Discovered after bug 1377180 being solved, test_subprocess_environmentAppend()[1] fails[2] on the new taskcluster windows instances[3] (still tier 3): > 02:39:50 WARNING - TEST-UNEXPECTED-FAIL | toolkit/modules/subprocess/test/xpcshell/test_subprocess.js | test_subprocess_environmentAppend - [test_subprocess_environmentAppend : 708] Got expected $PATH value - "Z:\\\\task_1499912358\\\\build\\\\venv\\\\Scripts;C:\\\\Program Files\\\\Mercurial;C:\\\\mozilla-build\\\\7zip;C:\\\\mozilla-build\\\\info-zip;C:\\\\ == "Z:\\\\task_1499912358\\\\build\\\\venv\\\\Scripts;C:\\\\Program Files\\\\Mercurial;C:\\\\mozilla-build\\\\7zip;C:\\\\mozilla-build\\\\info-zip;C:\\\\ At the moment, we don't have the full content of the variables. This will be fixed in bug 1380627. I'm filing this bug to keep track of that failure, so we can make XPCShell on TC Windows tier1 (bug 1293783) [1] https://dxr.mozilla.org/mozilla-central/rev/03bcd6d65af62c5e60a0ab9247ccce43885e707b/toolkit/modules/subprocess/test/xpcshell/test_subprocess.js#708 [2] https://treeherder.mozilla.org/logviewer.html#?job_id=113865559&repo=mozilla-central&lineNumber=10868 [3] https://treeherder.mozilla.org/#/jobs?repo=mozilla-central&revision=30ea2905130e85f9e1d8d56fa3097901eec6514b&filter-job_group_symbol=tc-X&filter-tier=3

Per [1], there is one extra occurrence of Z:\\task_1499958059\\build\\venv\\lib\\site-packages\\pywin32_system32 in the JS variable path: > 15:38:36 INFO - "Z:\\task_1499958059\\build\\venv\\Scripts;C:\\Program Files\\Mercurial;C:\\mozilla-build\\7zip;C:\\mozilla-build\\info-zip;C:\\mozilla-build\\kdiff3;C:\\mozilla-build\\moztools-x64\\bin;C:\\mozilla-build\\mozmake;C:\\mozilla-build\\msys\\bin;C:\\mozilla-build\\msys\\local\\bin;C:\\mozilla-build\\nsis-3.0b3;C:\\mozilla-build\\nsis-2.46u;C:\\mozilla-build\\python;C:\\mozilla-build\\python\\Scripts;C:\\mozilla-build\\upx391w;C:\\mozilla-build\\wget;C:\\mozilla-build\\yasm;C:\\windows\\system32;C:\\windows;C:\\windows\\System32\\Wbem;C:\\windows\\System32\\WindowsPowerShell\\v1.0\\;C:\\mozilla-build\\python27;C:\\mozilla-build\\python27\\Scripts;C:\\mozilla-build\\vim\\vim72;C:\\CoreUtils\\bin;C:\\mozilla-build\\buildbotve\\scripts;c:\\Program Files\\Microsoft Windows Performance Toolkit\\;C:\\Users\\cltbld\\AppData\\Local\\Programs\\Common\\Microsoft\\Visual C++ for Python\\9.0;C:\\ProgramData\\chocolatey\\bin;C:\\Program Files\\Puppet Labs\\Puppet\\bin;C:\\mozilla-build\\hg;Z:\\task_1499958059\\build\\venv\\Lib\\site-packages\\pywin32_system32;Z:\\task_1499958059\\build\\venv\\lib\\site-packages\\pywin32_system32;Z:\\task_1499958059\\build\\application\\firefox;Z:\\task_1499958059\\build\\venv\\lib\\site-packages\\pywin32_system32" 15:38:36 INFO - "Z:\\task_1499958059\\build\\venv\\Scripts;C:\\Program Files\\Mercurial;C:\\mozilla-build\\7zip;C:\\mozilla-build\\info-zip;C:\\mozilla-build\\kdiff3;C:\\mozilla-build\\moztools-x64\\bin;C:\\mozilla-build\\mozmake;C:\\mozilla-build\\msys\\bin;C:\\mozilla-build\\msys\\local\\bin;C:\\mozilla-build\\nsis-3.0b3;C:\\mozilla-build\\nsis-2.46u;C:\\mozilla-build\\python;C:\\mozilla-build\\python\\Scripts;C:\\mozilla-build\\upx391w;C:\\mozilla-build\\wget;C:\\mozilla-build\\yasm;C:\\windows\\system32;C:\\windows;C:\\windows\\System32\\Wbem;C:\\windows\\System32\\WindowsPowerShell\\v1.0\\;C:\\mozilla-build\\python27;C:\\mozilla-build\\python27\\Scripts;C:\\mozilla-build\\vim\\vim72;C:\\CoreUtils\\bin;C:\\mozilla-build\\buildbotve\\scripts;c:\\Program Files\\Microsoft Windows Performance Toolkit\\;C:\\Users\\cltbld\\AppData\\Local\\Programs\\Common\\Microsoft\\Visual C++ for Python\\9.0;C:\\ProgramData\\chocolatey\\bin;C:\\Program Files\\Puppet Labs\\Puppet\\bin;C:\\mozilla-build\\hg;Z:\\task_1499958059\\build\\venv\\Lib\\site-packages\\pywin32_system32;Z:\\task_1499958059\\build\\venv\\lib\\site-packages\\pywin32_system32;Z:\\task_1499958059\\build\\application\\firefox" I'm not sure what adds it. What do you think, Kris? [1] https://public-artifacts.taskcluster.net/Y4NPQescSL2qe_LtciU8nw/0/public/logs/live_backing.log

I guess it's added by the python wrapper for the virtualenv. We should probably just switch this test to check a custom variable, rather than $PATH, like we did in test_subprocess_environment.

> We should probably just switch this test to check a custom variable, rather than $PATH, like we did in test_subprocess_environment. Sounds good. Here's the patch attached, and here's its try run: https://treeherder.mozilla.org/#/jobs?repo=try&revision=e39bbcd4e58e6ca105b918ec9ad844b24ff68947&filter-tier=1&filter-tier=2&filter-tier=3&selectedJob=114791465

Comment on attachment 8887007 [details] Bug 1380628 - test_subprocess_environmentAppend fails on the new taskcluster windows instances https://reviewboard.mozilla.org/r/157768/#review163032 ::: toolkit/modules/subprocess/test/xpcshell/test_subprocess.js:701 (Diff revision 1) > + // Our Windows environment can't handle launching python without > + // PATH variables. > + if (AppConstants.platform == "win") { > + Object.assign(environment, { > + PATH: env.get("PATH"), > + PATHEXT: env.get("PATHEXT"), > + }); > + } > + This isn't necessary with `environmentAppend: true` ::: toolkit/modules/subprocess/test/xpcshell/test_subprocess.js:712 (Diff revision 1) > + }); > + } > + > let proc = await Subprocess.call({ > command: PYTHON, > - arguments: ["-u", TEST_SCRIPT, "env", "PATH", "FOO"], > + arguments: ["-u", TEST_SCRIPT, "env", "VALUE_ERASED_SOON", "UNTOUCHED_VALUE"], We need to test a value that gets inherited from the base environment, which means that we need to set one of these variables via `env.set` and only pass the other via the environment object.

Green try of the latest patch at: https://treeherder.mozilla.org/#/jobs?repo=try&revision=091bdcc1f53c018bbda8542ac366bb68634223cd&filter-tier=1&filter-tier=2&filter-tier=3

Comment on attachment 8887007 [details] Bug 1380628 - test_subprocess_environmentAppend fails on the new taskcluster windows instances https://reviewboard.mozilla.org/r/157768/#review164700

Pushed by jlorenzo@mozilla.com: https://hg.mozilla.org/integration/autoland/rev/4c64ef2265d5 test_subprocess_environmentAppend fails on the new taskcluster windows instances r=kmag

https://hg.mozilla.org/releases/mozilla-esr52/rev/ab62cf449772 https://hg.mozilla.org/releases/mozilla-esr52/rev/497079ac6d06