.144.

Change the definition of gfx::ColorSpace::CreateJPEG to use the Rec601
transfer matrix, not Rec709. This is necessary for the resulting YUV to
RGB conversion to match the previous behavior.

BUG=667966
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2670773002
Cr-Commit-Position: refs/heads/master@{#447893}

Traditionally goma for Chrome has been less useful on Windows than on
other platforms because it was incompatible with full debug information.
Building with goma requires using /Z7 instead of /Zi, and this causes
the linker's memory usage and runtime to blow up as all of the debug
information is merged.

However, /debug:fastlink makes this work. Because it doesn't merge all
of the debug information it makes goma and /Z7 practical. Full release
component builds can be done in less than fifteen minutes, with
incremental builds taking just a few seconds. Without goma a full
release component build of Chrome can easily take 40+ minutes, even on a
Z840.

Goma's speedup comes from massively parallelizing the compile phase,
however even with /debug:fastlink the linking phases are longer with the
/Z7 switch that is required by goma. A /debug:fastlink of chrome.dll in
a component build goes from ~32 seconds to ~110 seconds on a Z620 when
/Z7 (goma) is selected. This penalty will be reduced by VC++ 2017 which
claims 30% speed improvements on /debug:fastlink.

So, if you frequently need to do full relinks then goma may still be a
bad choice. However for most scenarios this change should make goma a
good choice for component builds of Chrome, even with full debug
information enabled. To make use of this ability you need to explicitly
specify the switches below - symbol_level will otherwise default to 1
when use_goma == true.

    use_goma = true
    is_win_fastlink = true
    symbol_level = 2

The fastlink PDBs work well for most scenarios but there are a few
scenarios (rare for most people) where they do not work:
- Copying PDBs to another machine (fails due to references to the .obj
files)
- Reporting on all symbols/globals with SymbolSort or similar fails
- ETW tracing fails
- VS heap profiling fails
Ideally mspdbcmf.exe would let us create "normal" PDBs when needed but
in practice this appears to be unusable with /Z7 created PDBs.

R=scottmg@chromium.org

Review-Url: https://codereview.chromium.org/2661023010
Cr-Commit-Position: refs/heads/master@{#447892}

Convert make_qualified_names.py, make_element_factory.py,
make_element_type_helpers.py, and make_element_lookup_trie.py
to use JSON5 config format.

Converted these config files accordingly:
  Source/core/html/HTMLAttributeNames.in
  Source/core/html/HTMLTagNames.in
  Source/core/html/parser/MathMLAttributeNames.in
  Source/core/html/parser/MathMLTagNames.in
  Source/core/svg/SVGAttributeNames.in
  Source/core/svg/SVGTagNames.in
  Source/core/svg/xlinkattrs.in
  Source/core/xml/xmlattrs.in
  Source/core/xml/xmlnsattrs.in

Note that these files are updated together because:
1. make_element_factory.py depends on make_qualified_names.py
2. HTMLTagNames.in is common input to the scripts make_element_factory,
make_element_type_helpers and make_element_lookup_trie

Removed FIXME in make_element_type_helpers.py since that case
no longer occurs.

Also fixed bug in json5_generator where parameters should not be
validated if default_parameters is not provided.
(caught when I had typo on default_parameters in the script)

BUG=677884

Review-Url: https://codereview.chromium.org/2645283006
Cr-Original-Commit-Position: refs/heads/master@{#447450}
Committed: https://chromium.googlesource.com/chromium/src/+/f5ca5654596eeda38c2a796bbbad299d92b7f008
Review-Url: https://codereview.chromium.org/2645283006
Cr-Commit-Position: refs/heads/master@{#447891}

Adds some comments that would have been helpful to me as I traced the code to understand it. I think they will be helpful to future readers of the code.

BUG=687424

Review-Url: https://codereview.chromium.org/2667023004
Cr-Commit-Position: refs/heads/master@{#447890}

https://skia.googlesource.com/skia.git/+log/78b6786beccb..1db215a8efb4

$ git log 78b6786be..1db215a8e --date=short --no-merges --format='%ad %ae %s'
2017-02-02 mtklein iOS: dequote args one level
2017-02-02 mtklein iOS: try running the binary
2017-02-02 bsalomon Revert "Revert "Spot shadow cleanup and optimization""
2017-02-02 halcanary bin/list-skia-bots
2017-02-02 msarett Blacklist large images on 32-bit Win 2k8

Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls

CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel
TBR=halcanary@google.com

Review-Url: https://codereview.chromium.org/2669203004
Cr-Commit-Position: refs/heads/master@{#447889}

This change:
1) Changes the title string to not include the trailing colon (this affects the
   non-Harmony version too)
2) Switches from 9px to 8px for the spacing between items (affects non-Harmony)
3) Makes the indent on permission entries conditional on non-Harmony

BUG=605667

Review-Url: https://codereview.chromium.org/2654123004
Cr-Commit-Position: refs/heads/master@{#447888}

Small code clean-up to avoid verbose standalone vector init.

R=vmpstr@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2672553002
Cr-Commit-Position: refs/heads/master@{#447887}

BUG=684152
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:closure_compilation

Review-Url: https://codereview.chromium.org/2669923002
Cr-Commit-Position: refs/heads/master@{#447886}

Previously, we were using the term 'whitelist' to refer to what is more
correctly a feature policy declaration (a feature name and a whitelist of
origins), and overloading the term 'feature policy' to refer to the contents of
a feature policy header (a collection of declarations). This CL tries to
rationalize the terminology used for class naming in blink and in content.

