January 2018 - Wine-bugs - list.winehq.org

[Bug 39441] New: The reverts keep getting slower
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

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

[Bug 37104] New: Infinite revert loop
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

https://bugs.winehq.org/show_bug.cgi?id=37104 Bug ID: 37104 Summary: Infinite revert loop 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 The current task scheduling algorithm can enter into an infinite revert loop while trying to prepare VMs for the next tasks. Assume the following settings: $MaxRevertingVMs = 2; $MaxRevertsWhileRunningVMs = 0; $MaxActiveVMs = 2; Then the following sequence can play out: | Steps VM | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 ----+-----+-----+-----+-----+-----+-----+-----+----- vm1 | rev | idl | off | off | rev | rev | rev | ... vm2 | rev | rev | rev | idl | off | off | rev | ... vm3 | off | off | rev | rev | rev | idl | off | ... The issue happens in steps 2, 4 and 6. The scheduler can shut down idle VMs to replace them with VMs that are more appropriate for the upcoming tasks. This is what happens in these steps: it decides the idle VM it just prepared is not what it wants after all, and thus shuts it down and prepares another one. The problem is it keeps changing its mind over and over and can never actually start a task because there is always a reverting VM and $MaxRevertsWhileRunningVMs = 0. Another prerequisite for this scenario to play out is probably to have multiple tasks have the exact same priority, so that their order is undefined. But regardless, the scheduler should probably not be shutting down an idle VM that has an actual 'pending' task. -- 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 39470] New: https://bugs.winehq.org/page.cgi?id=fields.html needs to be updated to include new statuses
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

https://bugs.winehq.org/show_bug.cgi?id=39470 Bug ID: 39470 Summary: https://bugs.winehq.org/page.cgi?id=fields.html needs to be updated to include new statuses Product: WineHQ Bugzilla Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: bugzilla-unknown Assignee: wine-bugs(a)winehq.org Reporter: dimesio(a)earthlink.net Distribution: --- NEEDINFO and STAGED need to be added and UPSTREAM needs to be changed to NOTOURBUG. -- 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 7

[Bug 42160] New: Need ddraw component
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

https://bugs.winehq.org/show_bug.cgi?id=42160 Bug ID: 42160 Summary: Need ddraw component Product: WineHQ Bugzilla Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: bugzilla-unknown Assignee: wine-bugs(a)winehq.org Reporter: alexhenrie24(a)gmail.com Distribution: --- I just tried to file a bug in ddraw.dll and couldn't find any component option for it. Something like "directx-ddraw" would be great. -- 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 3

[Bug 44080] New: Update descriptions in https://bugs.winehq.org/ enter_bug.cgi to use HTTPS URLs
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

https://bugs.winehq.org/show_bug.cgi?id=44080 Bug ID: 44080 Summary: Update descriptions in https://bugs.winehq.org/enter_bug.cgi to use HTTPS URLs 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: pmenzel+bugs.winehq.org(a)molgen.mpg.de CC: austinenglish(a)gmail.com Distribution: --- Please use the more secure HTTPS URLs in the descriptions of the Bugzilla components. -- 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 39411] New: Provide a way to monitor VM revert times
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

https://bugs.winehq.org/show_bug.cgi?id=39411 Bug ID: 39411 Summary: Provide a way to monitor VM revert times 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: --- It would be nice to have a graph of the VM revert times to verify that everything is working fine (or conversely to get a better sense that things are wrong and action is needed). Currently the way to do so is to ask Alexandre for the WineTestBot log, run a perl script on it to get a CSV, and import the CSV in a spreadsheet to graph the data. That's pretty cumbersome and obviously does not permit regular monitoring. Providing per-VM revert time graphs, maybe using something like RRDtool would be really nice. But then one could argue that revert times should just be low and stay that way and that then monitoring them may not be needed. -- 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 3

[Bug 31784] New: Add support for VM snapshots
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

http://bugs.winehq.org/show_bug.cgi?id=31784 Bug #: 31784 Summary: Add support for VM snapshots 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: fgouget(a)codeweavers.com Classification: Unclassified As we expand the number of configurations to test it seems wasteful to have to create and maintain a brand new VM just for them. In a lot of cases all we need is to slightly reconfigure a VM to add a CD image, add network interfaces or change the locale and run the test with these changes settings. That's where snapshots can be useful: take a snapshot for each configuration we need to test and then let the TestBot start the VM from the right snapshot. There is a catch though: while it's possible to run the tests in two VMs in parallel, for a given VM only one snapshot can be active at a time. There's two consequences: * TestBot must know about snapshot and not try to simultaneously run the tests in two snapshots belonging to the same VM. * If in the base configurations (those all Wine patches run on) we have a VM with ten snapshots, it means we will need to run the tests ten times *sequentially*. That VM is quickly going to become a bottleneck. So we should still limit the number of snapshots per VM, at least in the base configurations. -- 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 8

[Bug 39927] New: Update the build VM
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

https://bugs.winehq.org/show_bug.cgi?id=39927 Bug ID: 39927 Summary: Update the build VM 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 current build VM is based on Debian 7.0 which is pretty old. It would be nice to upgrade it to Debian 8.0. * This might solve some issues with the VM sometimes going dead after a Reconfig job. See: https://www.winehq.org/pipermail/wine-devel/2016-January/111059.html * This would also get us a newer MinGW compiler. This would get us newer headers, in particular for GetConsoleScreenBufferInfoEx(). Note however that since this is a Vista+ API and the tests must still run on Windows XP it means that specific API would still have to be loaded dynamically in order for the test to be accepted in Wine. But it would make things easier for quick / preliminary tests. See: https://www.winehq.org/pipermail/wine-devel/2015-December/110942.html -- 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 3

[Bug 31789] New: Fix handling of jscript/vbscript tests
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

http://bugs.winehq.org/show_bug.cgi?id=31789 Bug #: 31789 Summary: Fix handling of jscript/vbscript tests 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: fgouget(a)codeweavers.com Classification: Unclassified The Wine TestBot reports failures for all patches to the jscript/vbscript tests: === W7PROX64 (32 bit) === No test summary line found This generates quite a bit of spam on wine-devel and is annoying and needs to be fixed. I believe it also happens with some other tests. -- 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 9

[Bug 39410] New: Provide a way for the WineTestBot administrator to get the WineTestBot log
by wine-bugs＠winehq.org 04 Nov '18

04 Nov '18

https://bugs.winehq.org/show_bug.cgi?id=39410 Bug ID: 39410 Summary: Provide a way for the WineTestBot administrator to get the WineTestBot log 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: --- When a task fails it logs the failure into the WineTestBot log(*). The WineTestBot Engine may also log issues into the log. Analyzing the log also provides a way to figure out the evolution of the revert times. But accessing the log requires having login access to the WineTestBot server which the administrator may not have. So it would be nice to have access to the log (or portions thereof) through an administrator-only section of the website. (*) In the case of WineRunBuild and WineRunTask one will get most of that in the corresponding task's log on the website, but that's not the case for RevertVM for instance. -- 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