Re: libpthread.so.2 compatibility

On Sun, 4 Jun 2006, Daniel Eischen wrote:

> On Sun, 4 Jun 2006, John Hay wrote:
>
>> On Sun, Jun 04, 2006 at 02:26:51PM -0400, Daniel Eischen wrote:
>>> On Sun, 4 Jun 2006, John Hay wrote:
>>> 
>>>> Actually one does not even need a big complex app to see the problem.
>>>> Just copy /sbin/ggatec from 6.1 or 6.1-stable to a current box and you
>>>> will see it happen:
>>>> 
>>>> #######
>>>> angel:~ > uname -a
>>>> FreeBSD angel.cids.org.za 7.0-CURRENT FreeBSD 7.0-CURRENT #2: Sun May 28
>>>> 11:06:16 SAST 2006
>>>> jhay_at_angel.cids.org.za:/usr/src/sys/i386/compile/ANGEL  i386
>>>> angel:~ > ssh zibbi "uname -a"
>>>> FreeBSD zibbi.meraka.csir.co.za 6.1-STABLE FreeBSD 6.1-STABLE #4: Thu May
>>>> 25 06:11:44 SAST 2006
>>>> jhay_at_zibbi.meraka.csir.co.za:/usr/src/sys/i386/compile/ZIBBI  i386
>>>> angel:~ > scp -p zibbi:/sbin/ggatec /tmp/
>>>> ggatec                                        100%   16KB   8.1KB/s 
>>>> 00:02
>>>> angel:~ > /tmp/ggatec
>>>> Segmentation fault (core dumped)
>>>> #######
>>> 
>>> It is probably the networking ABI changes in libc.  There was a short 
>>> period
>>> of time when there were ABI changes in libc.so.6 in -current -- before 
>>> libc
>>> was bumped to libc.so.7.  What happens when you try moving a -stable
>>> libc.so.6
>>> to the -current machine?
>> 
>> Ok, I did that but it still core dump in pthread_setcancelstate()
>
> I don't know then.  If recompiling it fixes the problem, then
> something in /usr/include changed.  All the pthread_foo_t types
> are pointers to things allocated by the library, and I don't
> think any of them changed anyways.

How old was your system when you upgraded to -current?  There
were changes to malloc() in libc.so.6 which have been in -current
for a while, and libpthread is dependent on some internal locks
in libc.  If you are using a libc.so.6 before jasone's malloc()
changes and a newer libpthread, then that won't work.  When you
recompile, your binaries will be linked to libc.so.7, and
libpthread.so.2 will find the correct locks.  If you don't
find the following:

   $ readelf -s /lib/libc.so.6 | grep _malloc
      275: 0005f65c   139 FUNC    GLOBAL DEFAULT    8 _malloc_postfork
      299: 000e96d0     4 OBJECT  GLOBAL DEFAULT   19 _malloc_options
      870: 0005f5d0   139 FUNC    GLOBAL DEFAULT    8 _malloc_prefork
     2486: 000d1fd8     4 OBJECT  GLOBAL DEFAULT   11 _malloc_message

_malloc_postfork and _malloc_prefork in libc.so.6, then that is
probably why libpthread is failing.

-- 
DE