We're not quite ready for this, but almost so, according to the latest try run: https://treeherder.mozilla.org/#/jobs?repo=try&revision=f439ce85793b I'll make the patch so it's ready to go as soon as the tests are green.

Just prep work; won't land this just yet.

Comment on attachment 8683139 [details] [diff] [review] Schedule e10s mochitest-devtools-chrome on Windows 7, Review of attachment 8683139 [details] [diff] [review]: ----------------------------------------------------------------- can we combine this with the section where we turned on m-e10s(gl) ? then we have one line to check the platform/twigs, and we just have more += statements. As this patch is, it looks good, fix the small nit and/or combine with the other section if it makes sense. ::: mozilla-tests/config.py @@ +2473,5 @@ > + if platform == 'win32' and slave_platform == 'win7-ix': > + if name in TWIGS or ('gecko_version' in branch and branch['gecko_version'] != trunk_gecko_version): > + pass > + else: > + branch['platforms'][platform][slave_platform]['debug_unittest_suites'] += MOCHITEST_DT_8_E10S nit: trailing whitespace.

Comment on attachment 8683139 [details] [diff] [review] Schedule e10s mochitest-devtools-chrome on Windows 7, Review of attachment 8683139 [details] [diff] [review]: ----------------------------------------------------------------- ::: mozilla-tests/config.py @@ +2470,5 @@ > branch['platforms'][platform][slave_platform]['debug_unittest_suites'] += WEB_PLATFORM_TESTS_CHUNKED_MORE_E10S[:] + WEB_PLATFORM_REFTESTS_E10S > branch['platforms'][platform][slave_platform]['opt_unittest_suites'] += WEB_PLATFORM_TESTS_CHUNKED_E10S[:] + WEB_PLATFORM_REFTESTS_E10S[:] > + # Windows 7 > + if platform == 'win32' and slave_platform == 'win7-ix': > + if name in TWIGS or ('gecko_version' in branch and branch['gecko_version'] != trunk_gecko_version): NOTE: not all twigs have gecko_version === trunk, e.g. cypress doesn't track trunk. And I thought date didn't, but it seems to not set it right now.

Update to use same block as is used to turn on mochitest-gl-e10s

Comment on attachment 8683630 [details] [diff] [review] Schedule e10s mochitest-devtools-chrome on Windows 7, Carry r+ forward

There are a few failures yet on the latest try run: https://treeherder.mozilla.org/#/jobs?repo=try&revision=fcc0bbf423dd I'll disable a few on opt and see if this greens it up sufficiently to at least enable it there.

latest try run with a few more disabled tests: https://treeherder.mozilla.org/#/jobs?repo=try&revision=bd8676c74163

we're almost at green, one more try run: https://treeherder.mozilla.org/#/jobs?repo=try&revision=37a2925c56fc

really close! once more, with feeling: https://treeherder.mozilla.org/#/jobs?repo=try&revision=1d734e6e8f7e

Bug 1221499 - Disable tests that fail on e10s windows, r=jmaher

Comment on attachment 8688737 [details] MozReview Request: Bug 1221499 - Disable tests that fail on e10s windows, r=jmaher https://reviewboard.mozilla.org/r/25389/#review22945 I would really like to see bug numbers associated with these skip-if statements. Normally we do something like this: skip-if = e10s # bug 3141592 I know we should land this sooner than later, so if getting bugs with actual links to the failures will take some time, then lets land and update the manifests after the fact.

(In reply to Joel Maher (:jmaher) from comment #11) > Comment on attachment 8688737 [details] > MozReview Request: Bug 1221499 - Disable tests that fail on e10s windows, > r=jmaher > > https://reviewboard.mozilla.org/r/25389/#review22945 > > I would really like to see bug numbers associated with these skip-if > statements. Normally we do something like this: > skip-if = e10s # bug 3141592 > > I know we should land this sooner than later, so if getting bugs with actual > links to the failures will take some time, then lets land and update the > manifests after the fact. I talked with the e10s guys and they didn't think it would be too useful to file individual bugs for these, since we don't know the root causes...they were OK with a generic disable.

ok, that sounds good

Comment on attachment 8683630 [details] [diff] [review] Schedule e10s mochitest-devtools-chrome on Windows 7, https://hg.mozilla.org/build/buildbot-configs/rev/3f56adbd971a

Merged to production at http://hg.mozilla.org/build/buildbot-configs/rev/61e319b19e0b

Skipping at least two tests so far in https://hg.mozilla.org/mozilla-central/rev/8d9c87b4c312 Remains to be seen how deep this hole will go.

Will green up a failing chunk separately in bug 1227394

Removing leave-open keyword from resolved bugs, per :sylvestre.