April 2016 - Wine-Bugs - list.winehq.org

[Bug 39487] New: Make step dependencies more flexible
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

https://bugs.winehq.org/show_bug.cgi?id=39487 Bug ID: 39487 Summary: Make step dependencies more flexible Product: Wine-Testbot Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- The TestBot jobs are split into steps and tasks. * The steps are numbered and must be executed in order. Each step has an associated file. * Each step has one or more tasks which can all be run in parallel. Typically the first step is the build step associated to the patch. It produces the 32 and 64 bit executables for the following steps hence why it has to be executed first. When it fails the other steps are skipped which is not the case for the other steps. So the general structure is: Step 1 - patch.diff - Build Step 2 - test_32.exe - One task per VM Step 3 - test_64.exe - One task per VM So one source of inefficiency is that the 64 bit tests cannot start until all the 32 bit tests have completed. It does not necessarily make the overall job execution longer but it can leave a VM host idle while it waits for the remaining 32 bit tests to complete. And if a developer is watching the job's page he will not know of 64 bit failures until then. So while the build step has to come before all others, that condition should be relaxed for the other steps. The database schema presented in bug 39412 adds a DependencyNo field to the Step table so that step dependencies can be made explicit. This would allow making both test steps depend on build one. Note that with this scheme a given step could not directly depend on multiple other steps but that does not seem needed. It would also make more explicit the fact that test steps cannot be run if the build one fails instead of hard-coding that behavior in Jobs::UpdateStatus(). -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1 2

[Bug 39425] New: Improve resilience to VM host outages
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

https://bugs.winehq.org/show_bug.cgi?id=39425 Bug ID: 39425 Summary: Improve resilience to VM host outages Product: Wine-Testbot Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- Currently the WineTestBot gets stuck when the connection to the libvirt server on the VM hosts is broken. This includes cases where the libvirt server is restarted, the VM hosts is rebooted or cases where there's a network outage. The reason is that the Engine queries the status of the VMs itself in some circumstances. This creates a TCP connection which is never recreated in case it breaks. The proper fix is to banish all such queries from the Engine: not just to fix this issue but also because some of these operation can be long (a few seconds) and block the main loop of the single-threaded Engine, which can in turn cause the website to lag. -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1 2

[Bug 39385] New: Make the VM descriptions visible
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

https://bugs.winehq.org/show_bug.cgi?id=39385 Bug ID: 39385 Summary: Make the VM descriptions visible Product: Wine-Testbot Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- Each VM has a long description of its hardware and software configuration. They are meant to help developers figure out what's special about a specific VM and how it may explain its results. The long descriptions can be found on the test.winehq.org site by looking at individual reports information. But they cannot be seen on the testbot.winehq.org site except by the administrator when editing the VM configuration. They should be visible as a tooltip/infotip in most places when the VM's name is displayed. -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1 2

[Bug 34256] New: Please add "See Also" support for NetBSD, FreeBSD, etc.
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

http://bugs.winehq.org/show_bug.cgi?id=34256 Bug #: 34256 Summary: Please add "See Also" support for NetBSD, FreeBSD, etc. Product: WineHQ Bugzilla Version: unspecified Platform: x86-64 URL: http://source.winehq.org/git/bugzilla.git/blob/HEAD:/B ugzilla/Bug.pm#l2854 OS/Version: Linux Status: NEW Keywords: download, source Severity: enhancement Priority: P2 Component: bugzilla-unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: kennybobs(a)o2.co.uk Classification: Unclassified Created attachment 45590 --> http://bugs.winehq.org/attachment.cgi?id=45590 Error screen See attached screenshot. Could someone with the relevant skills please add support for adding bug trackers for the following to the "See Also" link? FreeBSD (http://www.freebsd.org/support/bugreports.html) Linux (https://bugzilla.kernel.org/) Illumos/OpenIndiana (https://www.illumos.org/issues/) NetBSD (http://www.netbsd.org/support/query-pr.html) Solaris (http://bugs.sun.com/) OpenBSD has shut down their bug tracking system. *shrugs* http://www.openbsd.org/query-pr.html Not urgent, but useful. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 5

[Bug 39414] New: Add a partway job status
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

https://bugs.winehq.org/show_bug.cgi?id=39414 Bug ID: 39414 Summary: Add a partway job status Product: Wine-Testbot Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- If a task is completed the corresponding job's status is 'Running' even if none or the remaining tasks is actually Running. So one can sometimes end up with dozens of 'Running' jobs when only one or two among them actually has an active task. So the idea would be to reserve the 'Running' status to jobs for which one task is actually running. Jobs where some tasks but not all have completed would instead have the 'Partway' status (suggestions for another name welcome). Another issue is that when multiple jobs are in the 'Running' or 'Partway' state it's hard to know which ones have been modified the most recently because the 'Ended' field is not set. This is probably less important though and a waterfall like presentation would help (see bug 39413). -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1 5

