Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 4 Jun 2006 09:54:14 +0200
From:      John Hay <jhay@meraka.org.za>
To:        current@freebsd.org
Subject:   libpthread.so.2 compatibility
Message-ID:  <20060604075414.GA47483@zibbi.meraka.csir.co.za>

next in thread | raw e-mail | index | archive | help
Hi Guys,

Should the -current libpthread.so.2 be compatible with apps of a few
months ago? The commit messages made me think that it should, but it
isn't.

I have a -current machine that was last compiled in January with
packages also from that time and things were all working. So I
upgraded (compiled from source) the FreeBSD part but did not
upgrade my ports/packages. I then found that the ones using
libpthread.so.2 was coredumping. Ones that I noticed was ruby18,
gnome-session and also the java programs from the
diablo-jdk-freebsd6-1.5.0.06.00.tbz package that I just downloaded
and installed from the FreeBSD Foundation site. If I just put the
January version of libpthread.so.2 back, they all work. They all
seem to die in pthread_setcancelstate according to gdb:

##################
angel:~ > gdb /usr/local/diablo-jdk1.5.0/bin/javac javac.core
GNU gdb 6.1.1 [FreeBSD]
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type "show copying" to see the conditions.
There is absolutely no warranty for GDB.  Type "show warranty" for details.
This GDB was configured as "i386-marcel-freebsd"...(no debugging symbols found)...
Core was generated by `javac'.
Program terminated with signal 11, Segmentation fault.
Reading symbols from /lib/libz.so.3...(no debugging symbols found)...done.
Loaded symbols for /lib/libz.so.3
Reading symbols from /lib/libpthread.so.2...(no debugging symbols found)...done.Loaded symbols for /lib/libpthread.so.2
Reading symbols from /lib/libc.so.6...(no debugging symbols found)...done.
Loaded symbols for /lib/libc.so.6
Reading symbols from /libexec/ld-elf.so.1...(no debugging symbols found)...done.Loaded symbols for /libexec/ld-elf.so.1
#0  0x280bb46e in pthread_setcancelstate () from /lib/libpthread.so.2
[New LWP 100270]
(gdb) bt
#0  0x280bb46e in pthread_setcancelstate () from /lib/libpthread.so.2
#1  0x280ac6d8 in pthread_create () from /lib/libpthread.so.2
#2  0x280b82a6 in pthread_join () from /lib/libpthread.so.2
#3  0x280bc1aa in pthread_testcancel () from /lib/libpthread.so.2
#4  0x280bd939 in __error () from /lib/libpthread.so.2
#5  0x280a3c19 in ?? () from /lib/libpthread.so.2
#6  0xbfbfe978 in ?? ()
#7  0x28078118 in ?? () from /libexec/ld-elf.so.1
#8  0xbfbfe928 in ?? ()
#9  0x280580d5 in _rtld_error () from /libexec/ld-elf.so.1
#10 0x2805ba93 in _rtld () from /libexec/ld-elf.so.1
#11 0x280578f2 in .rtld_start () from /libexec/ld-elf.so.1
(gdb)
################################

If this is the expected behaviour, that is fine, but otherwise this is
just a heads-up that the compatibility isn't working. Version 1.55 of
libpthread/Makefile makes one think compatibility was intended.

It is not a huge train smash for me. Between recompiling some and using
/etc/libmap.conf for others, I'm up and running again.

John
-- 
John Hay -- John.Hay@meraka.csir.co.za / jhay@FreeBSD.org



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060604075414.GA47483>