Tracker crash

[vidrep]

Tracker fails to close after trying to save a screnshot. Attached debugger to tracker and ran a back trace. syslog and debug report attached.

[waddlesplash]

USER: app application/x-vnd.haiku-screenshot-cli send to client failed: Bad port ID looks like the culprit.

[mmlr]

Can you clarify what you mean by "Tracker fails to close"? Do you mean one of the Tracker windows can't be closed anymore or did you try to quit the whole Tracker?

[vidrep]

After trying to save a screenshot I saw that tracker was still running in the Deskbar.

[vidrep]

hrev50830 x86_gcc2 Tracker crash after mounting another Haiku partition and doing a file copy. Crash occurred when attempting to unmount the partition. Debug report attached (not sure if the same as before or not)

[vidrep]

KERN: 924: DEBUGGER: Looper must be locked.
KERN: debug_server: Thread 924 entered the debugger: Debugger call: `Looper must be locked.'
KERN: stack trace, current PC 0x6067a114  commpage_syscall + 0x4:
KERN:   (0x7169cfe8)  0x1fcd7ad  check_lock__7BLooper + 0x45
KERN:   (0x7169d018)  0x20aeb05  _CheckLockAndSwitchCurrent__C5BView + 0x29
KERN:   (0x7169d048)  0x20aa641  Invalidate__5BViewG5BRect + 0xf9
KERN:   (0x7169d088)  0x204e6e6  Highlight__9BMenuItemb + 0x32
KERN:   (0x7169d0d8)  0x14c3a0c  Highlight__Q28BPrivate13ModelMenuItemb + 0x28
KERN:   (0x7169d108)  0x204edea  Select__9BMenuItemb + 0x62
KERN:   (0x7169d138)  0x204676d  _SelectItem__5BMenuP9BMenuItembN22 + 0x95
KERN:   (0x7169d168)  0x2044b8a  _UpdateStateClose__5BMenuP9BMenuItemRC6BPointRCUl + 0x13e
KERN:   (0x7169d1b8)  0x2044205  _Track__5BMenuPil + 0x5a1
KERN:   (0x7169d258)  0x2043151  Track__5BMenubP5BRect + 0x8d
KERN:   (0x7169d298)  0x205f19f  _StartTrack__10BPopUpMenuG6BPointbT2P5BRect + 0x97
KERN:   (0x7169d2e8)  0x205f083  _thread_entry__10BPopUpMenuPv + 0x4f
KERN:   (0x7169d338)  0x9bc053  thread_entry + 0x23

[diver]

The crash seems to be unrelated to the original issue. Seems to happen somewhere inside BMenuItem?

[vidrep]

I can't recall if I've seen this bug again since it was first reported. Closing.