Opened 10 years ago

Last modified 4 months ago

#4350 assigned bug

App server heap corruption-related crashes

Reported by: anevilyak Owned by: nobody
Priority: normal Milestone: R1
Component: Servers/app_server Version: R1/pre-alpha1
Keywords: Cc:
Blocked By: Blocking:
Has a Patch: no Platform: All

Description

On several different occasions in the past week or two I've encountered crashes that look something like the following:

#1: debugger()
#2: __assert_fail()
#3: BPrivate::hoardHeap:findAvailableSuperblock()
#4: BPrivate::threadHeap::malloc()
#5: malloc()
#6; operator new()
#7-#13 - various ops chaining off an std::map<long,
BPrivate::BTokenSpace::token_info> insertion.
#14: BPrivate::BTokenSpace::SetToken(this=0x18055f58, token=35,
type=1, object=0x1861a048)
#15: View::AttachedToWindow()
#16: View::AddChild()
#17: ServerWindow::_DispatchMessage()
#18: ServerWindow::_MessageLooper()
#19: MessageLooper::_message_thread()
#20: thread_entry()

Pardon the lack of detail but alt+sysrq+d appears to be unusable from the gdb session used for app_server crashes. I've encountered this one while trying to reproduce #4343, and #698.

Change History (5)

comment:1 Changed 10 years ago by jackburton

tried F12 by chance ? I think it's still mapped to it while in the app_server gdb session.

comment:2 Changed 10 years ago by anevilyak

Was not aware of that, will try next time.

comment:3 Changed 9 years ago by bonefish

Assertion in the same function reported in #4164. Doesn't contain any more helpful info either, though.

comment:4 Changed 2 years ago by axeld

Owner: changed from axeld to nobody
Status: newassigned

comment:5 Changed 4 months ago by waddlesplash

Anyone seen these since?

Note: See TracTickets for help on using tickets.