#21767 closed defect (invalid)
Can't restore the VM state saved before the Windows host reboot caused by host OS Updates
| Reported by: | boxer01 | Owned by: | |
|---|---|---|---|
| Component: | other | Version: | VirtualBox-7.0.10 |
| Keywords: | reboot save state | Cc: | |
| Guest type: | Windows | Host type: | Windows |
Description
If one runs the Windows updates on the host system and then saves the state of the VM before the restart of the host, this state can't be restored after the reboot of the host. I put the logs here. It's an old behavior, I link to the forum topic and put an old logs from 6.1 with the same bug here.
Sometimes there are even problems to save the state just before such reboot.
Attachments (3)
Change History (8)
by , 9 months ago
| Attachment: | Vbox_logs_update_16092022.7z added |
|---|
follow-up: 2 comment:1 by , 9 months ago
@boxer01: Both 7z files contain VBox.log files from other VirtualBox versions (7.0.0_BETA2 and 7.0.8).
comment:2 by , 9 months ago
Replying to fth0:
@boxer01: Both 7z files contain VBox.log files from other VirtualBox versions (7.0.0_BETA2 and 7.0.8).
Thank you for your answer. You are right, it isn't from the current version. Since the discussion / topic in forum in September 2022 I try to provoke this situation. And this happened with the Windows update last week. So I put it here.
I don't think that this is version dependent, because it also happened with 6.1 versions previously. So I think, unfortunately, it could probably also happen with current 7.0.10 version. If the version in the ticket is important for you - then you should change it, because I can't do it. I just left the one which already was here on ticket's creation, sorry for that. ;-(
comment:3 by , 9 months ago
FWIW, the exception took place in the VMSVGA FIFO thread, so it's probably graphics related, and there have been many graphics related bugfixes in VirtualBox 7.0.10.
If the version in the ticket is important for you [...]
It's just misleading information, because it indicates a crash that did not happen yet. I'm not saying that it cannot happen, though. Since I'm a VirtualBox user like you, I cannot change the ticket either. ;)
comment:4 by , 9 months ago
FWIW, the crash seems to happen in the TPM-related code, so this issue could be related to #21773.
Sorry, I got mislead by several occurrences of _plat_ACT_GetSignaled() in a large structure with function pointers, perhaps for Intel's CET.
comment:5 by , 9 months ago
| Resolution: | → invalid |
|---|---|
| Status: | new → closed |
The crash is in the VMSVGA FIFO thread, so not TPM related. What I noticed is that this doesn't seem to be an officially supported VirtualBox installation but some portable version from here. We don't support such variants. If you are able to reproduce this with an officially supported VirtualBox distribution from Oracle please reopen this defect.
Logs from 6.1