Updating the SwarmingIsolatedScriptTest to upload chartjson results to the

Updating the SwarmingIsolatedScriptTest to upload chartjson results to the
perf dashboard when present.  I am trying to replicate how we currently do
this for perf tests in scripts/slave/runtest.py, but I am not sure how
close I am.  I will iterate once this recipe is running on the fyi
waterfall.  It *shouldn't* break any existing behavior given that these
code paths will not execute unless chartjson is present in the results.

BUG=chromium:633253
Committed: https://chromium.googlesource.com/chromium/tools/build/+/d068176dade2dfc687b0625dd976b92737956843

[Ken Russell (switch to Gerrit), 2016-09-12 19:35:40]

kbr@chromium.org changed reviewers:
+ dpranke@chromium.org, kbr@chromium.org

[Ken Russell (switch to Gerrit), 2016-09-12 19:35:41]

Adding dpranke@ for opinions on this.

https://codereview.chromium.org/2330133002/diff/20001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/steps.py (right):

https://codereview.chromium.org/2330133002/diff/20001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1087: if 'chartjson' in
results and 'charts' in results['chartjson']:
Per comment on dependent CL https://codereview.chromium.org/2331993003/ : I
think that the chartjson results should be written to a separate file, by adding
another command line argument --isolated-script-chartjson-output similar to the
existing --isolated-script-test-output. It's known that we are going to need at
least one more output file format here for the flakiness dashboard, so instead
of overloading this simplified json format we should start enhancing the
interface between IsolatedScriptTest and recipes now.

https://codereview.chromium.org/2330133002/diff/20001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/swarming/api.py (right):

https://codereview.chromium.org/2330133002/diff/20001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/swarming/api.py:755:
merged_results[key].extend(results_json[key])
See above. I think we should write a new merge function which handles chartjson
results separately.

[eyaich1, 2016-09-13 16:52:59]

This CL is now dependent on https://codereview.chromium.org/2336293002/ which
updates the swarming api, but will not break anything if checked in first since
it does a check for the attribute first.

https://codereview.chromium.org/2330133002/diff/20001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/steps.py (right):

https://codereview.chromium.org/2330133002/diff/20001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1087: if 'chartjson' in
results and 'charts' in results['chartjson']:
On 2016/09/12 19:35:40, Ken Russell wrote:
> Per comment on dependent CL https://codereview.chromium.org/2331993003/ : I
> think that the chartjson results should be written to a separate file, by
adding
> another command line argument --isolated-script-chartjson-output similar to
the
> existing --isolated-script-test-output. It's known that we are going to need
at
> least one more output file format here for the flakiness dashboard, so instead
> of overloading this simplified json format we should start enhancing the
> interface between IsolatedScriptTest and recipes now.

Done.

https://codereview.chromium.org/2330133002/diff/20001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/swarming/api.py (right):

https://codereview.chromium.org/2330133002/diff/20001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/swarming/api.py:755:
merged_results[key].extend(results_json[key])
On 2016/09/12 19:35:41, Ken Russell wrote:
> See above. I think we should write a new merge function which handles
chartjson
> results separately.

Done.

[Ken Russell (switch to Gerrit), 2016-09-13 19:41:00]

Looks like a better direction. A few questions, mainly about how this will be
tested.

https://codereview.chromium.org/2330133002/diff/40001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/steps.py (right):

https://codereview.chromium.org/2330133002/diff/40001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1088:
self._output_charjson_results_if_present(api, step_result)
charjson -> chartjson

https://codereview.chromium.org/2330133002/diff/40001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1099: def
_output_charjson_results_if_present(self, api, step_result):
Is this code path tested? Have you sent a CQ dry run and/or run the recipe
simulation tests? They require 100% code coverage and I would be surprised if
this new code is tested. See for example scripts/slave/recipes/chromium.py and
several of the tests at the bottom, like
dynamic_swarmed_sharded_passed_isolated_script_test. I think you'll need to
write some canned output and simulate passes, failures, etc.

https://codereview.chromium.org/2330133002/diff/40001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1100: results =
getattr(step_result, 'isolated_script_chart_results', None) or {}
Also, for clarity, how about chart -> chartjson?

[eyaich1, 2016-09-14 18:42:30]

https://codereview.chromium.org/2330133002/diff/40001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/steps.py (right):

https://codereview.chromium.org/2330133002/diff/40001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1088:
self._output_charjson_results_if_present(api, step_result)
On 2016/09/13 19:41:00, Ken Russell wrote:
> charjson -> chartjson

Done.

https://codereview.chromium.org/2330133002/diff/40001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1099: def
_output_charjson_results_if_present(self, api, step_result):
On 2016/09/13 19:40:59, Ken Russell wrote:
> Is this code path tested? Have you sent a CQ dry run and/or run the recipe
> simulation tests? They require 100% code coverage and I would be surprised if
> this new code is tested. See for example scripts/slave/recipes/chromium.py and
> several of the tests at the bottom, like
> dynamic_swarmed_sharded_passed_isolated_script_test. I think you'll need to
> write some canned output and simulate passes, failures, etc.

Ah, so this is where they are called!  Ok making slightly more sense now.  As
per the comments in my other CL that updates the swarming api
(crrev.com/2336293002) I think once I create a test there it will also exercise
this code since it will write the output to the right place.  I am going to wait
on this test until I have it successful in the other CL.

https://codereview.chromium.org/2330133002/diff/40001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1100: results =
getattr(step_result, 'isolated_script_chart_results', None) or {}
On 2016/09/13 19:41:00, Ken Russell wrote:
> Also, for clarity, how about chart -> chartjson?

Done.

[Dirk Pranke, 2016-09-14 19:37:19]

I defer to kbr@ here, so lgtm when he's happy. 

(Unless kbr@ really wants me to also review it).

[Ken Russell (switch to Gerrit), 2016-09-17 03:34:44]

Since I gather this is dependent on https://codereview.chromium.org/2336293002/,
let's defer re-review until that's sorted out.

[eyaich1, 2016-09-20 18:01:18]

On 2016/09/17 03:34:44, Ken Russell wrote:
> Since I gather this is dependent on
https://codereview.chromium.org/2336293002/,
> let's defer re-review until that's sorted out.

Dependent cl has landed.  PTAL

[Ken Russell (switch to Gerrit), 2016-09-21 00:26:46]

On 2016/09/20 18:01:18, eyaich1 wrote:
> On 2016/09/17 03:34:44, Ken Russell wrote:
> > Since I gather this is dependent on
> https://codereview.chromium.org/2336293002/,
> > let's defer re-review until that's sorted out.
> 
> Dependent cl has landed.  PTAL

Let's revisit this once the revert of the other one's been sorted out.

[eyaich1, 2016-09-27 14:22:12]

On 2016/09/21 00:26:46, Ken Russell wrote:
> On 2016/09/20 18:01:18, eyaich1 wrote:
> > On 2016/09/17 03:34:44, Ken Russell wrote:
> > > Since I gather this is dependent on
> > https://codereview.chromium.org/2336293002/,
> > > let's defer re-review until that's sorted out.
> > 
> > Dependent cl has landed.  PTAL
> 
> Let's revisit this once the revert of the other one's been sorted out.

Dependent CL is checked in and not breaking anything.  This is now ready for
review.  PTAL

[Ken Russell (switch to Gerrit), 2016-09-28 20:34:26]

Looks good overall. A few questions.

Could you update the Subject and the commit message so there's a 1-line summary
at the top, followed by paragraphs? Thanks.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/steps.py (right):

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1119: blink_revision =
GetBlinkRevision(api.chromium.c.build_dir)
Blink's now folded into Chromium and doesn't have a separate revision. Is it
necessary to still maintain the concept of blink_revision? If so, can this just
use main_revision and can GetBlinkRevision be deleted?

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1120: revisions =
slave_utils.GetTelemetryRevisions(
Does this function need to be specialized to Telemetry? Could it be called
something like GetPerfDashboardRevisions or similar?

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1136: print 'Error: No json
output from telemetry.'
Perhaps 'Error: No perf dashboard JSON was produced'?

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/runtest.py
File scripts/slave/runtest.py (right):

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/runtest.p...
scripts/slave/runtest.py:421: def _GetBlinkRevision(options):
See above; I think this should no longer be needed.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/slave_uti...
File scripts/slave/slave_utils.py (right):

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/slave_uti...
scripts/slave/slave_utils.py:753: versions['webkit_rev'] = blink_revision
See above; I think this should either no longer be needed, or, at minimum, can
just be main_revision.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/slave_uti...
scripts/slave/slave_utils.py:788: def GetBlinkRevision(build_dir,
webkit_revision=None):
This logic is definitely no longer needed.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/unittests...
File scripts/slave/unittests/slave_utils_test.py (right):

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/unittests...
scripts/slave/unittests/slave_utils_test.py:24: CHROMIUM_LOG = """
These example logs are out of date; for example the source of truth is now git
so the git-svn-id entry is no longer present. Can they be updated? Or would
doing that in this CL cause too much churn in the recipe expectations?

