VirtualBox

Changes between Initial Version and Version 1 of Ticket #19226, comment 69


Ignore:
Timestamp:
11/13/2020 05:15:50 PM (23 months ago)
Author:
GnomeUser
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #19226, comment 69

    initial v1  
    33There are three separate issues we discuss here: 
    44 
    5 1) segfaults and crashes on the 6.1.16. This is what this issue about and it seams to be fixed in 6.1.17.141128 testbuild. But more people should download the testbuild or the next VB version to test it. 
     51) segfaults and crashes on the 6.1.16. This is what this issue about and it seems to be fixed in 6.1.17.141128 testbuild. But more people should download the testbuild or the next VB version to test it. 
    66 
    7 2) a permanent freeze of the testbuild after few hours of work. I am not sure if it is relevant to this issue because it happens without any copy&paste operation. But it may be caused by your patches. The 6.1.16 didn't freeze like that. Does the testbuild only contain your patch and do not have any other current development changes? May be it is caused by changes in 3d acceleration driver or whatever. But it seams that I am the only one who downloaded the testbuild and other users here didn't face this bug. The guest doesn't crash and I can't give you the coredump. I can attach gdb to the running guest when this type of freeze happens in testbuild and give you the stacktraces of all the running threads if you want. 
     72) a permanent freeze of the testbuild after few hours of work. I am not sure if it is relevant to this issue because it happens without any copy&paste operation. But it may be caused by your patches. The 6.1.16 didn't freeze like that. Does the testbuild only contain your patch and do not have any other current development changes? May be it is caused by changes in 3d acceleration driver or whatever. But it seems that I am the only one who downloaded the testbuild and other users here didn't face this bug. The guest doesn't crash and I can't give you the coredump. I can attach gdb to the running guest when this type of freeze happens in testbuild and give you the stacktraces of all the running threads if you want. 
    88 
    9 3) a temporal freeze with "formats 0x1 to 'UTF8_STRING'" and "VERR_TIMEOUT" in the log, mentioned by me, @Borrelworst, @simonc and everybody else. It doesn't cause guest to abort or crash on the 6.1.17.141128 testbuild. It is not a permanent freeze, it is just a slight delay (i.e. TIMEOUT) guest continue to run absolutely normal before and after it. This VERR_TIMEOUT was in the logs all the time, 6.1.16 and the testbuild. It seams that you didn't touch it by your patches. To reproduce I just run the WinXP guest, abuse it in some way (kill and run VBoxTray.exe etc) and then copy unicode string from guest to host. 
     93) a temporal freeze with "formats 0x1 to 'UTF8_STRING'" and "VERR_TIMEOUT" in the log, mentioned by me, @Borrelworst, @simonc and everybody else. It doesn't cause guest to abort or crash on the 6.1.17.141128 testbuild. It is not a permanent freeze, it is just a slight delay (i.e. TIMEOUT) guest continue to run absolutely normal before and after it. This VERR_TIMEOUT was in the logs all the time, 6.1.16 and the testbuild. It seems that you didn't touch it by your patches. To reproduce I just run the WinXP guest, abuse it in some way (kill and run VBoxTray.exe etc) and then copy unicode string from guest to host. 
    1010 
    1111I think you should try to run WinXP guest with installed guest additions on linux host (Ubuntu 20.04.1 X.org Gnome in my case) running 6.1.17.141128 testbuild, and leave it running in the background for 12 hours or so. You would observe either (2) or (3) freezes. If not then... well... I don't know, the (3) seems like a very easy reproducible bug and everyone here is mentioning it. 

www.oracle.com
ContactPrivacy policyTerms of Use