[Bug 40240] New: ntdll:exception causes the Windows 10 VM to crash
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

https://bugs.winehq.org/show_bug.cgi?id=40240 Bug ID: 40240 Summary: ntdll:exception causes the Windows 10 VM to crash Product: Wine-Testbot Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- The ntdll:exception test causes the Windows 10 VM (w1064) to crash. Given that this does not happen on real hardware it is a TestBot bug. Incidentally this is why we don't get any WineTest result from that VM. Either there is a way to avoid the crash by adjusting the QEMU configuration or a QEMU bug should be filed. -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1 13

[Bug 39413] New: Make it easier to monitor the TestBot activity
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

https://bugs.winehq.org/show_bug.cgi?id=39413 Bug ID: 39413 Summary: Make it easier to monitor the TestBot activity Product: Wine-Testbot Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- At the bottom of the main page there is a list of the VMs and their status. However the information there is incomplete: * When a VM is reverting there is no way to know for how long it has been reverting. * When a VM is running a task there is no link to the corresponding task. * There is no history of events so if someone reports that something strange happened there is no way to check(*). So it would be nice to have something like BuildBot's Waterfall display. http://buildbot.buildbot.net/waterfall (*) Actually one could check the WineTestBot log but that has its own problems and requires poring over lots of traces. See bug 39410. -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1 2

[Bug 32354] New: testbot: A crashing test is not detected
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

http://bugs.winehq.org/show_bug.cgi?id=32354 Bug #: 32354 Summary: testbot: A crashing test is not detected Product: Wine-Testbot Version: unspecified Platform: x86 OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: wine.dev(a)web.de Classification: Unclassified A 64bit test crashed on all 64bit testbot machines, but the summary has "0" as "Number of failures". Test run for the broken patch: http://testbot.winehq.org/JobDetails.pl?Key=22963&log_301=1#k301 Later test: http://testbot.winehq.org/JobDetails.pl?Key=23106&log_302=1#k302 (I send a fix for the broken code in wine in some minutes) -- By by ... Detlef -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 10

[Bug 40237] New: account not deletable
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

https://bugs.winehq.org/show_bug.cgi?id=40237 Bug ID: 40237 Summary: account not deletable Product: WineHQ Bugzilla Version: unspecified Hardware: x86 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: bugzilla-unknown Assignee: wine-bugs(a)winehq.org Reporter: BerlinerPaul(a)web.de Distribution: --- hi, I want to delete my account but I'm unable to find some link in the preferences for this. best -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1 4

[Bug 39441] New: The reverts keep getting slower
by wine-bugs＠winehq.org Nov. 4, 2018

Nov. 4, 2018

https://bugs.winehq.org/show_bug.cgi?id=39441 Bug ID: 39441 Summary: The reverts keep getting slower Product: Wine-Testbot Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: unknown Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- A revert that would take under 10 seconds right after creating the snapshot would take over 7 minutes 6 months later. So every few months this would cause the TestBot to become really sluggish and barely able to keep up with the patch influx. Restarting libvirt, rebooting the host, restoring the VM from backup or even transferring it to another host had no effect on the revert time. While the revert is taking place the QEmu process fully occupies one core, no disk I/O is performed and the VM is not running. The exact reason is not yet known exactly but it seems to have to do with the VM's timer devices, particularly the rtc one. To confuse matters further not all VMs are affected: only Windows 2000, XP, 2003, 2008 and 10 suffer from this. The other post Windows Vista are immune. Yet the guest is not active while the revert is taking place so it should not have an impact on it. Finally at WineConf 2015 it was discovered that the revert time of a live snapshot is simply proportional to the snapshot's age. This yielded a first workaround which is to refresh the live snapshots regularly. Further investigation showed that the common point between the impacted live snapshots is that they all have the following clock settings <clock offset='localtime'> <timer name='rtc' tickpolicy='delay'/> <timer name='pit' tickpolicy='delay'/> <timer name='hpet' present='no'/> </clock> while the unaffected VMs have <clock offset='localtime'/> But switching from the former to the latter does not fix the affected VMs. Still this lead to a better fix which has now been put in place: setting track='guest' on the rtc timer. Regardless, something is wrong with the way QEmu handles timers and live snapshots so a bug was reported: https://bugs.launchpad.net/qemu/+bug/1505041 Maybe this will shed some light on what's really happening and what the correct timer settings are. -- Do not reply to this email, post in Bugzilla using the above URL to reply. You are receiving this mail because: You are watching all bug changes.

1 4