Opened 5 years ago

Last modified 3 years ago

#15779 new bug

NetPositive cause unkillable thread and make shutdown impossible

Reported by: X512 Owned by: nobody
Priority: normal Milestone: Unscheduled
Component: System/Kernel Version: R1/Development
Keywords: Cc:
Blocked By: Blocking:
Platform: All

Description

This is hrev53937.

Steps to reproduce:

  1. Open NetPositive.
  2. Drop /boot/system/documentation/BeBook/index.html that is located in be_book package on NetPositive window.
  3. One of NetPositive threads will cause 100% CPU core load in kernel space.
  4. After attempt to kill NetPositive, thread move to "Gone teams...".
  5. It attempt to shutdown system, it will lock on "Clean up a bit".

100% reproducible on my environment.

Attachments (1)

CIMG4614.JPG (3.4 MB ) - added by X512 5 years ago.
Backtrace of NetPositive thread that cause problem.

Change History (4)

by X512, 5 years ago

Attachment: CIMG4614.JPG added

Backtrace of NetPositive thread that cause problem.

comment:1 by X512, 5 years ago

previous_syslog contains a lot of repeating records like this:

vm_soft_fault: va 0x185d0000 not covered by area in address space
vm_page_fault: vm_soft_fault returned error 'Bad address' on fault at 0x185d0000, ip 0x8013e02b, write 0, user 0, thread 0x43a9

comment:2 by waddlesplash, 5 years ago

Very strange, do we have some sort of infinite loop in the fault handler?

comment:3 by cocobean, 3 years ago

X512: Be Book uses PNG images which may trigger the NetPositive issue.

Last edited 3 years ago by cocobean (previous) (diff)
Note: See TracTickets for help on using tickets.