New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 685414 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner:
Last visit > 30 days ago
Closed: Feb 2017
Cc:
Components:
EstimatedDays: ----
NextAction: ----
OS: ----
Pri: 3
Type: Bug



Sign in to add a comment

cbuildbot --remote cannot detect videotest failure

Project Member Reported by warx@chromium.org, Jan 25 2017

Issue description

ChromePFQ has build failure video_ChromeHWDecodeUsed failures. I tried bisecting using cbuildbot --remote tricky-tot-chrome-pfq-informational --chrome_version=[xxx]

The lower bound chromium revision of bad build still passes. Local device test also shows the test failure.

So, there are something different betwen chrome-pfq builds and cbuildbot. Yesterday, we got some time wasted on waiting for the results from cbuildbot.


Link to the fact: https://bugs.chromium.org/p/chromium/issues/detail?id=674209#c12
 
Cc: ihf@chromium.org
Maybe related to official build vs. non-official builds and which codecs are available?

Comment 2 by warx@chromium.org, Jan 25 2017

if non-official builds on cbuildbot, they should fail on all h264 videotests. My guess is they just skipped videotests as the log is very very sparse: https://uberchromegw.corp.google.com/i/chromiumos.tryserver/builders/chrome_pfq_informational/builds/29/steps/HWTest%20%5Bbvt-cq%5D/logs/stdio

Comment 3 by ihf@chromium.org, Jan 26 2017

Yes, it didn't run anything on hardware. Only VMTest. I find it difficult to convince tryjobs to use actual hardware, at least in one pass. It would be very nice if it did though.
Components: -Infra>Platform>Buildbot>TryServer Infra>Client>ChromeOS

Comment 5 by aut...@google.com, Feb 9 2017

Owner: warx@chromium.org
Did you use the --hwtest flag? If not, this skip is expected

Comment 6 by warx@chromium.org, Feb 10 2017

I didn't. So this could be a Wontfix and be updated in YAQS?

Comment 7 by warx@chromium.org, Feb 16 2017

Status: WontFix (was: Untriaged)
Verified in a recent run with --hwtest

Sign in to add a comment