This is part 1 of 5 in the effort to move the FeaturePolicy implementation
into the content layer, which will facilitate use of the framework for browser-
based policy decisions.

See the other CLs in this series:
[1/5] (This CL)                                   (Rename classes)
[2/5] https://codereview.chromium.org/2654873004/ (Centralize content-side code)
[3/5] https://codereview.chromium.org/2655663004/ (Maintain parallel FP in browser)
[4/5] https://codereview.chromium.org/2651883008/ (Use content/ FP in blink)
[5/5] https://codereview.chromium.org/2656533004/ (Remove unused blink code.)

BUG=685195
CQ_INCLUDE_TRYBOTS=master.tryserver.chromium.linux:linux_site_isolation

Review-Url: https://codereview.chromium.org/2648423002
Cr-Commit-Position: refs/heads/master@{#447885}

Add support for --enable-logging and --logging-level for headless mode.

BUG=546953
TEST=cr run chrome --headless --enable-logging --v=2

Review-Url: https://codereview.chromium.org/2671713004
Cr-Commit-Position: refs/heads/master@{#447884}

BUG=624694

Review-Url: https://codereview.chromium.org/2648433003
Cr-Commit-Position: refs/heads/master@{#447883}

This patch reduces the time spent in ::FromDrawImage call by 50%. Comes
with a corresponding perftest. This is a frequently used function
when it comes to dealing with software images, so any savings should
be good.

Perftest result without the sidc change (z620):
*RESULT software_image_decode_cache_fromdrawimage: result= 318435.53125 runs/s

With the change:
*RESULT software_image_decode_cache_fromdrawimage: result= 657221.0625 runs/s (+106%)

R=enne@chromium.org, ericrk@chromium.org
CQ_INCLUDE_TRYBOTS=master.tryserver.blink:linux_trusty_blink_rel

Review-Url: https://codereview.chromium.org/2669933002
Cr-Commit-Position: refs/heads/master@{#447882}

I also renamed some stuff to improve clarity (I couldn't initially
determine why GetBorderColor didn't return the same thing as
kBorderColor). This shouldn't cause any behavioral change. The unneeded
complexity here was probably left over from pre-MD codepaths.

BUG=648281

Review-Url: https://codereview.chromium.org/2673503002
Cr-Commit-Position: refs/heads/master@{#447881}

BUG=688127

Review-Url: https://codereview.chromium.org/2671713005
Cr-Commit-Position: refs/heads/master@{#447880}

Whenever visibility of device type icon in the audio row of system menu
changes, layout needs to be invalidated so that the row is re-layed out.

BUG=680171
TEST=manual

Review-Url: https://codereview.chromium.org/2671593003
Cr-Commit-Position: refs/heads/master@{#447879}

BUG=652781

+rogerta@ for '+google_apis', DEPS change.

Review-Url: https://codereview.chromium.org/2661153003
Cr-Commit-Position: refs/heads/master@{#447878}

for V8 roll.

BUG=chromium:688167
TBR=dgozman@chromium.org
NOTRY=true

Review-Url: https://codereview.chromium.org/2673793002
Cr-Commit-Position: refs/heads/master@{#447877}

We need to check if primaryFont() returns null.

BUG=687701

Review-Url: https://codereview.chromium.org/2672743002
Cr-Commit-Position: refs/heads/master@{#447876}

After this, `git cl format` will reorder includes in blocks of #includes
that are not separated by newlines.

This works in almost all cases, but it can break some code e.g.

  #include <windows.h>
  #include <shellapi.h>

clang-format will reorder these now, but shellapi.h only compiles if
windows.h was included first. Relying on this is brittle, so replace
code like this with

  #include <windows.h>

  // Must be after windows.h:
  #include <shellapi.h>

Since clang-format doesn't reorder across blocks, this will do the right
thing.

This also means you're still on the hook of putting blocks with user headers,
C++ headers, and C headers in the right order.

This will hopefully replace src/tools/sort-headers.py which contains
some hacky heuristics -- but just inserting newlines between includes
when needed (with a comment) seems like a better tradeoff anyhow.
And the automatic integration with `git cl format` is nice.

(clang-format has IncludeIsMainRegex and IncludeCategories for adding
heuristics, but we shouldn't use these, they're too complicated.)

BUG=688155
TBR=brettw

Review-Url: https://codereview.chromium.org/2669263003
Cr-Commit-Position: refs/heads/master@{#447875}

ICCProfiles are given an id when they are created, and are assumed to
be unique.

This breaks down in some tests, where we create ICCProfiles in the
browser and send them to the renderer (as usual), but then also create
test ICCProfiles in the in the renderer, that can conflict with the
ids provided by the browser.

Fix this by using hard-coded ids for the test ICCProfiles that are
created by the renderer.

BUG=634102

Review-Url: https://codereview.chromium.org/2663153002
Cr-Commit-Position: refs/heads/master@{#447874}

https://pdfium.googlesource.com/pdfium.git/+log/4bd0d99c1110..44b297bb2c94

$ git log 4bd0d99c1..44b297bb2 --date=short --no-merges --format='%ad %ae %s'
2017-02-02 dsinclair Clean up CSS code to match styles.

Documentation for the AutoRoller is here:
https://skia.googlesource.com/buildbot/+/master/autoroll/README.md

If the roll is causing failures, see:
http://www.chromium.org/developers/tree-sheriffs/sheriff-details-chromium#TOC-Failures-due-to-DEPS-rolls

TBR=dsinclair@chromium.org

Review-Url: https://codereview.chromium.org/2670243002
Cr-Commit-Position: refs/heads/master@{#447873}