VirtualBox

Opened 8 years ago

Closed 8 years ago

#15315 closed defect (invalid)

VirtualBox SEGV:s on s12_96 host, no problems on s12_95

Reported by: Thomas Törnblom Owned by:
Component: host support Version: VirtualBox 5.0.16
Keywords: Cc:
Guest type: Solaris Host type: Solaris

Description (last modified by Frank Mehnert)

After upgrading my host and guest to s12_96 VirtualBox SEGV:s within minutes after starting the guest. Appears to be networking related.

Stack trace:

# mdb core.VirtualBox.2715 
Loading modules: [ libc.so.1 libnvpair.so.1 libuutil.so.1 ld.so.1 ]
> $C
ffff80ffbe24b920 libc.so.1`mutex_lock_impl+0x4c()
ffff80ffbe24b930 libc.so.1`mutex_lock+0xc()
ffff80ffbe24b960 VBoxRT.so`RTSemEventSignal+0x7c()
ffff80ffbe24b980 VBoxRT.so`rtReqQueueSubmit+0x48()
ffff80ffbe24b9b0 VBoxRT.so`RTReqSubmit+0xc8()
ffff80ffbe24ba00 VBoxRT.so`RTReqQueueCallV+0x1a5()
ffff80ffbe24baf0 VBoxRT.so`RTReqQueueCallEx+0x85()
ffff80ffbe24bb30 VBoxDD.so`slirp_output+0x62()
ffff80ffbe24bbc0 VBoxDD.so`ip_output0+0x482()
ffff80ffbe24bc00 VBoxDD.so`udp_output+0x68()
ffff80ffbe24be30 VBoxDD.so`sorecvfrom+0x203()
ffff80ffbe24be90 VBoxDD.so`slirp_select_poll+0x3b4()
ffff80ffbe24bef0 VBoxDD.so`_Z19drvNATAsyncIoThreadP9PDMDRVINSP9PDMTHREAD+0x134()
ffff80ffbe24bf30 VBoxVMM.so`_Z15pdmR3ThreadMainP11RTTHREADINTPv+0x6a()
ffff80ffbe24bf50 VBoxRT.so`rtThreadMain+0x2c()
ffff80ffbe24bf80 VBoxRT.so`_Z18rtThreadNativeMainPv+0x51()
ffff80ffbe24bfa0 libc.so.1`_thrp_setup+0xa4()
ffff80ffbe24bfb0 libc.so.1`_lwp_start()
> 

Change History (6)

comment:1 by Thomas Törnblom, 8 years ago

Appears there is a 256k limit on attachments. My bzipped core is 33M.

comment:2 by Frank Mehnert, 8 years ago

Description: modified (diff)

comment:3 by Thomas Törnblom, 8 years ago

Tried it with 5.0.18, and it fails the same.

comment:4 by Thomas Törnblom, 8 years ago

Tried switching from NAT to bridged networking and back, and it appears to stay up now.

May have been some old crud somewhere.

Running 5.0.18 on a s12_96 host with a s12_96 guest now.

comment:5 by Thomas Törnblom, 8 years ago

No, it just took longer, but a few hours after start it crashed with the same stack.

comment:6 by Valery Ushakov, 8 years ago

Resolution: invalid
Status: newclosed

Solaris regression.

Note: See TracTickets for help on using tickets.

© 2023 Oracle
ContactPrivacy policyTerms of Use