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

#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):

 It comes from USB bus manager's {{{new(std::nothrow) Stack()}}}.

 Before [source:haiku/trunk/src/add-
 ons/kernel/bus_managers/usb/usb.cpp?rev=36247#L75 that call] I can
 `dprintf()` a test message just fine, while as first instruction of the
 {{{Stack()}}} constructor I get garbage printed (same for
 {{{TRACE_USB}}}).

 Seeing nothing fundamentally wrong with the constructor itself, I could
 think of two possible reasons for this failure:
  * {{{new(std::nothrow)}}} might be broken in this context (it looks like
 a simple {{{malloc}}} though if I read
 [source:haiku/trunk/headers/private/kernel/util/kernel_cpp.h?rev=33670#L43
 kernel_cpp.h] right), or
  * something could be going wrong with the ELF sections (ctors or data).
 There is a special add-on linker script but it seems
 [source:haiku/trunk/build/jam/KernelRules?rev=30825#L112 it is not used].

 Concerning your PLT observations, the
 [source:haiku/trunk/src/system/ldscripts/ppc/kernel.ld?rev=34857 kernel
 linker script] does include multiple mentions of {{{plt}}} sections. And
 then there's the
 [source:haiku/trunk/src/system/kernel/Jamfile?rev=36158#L85 removal of ppc
 linkhack.so -shared flag] that might have undesired effects?

-- 
Ticket URL: <http://dev.haiku-os.org/ticket/6129#comment:11>
Haiku <http://dev.haiku-os.org>
Haiku - the operating system.

Other related posts: