Summary: | [JSCOnly Add an armv7 JSCOnly EWS that runs tests | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Guillaume Emont <guijemont> | ||||
Component: | Tools / Tests | Assignee: | Nobody <webkit-unassigned> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | aakash_jain, abarth, commit-queue, eric, ews-watchlist, glenn, guijemont+jsc-armv7-ews, lforschler, mcatanzaro, webkit-bug-importer | ||||
Priority: | P2 | Keywords: | InRadar | ||||
Version: | WebKit Nightly Build | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Bug Depends on: | 190190 | ||||||
Bug Blocks: | |||||||
Attachments: |
|
Description
Guillaume Emont
2018-10-02 01:05:57 PDT
Created attachment 351597 [details]
Patch
WIP patch for early review, not to be committed as I am still working on preparing the bot.
(In reply to Guillaume Emont from comment #1) > Created attachment 351597 [details] > Patch > > WIP patch for early review, not to be committed as I am still working on > preparing the bot. Also, this will need #190190 to be fixed to have a chance to work. *** Bug 190189 has been marked as a duplicate of this bug. *** Comment on attachment 351597 [details] Patch Clearing flags on attachment: 351597 Committed r237001: <https://trac.webkit.org/changeset/237001> All reviewed patches have been landed. Closing bug. Reading a little late that I was supposed[1] to add people to the bug ...oops! Does either of you need to do some manual operation to enable the bot, or did I forget something in the patch? The ews seems to be running on the builder, but never gets patches to process: 2018-10-10 07:23:28,109 - Fetching next work item for jsc-armv7-ews 2018-10-10 07:23:28,838 - No work item. Sleeping until 2018-10-10 07:25:28 (120 seconds). Meanwhile this EWS does not appear on https://webkit-queues.webkit.org/ or on test bubbles (tried with Bug 190428 ). [1] https://trac.webkit.org/wiki/EWS Aakash, do you need to restart the ews server? (In reply to Lucas Forschler from comment #8) > Aakash, do you need to restart the ews server? Yes, Restarted it. The change is now live: https://webkit-queues.webkit.org/queue-status/jsc-armv7-ews New issue (log from bot): 2018-10-12 09:39:26,014 - Running: webkit-patch --status-host=webkit-queues.webkit.org --bot-id=igalia-jsc-armv7-ews apply-attachment --no-update --non-interac tive 352173 --port=jsc-only 2018-10-12 09:47:42,032 - Failed to run "['/home/buildbot/ews/WebKit/Tools/Scripts/webkit-patch', '--status-host=webkit-queues.webkit.org', '--bot-id=igalia-js c-armv7-ews', 'apply-attachment', '--no-update', '--non-interactive', 352173, '--port=jsc-only']" exit_code: 1 cwd: /home/buildbot/ews/WebKit No handlers could be found for logger "webkitpy.common.system.user" Logging in as guijemont+jsc-armv7-ews@igalia.com... Received URLError: "unknown url type: https" while loading https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1. Retrying in 10 seconds... Received URLError: "unknown url type: https" while loading https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1. Retrying in 15.0 seconds... Received URLError: "unknown url type: https" while loading https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1. Retrying in 22.5 seconds... Received URLError: "unknown url type: https" while loading https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1. Retrying in 33.75 seconds... Received URLError: "unknown url type: https" while loading https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1. Retrying in 50.625 seconds... Received URLError: "unknown url type: https" while loading https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1. Retrying in 75.9375 seconds... Received URLError: "unknown url type: https" while loading https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1. Retrying in 113.90625 seconds... Received URLError: "unknown url type: https" while loading https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1. Retrying in 170.859375 seconds... Traceback (most recent call last): File "/home/buildbot/ews/WebKit/Tools/Scripts/webkit-patch", line 84, in <module> main() File "/home/buildbot/ews/WebKit/Tools/Scripts/webkit-patch", line 79, in main WebKitPatch(os.path.abspath(__file__)).main() File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/tool/multicommandtool.py", line 305, in main result = command.check_arguments_and_execute(options, args, self) File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/tool/multicommandtool.py", line 123, in check_arguments_and_execute return self.execute(options, args, tool) or 0 File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/tool/commands/download.py", line 180, in execute patches = self._fetch_list_of_patches_to_process(options, args, tool) File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/tool/commands/download.py", line 224, in _fetch_list_of_patches_to_process patch = tool.bugs.fetch_attachment(patch_id, throw_on_access_error=True) File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/bugzilla/bugzilla.py", line 552, in fetch_attachment bug_id = self.bug_id_for_attachment_id(attachment_id, throw_on_access_error) File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/bugzilla/bugzilla.py", line 530, in bug_id_for_attachment_id return NetworkTransaction().run(lambda: self.get_bug_id_for_attachment_id(attachment_id, throw_on_access_error)) File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/networktransaction.py", line 53, in run return request() File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/bugzilla/bugzilla.py", line 530, in <lambda> return NetworkTransaction().run(lambda: self.get_bug_id_for_attachment_id(attachment_id, throw_on_access_error)) File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/bugzilla/bugzilla.py", line 533, in get_bug_id_for_attachment_id self.authenticate() File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/bugzilla/bugzilla.py", line 576, in authenticate "index.cgi?GoAheadAndLogIn=1") File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/bugzilla/bugzilla.py", line 335, in open_url return NetworkTransaction().run(lambda: self.browser.open(url), url) File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/networktransaction.py", line 61, in run self._check_for_timeout() File "/home/buildbot/ews/WebKit/Tools/Scripts/webkitpy/common/net/networktransaction.py", line 67, in _check_for_timeout raise NetworkTimeout() webkitpy.common.net.networktransaction.NetworkTimeout: NetworkTimeout 2018-10-12 09:47:42,033 - Patch does not apply 2018-10-12 09:47:43,757 - Error: jsc-armv7-ews unable to apply patch. 2018-10-12 09:47:45,175 - Releasing work item 352173 from jsc-armv7-ews (In reply to Guillaume Emont from comment #10) > New issue (log from bot): Actually visible in https://webkit-queues.webkit.org/results/9552471 Double checked, and the credentials in .git/config seem correct for the account guijemont+jsc-armv7-ews@igalia.com > Actually visible in https://webkit-queues.webkit.org/results/9552471 Need to debug. Might be related to: https://bugs.webkit.org/show_bug.cgi?id=187061 Also, please make sure that this bot(igalia-jsc-armv7-ews) can actually communicate with Bugzilla? (ping, curl etc.) (In reply to Aakash Jain from comment #13) > > Actually visible in https://webkit-queues.webkit.org/results/9552471 > > Need to debug. Might be related to: > https://bugs.webkit.org/show_bug.cgi?id=187061 > > Also, please make sure that this bot(igalia-jsc-armv7-ews) can actually > communicate with Bugzilla? (ping, curl etc.) Yes, I can wget https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1 from the host no problem. Looking at this bug and if there is any issue with our python install. (In reply to Guillaume Emont (jsc-armv7-ews) from comment #14) Manually running this works fine on the host FWIW: urllib2.urlopen("https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1").read() (In reply to Guillaume Emont (jsc-armv7-ews) from comment #15) > (In reply to Guillaume Emont (jsc-armv7-ews) from comment #14) > Manually running this works fine on the host FWIW: > > urllib2.urlopen("https://bugs.webkit.org/index.cgi?GoAheadAndLogIn=1").read() Oops. We have another python in the PATH, that is part of the buildroot "host" image used to generate the toolchain. This python indeed fails when trying the above, so I guess the script ends up using the wrong python (it is before in the path). I'll see what I can do about that. Changing the order of things in PATH seems to have fixed the python issue. Now to the next issue: https://webkit-queues.webkit.org/results/9553802 (In reply to Guillaume Emont (jsc-armv7-ews) from comment #17) > Changing the order of things in PATH seems to have fixed the python issue. > Now to the next issue: https://webkit-queues.webkit.org/results/9553802 Can you manually run this command on the bot and see if that works? ['/home/buildbot/ews/WebKit/Tools/Scripts/webkit-patch', '--status-host=webkit-queues.webkit.org', '--bot-id=igalia-jsc-armv7-ews', 'build-and-test', '--no-clean', '--no-update', '--test', '--non-interactive', '--build-style=release', u'--group=jsc', '--port=jsc-only'] Also please run without 'u' to check if the unicode in u'--group=jsc' is causing any issues. manually added --jsc-only as a parameter in the run script, though I suspect I shouldn't have to. It is now building its first patch. (In reply to Guillaume Emont (jsc-armv7-ews) from comment #19) > manually added --jsc-only as a parameter in the run script, though I suspect > I shouldn't have to. It is now building its first patch. And the bug we have now is that my dealing of JSCTESTS_OPTIONS is not correct, I should split the arguments into a list. Patch upcoming (In reply to Guillaume Emont (jsc-armv7-ews) from comment #20) > (In reply to Guillaume Emont (jsc-armv7-ews) from comment #19) > > manually added --jsc-only as a parameter in the run script, though I suspect > > I shouldn't have to. It is now building its first patch. > > And the bug we have now is that my dealing of JSCTESTS_OPTIONS is not > correct, I should split the arguments into a list. Patch upcoming Patch went to Bug 190537. Now I need to understand the weird build errors that I can't reproduce outside of the EWS, such as https://webkit-queues.webkit.org/results/9559484 EWS seems to be running fine now. I might have found an issue (Bug 190662) while debugging it though. |