[haiku-bugs] Re: [Haiku] #6129: Kernel hang after allocate_commpage_entry

  • From: "andreasf" <trac@xxxxxxxxxxxx>
  • Date: Sun, 06 Jun 2010 05:05:40 -0000

#6129: Kernel hang after allocate_commpage_entry
---------------------------+------------------------------------------------
 Reporter:  andreasf       |       Owner:  axeld         
     Type:  bug            |      Status:  new           
 Priority:  normal         |   Milestone:  R1            
Component:  System/Kernel  |     Version:  R1/Development
 Keywords:                 |    Platform:  PowerPC       
Blockedby:  5193           |       Patch:  0             
 Blocking:  1048           |  
---------------------------+------------------------------------------------

Comment(by andreasf):

 In #5193 I've attached patches that no longer require the above
 workaround.

 At r37025 with those patches I now get:
 {{{
 INIT: Init Device Mapper
 PANIC: remove page 0x8681c098 from cache 0x85825c30: page still has
 mappings!

 Welcome to Kernel Debugging Land...
 Running on CPU 0
 Current thread pointer is 0x86051000, which is an address we can't read
 from.
 stack trace for thread 0x2 "undertaker"
     kernel stack: 0xffff8000 to 0xffffc000
 frame            caller     <image>:function + offset
 ffffbb70 (+ 240) 80164394   <kernel_ppc>:_ZL11stack_traceiPPc + 0x0350
 ffffbc60 (+  16) 8016450c   <kernel_ppc>:arch_debug_stack_trace + 0x0018
 ffffbc70 (+  16) 800b8208   <kernel_ppc>:_ZL22stack_trace_trampolinePv +
 0x0010
 ffffbc80 (+  16) 8016a520
 <kernel_ppc>:arch_debug_call_with_fault_handler + 0x002c (nearest)
 ffffbc90 (+  48) 800b89dc   <kernel_ppc>:debug_call_with_fault_handler +
 0x0078
 ffffbcc0 (+ 112) 800b9f3c
 <kernel_ppc>:_ZL20kernel_debugger_loopPKcS0_P13__va_list_tagl + 0x0264
 ffffbd30 (+  80) 800ba300
 <kernel_ppc>:_ZL20kernel_debugger_internalPKcS0_P13__va_list_tagl + 0x01c8
 ffffbd80 (+ 144) 800ba68c   <kernel_ppc>:panic + 0x00ac
 ffffbe10 (+  64) 8015c3d4   <kernel_ppc>:_ZN7VMCache6DeleteEv + 0x00d0
 ffffbe50 (+  48) 8015b144   <kernel_ppc>:_ZN7VMCache6UnlockEb + 0x0188
 ffffbe80 (+  48) 801434b8
 <kernel_ppc>:_ZL11delete_areaP14VMAddressSpaceP6VMAreab + 0x027c
 ffffbeb0 (+ 160) 801434b8   <kernel_ppc>:vm_delete_area + 0x0224
 ffffbf50 (+  16) 80143940   <kernel_ppc>:delete_area + 0x0038
 ffffbf60 (+ 128) 800a87dc   <kernel_ppc>:_ZL10undertakerPv + 0x01c8
 ffffbfe0 (+  16) 800a484c
 <kernel_ppc>:_ZL28_create_kernel_thread_kentryv + 0x0020
 ffffbff0 (+16400) 8016a4e4   <kernel_ppc>:ppc_kernel_thread_root + 0x0010
 (nearest)
 kdebug> info
 kernel build: Jun  3 2010 12:49:23
 SVN revision: 37025

 cpu count: 1, active times:
   [1] 2002226
 pages:          10333 (131072 max)
 sems:           42 (65536 max)
 ports:          0 (4096 max)
 threads:        12 (4096 max)
 teams:          1 (2048 max)
 kdebug> page 0x8681c098
 queue_next,prev: 0x8686c06c, 0x00000000
 physical_number: 0xa32
 cache:           0x85824c30
 cache_next:      0x00000000
 state:           wired
 wired_count:     1
 usage_count:     0
 busy:            0
 busy_writing:    0
 accessed:        0
 modified:        0
 accessor:        -1
 area mappings:
 kdebug> cache 0x85824c30
 CACHE 0x85824c30:
   ref_count:    0
   source:       0x00000000
   type:         RAM
   virtual_base: 0x0
   virtual_end:  0x4000
   temporary:    1
   scan_skip:    1
   lock:         0x85824c84
   lock.holder:  2
   areas:
   consumers:
   pages:
         3 in cache
 kdebug>
 }}}
 ''Note that the kernel debugger is now in fact working interactively!''

 I see no direct relation to #5193, since the {{{virtual_base}}} is below
 the kernel area and thus no area would be created for such a range...

-- 
Ticket URL: <http://dev.haiku-os.org/ticket/6129#comment:5>
Haiku <http://dev.haiku-os.org>
Haiku - the operating system.

Other related posts: