Re: Problem with 1.1.9?

Chris Pickett wrote:
> Carl Lebsack wrote:
> 
>> On Sun, 6 Feb 2005, Chris Pickett wrote:
>>
>>
>>> It's dying
>>>
>>> Carl Lebsack wrote:
>>>
>>>> Sure enough a segfault.  I still don't understand what's going on.
>>>>
>>>> Carl lebsack
>>>>
>>>> #0  0x40136a69 in _int_free () from /lib/libc.so.6
>>>> #1  0x4013585c in free () from /lib/libc.so.6
>>>> #2  0x400176aa in _svmf_free (ptr=0x809f520) at system.c:422
>>>> #3  0x4001c607 in _svmh_gmfree_ints (ptr=0x804d698) at 
>>>> global_alloc.c:750
>>>> #4  0x40036961 in _svmf_initialize_instruction_indices 
>>>> (env=0x804dcd8,    method=0x412b2e50) at prepare_code.c:34
>>>
>>>
>>>
>>> It's presumably dying trying to free something that isn't allocated.
>>>
>>>   /* allocate instruction_indices array */
>>>   if (vm->class_loading.method_preparation.instruction_indices != NULL)
>>>     {
>>>       _svmm_gmfree_ints (vm->class_loading.method_preparation.
>>>                          instruction_indices);
>>>     }
>>>
>>> What hardware, OS, kernel, distribution, gcc, and glibc are you 
>>> using? Output of uname -a?  Did you reinstall sablevm-classpath as 
>>> well?  It's changed in the meantime.
>>
>>
>>
>> Intel Pentium M, Slackware Linux, 2.6.0, gcc 3.3.2, glibc 2.3.1.  I 
>> did reinstall classpath.
> 
> 
> Upgrade your kernel + kernel headers is my advice.  Even if it doesn't 
> help you (I have no idea if it will), there have been fairly significant 
> improvements since 2.6.0.

And if that doesn't work, you might want to try updating your glibc 
since that's where it's actually dying (hard to say if it's really 
SableVM's fault in this case), although I'm not sure how painful that is 
in terms of updating everything else, I've never used Slackware.

Chris