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/unittests...
scripts/slave/unittests/slave_utils_test.py:39: BLINK_LOG = """
A separate Blink example log should no longer be needed.

[eyaich1, 2016-09-29 12:43:35]

I responded to most of your comments with TODOs as I would like to limit the
number of changes in this CL so I can just compare them side by side.  I am
happy to follow up in a cleanup CL once I know this recipe is doing the same
upload as the current perf recipe.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/recipe_mo...
File scripts/slave/recipe_modules/chromium_tests/steps.py (right):

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1119: blink_revision =
GetBlinkRevision(api.chromium.c.build_dir)
On 2016/09/28 20:34:26, Ken Russell wrote:
> Blink's now folded into Chromium and doesn't have a separate revision. Is it
> necessary to still maintain the concept of blink_revision? If so, can this
just
> use main_revision and can GetBlinkRevision be deleted?

That makes sense, but like I said in the notes I was basically just duplicating
what was in the existing recipe.  While I agree some cleanup needs to be done, I
would like to reduce the number of changes so I can compare the results of both
without too many additional changes.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1120: revisions =
slave_utils.GetTelemetryRevisions(
On 2016/09/28 20:34:26, Ken Russell wrote:
> Does this function need to be specialized to Telemetry? Could it be called
> something like GetPerfDashboardRevisions or similar?

Done.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/recipe_mo...
scripts/slave/recipe_modules/chromium_tests/steps.py:1136: print 'Error: No json
output from telemetry.'
On 2016/09/28 20:34:26, Ken Russell wrote:
> Perhaps 'Error: No perf dashboard JSON was produced'?

Done.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/runtest.py
File scripts/slave/runtest.py (right):

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/runtest.p...
scripts/slave/runtest.py:421: def _GetBlinkRevision(options):
On 2016/09/28 20:34:26, Ken Russell wrote:
> See above; I think this should no longer be needed.

Are you ok with me waiting to remove it for reasons in comment above?

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/slave_uti...
File scripts/slave/slave_utils.py (right):

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/slave_uti...
scripts/slave/slave_utils.py:753: versions['webkit_rev'] = blink_revision
On 2016/09/28 20:34:26, Ken Russell wrote:
> See above; I think this should either no longer be needed, or, at minimum, can
> just be main_revision.

Same as previous comments I would prefer to wait to remove. but added a todo
below.

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/slave_uti...
scripts/slave/slave_utils.py:788: def GetBlinkRevision(build_dir,
webkit_revision=None):
On 2016/09/28 20:34:26, Ken Russell wrote:
> This logic is definitely no longer needed.

Added a TODO

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/unittests...
File scripts/slave/unittests/slave_utils_test.py (right):

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/unittests...
scripts/slave/unittests/slave_utils_test.py:24: CHROMIUM_LOG = """
On 2016/09/28 20:34:26, Ken Russell wrote:
> These example logs are out of date; for example the source of truth is now git
> so the git-svn-id entry is no longer present. Can they be updated? Or would
> doing that in this CL cause too much churn in the recipe expectations?

Added a TODO

https://codereview.chromium.org/2330133002/diff/80001/scripts/slave/unittests...
scripts/slave/unittests/slave_utils_test.py:39: BLINK_LOG = """
On 2016/09/28 20:34:26, Ken Russell wrote:
> A separate Blink example log should no longer be needed.

Added a TODO

[Ken Russell (switch to Gerrit), 2016-09-30 15:03:48]

TODOs and later cleanups are fine. Sorry if I missed earlier comments about this
intending to maintain the perf recipe's functionality. LGTM

[eyaich1, 2016-09-30 15:09:25]

The CQ bit was checked by eyaich@chromium.org

[eyaich1, 2016-09-30 15:09:25]

The patchset sent to the CQ was uploaded after l-g-t-m from dpranke@chromium.org
Link to the patchset: https://codereview.chromium.org/2330133002/#ps100001
(title: "Responding to review comments")

[commit-bot: I haz the power, 2016-09-30 15:09:33]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-30 15:16:30]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-09-30 15:16:31]

Try jobs failed on following builders:
  Build Presubmit on luci.infra.try (JOB_FAILED,
https://luci-milo.appspot.com/swarming/task/3194769d24e4a210)

[eyaich1, 2016-09-30 17:50:25]

I forgot to update the expectations and coverage.  I had to add no cover for now
due to a bug in the expectations framework where I can't set mock properties
that have a '-' in them and perf-id is a necessary property for the dashboard
results.  

PTAL

[Ken Russell (switch to Gerrit), 2016-09-30 21:45:32]

LGTM again with a couple of comments.

https://codereview.chromium.org/2330133002/diff/120001/scripts/slave/recipe_m...
File scripts/slave/recipe_modules/chromium_tests/steps.py (right):

https://codereview.chromium.org/2330133002/diff/120001/scripts/slave/recipe_m...
scripts/slave/recipe_modules/chromium_tests/steps.py:1129: if dashboard_json: #
pragma: no cover
Could you add a TODO or comment related to the addition of the "no cover"
pragma? From your description in the CL it's a pretty serious problem and should
be documented.

https://codereview.chromium.org/2330133002/diff/120001/scripts/slave/results_...
File scripts/slave/results_dashboard.py (right):

https://codereview.chromium.org/2330133002/diff/120001/scripts/slave/results_...
scripts/slave/results_dashboard.py:239: print "1"
Remove debugging prints -- unless you really meant to leave them in in
production, in which case add TODOs about removing them.

[eyaich1, 2016-10-03 13:32:56]

Thanks!

https://codereview.chromium.org/2330133002/diff/120001/scripts/slave/recipe_m...
File scripts/slave/recipe_modules/chromium_tests/steps.py (right):

https://codereview.chromium.org/2330133002/diff/120001/scripts/slave/recipe_m...
scripts/slave/recipe_modules/chromium_tests/steps.py:1129: if dashboard_json: #
pragma: no cover
On 2016/09/30 21:45:32, Ken Russell wrote:
> Could you add a TODO or comment related to the addition of the "no cover"
> pragma? From your description in the CL it's a pretty serious problem and
should
> be documented.

Done.  It was just a small bug in the infrastructure code for the tests to
accept additional formattting names in the properties.  that might have even
already landed.  I will update the coverage asap in a follow on cl.

https://codereview.chromium.org/2330133002/diff/120001/scripts/slave/results_...
File scripts/slave/results_dashboard.py (right):

https://codereview.chromium.org/2330133002/diff/120001/scripts/slave/results_...
scripts/slave/results_dashboard.py:239: print "1"
On 2016/09/30 21:45:32, Ken Russell wrote:
> Remove debugging prints -- unless you really meant to leave them in in
> production, in which case add TODOs about removing them.

Done.

[eyaich1, 2016-10-03 13:33:07]

The CQ bit was checked by eyaich@chromium.org

[eyaich1, 2016-10-03 13:33:08]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, kbr@chromium.org
Link to the patchset: https://codereview.chromium.org/2330133002/#ps140001
(title: "Responding to review comments")

[commit-bot: I haz the power, 2016-10-03 13:33:11]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-03 13:41:19]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-10-03 13:41:20]

Try jobs failed on following builders:
  Build Presubmit on luci.infra.try (JOB_FAILED,
https://luci-milo.appspot.com/swarming/task/31a3913c8f8c1210)

[eyaich1, 2016-10-03 14:05:55]

The CQ bit was checked by eyaich@chromium.org

[eyaich1, 2016-10-03 14:05:56]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, kbr@chromium.org
Link to the patchset: https://codereview.chromium.org/2330133002/#ps160001
(title: "Adding an expectation")

[commit-bot: I haz the power, 2016-10-03 14:05:59]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-03 14:14:08]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-10-03 14:14:09]

Try jobs failed on following builders:
  Build Presubmit on luci.infra.try (JOB_FAILED,
https://luci-milo.appspot.com/swarming/task/31a3af56718ed910)

[eyaich1, 2016-10-04 15:55:52]

On 2016/10/03 14:14:09, commit-bot: I haz the power wrote:
> Try jobs failed on following builders:
>   Build Presubmit on luci.infra.try (JOB_FAILED,
> https://luci-milo.appspot.com/swarming/task/31a3af56718ed910)

OK, Another iteration of this CL.  Needed to configure the recipe differently
since it can't access the slave module directly.  Added a script that executes
the upload instead. 

PTAL

[Ken Russell (switch to Gerrit), 2016-10-05 01:07:21]

On 2016/10/04 15:55:52, eyaich1 wrote:
> On 2016/10/03 14:14:09, commit-bot: I haz the power wrote:
> > Try jobs failed on following builders:
> >   Build Presubmit on luci.infra.try (JOB_FAILED,
> > https://luci-milo.appspot.com/swarming/task/31a3af56718ed910)
> 
> OK, Another iteration of this CL.  Needed to configure the recipe differently
> since it can't access the slave module directly.  Added a script that executes
> the upload instead. 
> 
> PTAL

LGTM as far as I can tell from code examination.

[eyaich1, 2016-10-05 13:44:13]

The CQ bit was checked by eyaich@chromium.org

[eyaich1, 2016-10-05 13:44:14]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, kbr@chromium.org
Link to the patchset: https://codereview.chromium.org/2330133002/#ps200001
(title: "Adding additional test coverage")

[commit-bot: I haz the power, 2016-10-05 13:44:24]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-05 13:52:52]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-10-05 13:52:53]

Try jobs failed on following builders:
  Build Presubmit on luci.infra.try (JOB_FAILED,
https://luci-milo.appspot.com/swarming/task/31ade83d86f78b10)

[eyaich1, 2016-10-05 13:57:42]

The CQ bit was checked by eyaich@chromium.org

[eyaich1, 2016-10-05 13:57:43]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, kbr@chromium.org
Link to the patchset: https://codereview.chromium.org/2330133002/#ps220001
(title: "Updating slave utils")

[commit-bot: I haz the power, 2016-10-05 13:57:53]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-05 14:06:25]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-10-05 14:06:27]

Try jobs failed on following builders:
  Build Presubmit on luci.infra.try (JOB_FAILED,
https://luci-milo.appspot.com/swarming/task/31adf8433a2f9910)

[eyaich1, 2016-10-05 14:25:41]

The CQ bit was checked by eyaich@chromium.org

[eyaich1, 2016-10-05 14:25:42]

The patchset sent to the CQ was uploaded after l-g-t-m from
dpranke@chromium.org, kbr@chromium.org
Link to the patchset: https://codereview.chromium.org/2330133002/#ps240001
(title: "Fixing presubmit issues")

[commit-bot: I haz the power, 2016-10-05 14:25:47]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-10-05 14:32:33]

Description was changed from

==========
Updating the SwarmingIsolatedScriptTest to upload chartjson results to the
perf dashboard when present.  I am trying to replicate how we currently do
this for perf tests in scripts/slave/runtest.py, but I am not sure how
close I am.  I will iterate once this recipe is running on the fyi
waterfall.  It *shouldn't* break any existing behavior given that these
code paths will not execute unless chartjson is present in the results.

BUG=chromium:633253
==========

to

==========
Updating the SwarmingIsolatedScriptTest to upload chartjson results to the
perf dashboard when present.  I am trying to replicate how we currently do
this for perf tests in scripts/slave/runtest.py, but I am not sure how
close I am.  I will iterate once this recipe is running on the fyi
waterfall.  It *shouldn't* break any existing behavior given that these
code paths will not execute unless chartjson is present in the results.

BUG=chromium:633253

Committed:
https://chromium.googlesource.com/chromium/tools/build/+/d068176dade2dfc687b0...
==========

[commit-bot: I haz the power, 2016-10-05 14:32:34]

Committed patchset #13 (id:240001) as
https://chromium.googlesource.com/chromium/tools/build/+/d068176dade2dfc687b0...

[eyaich1, 2016-10-05 18:42:04]

A revert of this CL (patchset #13 id:240001) has been created in
https://codereview.chromium.org/2391353002/ by eyaich@chromium.org.

The reason for reverting is: Broke the cronet bots

https://build.chromium.org/p/chromium.android/waterfall?builder=Android+Crone...

See crbug.com/653167

Will break up this CL into the refactor of slave utils and run test (which is
what I think broke the existing waterfall since it doesn't run through the new
code in chromium_tests).