diff options
Diffstat (limited to '')
| -rw-r--r-- | results/classifier/gemma3:12b/hypervisor/864 | 16 | ||||
| -rw-r--r-- | results/classifier/gemma3:12b/hypervisor/864490 | 11 |
2 files changed, 27 insertions, 0 deletions
diff --git a/results/classifier/gemma3:12b/hypervisor/864 b/results/classifier/gemma3:12b/hypervisor/864 new file mode 100644 index 000000000..ca986b2a6 --- /dev/null +++ b/results/classifier/gemma3:12b/hypervisor/864 @@ -0,0 +1,16 @@ + +HVF virtual counter diverges from CLOCK_VIRTUAL when the host sleeps +Description of problem: +HVF's virtual counter diverges from `CLOCK_VIRTUAL` when the host sleeps and causes the inconsistency between Linux's system counter and everything else. + +HVF's virtual counter apparently relies on something similar to `mach_absolute_time`, which stops when the host sleeps and resumes after it wakes up. However, `CLOCK_VIRTUAL` is implemented with `mach_continuous_time`, which continues even while the host sleeps. Linux uses the virtual counter as the source of the system counter and sees inconsistencies between the system counter and the other devices. +Steps to reproduce: +1. Launch Fedora. +2. Compare the time shown at the top of the guest display and one at the top of the host display. The difference should be less than 2 minutes. +3. Let the host sleep for 3 minutes. +4. Compare the times again. The difference is now greater than 2 minutes. +Additional information: +Here are solutions I've came up with so far. There are trade-offs but any of them should be better than the current situation. I'm happy to implement one if the maintainers have decided which one is the best or figure out a superior alternative. +- Implement `cpus_get_virtual_clock` of `AccelOpsClass` with `mach_absolute_time`. It would make HVF inconsistent with the other accelerators. Linux also expects the virtual clock is "continuous" and it leaves the divergence from the real time. +- Request XNU `HOST_NOTIFY_CALENDAR_CHANGE` to update the virtual clock with the continuous time. The interface is undocumented. +- Use `IORegisterForSystemPower` to update the virtual clock with the continuous time. It is undocumented that the interface handles every cases where `mach_absolute_time` and `mach_continuous_time`, but it actually does if I read XNU's source code correctly. diff --git a/results/classifier/gemma3:12b/hypervisor/864490 b/results/classifier/gemma3:12b/hypervisor/864490 new file mode 100644 index 000000000..fb497cda4 --- /dev/null +++ b/results/classifier/gemma3:12b/hypervisor/864490 @@ -0,0 +1,11 @@ + +Windows 2008 x64 (SBS Server) freezes randomly when using more than 1 CPU core + +This issue has been giving headache to us since a long time. +Difficult to reproduce as it happens randomly. +We had this issue when we ran Windows 2008 x64 or Windows SBS Server guests in either XEN 3.3 or Proxmox environments. +When only one CPU core is assigned to the guest, everything is fine. If 2 or more cores are assigned, the guest stops responding after several hours - and in the host machine one of the cores is using 100%. The only thing that helps is resetting the guest. + +I am ready to provide logs/crashdumps if needed, because we want to help resolve this issue. I saw some posts on the web of people having the same problems - for some of the workaround was to fix some BIOS settings, but we did not have success with those (e.g. disabling C1E Support and Intel C-State ) + +Server is running on Intel® Core™ i7-920 Quad-Core, 24 Gig RAM. \ No newline at end of file |